uva-dare (digital academic repository) advancements in … · 2 3 ph.d thesis of thierry g....

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Advancements in classification, treatment and outcome of radial head fracturesGuitton, T.G.

Link to publication

Citation for published version (APA):Guitton, T. G. (2011). Advancements in classification, treatment and outcome of radial head fractures

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 17 Feb 2019

http://dare.uva.nl/personal/pure/en/publications/advancements-in-classification-treatment-and-outcome-of-radial-head-fractures(9a7218aa-38b8-41e4-a261-f76dd639e6fd).html

This thesis shows that advancements in technical analysis, imaging modalities, increased interest in psychosocial aspects of treatment and the availa-bility of long-term outcome data can help improve classification, treatment and outcome in fractures of the radial head. It is science that created these advancements and through adequate scientific evaluation of these advancements we can continue creating more effective treatments for patients.

Thierry G. Guitton

Uitnodiging

Hierbij nodig ik u van harte uit voor het bijwonen van de openbare verdediging vanmijn proefschrift

Donderdag 21 april 2011 14.00 uur AgnietenkapelOudezijds Voorburgwal 231Amsterdam

Aansluitend is er een receptie.

Thierry G. GuittonEemsstraat 40 H1079 TJ Amsterdam06 81 43 03 [email protected]

PArAnimfEnSasja Heetveld (06 30 61 90 88)ralph v.d. Houten (06 11 36 38 45)

[email protected]

Advancements in Classification, Treatment and Outcome of Radial Head Fractures


Ad

vAn

cemen

ts in clA

ssificAtion

, treAtmen

t An

d o

Utco

me o

f rAd

iAl H

eAd

frActU

res tHierry g

. gU

itton

omslag 144 pag_correctie.indd 1 10-02-11 23:05

Behorende Bij het proefschrift

“Advancements in Classification, Treatment and Outcome of Radial Head Fractures”

1. Everygreatadvanceinsciencehasissuedfromanewaudacityoftheimagination.(John Dewey, 1859 – 1952)

2. Ifindoubt,resect.(B.F. Morrey)

3. Thevolumeandproximalarticularsurfaceareaoftheradialheadcanbeestimatedbasedonanatomicalmeasurementsandgender.(This thesis)

4. Smallfragmentsaremorecommoninpartialradialheadthaninwholeradialheadfractures.(This thesis)

5. AbouthalfoftheMasonType2radialheadfractureswouldnotsatisfytheBrobergandMorreycriteriatobeconsideredType2fractures.(This thesis)

6. Thethresholdof3fragmentsinthedecisionbetweenopenreduction-internalfixationandprostheticreplacementmaynothelpguidemanagement.(This thesis)

7. Three-dimensionalCTimagesareeasierforsurgeonstointerpret.(This thesis)

8. Increasinglevelsofsophisticationinimagingandmodelingimprovedthesensitivityfordiagnosisoffracturecharacteristicsanddecreasedobservervariationbetweensurgeonandfirstassistant.(This thesis)

9. Fractureclassificationandcharacterizationbasedon3Dimagingandmodelsaremoreaccurateandreliable.(This thesis)

10. Patientsrecovergreatermotionanddosomorerapidlyafterinjurywhentheyareconfidentandfeelgoodaboutstretchingtheirarm.(This thesis)

11. Postinjuryactivitiesandoccupationarenotimportantriskfactorsforthedevelop-mentoradvancementofradiographicarthrosis.(This thesis)

12. Ifyounevermissedanairplane,youwastedtoomuchtimeofyourlifeatairports.(Andrei Shleifer)

13. It’sallinthehead,theradialhead.

Thierry G. Guitton

Amsterdam,21April2011

Stellingen

stellingen_02.indd 1 22-02-11 08:34

1

Thierry G. Guitton


BW_def.indd 1 07-02-11 22:25

2 3

Ph.D Thesis of

Thierry G. Guitton, MScPhD Research FellowHarvard Medical SchoolOrthopaedic Hand and Upper Extremity ServiceMassachusetts General Hospital

ADDress of corresPonDence

Thierry G. Guitton, MScMassachusetts General HospitalOrthopaedic Hand and Upper Extremity ServiceYawkey Center Suite 210055 Fruit StreetBoston, MA 02114, USAE-mail: [email protected]

©2011 Thierry G. Guitton, Amsterdam, the Netherlands

This thesis was prepared at the Orthopaedic Hand and Upper Extremity Service, Massachusetts

General Hospital, Harvard Medical School, Boston, Massachusetts, United States of America and

the Orthotrauma Research Center Amsterdam, Academic Medical Center, University of Amster-

dam, the Netherlands.

This thesis was supported by: Harvard Orthopedic Research Fund, Stichting Wetenschappelijk

Onderzoek Orthopaedische Chirurgie, Stichting Vreedefonds, Stichting Fundatie van de Vrijvrouwe

van Renswoude te ’s-Gravenhage, Stichting Bekker-La Bastide Fonds, Stichting Anna Fonds,

Stichting Algemeen Studiefonds, Fonds van Beuningen van Helsdingen, AUV Spinozafonds,

Noorthey Fonds and the Netherland-America Foundation.

The publication of this thesis was supported by: Nederlandse Orthopaedische Vereniging, Anna

Fonds, Synthes, DePuy and Vereniging voor Sportgeneeskunde.

ISBN/EAN: 978-94-91222-01-6

Cover Illustration: Thierry G. Guitton

Layout: Bianca Ruygrok

Print: OCC de Hoog, Oosterhout, the Netherlands


AcAdemisch proefschrift

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. D.C. van den Boomten overstaan van een door het college voor promoties ingestelde

commissie, in het openbaar te verdedigen in de Agnietenkapelop donderdag 21 april 2011, te 14.00 uur

door

Thierry Guillaume Guitton

geboren te Delft

BW_def.indd 2-3 07-02-11 22:25

4 5

Promotie commissie

PromoTores

prof. dr. C.N. van DijkProf. dr. J.B. Jupiter

coPromoTores

dr. P. Kloendr. D.C. Ring

overige leDen

Prof. dr. P.M.M. Bossuyt dr. D. EygendaalProf. dr. J.C. GoslingsProf. dr. ir. C.A. GrimbergenProf. dr. E. Schadédr. S.D. Strackee

Faculteit der Geneeskunde

A mes êtres les plus chers: mes parents, mes frères et Louise

BW_def.indd 4-5 07-02-11 22:25

6 7

Preface

Thierry continues the tradition of the Amsterdam-Boston collaborative. I know that my PhD students don’t arrive with the skills they leave with, but as I write this preface I can’t clearly remember what things were like when Thierry started–per-haps because my pride is so great in how they are finishing. Thierry is juggling a dozen active projects at a time–I can rely on him for energy, skill, and enthusiasm. He will end up with over double the number of publications than will be part of his PhD thesis including papers that branch in the appropriate use of pathology testing of ganglion cysts, surgical site infections, and measures of quality and safety and how they are influenced by patient complexity including transfer from another hospital. To complete the PhD work Thierry had to bring to fruition the quantitative 3D-CT process. That required fine-tuning both the technical aspects and the methods of statistical analysis. It also required the invention of a method for estimating in-tact bone volumes in which Thierry all succeeded. Thierry single handedly brought the Science of Variation Group to frui-tion. When the web developer hit problem after problem, Thierry discovered a way to use readily available commercial web survey tools to do the job cheaper, easier, and very reliably. The international enthusiasm for the work of this group is re-ally heartening to dedicated scientists such as Thierry and I expect our collabora-tive will make quick work of many of the current questions about observer varia-tion. Which of course will only raise new questions–but that’s what we’ve trained Thierry for. The process of executing so many scientific experiments, presenting them publically and then getting them published in peer-reviewed scientific journals hones the scientific skills. The mark of success is the enthusiasm and confidence with which a graduating PhD student such as Thierry takes on new projects and also takes on the role of a leader and manager for new PhD students, short-term visitors, and even some of the local residents and fellows.

It gives me great pride to see Thierry’s abilities in action and of course my hope is that we will continue to collaborate throughout our careers, and that he will always share my love for science. Science is what humans developed to keep from fooling themselves and to keep from being fooled by others. A useful tool to be sure and science’s accomplishments are undeniable, but when the data are counterintuitive–when they challenge our preferences and customs–many of us become uncomfortable. Not so Thierry. Having completed his PhD, Thierry now has the confidence, the comfort, and the capability to pursue more experiments, and I expect his contributions to our understanding and management of illness to be sub-stantial, and his efforts are greatly admired and appreciated.

David Ring, MD PhDAssociate Professor of Orthopaedic Surgery

Harvard Medical School

Orthopaedic Hand and Upper Extremity Service

Massachusetts General Hospital

Yawkey Center, Suite 2100

55 Fruit Street

Boston, MA 0211, USA

Tel: 617-724-3953

Fax: 617-724-8532

[email protected]

BW_def.indd 6-7 07-02-11 22:25

8 9

Table of Contents

PArT i currenT issues

Chapter 1 General Introduction GuittonTG

PArT i i clAssificATion

Chapter 2 Quantitative Measurements of the Volume and Surface Area of the Radial Head GuittonTG,vanderWerfHJ,RingD.JHandSurgAm.2010;35(3):457-63.

Chapter 3 Quantitative Three-Dimensional Computed Tomography Measurement of Radial Head Fractures GuittonTG,vanderWerfHJ,RingD.JShoulderElbowSurg.2010;19(7):973-7.

Chapter 4 Diagnostic Accuracy of Two-Dimensional and Three- Dimensional Imaging and Modeling of Radial Head Fractures GuittonTG,BrouwerKM,DyerG,ZurakowskiD,MudgalC,RingD.

Submitted

Chapter 5 Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed Tomography GuittonTG,RingD,ScienceofVariationGroup.

InrevisionJBoneJointSurgAm

PArT i i i TreATmenT

Chapter 6 Attitude Towards Stretch Pain of the Elbow After Radial Head Fracture GuittonTG,VranceanuA,RingD.

Submitted

PArT iv ouTcome

Chapter 7 Incidence and Risk Factors for the Development of Radiographic Arthrosis After Traumatic Elbow Injuries GuittonTG,ZurakowskiD,vanDijkNC,RingD.JHandSurgAm.2010;35(12):1976-80.

PArT v generAl Discussion

Chapter 8 Discussion GuittonTG

PArT vi summAry

Summary Samenvatting Glossary

APPenDices

Bibliography Acknowledgements Curriculum vitae

10

11

26

27

41

53

65

78

79

90

91

100

101

112

113119125

128

129139143

BW_def.indd 8-9 07-02-11 22:25

10 part i | current issues 11 introduction | chapter 1

Part I: Current Issues

CHAPTER 1Introduction

Thierry G. Guitton

BW_def.indd 10-11 07-02-11 22:25


IntroductionWith the evolution from homo-sapiens to the human being of today the elbow evolved from a weight bearing joint to a complex non-weight bearing articulation. The elbow joint is now a so-called ginglymus or hinge joint and consists of the hu-merus, radiar and ulnar bones. In combination with the development of the brain, the human became able to perform complex motions with this joint. From an ana-tomical point of view the elbow is very complex; it consists of relatively small bones, complex shapes of articulations and numerous adjacent neurovascular structures that all increase the difficulty of operative treatment. Paul of Aegina (625-690 A.D.) made the first description of radial head frac-tures: “The ulna and radius are sometimes fractured together and sometimes one of them only, either in the middle or at one end as at the elbow or the wrist” 1. However, Sir Astley Cooper, an English surgeon and anatomist who made historical contributions to medicine, never recognized this injury in his book titled: “Dislocations and Frac-tures” in 1822 15. Earlier recognition of this injury was probably hindered because of “thick muscle covering” 16, 58. On the other hand, the often poor results of radial head fracture treatment were noted early: Helferich recommended resection of the radial head to prevent late deformity in 1899 28. The incidence of radial head fractures is reported to be between 1.7 to 5.4 percent of all fractures 14, 34, 56. Radial head fractures are considered common injuries, found in nearly 20% of all elbow fractures 34, 67, 78. Loss of elbow motion after a radial head fracture can have a substantial impact on upper extremity function and peo-ple’s overall health status 44. A relatively simple radial head fracture can cause great impairment and disability. With the increased emphasis in healthcare on quality of life, there has been an increased interest in development of treatments for radial head fractures. As life expectancy increases, the older population continues

to grow and at the same time, live more active lives. As a consequence, the incidence of radial head fractures can be expected to increase 13. As a result there will be an increased scientific, clinical, and economic interest in the treatment of radial head fractures. The classification of radial head fractures underwent several changes over the past years. Scharplatz and Allgower based the classification of elbow fractures on the direction of force of the injury 66. Early classification of radial head fractures by Carstam 11, Bakalim 3 and Mason 48 were only based on radiographs. They failed to take associated injuries into account. At this point, the most commonly used classi-fication system is that proposed by Mason 48. Mason classified fractures of the radial head as nondisplaced (Type 1), displaced partial head (Type 2), and displaced whole head fractures (Type 3). Several authors suggested a fourth category in which there is involvement of an elbow dislocation 8, 14, 22, 54. More complex sub classifications are

proposed based on different degrees of displacement and comminution 56, 57, 72. Bro-berg and Morrey modified Mason’s classification as follows 8: Type 1 fractures involve less than 30% of the articular surface or are displaced fewer than 2 millimeters; Type 2 fractures are partial head fractures involving at least 30% of the articular surface and displaced at least 2 mm; Type 3 fractures are displaced articular fractures in-volving the entire head of the radius; and Type 4 fractures have an associated elbow dislocation 53. As most classification systems are imperfect, there is often debate in distinguishing between fracture types. For example, the percentage of involvement of the radial head or the amount of displacement that should be present for a Mason 2 fracture is arbitrary. As of today, no “ideal” classification for radial head fractures exists 5. More detailed analysis with sophisticated techniques may help to clarify these issues. To my knowledge, measurement of proximal articular surface area and ra-dial head volume has not been attempted. If a system was developed which quanti-fied fracture fragment size and injury patterns, and added the ability to estimate percentage involvement, it would make classification systems more intuitive for cli-nicians. The technologic advances in imaging of the upper-extremity have taken an immense leap in the last decade. Our group has developed a technique to quanti-tatively investigate broken bones with the use of Computed Tomography (CT). This Quantitative Three-dimensional CT (Q3D-CT) modeling technique creates a polygon mesh. This is a collection of vertices, edges and faces that defines the shape of a poly-hedral object in 3D computer graphics and solid modeling, consisting of triangles, only explicitly representing the surface. In other words, a hollow 3D model of solely the outer surface of the bony structures and fragments can be generated. This Q3D-CT modeling technique has several potential opportunities. First, it provides the opportunity to learn more about fracture patterns. For example, one can calculate volume and articular surfaces of bones. Secondly, more detailed and quantitative information can be derived from this technique concerning the specific anatomical aspects of bone that could assist the clinician in reconstruction surgery. Anteroposterior (AP) and lateral radiographs may not provide an accurate representation of the individual fracture pattern of radial head fractures 9. The addi-tion of a 45 degree oblique radiograph is helpful in recognizing the size and orienta-tion of the fracture fragments. However, the value of Magnetic Resonance Imaging (MRI) and CT scans remain under continuous scrutiny 23. Several retrospective studies have demonstrated improved injury characterization with two-dimensional (2D) CT images as compared to standard radiographs alone. These studies found that 3D re-constructions of CT scans may have advantages over standard 2D images 18, 27, 48. Three-dimensional reconstructions are relatively new, and are becoming more read-

BW_def.indd 12-13 07-02-11 22:25


ily available in most hospitals. Three-dimensional reconstructions might be more intuitive and may lead to improved identification of fracture characteristics such as fragments, fracture edges and articular surfaces. Three-dimensional physical models of elbow fractures can now be created with the use of special 3D printers. Three-dimensional physical models can even increase the advantages of the 3D reconstruc-tions. Three-dimensional imaging and 3D physical models should allow better pre-operative planning in terms of implants and equipment. Additionally, the surgeon will have better mental and psychomotor preparation. Three-dimensional imaging is also more intuitive for patients and could lead to better understanding and im-proved decision making and compliance. High quality prospective and multi-rater studies could identify the potential advantages from 3D imaging and 3D physical models over radiographs and 2D-CT. Immobilization of three to four weeks 74, passive motion and avoidance of “operative treatment” 26, removal of the fracture fragment 60 and excision of the entire head for severe comminution 30 were all recommended treatments for radial head fractures in the early 1900’s. Evidence can now be found for nearly any type of treatment. Although the radial head has been subject to research in the past, the majority consists of retrospective studies and case series. There is a lack of high qual-ity randomized and comparative trials available. Therefore the debate regarding the best treatment continues 50, 79. The focus has mainly been on the technical side of management. There is an increased interested in the recent literature in orthopedics on the psychosocial aspect in treatment of elbow trauma. The psychosocial aspects in treatment of radial head fractures could help unveil the ideal management. It was found that psychosocial factors (depression in particular) may best explain the dis-crepancy between impairment and disability 44. As many psychosocial factors are amenable to treatment, additional research along these lines is merited. For exam-ple, it can be counter-intuitive to intentionally cause pain in the setting of an injury. Vulnerability is enhanced by automatic thoughts such as “pain indicates harm”, “the pain is permanent”, or other aspects of a maladaptive pain response that psycholo-gists have termed pain catastrophizing. There is a strong interaction between de-pression and pain catastrophizing and both may be important. This needs further investigation. Post-traumatic arthritis is a form of arthritis that is caused by forced inap-propriate motion of a joint or ligament that is damaged because of a fracture. An intra-articular fracture such as a radial head fracture may increase the forces on the articular cartilage, and the articular surface will wear out faster, finally leading to arthritis. Little has been published regarding risk factors for arthrosis after elbow injury, especially in the long term. Data from multiple long-term follow-up studies of injured elbows provide the opportunity to assess the risk factors for post-traumatic elbow arthrosis after radial head fracture.

The advancements in analyzing techniques, 3D imaging and modeling, increased interest in psychosocial aspects of treatment and the recent availability of multiple long term outcome studies gives us the opportunity to further inves-tigate the classification, treatment and outcome of radial head fractures. This all could lead to improved treatment and possible better outcomes for patients. The aim of this thesis is to apply these advancements to the radial head 1) to gain fur-ther insight in classification, treatment and outcome of radial head fractures and 2) to function as a model for general improvements in orthopedic trauma surgery. In conclusion, the purpose of this thesis is: 1) to validate the Q3D-CT model-ing technique; 2) to apply this new Q3D-CT modeling technique to improve the un-derstanding of radial head fracture morphology; 3) to evaluate prospectively with a multi-rater study the influence of 3D images on classification and treatment of radial head fractures; 4) to further investigate the psychosocial aspects in radial head frac-ture treatment; 5) to identify predictors for long term consequences of radial head fractures.

Outline of the Chapters

chAPTer 2

QuAnTiTATive meAsuremenTs of The volume AnD surfAce AreA of The rADiAl

heAD

ThierryG.Guitton,MSc,HuubJ.vanderWerf,MD,DavidRing,MDPhD

The morphology of the healthy radial head has been investigated with calliper ruler, osteometric board, coordinator measuring machine (CMM), X-ray, Computed Tomog-raphy (CT), Magnetic Resonance Imaging (MRI) and Computer-Aided Design (CAD) software in the past 4, 6, 10, 19, 24, 31, 33, 37, 38, 40, 41, 46, 47, 59, 65, 73, 76, 77. We developed a quantitative 3-dimensional computed tomography (Q3D-CT) modeling technique that can mea-sure size, shape and proximal articular surface area. In chapter 2, we will validate our Q3D-CT modeling technique and inves-tigate the hypothesis that analysis of normal, unfractured radial heads in patients with CT scan obtained for other reasons (intact radial head) will allow us to develop a linear regression model capable of estimating the volume and proximal articular surface area of the radial head prior to fracture based on 1 or more of the following: radial head diameter, radial neck diameter, coronoid length diameter, height, weight and gender.

BW_def.indd 14-15 07-02-11 22:25


chAPTer 3

QuAnTiTATive Three-DimensionAl comPuTeD TomogrAPhy meAsuremenT of

rADiAl heAD frAcTures

ThierryG.Guitton,MSc,HuubJ.vanderWerf,MD,DavidRing,MDPhD

Radial head fractures are usually classified according to the size and displacement of the fracture fragments into partial and whole head fractures as per the Mason clas-sification system 34, 48, 53. The quantitative aspects of the classification of these injuries, such as the thresholds of 30% surface area and 2 millimeter displacement are rela-tively arbitrary and based on radiographs 8, 48. Additionally, the classification of radial head fractures according to Broberg and Morrey’s modification of the Mason classifi-cation has substantial observer variation 49, 52. Three-dimensional computed tomogra-phy models provide more detailed information of the fractured bone and provide an opportunity to quantify fracture characteristics better than radiographs 64. In chapter 3, we applied the Q3D-CT analysis technique to a consecutive se-ries of adult patients with a fracture of the radial head with the objective of devel-oping quantitative assessments of radial head fracture fragments that might help clarify current classification systems and decision-making.

chAPTer 4

DiAgnosTic AccurAcy of Two-DimensionAl AnD Three-DimensionAl imAging

AnD moDeling of rADiAl heAD frAcTures

ThierryG.Guitton,MSc,KimBrouwer,MSc,GeorgeDyer,MD,DavidZurakowski,PhD,ChaitanyaMudgal,MD,David

Ring,MDPhD

Optimal management of radial head fractures is debated, but accurate preoperative radiological characterization of the fracture may facilitate management. Prior stud-ies have demonstrated improved agreement in characterization and classification of various fractures with 3D-CT compared to 2D-CT images and radiographs 7, 12, 21, 35, 36, 42,

43, 70, 71. These studies were based upon retrospective data and the reference standard was based upon surgeon recollection and the medical record (e.g. operative notes). We believe that we can measure the accuracy of 3D-CT imaging better prospectively. In addition, 3D physical models, that are constructed based on CT images, can actu-ally be held in the hand and may add even more to the evaluation of fracture charac-teristics and surgical planning. In chapter 4, we will investigate if the classification and characterization of fractures of the radial head is more accurate with 3D than 2D-CT images and radio-graphs, using a prospective study design with intraoperative inspection as the refer-ence standard.

chAPTer 5

inTerobserver reliAbiliTy of rADiAl heAD frAcTure clAssificATion:

Two-DimensionAl vs. Three-DimensionAl comPuTeD TomogrAPhy

ThierryG.Guitton,MSc,Ring,MDPhD,ScienceofVariationGroup

The classification of radial head fractures according to Broberg and Morrey’s modi-fication of the Mason classification 8 has substantial interobserver variation 49, 52. Treatment decisions for radial head fractures are often based on radiological cri-teria and measurements according to Broberg and Morrey’s modification of the Mason classification 34, 48. Evidence suggests that more sophisticated images such as 3D-CT improve intraobserver reliability more than interobserver reliability 18, 27. A major limitation of most studies of observer variation is the use of only a few ob-servers, most of them typically relatively junior surgeons. In chapter 5, a new collaboration motivated to better understand interob-server variation 51, consisting of observers who have completed all training and are independently treating patients, provides an opportunity to further investigate in-terobserver variability and how to reduce it. We will investigate if 3D-CT images im-prove the interobserver reliability of the classification and characterization of radial head fractures over 2D-CT and radiographs.

chAPTer 6

ATTiTuDe TowArDs sTreTch PAin of The elbow AfTer rADiAl heAD frAcTure

ThierryG.Guitton,MSc,Ana-MariaVranceanu,PhD,DavidRing,MDPhD

Isolated stable and minimally displaced fractures of the radial head (Types 1 and 2 of the Broberg-Morrey modification of the Mason Classification) are common fractures that are usually treated non-operatively. The most common sequel of these fractures is elbow stiffness 25, 29, 32, 61, 68, 75, 80. In our experience, the elbow stiffness may be a result of excessive immobilization or ineffective stretching exercises. Research suggests that fear of pain, thinking the worst in response to nociception (pain catastroph-izing) and pain anxiety may be important determinants of recovery after an acute fracture 2, 45. Similarly, depression hinders recovery after fracture 55, 69. In chapter 6, the influence of patients who agree or disagree that pain is useful for recovery will be evaluated. This prospective study was designed to test the hypothesis that agreement with the idea that “stretching of the elbow beyond the point were it becomes painful is important in recovery” leads to greater elbow range of motion one month after injury.

BW_def.indd 16-17 07-02-11 22:25


chAPTer 7

inciDence AnD risk fAcTors for The DeveloPmenT of rADiogrAPhic

ArThrosis AfTer TrAumATic elbow injuries

ThierryG.Guitton,MSc,DavidZurakowski,PhD,C.NiekvanDijk,MDPhD,DavidRing,MDPhD

Radiographic arthrosis is a common sequela of elbow trauma resulting from direct cartilage injury, instability, and articular incongruity 17, 20, 62, 63. It is understood that over the long term, many patients develop radiographic signs of arthrosis after el-bow trauma, although symptoms vary and few patients present for treatment 29, 39. Not much has been published regarding risk factors for arthrosis after elbow injury, especially in the long term. In chapter 7, data from multiple long-term follow-up studies of injured el-bows provide the opportunity to assess the risk factors for posttraumatic elbow ar-throsis on radiographs.

Summary of IntroductionThe purpose of this doctoral thesis is to apply the advancements in technical analy-sis, imaging modalities, psychosocial aspects and long term data to the treatment of radial head fractures. More specifically, it is my goal to increase our knowledge on classification, treatment and outcome of radial head fractures. This goal will be achieved by addressing the following study questions:

chAPTer 2

General aim: To investigate if Q3D-CT modeling technique based on anatomical and demographic data, that can measure size, shape, and proximal articular surface area, can be used to develop formulas that could predict the volume and proximal surface area of the intact radial head in patients with fractures of the radial head.Specific study question: Are linear regression models capable of estimating the vol-ume and proximal articular surface area of the radial head prior to fracture based on one or more of the following: radial head diameter, radial neck diameter, coronoid length diameter, height, weight and gender?

chAPTer 3

General aim: To quantitatively analyze radial head fracture fragment morphology on Q3D-CT images in terms of size, shape, and articular surface area. Specific study question: Do partial head (Mason 2) fractures and whole head frac-tures (Mason 3) have the same percentage of small fracture fragments by volume and surface area criteria?

chAPTer 4

General aim: To investigate if classification and characterization of fractures of the radial head is more accurate with 3D-CT images and 3D models than 2D-CT images and radiographs, using a prospective study design with intraoperative inspection as the reference standard.Specific study question: Do 3D-CT images and 3D models predict fracture character-istics more accurately than 2D-CT images and radiographs?

chAPTer 5

General aim: To investigate in a large web-based collaborative of experienced ortho-paedic surgeons if 3D-CT images improve the interobserver reliability of the classifi-cation of radial head fractures according to the Broberg and Morrey modification of the Mason classification.Specific study question: Do 3D-CT images improve the interobserver reliability of the classification and characterization of radial head fractures over 2D-CT and radiographs.

chAPTer 6

General aim: To investigate if agreement with the idea that “stretching of the elbow beyond the point were it becomes painful is important in recovery” leads to greater elbow range of motion one month after injury.Specific study question: Does agreement that painful stretches are an important part in recovery leads to greater motion one month after injury?

chAPTer 7

General aim: To assess the risk factors for posttraumatic elbow arthrosis on radiographs after elbow injury in the long term.Specific study question: Do different types of elbow injuries have rates of radiograph-ic arthrosis (independent of function or outcome) that are comparable at equivalent follow-up times?

BW_def.indd 18-19 07-02-11 22:25


References

1. Aegina P. Fractures and Dislocations. Translation by Adams F, editor: New Sydenham Society;

1846.

2. Arinzon Z, Gepstein R, Shabat S, Berner Y. Pain perception during the rehabilitation phase fol-

lowing traumatic hip fracture in the elderly is an important prognostic factor and treatment

tool. Disability and Rehabilitation. 2007;29(8):651-8.

3. Bakalim C. Fractures of radial head and their treatment. Acta Orthop Scand. 1970;42:320-31.

4. Beredjiklian PK, Nalbantoglu U, Potter HG, Hotchkiss RN. Prosthetic radial head components

and proximal radial morphology: a mismatch. J Shoulder Elbow Surg. 1999;8(5):471-5.

5. Bernstein J. Fracture classification systems: do they work and are they useful? J Bone Joint Surg

Am. 1994;76(5):792-3.

6. Berrizbeitia EL. Sex determination with the head of the radius. J Forensic Sci. 1989;34(5):1206-13.

7. Borrelli J, Jr., Goldfarb C, Catalano L, Evanoff BA. Assessment of articular fragment displace-

ment in acetabular fractures: a comparison of computerized tomography and plain radio-

graphs. J Orthop Trauma. 2002;16(7):449-56; discussion 56-7.

8. Broberg MA, Morrey BF. Results of treatment of fracture-dislocations of the elbow. Clin Orthop

Relat Res. 1987(216):109-19.

9. Browner BD, Jupiter, J.B., Levine, A.M., Trafton, P.G. Trauma to the adult elbow and fractures of

the distal humerus. In: McKee MD, Jupiter JB, editors. Skeletal Trauma. 4th ed. Philadelphia:

W.B. Saunders Company; 2008. p. 1507.

10. Captier G, Canovas F, Mercier N, Thomas E, Bonnel F. Biometry of the radial head: biomechani-

cal implications in pronation and supination. Surg Radiol Anat. 2002;24(5):295-301.

11. Carstam N. Operative treatment of fractures of the upper end of the radius. Acta Orthop

Scand. 1950;19:502-26.

12. Castagno AA, Shuman WP, Kilcoyne RF, Haynor DR, Morris ME, Matsen FA. Complex fractures of

the proximal humerus: role of CT in treatment. Radiology. 1987;165(3):759-62.

13. Chen NC, Jupiter JB. Management of distal radial fractures. J Bone Joint Surg Am.

2007;89(9):2051-62.

14. Conn J, Wade PA. Injuries of the elbow (a ten year review). J Trauma. 1961;1:248-61.

15. Cooper A. Dislocations and fractures of the joints. Cooper RB, editor. Boston: T.R. Marvin; 1844.

16. Desault PJ. A treatise on fractures, luxations and affection of the bones. Bichat X, editor. Trans-

lated by C. Caldwell. Philadelphia: Kimber & Conrad; 1811.

17. Doornberg J, Ring D, Jupiter JB. Effective treatment of fracture-dislocations of the olecranon

requires a stable trochlear notch. Clin Orthop Relat Res. 2004(429):292-300.

18. Doornberg J, Lindenhovius A, Kloen P, van Dijk CN, Zurakowski D, Ring D. Two and three-dimen-

sional computed tomography for the classification and management of distal humeral fractures.

Evaluation of reliability and diagnostic accuracy. J Bone Joint Surg Am. 2006;88(8):1795-801.

19. Doornberg JN, Linzel DS, Zurakowski D, Ring D. Reference points for radial head prosthesis size.

J Hand Surg Am. 2006;31(1):53-7.

20. Doornberg JN, Ring DC. Fracture of the anteromedial facet of the coronoid process. J Bone Joint

Surg Am. 2006;88(10):2216-24.

21. Edelson G, Kelly I, Vigder F, Reis ND. A three-dimensional classification for fractures of the

proximal humerus. J Bone Joint Surg Br. 2004;86(3):413-25.

22. Gaston SR, Smith FM, Baab DD. Adult injuries of the radial head and neck: importance of time

element in treatment. Am J Surg. 1949;78:631-5.

23. Geel C. Fractures of the radial head. In: McQueen MM, Jupiter JBKJF, editors. Radius and Ulna.

Boston: Butterworth-Heinemann; 1999. p. 159-68.

24. Gupta GG, Lucas G, Hahn DL. Biomechanical and computer analysis of radial head prostheses.

J Shoulder Elbow Surg. 1997;6(1):37-48.

25. Hammacher ER, van der Werken C. Radial head fractures: operative or conservative treatment?

The Greek temple model. Acta Orthop Belg. 1996;62 Suppl 1:112-5.

26. Hammond R. Fracture of the head and neck of the radius. Ann Surg. 1910;53:207.

27. Harness NG, Ring D, Zurakowski D, Harris GJ, Jupiter JB. The influence of three-dimensional

computed tomography reconstructions on the characterization and treatment of distal radial

fractures. J Bone Joint Surg Am. 2006;88(6):1315-23.

28. Helferich H. Fractures and dislocations. Translated by J. Hutchinson.: New Sidenham Society;

1899. p. 96-7.

29. Herbertsson P, Josefsson PO, Hasserius R, Karlsson C, Besjakov J, Karlsson M. Uncomplicated

Mason type-II and III fractures of the radial head and neck in adults. A long-term follow-up

study. J Bone Joint Surg Am. 2004;86-A(3):569-74.

30. Hizrot J. The treatment of simlpe fractures: a study of some end results. Ann Surg. 1912;55:338.

31. Holman DJ, Bennett KA. Determination of sex from arm bone measurements. Am J Phys An-

thropol. 1991;84(4):421-6.

32. Hotchkiss RN. Fractures of the Radial Head and Related Instability and Contracture of the

forearm. Instructional Course Lectures; 1998. p. 173-7.

33. Itamura JM, Roidis NT, Chong AK, Vaishnav S, Papadakis SA, Zalavras C. Computed tomography

study of radial head morphology. J Shoulder Elbow Surg. 2008;17(2):347-54.

34. Johnston GW. A follow-up of one hundred cases of fracture of the head of the radius with a

review of the literature. Ulster Med J. 1962;31:51-6.

35. Katz MA, Beredjiklian PK, Bozentka DJ, Steinberg DR. Computed tomography scanning of intra-

articular distal radius fractures: does it influence treatment? J Hand Surg Am. 2001;26(3):415-21.

36. Kilcoyne RF, Shuman WP, Matsen FA, 3rd, Morris M, Rockwood CA. The Neer classification of

displaced proximal humeral fractures: spectrum of findings on plain radiographs and CT scans.

AJR Am J Roentgenol. 1990;154(5):1029-33.

37. King GJ, Zarzour ZD, Patterson SD, Johnson JA. An anthropometric study of the radial head:

implications in the design of a prosthesis. J Arthroplasty. 2001;16(1):112-6.

38. Knight DJ, Rymaszewski LA, Amis AA, Miller JH. Primary replacement of the fractured radial

head with a metal prosthesis. J Bone Joint Surg Br. 1993;75(4):572-6.

BW_def.indd 20-21 07-02-11 22:25


39. Konrad GG, Kundel K, Kreuz PC, Oberst M, Sudkamp NP. Monteggia fractures in adults: long-

term results and prognostic factors. J Bone Joint Surg Br. 2007;89(3):354-60.

40. Koslowsky TC, Beyer F, Germund I, Mader K, Jergas M, Koebke J. Morphometric parameters of

the radial neck: an anatomical study. Surg Radiol Anat. 2007;29(4):279-84.

41. Koslowsky TC, Germund I, Beyer F, Mader K, Krieglstein CF, Koebke J. Morphometric parameters

of the radial head: an anatomical study. Surg Radiol Anat. 2007;29(3):225-30.

42. Kuhlman JE, Fishman EK, Ney DR, Magid D. Two-and three-dimensional imaging of the painful

shoulder. Orthop Rev. 1989;18(11):1201-8.

43. Lindenhovius A, Karanicolas PJ, Bhandari M, van Dijk N, Ring D. Interobserver reliability of coro-

noid fracture classification: two-dimensional versus three-dimensional computed tomogra-

phy. J Hand Surg Am. 2009;34(9):1640-6.

44. Lindenhovius AL, Buijze GA, Kloen P, Ring DC. Correspondence between perceived disability and

objective physical impairment after elbow trauma. J Bone Joint Surg Am. 2008;90(10):2090-7.

45. Linton SJ, Buer N, Samuelsson L, Harms-Ringdahl K. Pain-related fear, catastrophizing and pain

in the recovery from a fracture. Scandinavian Journal of Pain. 2010;1(1):38-42.

46. Mahaisavariya B, Saekee B, Sitthiseripratip K, Oris P, Tongdee T, Bohez EL, et al. Morphology of

the radial head: a reverse engineering based evaluation using three-dimensional anatomical

data of radial bone. Proc Inst Mech Eng [H]. 2004;218(1):79-84.

47. Mall G, Hubig M, Buttner A, Kuznik J, Penning R, Graw M. Sex determination and estimation of

stature from the long bones of the arm. Forensic Sci Int. 2001;117(1-2):23-30.

48. Mason ML. Some observations on fractures of the head of the radius with a review of one

hundred cases. Br J Surg. 1954;42(172):123-32.

49. Matsunaga FT, Tamaoki MJ, Cordeiro EF, Uehara A, Ikawa MH, Matsumoto MH, et al. Are clas-

sifications of proximal radius fractures reproducible? BMC Musculoskelet Disord. 2009;10:120.

50. McLaughlin HL. Some fractures with a time limit. Surg Clin North Am. 1955;35:555.

51. Morgan DL, Allen DG. Early events in stretch-induced muscle damage. J Appl Physiol.

1999;87(6):2007-15.

52. Morgan SJ, Groshen SL, Itamura JM, Shankwiler J, Brien WW, Kuschner SH. Reliability evaluation

of classifying radial head fractures by the system of Mason. Bull Hosp Jt Dis. 1997;56(2):95-8.

53. Morrey B. Radial head fractures. In: BF M, editor. The elbow and its disorders. Philadelphia: WB

Saunders; 1985. p. 355-81.

54. Morrey BF, An KN. Articular and ligamentous contribution to the stability of the elbow joint.

J Sports Med. 1983;11:315-9.

55. Mossey JM, Mutran E, Knott K, Craik R. Determinants of recovery 12 months after hip fracture:

the importance of psychosocial factors. Am J Public Health. 1989;79(3):279-86.

56. Murray R. Fractures of the head and neck of the radius. Br J Surg. 1940;28:106.

57. Paulsen JO, Tophoj K. Fracture of the head and neck of the radius. Acta Orthop Scand.

1974;45:66.

58. Petit JL. A treatise of the disease of the bones; containing an extract and complete account of

all their various kinds. Woodward T, editor. London; 1720.

59. Popovic N, Djekic J, Lemaire R, Gillet P. A comparative study between proximal radial morphol-

ogy and the floating radial head prosthesis. J Shoulder Elbow Surg. 2005;14(4):433-40.

60. Rabourdin AN. Fractures of the head of the radius. Paris: Steinheill; 1910.

61. Radin EL, Riseborough EJ. Fractures of the radial head. A review of eighty-eight cases and

analysis of the indications for excision of the radial head and non-operative treatment. J Bone

Joint Surg Am. 1966;48(6):1055-64.

62. Regel G, Weinberg AM, Seekamp A, Blauth M, Tscherne H. [Complex trauma of the elbow].

Orthopade. 1997;26(12):1020-9.

63. Ring D. Elbow stiffness associated with malunion or nonunion. In: JB J, editor. The stiff elbow.

1st ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2006. p. 41-9.

64. Rodt T, Bartling SO, Zajaczek JE, Vafa MA, Kapapa T, Majdani O, et al. Evaluation of surface and

volume rendering in 3D-CT of facial fractures. Dentomaxillofacial Radiol. 2006;35(4):227-31.

65. Ryan JR. The relationship of the radial head to radial neck diameters in fetuses and adults with

reference to radial-head subluxation in children. J Bone Joint Surg Am. 1969;51(4):781-3.

66. Scharplatz D, Allgower M. Fracture-dislocations of the elbow. Injury. 1975;7(2):143-59.

67. Schwartz RP, Young F. Treatment of fractures of the head and neck of the radius and slipped

radial epihysis in children. SurgGynecolObstet. 1933;57:528-37.

68. Schwarz N. [Conservative therapy of isolated radius head fracture--immobilization or early

functional movement treatment?]. Aktuelle Traumatol. 1983;13(3):97-102.

69. Shyu YI, Chen MC, Cheng HS, Deng HC, Liang J, Wu CC, et al. Severity of depression risk predicts

health outcomes and recovery following surgery for hip-fractured elders. Osteoporosis Inter-

national. 2008;19(11):1541-7.

70. Siebenrock KA, Gerber C. The reproducibility of classification of fractures of the proximal end

of the humerus. J Bone Joint Surg Am. 1993;75(12):1751-5.

71. Smith RM. The classification of fractures. J Bone Joint Surg Br. 2000;82(5):625-6.

72. Stankovic P. Uber die operative versorgung von fractures des proximalen radiussendes. Chirurg.

1978;49:377.

73. Swieszkowski W, Skalski K, Pomianowski S, Kedzior K. The anatomic features of the radial head

and their implication for prosthesis design. Clin Biomech (Bristol, Avon). 2001;16(10):880-7.

74. Thomas TT. Fractures of the head and the radius. Univ Pa Med Bull. 1905;18:184-97.

75. Unsworth-White J, Koka R, Churchill M, D’Arcy JC, James SE. The non-operative management of

radial head fractures: a randomized trial of three treatments. Injury. 1994;25(3):165-7.

76. van Riet RP, van Glabbeek F, Neale PG, Bortier H, An KN, O’Driscoll SW. The noncircular shape of

the radial head. J Hand Surg [Am]. 2003;28(6):972-8.

77. van Riet RP, van Glabbeek F, Neale PG, Bimmel R, Bortier H, Morrey BF, et al. Anatomical consid-

erations of the radius. Clin Anat. 2004;17(7):564-9.

78. Watson-Jones R. Fractures and other bone and joint injuries. 2nd ed. Baltimore: Williams &

Wilkins; 1941.

BW_def.indd 22-23 07-02-11 22:25


79. Watson-Jones R. Fractures and other bone and joint injuries. 4th ed. Baltimore: Williams &

Wilkins; 1955.

80. Weseley MS, Barenfeld PA, Eisenstein AL. Closed treatment of isolated radial head fractures.

J Trauma. 1983;23(1):36-9.

BW_def.indd 24-25 07-02-11 22:25

26 part ii | classification 27 quantitative measurements of the radial head | chapter 2

Part II: Classification CHAPTER 2Quantitative Measurements of the Volume and Surface Area of the Radial Head

Thierry G. Guitton, MSc

Huub J. van der Werf, MD

David Ring, MD PhD

JHandSurgAm.2010;35(3):457-63.

BW_def.indd 26-27 07-02-11 22:25


Abstract

Purpose: We investigated the hypothesis that a quantitative 3-dimensional comput-ed tomography (Q3D-CT) modeling technique based on anatomical and demographic data that can measure size, shape, and proximal articular surface area can be used to develop formulas that could predict the volume and proximal surface area of the intact radial head in patients with fractures of the radial head.Methods: We used a consecutive series of 50 computed tomography (CT) scans with a slice thickness of 1.25 mm or less obtained in patients with fracture of the distal humerus, but no injury to the radial head, to create 3-dimensional models (3D). The volume and proximal articular surface area of the radial head were measured, and predictive formulas based on anatomical measurements and genders were calculat-ed using multiple linear regression.Results: There were significant correlations between total radial head volume and proximal radial head articular surface area for height, weight, radial head diame-ter, radial neck diameter, coronoid diameter, and gender. Multiple linear regression modeling resulted in formulas that could account for 89% of the variation in radial head volume and 75% of the variation in proximal articular surface area.Conclusions: The volume and proximal articular surface area of the radial head can be estimated based on anatomical measurements and gender. This may lead to bet-ter estimates of lost fragments when it is not possible to directly model the fractured radial head and CT scan of the opposite limb is not available.Levelofevidence:LevelIV,DiagnosticStudy

IntroductionThe morphology of the healthy radial head has been investigated with calliper ruler, osteometric board, coordinator measuring machine (CMM), X-ray, CT scan, Magnetic Resonance Imaging (MRI) and Computer-Aided Design (CAD) software in the past 1, 2,

4-8, 10-15, 17, 19, 20, 22, 23. Prior studies found differences between genders 6, 15, 22 and no side-to-side differences 20. Some studies have small numbers of subjects 1, 5, 10, 14, 15, 19, 20, 23, some used non-digital measurement tools 15, 19, 23, or old skeletons 2 and others use non-stan-dard software with low quality CT scans 20. Three-dimensional CT models provide more detailed information 18. We developed a quantitative 3-dimensional computed tomography (Q3D-CT) modeling technique that can measure size, shape and proxi-mal articular surface area. When this Q3D-CT method is used to analyze fractured radial heads, and in the absence of a CT scan of the opposite elbow, a method for estimating the total volume and proximal articular surface area of the unfractured head of the radius will allow estimation of the percentage of head that is fractured. Such percentages are used in classifications and affect management decisions on the basis of inter-pretation or measurements from plain radiographs, which may be less precise than calculations made from CT images. We investigated the hypothesis that analysis of normal, unfractured radi-al heads in patients with a CT scan obtained for other reasons (intact radial head) would allow us to develop a linear regression model capable of estimating the volume and proximal articular surface area of the radial head based on one or more of the following: radial head diameter, radial neck diameter, coronoid length diameter, height, weight, and gender.

Materials and MethodsInclusion and Exclusion CriteriaA search of a billing database between 2002 and 2008 identified 228 patients with a fracture of the distal humerus but an intact radial head. The 50 adult patients with a CT scan that had a slice thickness between 0.62 and 1.25 mm were included in our analysis. There were 19 men and 31 women, with an average age of 54 years (range, 19-92 years). A total of 26 patients injured their right elbow and 24 injured their left elbow.

Modelling TechniqueDigital Imaging and Communications in Medicine (DICOM) is a standard for han-dling, storing, printing, and transmitting information of CT scans. Several different CT scanners were used and scanners could scan up to 140 Kv and 500 to 700 mA

BW_def.indd 28-29 07-02-11 22:25


with slices from 8 to 64/Dual Source. The DICOM files were obtained through Vitrea 2 software (Vital Images, Plymouth, MN). Vitrea is a visualization solution that cre-ates 3-dimensional reconstructions from CT scans. The DICOM files were exported for further processing into Matlab 7.7 (MathWorks, Natick, MA), a numerical com-puting environment. With Matlab the CT slices (DICOM) were converted into regu-lar pictures so they were suitable for further processing. A special code written by the Massachusetts General Hospital 3D Imaging Service aids in this process and identifies higher densities with a consistent algorithm in the CT slices (in essence, bony structures). In addition, data describing the relationship between the slides and the higher densities were saved. The created images and the additional cre-ated data were then uploaded into Rhinoceros 4.0 (McNeel North America, Seattle, WA). Rhinoceros is a 3-dimensional modeling tool based on Non-Uniform Rational B-Spline (NURBS), a mathematical model commonly used in computer graphics for generating and representing curves and surfaces. Rhinoceros stacked the images on top of each other, taking their relationship into account. During the image pro-cessing in Matlab, the higher densities (bony structures) are automatically high-lighted with points on every single CT slice. The actual CT slice is depicted behind the pointwise representation of the bone in the software. Depiction of the CT slice with the points on top of them allows precise identification of all bony structures and fragments, even if they were impacted. The software puts new points in each CT slice, keeping them at the same level as the automatically generated points. After all points were set, we drew lines that then represented the actual outer border of the bone, and so created a wire model (Figure 1A). The line drawings are an auto-mated feature in the software that follows the automatically generated pointwise representation of bony structures. We then used this wire model to create a poly-gon mesh (Figure 1B). This is a collection of vertices, edges, and faces that defines the shape of a polyhedral object in 3D computer graphics and solid modeling, con-sisting of triangles only explicitly representing the surface. In other words, a hollow 3D model of solely the outer surface of the bony structures and fragments was generated.

EvaluationAfter the 3-dimensional models were created, we measured the volume and proxi-mal articular surface area from the radial head. Volumetric measurements and sur-face area measurements are a standard feature in Rhinoceros 4.0. To calculate the radial head volume, we separated the radial head from the shaft with a plane per-pendicular to the proximal articular surface (articulating with the capitellum) 21 and placed a cutting plane at the distal border of the cartilage of the articular circumfer-ence of the radial head 24, because we thought this was in the most distinct land-

mark of the radial head-neck margin (Figure 1C). The proximal articular surface area was calculated using the same cut-off points and done by selecting all the meshes that represented the proximal articular surface area of the radial head (Figure 1D). In addition to these volumetric and proximal articular surface area measurements, the diameter of the radial head and neck was measured. This was also done for the coronoid process by measuring the distance between the medialmost aspect of the coronoid (parallel to the radial-ulnar articulation) and the lateralmost aspect of the ulna (proximal to the radio-ulnar articulation at the origin of the radio-ulnar notch) 16. Hereon, we refer to this distance as the coronoid diameter. Given that the radial head and neck are slightly elliptical, we used the maximum diameter of the radial head and neck. It is often possible to measure the maximum radial head and neck diameters in the fractured radius, which provides additional parameters that can be used to predict the volume or proximal articular surface area of the intact radial head to estimate the percentage of the radial head that is fractured.

Figure1A:Wiremodelbasedonstackingof2-dimen-sionalimages.

Figure1C:Cuttingplaneoftheradialheadtomea-surevolume.

Figure1B:Final3-dimensionalmeshmodeloftheradialhead.

Figure1D:Articularsurfaceoftheradialheadwasquantified

Figure 1: Creation of the three-dimensional model from CT data.

BW_def.indd 30-31 07-02-11 22:25


Statistical AnalysisContinuous data are presented as the mean standard deviation (SD) and are reported in millimeters. Volumes are reported in cubic millimeters (mm3) and proximal ar-ticular surface area in square millimeters (mm2). Differences in continuous variables were evaluated using Student’s t-test for independent groups. For bivariate analysis, the relationships between the radial head volume and articular surface, and continu-ous variables (age, height, weight, radial head diameter, radial neck diameter, and coronoid diameter), were evaluated one at a time using Pearson r correlations. We evaluated dichotomous variables (gender and side) using the Mann-Whitney U-test. Linear regression analysis was used to determine formulas for the prediction of ra-dial head volume and surface areas based on basic measurements and demographic factors that would be available in patients with a fracture of the radial head. Two models were determined: a model with the strongest outcome and the strongest model without the variable radial head diameter. Multivariate analysis of variance was performed to identify the F statistic for the selected model, where the F ratio compares the variation of the dependent variable that is explained by the model to the part of variation that is not accounted for by the model. Significance of the F statistic below 0.05 indicates that the predictors in the selected model provide useful information about the dependent variable. Goodness-of-fit was assessed using adjusted R-squared, which measures the proportion of variation in the dependent variable that is explained by the model, with a correction for the number of explanatory variables. As a measure of the ac-curacy of the strongest model, the difference between actual (as calculated by the software in the reconstructed model) and predicted radial head volume and articu-lar surface for the 50 elbows was calculated and the middle 95% confidence interval (2.5% and 97.5% percentile ranks) was computed. In addition, to compare 2 calculated experimental values to each other as a quantitative indicator of quality, we calcu-lated the average relative percent difference for both strongest models. To quanti-fy the repeatability of the automated software algorithm, 1 observer built 1 radius bone model 5 times consecutively. The coefficient of variation, standard deviation (SD), mean, and 95% confidence interval (CI) are reported for the radial head volume, radial head articular surface, radial head diameter, and radial neck measurements. As another measure of accuracy, we tested our empiric formula for articular surface area to the basic mathematical formula assuming the radial head is round (πr2 for-mula). A power analysis indicated that a minimum sample size of 50 patients would provide 90% statistical power (b = 0.1, a = 0.05) to detect a moderate correla-tion (rho>0.40) of radial head diameter and radial head volume.

ResultsTable I shows the results from all subjects and the comparisons between men and women for age, height, weight, radial head, and neck diameters and volumes.

Men and women compared There was a significant statistical difference between man and women be-tween height (p<0.05), weight (p<0.05), radial head volume (p<0.01), radial head sur-face (p<0.01), radial head diameter (p<0.01), and radial neck diameter (p<0.01). There was no statistical difference between age (p<0.06) in men and women.

Radial head volume (Tables II, III)There were significant correlations between total radial head volume and height (r = 0.44; p<.0001), weight (r = 0.29; p<0.05), radial head diameter (r = 0.83; p<0.001), radial neck diameter (r = 0.74; p<0.001), coronoid diameter (r = 0.78; p<0.001), and gender (p<0.001). The strongest multivariable model (F = 118.3; p<0.001) consisted of the variables radial head diameter, coronoid diameter, and gender, and accounted for 88.5% of the variation in radial head volumes. The strongest model without the variable radial head diameter consisted of the variables radial neck diameter, coro-noid diameter, and gender, and accounted for 86.2% of the variation in radial head volumes (F =95.4; p<0.001).

Table I. Overall Results and Comparisons Between Men and Women

All Patients Men Women Men and Women

(N = 50 ) (N = 19 ) (N = 31) Compared

Mean SD Mean SD Mean SD P - Value

Age, (y) 53 22 46 22 58 21 0.06Height, (m) 1.68 0.13 1.74 0.07 1.64 0.14 < 0.05*Weight, (kg) 73 16.0 78 12 70 18 < 0.05*Radial Head diameter, (mm) 22 2.2 23 2.0 20 1.0 < 0.01*Radial Neck diameter, (mm) 15 1.6 16 2.0 14 1.0 < 0.01*Radial Head Volume, (mm) 3327 901 4301 585 2730 386 < 0.01*Radial Head Surface, (mm) 365 83 441 76 319 44 < 0.01*

*=Significantdifference

BW_def.indd 32-33 07-02-11 22:25


Radial head articular surfaceThere was significant correlation between total radial head surface and height (r = 0.40; p<0.001), weight (r = 0.37; p<0.001), radial head diameter (r = 0.82; p<0.001), ra-dial neck diameter (r = 0.68; p<0.001), coronoid diameter (r = 0.71; p<0.001), and gen-der (p<0.001). The strongest multivariable model (F = 44.9; p<0.001) consisted of the variables radial head diameter, coronoid diameter, and gender, and accounted for 74.5% of the variation in radial head articular surfaces. The strongest model without the variable radial head diameter consisted of the variables radial neck diameter, coronoid diameter, and gender, and accounted for 66.0% of the variation in radial head articular surfaces (F = 29.8; p<0.001).

Modeling/predictionsThe following fitted equation resulted for radial head volume:Radial Head Volume = -1926.64 + 146.50*Radial Head Ø + 70.72*Coronoid Ø + 769.78*Gender The following fitted equation resulted for radial head surface:Radial Head Surface = -212.15 + 19.81*Radial Head Ø + 5.43*Coronoid Ø + 32.85*Gender

The predictive linear model estimated the radial head volumes between 507 mm3 more and 575 mm3 less than the actual volumes based on 95% CI, and the aver-age relative difference was 0.53%. The predictive linear model estimated the radial head surface area between 75 mm2 more and 81 mm2 less than the actual volumes based on 95% CI, and the average relative percent difference was 0.51%. We tested the basic mathematical formula assuming the radial head is round. If we compare R2 (the variability explained) we found 0.75 for our empiric formula and 0.66 for the π r2 formula. The 5 consecutively built models had a mean radial head volume of 3179

Table II. Results of bivariate analysis of Radial Head

Radial Head Volume Radial Head SurfaceVariable r P-Value r P-ValueAge -0.10 0.5 0.00 0.99Side 0.98 0.80Gender < 0.001* < 0.001*Height 0.44 < 0.001* 0.40 < 0.001*Weight 0.29 < 0.05* 0.37 < 0.001*Radial Neck diameter 0.74 < 0.001* 0.68 < 0.001*Radial Head diameter 0.83 < 0.001* 0.82 < 0.001*Coronoid diameter 0.78 < 0.001* 0.71 < 0.001*

*=Statisticallysignificant.

Tabl

e 3.

Res

ults

of B

ivar

iate

and

Mul

tivar

iate

Ana

lysi

s of V

aria

nce

for R

adia

l Hea

d

R adi

al H

ead

Volu

me

FP-

Valu

eRa

dial

Hea

d Su

rfac

eF

P-Va

lue

Biva

riate

Mul

tivar

iate

Biva

riate

Mul

tivar

iate

(Cor

rela

tion

P-va

lue)

(Acc

ount

s for

% V

aria

bilit

y)(C

orre

latio

n P-

valu

e)(A

ccou

nts f

or %

Var

iabi

lity)

Varia

bles

A

ge0.

520.

7

Gen

der

< 0.

001^

†<

0.00

1^†

S

ide

0.98

0.80

H

eigh

t<

0.00

1<

0.00

1

Wei

ght

< 0.

05<

0.00

1

Rad

ial H

ead

diam

eter

< 0.

001^

< 0.

001^

R

adia

l Nec

k di

amet

er<

0.00

1†<

0.00

1†

Cor

onoi

d di

amet

er<

0.00

1^†

< 0.

001^

†M

odel

s

I88

.511

8.3

< 0.

001

74.5

44.9

< 0.

001

I

I86

.295

.4<

0.00

166

.029

.8<

0.00

1

^ =

Str

onge

st M

odel

† =

Str

onge

st M

odel

with

out R

adia

l Hea

d D

iam

eter

Ent

ered

in A

naly

sis

BW_def.indd 34-35 07-02-11 22:25


mm3 (SD = 16.4 mm3, 95% CI = 3159 - 3199 mm3) and a CV of 0.5%. The mean radial head articular surface was 592 mm2 (SD = 0.8 mm2, 95% CI = 591 - 593 mm2) and a CV of 0.14%. The radial head and radial neck diameter measurements variation was zero because all 5 readings were identical and thus, based on the sample data, the repeat-ability was 100% (Table IV).

DiscussionThe limitations of this investigation include the fact that the accuracy of this method depends on the quality of the CT scan. Because CT scans do not account for articular cartilage, our measurements will differ from those based on MRI 1 or direct measure-ments of fresh cadaveric bone 10, 13, 20. We did not thoroughly evaluate inter- and in-tra-observer variability in creation of the models because our method was time and resource intensive and, based on experience with 2 observers doing several models during training, the method leaves limited room for bias. When one person created the same model five times, we found very little variation in the measures of volume and surface area. The differences between surface area calculated using our formula and that using simple geometry are probably reflect the ovoid shape of the radial head. There was no correction for hypertrophy of the dominant elbow (due to ex-ercise) as described by Jones 9 because there were no known athletes in the cohort. In the final multivariable models height, weight and radial neck diameter did not significantly contribute to the fit of the model (as gauged by adjusted R2) and were therefore not used. The strong points of this investigation include the fact that we used a rela-tively large number of CT scans 1, 5, 10, 14, 15, 19, 20, 23 and a consistent algorithm was used for bone identification (on CT slides) and automated curve and polygon mesh creation,

which left limited room for judgment or bias on the part of the individual creating the model. The relatively small standard deviations of the measured volumes and surface areas, the relatively narrow 95% CI of the predictive linear models, and the fact that our multivariable models account for over 70% of the variability of volume and surface area, all indicate that we can make reasonable and useful estimation of these parameters in fractured radial heads. Our finding of a significant difference in radial head volume and surface area between men and women is consistent with Mall and colleagues 15. Further-more, we found correlations between radial head and radial neck diameters as did Ryan 19. We produced equations capable of estimating the volume and proximal articular surface area of the intact radial head–on the basis of parameters usually available in fractured radial heads–with an average relative percent difference of 0.5%. The ability to estimate the volume and surface area of the bone prior to frac-ture, provides useful information when we analyze a fractured radial head. For in-stance, it allows us to measure the percentage of the surface area involved in the fracture, which is 1 criterion in Broberg and Morrey’s modification of Mason’s clas-sification. Keeping in mind the many shortcomings of our approach, we believe that it will nonetheless improve our analysis and characterization of radial head fracture patterns. These Q3D-CT methods are, at least initially, more important for clinical re-search. We are using this technique to study fracture fragment size and injury pat-tern, and the ability to estimate percentage involvement helps make the results more intuitive for clinicians. Classifications and management decisions often refer to 30% of the surface area for instance 3, but it’s not clear that this is an important cutoff, that we can make this measurement accurately from radiographs, or that it is representa-tive of the fracture patterns that actually occur. More detailed analysis with these sophisticated techniques may help to clarify these issues. Additional work is needed to better define the accuracy and reliability of our method and determine how sensi-tive it is to the quality of the CT scan and the person doing the analysis.

Table IV. Measurements of Repeatability

Articular Radial Head Radial NeckModel Volume Surface Diameter Diameter1 3161.42 592.04 25.58 22.412 3199.95 591.27 25.58 22.413 3168.77 591.74 25.58 22.414 3172.43 593.36 25.58 22.415 3192.29 592.50 25.58 22.41Mean 3178.97 592.18 25.58 22.41SD 16.37 0.80 0.00 0.0095% CI 3159 - 3199 591 - 593 0.00 0.00CV 0.50% 0.14% 0.00 0.00Volume=mm3,articularsurface=mm2,diameter=mm,SD=standarddeviation,95%CI=confidenceinterval,CV=coefficientofvariation.

BW_def.indd 36-37 07-02-11 22:25


References



2. Berrizbeitia EL. Sex determination with the head of the radius. J Forensic Sci. 1989;34(5):1206-

13.


Relat Res. 1987(216):109-19.

4. Captier G, Canovas F, Mercier N, Thomas E, Bonnel F. Biometry of the radial head: biomechani-

cal implications in pronation and supination. Surg Radiol Anat. 2002;24(5):295-301.


J Hand Surg Am. 2006;31(1):53-7.

6. Gupta GG, Lucas G, Hahn DL. Biomechanical and computer analysis of radial head prostheses. J

Shoulder Elbow Surg. 1997;6(1):37-48.

7. Holman DJ, Bennett KA. Determination of sex from arm bone measurements. Am J Phys An-

thropol. 1991;84(4):421-6.

8. Itamura JM, Roidis NT, Chong AK, Vaishnav S, Papadakis SA, Zalavras C. Computed tomography

study of radial head morphology. J Shoulder Elbow Surg. 2008;17(2):347-54.

9. Jones HH, Priest JD, Hayes WC, Tichenor CC, Nagel DA. Humeral hypertrophy in response to

exercise. J Bone Joint Surg Am. 1977;59(2):204-8.



11. Knight DJ, Rymaszewski LA, Amis AA, Miller JH. Primary replacement of the fractured radial

head with a metal prosthesis. J Bone Joint Surg Br. 1993;75(4):572-6.

12. Koslowsky TC, Beyer F, Germund I, Mader K, Jergas M, Koebke J. Morphometric parameters of

the radial neck: an anatomical study. Surg Radiol Anat. 2007;29(4):279-84.








16. Matzon JL, Widmer BJ, Draganich LF, Mass DP, Phillips CS. Anatomy of the coronoid process. J

Hand Surg Am. 2006;31A(8):1272-8.

17. Popovic N, Djekic J, Lemaire R, Gillet P. A comparative study between proximal radial morphol-

ogy and the floating radial head prosthesis. J Shoulder Elbow Surg. 2005;14(4):433-40.








and their implication for prosthesis design. Clin Biomech. 2001;16(10):880-7.

22. van Riet RP, Van Glabbeek F, Neale PG, Bortier H, An KN, O’Driscoll SW. The noncircular shape of

the radial head. J Hand Surg [Am]. 2003;28(6):972-8.

23. van Riet RP, Van Glabbeek F, Neale PG, Bimmel R, Bortier H, Morrey BF, et al. Anatomical consid-


24. van Riet RP Van Glabbeek F, Neale PG, Bimmel R, Bortier H, Morrey BF, O’Driscoll SW, An KN.

Anatomical considerations of the radius. Clin Anat. 2004;17(7):564-9.

BW_def.indd 38-39 07-02-11 22:25

40 part ii | classification 41 quantitative measurements of radial head fractures | chapter 3

CHAPTER 3Quantitative Three-Dimensional Computed Tomography Measure-ment of Radial Head Fractures


Huub J. van der Werf, MD

David Ring, MD PhD

JShoulderElbowSurg.2010;19(7):973-7.

BW_def.indd 40-41 07-02-11 22:25


Abstract

Background: We developed a method to quantitatively analyze fracture fragment morphology on three-dimensional computed tomography images (Q3D-CT) in terms of size, shape, and articular surface area. Methods: We analyzed 46 adult patients with a computed tomography scan of a frac-tured radial head with Q3D-CT. We defined an unstable fracture as complete loss of cortical contact of at least 1 fragment. Of the patients, 3 had a Mason type 1 fracture (all stable), 26 had a type 2 fractures (7 stable [27%] and 19 unstable [73%]), and 17 had a type 3 fracture (all unstable). The volume and articular surface area of each articular fracture fragment were measured. A small fragment was defined as having a volume of less than 100 mm3 or an articular surface of less than 100 mm2.Results: Partial head fractures (Mason type 2) (26 fractures) are usually multi-frag-mented (19 of 26 [73%]) and often have small fragments by volume (32 fragments) and surface area (46 fragments) criteria, particularly when the fracture is displaced and unstable. Only 4 of the 17 patients (25%) with whole-head fractures (Mason type 3) had greater than 3 fragments, but 9 of 17 fractures (69%) with 3 or fewer fragments had small fragments.Conclusions: According to this initial application of Q3D-CT analysis, partial-head fractures are often complex and difficult to repair (small fragments), and most whole-head fractures have 3 or fewer fragments, but many of those fragments are small and may be difficult to repair.Levelofevidence:LevelIV,DiagnosticStudy

IntroductionMason classified fractures of the radial head as nondisplaced (Type 1), displaced par-tial head (Type 2), and displaced whole head fractures (Type 3) 4, 6, 9. Broberg and Mor-rey modified Mason’s classification as follows 1: Type 1 fractures involve less than 30% of the articular surface or are displaced fewer than 2 millimeters; Type 2 fractures are partial head fractures involving at least 30% of the articular surface and displaced at least 2 mm; Type 3 fractures are displaced articular fractures involving the entire head of the radius; and Type 4 fractures have an associated elbow dislocation 9. The thresholds of 30% surface area and 2 millimeter are relatively arbitrary and based on radiographs 1, 6 and the classification of radial head fractures according to Broberg and Morrey’s modification of the Mason classification has substantial observer variation 7, 8. Three-dimensional computed tomography (3D-CT) models provide more detailed information of the fractured bone and provide an opportunity to quantify fracture characteristics better than radiographs 14. We developed a Q3D-CT modeling technique that can measure fragment size, shape and articular surface area. We then used this technique to analyze frac-tures of the radial head with the objective of developing quantitative assessments of radial head fracture fragments that might help clarify current classification systems and decision-making. Specifically, we tested the null hypothesis that partial head (Mason 2) fractures and whole head fractures (Mason 3) have the same percentage of small fracture fragments that might be difficult to repair by volume and surface area criteria.

Materials and MethodsThe Massachusetts General Hospital Institutional Review Board has approved the human protocol for this investigation under No. 1999P008705. All investigations were conducted in conformity with ethical principles of research, and informed con-sent for participation in the study was obtained.

Inclusion and Exclusion CriteriaA search of billing records identified 72 patients with a fracture of the radial head that were evaluated with computed tomography (CT) between 2002 and 2008. Forty-six CT scans had a slice thickness between 0.62 and 1.25 mm deemed adequate for three-dimensional modelling. Several different CT scanners were used with up to 140Kv and 500-700 mAs and slices from 8 to 64/Dual Source. There were 24 men and 22 women with an average age of 47 years (range, 22 to 79 years). Eighteen patients injured their right elbow, 28 injured their left elbow. Fifteen patients had an isolated radial head injury, 16 patients dislocation with frac-

BW_def.indd 42-43 07-02-11 22:25


tures of the radial head and coronoid (terrible triad), 13 patients a posterior olecranon fracture-dislocation, and 2 patients had dislocation of the elbow with a fracture of the radial head. Twenty-four patients were treated with a radial head prosthesis, 7 with open reduction and internal fixation and 15 non-operatively.

ClassificationThe radial head fractures were classified based on radiographs and CT-scans taken immediately after injury according to Broberg and Morrey’s modification of the clas-sification of Mason, but excluding the Type 4 category 1, 6. Three patients had a type 1 fracture, 26 patients a type 2 fracture, and 17 patients a type 3 fracture. Unstable frac-tures are defined as having separation (complete loss of cortical contact) of at least one radial head fracture fragment 10.

Modelling TechniqueDICOM (Digital Imaging and Communications in Medicine) files were obtained through Vitrea (Vitrea 2 software; Vital Images, Inc., Plymouth, MN) and exported for further processing into Matlab (MATLAB 7.7; The MathWorks, Inc., Natick, MA) (Figure 1A). In this process higher densities were identified with a consistent algo-rithm in the CT-slices (in essence bony structures). The created images and the ad-ditional created data were then uploaded into Rhinoceros (Rhinoceros 4.0; McNeel North America, Seattle, WA). Rhinoceros stacked the images (jpegs) on top of each other taking their relationship into account. During the image processing in Matlab, the higher densities (in essence bony structures) are automatically highlighted with points on every single CT-slice (Figure 1B). The actual CT-slice is depicted behind the point-wise representation of the bone in the software. Lines were drawn which then would represent the actual outer border of the bone and so created a wire model (Figure 1C). This wire model was then used to create a hollow 3D model of solely the outer surface of the bony structures (Figure 1D).

EvaluationAfter the 3D models were created, the volume and articular surface area of each fracture fragment and the remaining un-fractured bone were measured. Volumetric measurements and surface area measurements are a standard feature in Rhinoceros (Rhinoceros 4.0; McNeel North America, Seattle, WA). The articular surface area cal-culations were done by selecting all the meshes which represented the proximal ar-ticular surface of the particular fragment. Only fragments that were separated from the rest of the radius were mea-sured. The volume of the intact radius was arbitrarily cutoff by the limit of the CT in a way that hindered meaningful measurement of this part 3. The volumetric mea-

surements were categorized into three different sizes; major fragment (>500 mm3), minor fragments (100-500 mm3) and small fragments (<100 mm3). We divided our ar-ticular surface measurments into two groups; major fragment (>100 mm2) and small with less than 100 mm2 surface area. The numbers of fragments reported are including any intact radial head as a separate fragment for mason type 1 and mason type 2 fractures. All mason type 3 fractures by definition have no intact radial head. The volumes and surfaces of the intact radial head are not incorporated in the fragment calculations. To calculate the percentage of articular surface area and volume that was fracture of the intact radial head a formula developed from volume and articular surface area measurements of 50 unfractured radial heads was used 3.

Figure1A:Three-dimensionalreconstructionsofwholeheadfracture(Masontype3).

Figure1C:Wiremodelof4radialheadfragmentsandshaftofradius.

Figure1B:Pointsandlinesrepresenttheouterborderof4radialheadfragments.

Figure1D:Final3-dimensionalmodelofMasontype3fracture.Theradialheadisfracturedin4differentfragments.

Figure 1: A 36-year-old woman dislocated her right elbow while rollerblading,

fracturing her radial head.

BW_def.indd 44-45 07-02-11 22:25


Statistical AnalysisContinuous data are presented as the mean (range). Mean fragment volume and articular surface area were calculated for every Mason fracture type. Relative refer-ences are reported as percentages.

ResultsIn this study, 17 fractures created 2 fracture fragments, 15 created 3 fracture frag-ments and 14 created 4 or more fracture fragments (Table I). The fractures were classified as unstable in 36 patients and stable in 10 patients. All Mason Type 1 frac-tures were stable. Seven Mason Type 2 fractures were stable and 19 were unstable (loss of cortical contact). (Figure 2) All Mason Type 3 fractures were unstable.

By use of the volume measurement criterion, 39 of 125 fracture fragments (31%) were classified as small (<100 mm3) (Table II). Moreover, 21 fractures (46%) had at least 1 small fracture fragment. The mean number of small fracture fragments was 0.85 per fracture (Table III).

By use of the surface area measurement criterion, 65 of 96 fracture frag-ments (68%) were classified as small (<100 mm2) (Table IV). Moreover, 31 fractures (67%) had at least one small fracture fragment. The mean number of small fracture fragments was 1.4 per fracture (Table V).

Table I. Mason Type and Number of Fracture Fragments

2 Fragments 3 Fragments ≥4 Fragments Total

Mason 1 3 0 0 3

Mason 2 7 9 10 26

Mason 3 7 6 4 17

Total 17 15 14 46

Table II. Fracture Fragment by Volumetric Criteria

Major (>500 mm3) Minor (100 – 500 mm3) Small (<100 mm3) Total

Mason 1

2 Fragments 3 0 0 3

Mason 2

2 Fragments 6 1 0 7

3 Fragments 3 7 8 18

≥4 Fragments 4 20 24 48

Mason 3

≤3 Fragments 23 6 3 32

>3 Fragments 7 6 4 17

Total 46 40 39 125

Table III. Volume-Specific Fracture Fragment Characteristics

Mean Estimated Bone Loss by Volume (mm3)

No. of Small Fragments by Volume Criterion

No. of Small Fragments by Volume Criterion per Fracture

Mason 1

2 Fragments 13 0

Mason 2

2 Fragments 81 0

3 Fragments 450 8 0.89

≥4 Fragments 514 24 2.40

Mason 3

≤3 Fragments 622 3 0.23

>3 Fragments 500 4 1.00

Total 39 0.85

No.=number

Figure 2: Pie chart depicting the distribution of Mason 2 fractures according to fracture “stability”

(where unstable is defined as lack of cortical contact) and number of fracture fragments.

1

4

93

7

2 2 fragments

3 fragments

>4 fragments

unstable

stable

BW_def.indd 46-47 07-02-11 22:25


Mason Type 2 (Partial Head Fractures)Among the 26 partial head fractures (Mason Type 2), the mean articular surface area of the fractured part of the radial head was 107 mm2 (range, 42 to 180 mm2) and the mean volume was 818 mm3 (range, 164 – 1956 mm3). These data correspond to an esti-mated percentage of total intact surface area of 31% (range, 11 to 53%) and volume of 25% (range, 4 to 59%), respectively.

The fracture fragments had a mean articular surface area of 61 mm2 (range, 21 to 136 mm2), and a mean volume of 444 mm3 (range, 45 to 1540 mm3). Of the frac-tures, 19 fractures (73%) had 3 or 4 fracture fragments. Unstable fractures were more often fragmented. Small fracture fragments were more common among fragmented fractures (Tables II and IV).

Mason Type 3 Among the 17 whole-head fractures (Mason Type 3), the mean articular surface area of the fractured part of the radial head was 286 mm2 (range, 204 to 414 mm2) and the mean volume was 2529 mm3 (range, 1353 – 3960 mm3). These data correspond to an estimated percentage of total intact surface area of 83% (range, 59 to 97%) and volume of 81% (range, 47 to 100%), respectively, meaning that there is and estimated 17% loss of bone (e.g. small unidentified fragments) by volume and 19% by surface area criteria. Only 4 fractures (24%) created more than 3 fragments (Figure 3). More fragmented fractures had small fragments and greater estimated bone loss.

DiscussionQuantitative analysis of CT scans provided measurements of the volume and articu-lar surface area of radial head fracture fragments. The strengths of our article include the fact that we developed this technique with widely used software. A consistent algorithm was used for bone identification (on CT slides) and automated curve and polygon mesh creation, which left limited room for judgment or bias on the part of the individual creating the model.

The limitations of this article include the fact that we could not use the op-posite arm to determine the volume and surface area of the unfractured radial head.

Figure 3: Pie chart depicting distribution of Mason 3 fractures according to the number of fracture

fragments.

Table V. Surface-Specific Fracture Fragment Characteristics

Estimated Bone Loss by Surface Area as % of Intact Radial Head

No. of Small fragments by Surface Area Criterion

No. of Small fragments by Surface Area per fracture

Mason 1

2 Fragments 17 0 .

Mason 2



≥4 Fragments 17 29 2.90

Mason 3

≤3 Fragments 16 9 0.69

>3 Fragments 20 10 2.50

Total 65 1.41

Table IV. Fracture Fragment by Surface Criteria

Major (>100 mm2) Small (<100 mm2) Total (n)

Mason 1

2 Fragments 3 0 3

Mason 2

2 Fragments 3 4 7

3 Fragments 3 13 16

≥4 Fragments 2 29 31

Mason 3

≤3 Fragments 18 9 27

>3 Fragments 2 10 12

Total 31 65 96

7

6

4

2 fragments

> 4 fragments

3 fragments

BW_def.indd 48-49 07-02-11 22:25


Therefore, all measures expressed as a percentage of the entire head are estimates based on formulae determined in a separate study of intact radial heads 3. In addi-tion, our definition of a small fracture fragment was arbitrary. Given these important limitations, this should be considered pilot work from which no management deci-sions can be made.

We found that partial-head fractures (Mason 2) are usually multi-fragment-ed (73%) and often have small fragments that may be difficult to repair using both volume and surface area criteria (23 fractures [88%]), particularly when the fracture is displaced and unstable. Small fragments are more common in partial than in whole head fractures, at least among the unstable fractures associated with elbow dislocation or fracture of the ulna. This finding is important because partial-head fractures are often considered straightforward to repair and irreparable partial-head fractures are rarely discussed 2, 5, 11-13.

Another important finding is that, according to the surface area criterion, many of the Mason 2 fractures involved less than a third of the radial head (12 out of 26). This means that about half of these fractures would not satisfy the Broberg and Morrey criterion of greater than 30% of the articular surface area to be considered Type 2 fractures 1, 6.

Among whole-head fractures (Mason 3), only 4 (25%) had greater than 3 fragments, but 9 (69%) of the fractures with 3 or fewer fragments had small frag-ments. Furthermore, 7 of 13 fractures were estimated to have more than 10% bone loss by volumetric criteria-3 of 13 fractures by articular surface criteria-which may indicate lost small fragments. Because even whole-head fractures with fewer than 3 fragments can have small fragments, the threshold of 3 fragments in the decision between open reduction-internal fixation and prosthetic replacement may not help guide management in all cases.

ConclusionsQuantitative analysis of 3D-CT scans is a useful technique for analyzing articular fracture pattern and morphology. Using this technique, we identified that that par-tial-head fractures (Mason 2) frequently involve less than a third of the radial head surface area, partial-head fractures have more small and difficult-to-repair frag-ments than whole-head fractures (Mason 3), and whole head fractures with more than 3 fragments are relatively uncommon but some 3-fragment fractures have small fragments. These findings may influence our conception and classification of radial head fractures. For now, this technique is primarily designed for research purposes and not for patient care, but with further development Q3D-CT might prove useful in management decisions for individual patients.

References


Relat Res. 1987(216):109-19.

2. Doornberg J, Elsner A, Kloen P, Marti RK, van Dijk CN, Ring D. Apparently isolated partial

articular fractures of the radial head: prevalence and reliability of radiographically diagnosed

displacement. J Shoulder Elbow Surg. 2007;16(5):603-8.

3. Guitton TG, van der Werf HJ, Ring D. Quantitative Measurements of the Volume and Surface

Area of the Radial Head. J Hand Surg Am 2010: 35(3): 457-63.



5. Lindenhovius AL, Felsch Q, Doornberg JN, Ring D, Kloen P. Open reduction and internal fixation

compared with excision for unstable displaced fractures of the radial head. J Hand Surg Am.

2007;32(5):630-6.







9. Morrey B. Radial head fractures. In: BF M, editor. The elbow and its disorders. Philadelphia: WB

Saunders; 1985. p. 355-81.

10. Rineer CA, Guitton TG, Ring D. Radial head fractures: loss of cortical contact is associated with

concomitant fracture or dislocation. J Shoulder Elbow Surg.19(1):21-5.

11. Ring D, Quintero J, Jupiter JB. Open reduction and internal fixation of fractures of the radial

head. J Bone Joint Surg Am. 2002;84-A(10):1811-5.

12. Ring D. Open reduction and internal fixation of fractures of the radial head. Hand Clin.

2004;20(4):415-27, vi.

13. Ring D. Displaced, unstable fractures of the radial head: fixation vs. replacement-what is the

evidence? Injury. 2008;39(12):1329-37.



BW_def.indd 50-51 07-02-11 22:25

52 part ii | classification 53 imaging and modeling of radial head fractures | chapter 4

CHAPTER 4Diagnostic Accuracy of Two-Dimensional and Three-Dimensional Imaging and Modeling of Radial Head Fractures


Kim M. Brouwer, MSc

George Dyer, MD

David Zurakowski, PhD

Chaitanya Mudgal, MD

David Ring, MD PhD

Submitted

BW_def.indd 52-53 07-02-11 22:25


Abstract

Background: This investigation tests the hypothesis that classification and charac-terization of fractures of the radial head is more accurate with (3D) than (2D) com-puted tomography (CT) images and radiographs, using a prospective study design with intraoperative inspection as the reference standard.Methods: Treating surgeons and first assistants completed a questionnaire assigning a fracture type according to the Broberg and Morrey modification of Mason’s classi-fication, evaluating selected fracture characteristics, and electing preferred manage-ment four times: Initially based upon radiographs and 2D images alone, a second time based on radiographs, 2D and 3D-CT images, a third time on radiographs, 2D, 3D-CT and 3D physical models, and a final time after surgery based on intra-operative visualization of the fracture. The agreement between surgeon and first assistant as well as the sensitivity and specificity were calculated for 2D-CT and radiographs, 3D-CT, and 3D physical models as compared to the intraoperative direct observation.Results: The addition of 3D-CT reconstructions and 3D physical models to standard radiographs and 2D-CT scans improved the reliability of fracture classification ac-cording to the Broberg and Morrey modification of the Mason classification (kappa values, 2D-CT = 0.23, 3D-CT = 0.26, and 3D model = 0.37; all p < 0.05). The addition of the 3D-CT and the 3D physical model significantly improved the sensitivity com-pared to 2D-CT (all p < 0.01) for fracture line separating the entire head from the neck, comminution of the radial neck, fracture involving the articular surface, articular fracture gap greater than 2 millimeters, impacted fracture fragments, greater than 3 articular fragments, and articular fragments judged too small to repair. Conclusion: Increasing levels of sophistication in imaging and modeling improved the sensitivity for diagnosis of fracture characteristics using the intraoperative in-terpretation of the operating surgeon as a reference standard. Fracture classification, characterization, and proposed treatment were also noted to be less variable with more sophisticated imaging and modeling.LevelofEvidence:Diagnostic,LevelI

IntroductionOptimal management of radial head fractures is debated, but accurate preoperative radiological characterization of the fracture may facilitate management. Prior stud-ies have demonstrated improved agreement in characterization and classification of various fractures with three-dimensional computed tomography (3D-CT) compared to two-dimensional computed tomography (2D-CT) images and radiographs 1, 3, 5, 8, 9, 11,

13, 19, 20. These studies were based upon retrospective data and the reference standard was based upon surgeon recollection and the medical record (e.g. operative notes). Three-dimensional models that are constructed based on CT images and can be held in the hand and, may facilitate fracture characterization and surgical plan-ning. Computer-generated bone models have been used in the planning of osteotomy of multidirectional distal radius malunions 7, 16. This investigation tests the hypothesis that 3D-CT images identify and pre-dict fracture characteristics more accurately than 2D-CT images and radiographs, using a prospective study design with intraoperative inspection as the reference standard. A secondary hypothesis was that 3D physical models predict fracture char-acteristics more accurately than 2D and 3D-CT images and radiographs.

Material and MethodsInclusion and Exclusion CriteriaUnder an Institutional Review Board (IRB) approved protocol, we prospectively in-cluded patients between 2007 and 2010 with a radial head fracture seen at two Level 1 trauma centers. Inclusion criteria were 1) fracture of the radial head; 2) election of operative treatment; 3) availability of CT scan; 4) age of 18 years or older. Exclusion criteria were pregnant women and patients unable to give informed consent. Forty-one patients satisfied the inclusion and exclusion criteria. Two patients were exclud-ed for incomplete questionnaires, resulting in a final cohort of 39 patients. Among the 39 patients, the mean age was 52 years (range, 23 to 92 years). There were 18 men (46%) and 21 (54%) women. The radial head fracture was an iso-lated injury in 4 patients (10%), and was associated with an elbow dislocation in 7 (18%) patients, an elbow dislocation and coronoid fracture (the so-called terrible triad injury) in 15 patients (38%), a posterior-olecranon-fracture dislocation (POFD) in 8 patients (21%), metaphyseal fracture of the proximal ulna (posterior Monteggia fracture) in 1 patient (3%), a complex fracture of the distal humerus in 1 patient (3%), an Essex-Lopresti lesion in 1 patient (3%), an anterior-transolecranon-fracture dislo-cation in 1 patient (3%), and a capitellum/trochlea fracture in 1 patient (3%). The left side was injured in 22 patients (56%) and the right side in 17 patients (44%). Twenty-three patients (59%) fractured their elbow in a fall from a standing height, 13 (33%)

BW_def.indd 54-55 07-02-11 22:25


from a greater height, 2 (5%) patients in a motor vehicle collision (MVC) and one (3%) in a crush injury.

Evaluation3D-CT reconstructions were ordered for all patients. CT scans were sent to Medical Modeling LLC (Golden, CO) for same-day manufacture of the 3D physical model re-constructions. The treating surgeons completed a questionnaire assigning a fracture type according to the Broberg and Morrey modification of Mason’s classification and important fracture characteristics and management. Broberg and Morrey modified Mason’s classification as follows: Type 1 fractures involve less than 30% of the articular surface or are displaced fewer than 2 millimeters; Type 2 fractures are partial head fractures involving at least 30% of the articular surface and displaced at least 2 mil-limeters; Type 3 fractures are displaced articular fractures involving the entire head of the radius; and Type 4 fractures have an associated elbow dislocation 2. The questionnaire was completed four times: initially based upon radio-graphs and 2D images alone; a second time based on radiographs, 2D and 3D-CT im-ages; a third time on radiographs, 2D, 3D-CT and 3D physical models; and a fourth time based on intra-operative visualization of the fracture characteristics. The fourth questionnaire completed by the surgeon represented the reference standard. Both the surgeon and the first assistant rated the fractures, allowing us to calculate in-terobserver agreement. Sensitivity and specificity were calculated for 2D-CT and ra-diographs, 3D-CT, and 3D physical models as compared to the intraoperative direct observation of the surgeon.

Statistical AnalysisInterobserver agreement regarding fracture characteristics and treatment proposal was measured for each method by the chance-corrected kappa (κ) coefficient with strength of agreement assessed using the benchmarks of Landis and Koch 12. Logistic regression was applied using a generalized estimating equations (GEE) strategy in order to account for the same 39 cases evaluated by multiple surgeons using each of 4 different methods (2D, 2D/3D, 2D/3D with physical model, direct operative view) with a binomial distribution used for binary yes/no fracture characteristics and a multinomial logit distribution for Broberg-Morrey classification (Types I-IV) and treatment plan (5 options: nonoperative management; Open Reduction and Internal Fixation (ORIF) with wires, screws or pins; ORIF with plate and screws; radial head excision; radial head replacement/arthroplasty). Differences between the methods were determined using the maximum likelihood Wald chi-square test with a two-tailed p < 0.05 as the criterion for statistical significance 6. Power analysis revealed that a minimum sample size of 30 fractures would provide 80% power (a = 0.05, b =

0.20) to detect significant intra- and inter-observer agreement using the kappa coef-ficient 12 as well as in comparing diagnostic characteristics between the two imaging modalities. Sensitivity, specificity, and accuracy for detection of each of the fracture characteristics and type of treatment with two-dimensional images, 3D reconstruc-tions and 3D Model was calculated with the intra operative findings of the attending surgeon as the gold standard. The statistical significance of these differences was evaluated using McNemar’s test for paired binary data 10. Statistical analysis was per-formed using SPSS version 18.0 (SPSS Inc./IBM, Chicag, IL).

Source of FundingNo funding was received in direct support of this study. An agreement approved by our Human Research Committee and Research Contracting Department, Medical Modeling LLC (Golden, CO) provided free 3D physical models.

ResultsThe addition of 3D-CT reconstructions and 3D models to standard radiographs and 2D- CT scans improved the reliability of fracture classification according to the Bro-berg and Morrey modification of the Mason classification, diagnosis of comminu-tion of the radial neck, involvement of the articular surface, articular gap or step of 2 millimeters or greater, central impaction of the articular surface, presence of more than 3 articular fragments, presence of articular fragments too small to repair, and proposed treatment (Table I).

Table I. Interobserver Agreement for Classification and Treatment of Radial Head and Neck Fractures for Each CT Method and Direct Operative View

Characteristic 2D CT 2D/3D CT 2D/3D CT With Model

Operative View(Gold Standard)

Broberg-Morrey 0.23* 0.26* 0.37* 0.38*Fracture line 0.69† 0.54† 0.59† 0.54†Comminution 0.31* 0.48† 0.57† 0.40*Articular surface 0.12 0.28 0.22 0.28Gap >2 mm 0.37* 0.59† 0.57† 0.37*Impaction 0.17 0.24 0.12 0.23>3 fragments 0.29 0.50† 0.64† 0.57†Small fragments 0.26 0.34* 0.33* 0.42*Proposed Treatment 0.47† 0.53† 0.67† 0.85†

Dataarekappa(k)valuesbasedon39casesevaluatedbytwoindependentsurgeons.Guidelinesforstrengthofobserveragreement:k=0-0.20(slight),k=0.21-0.40(fair),k=0.41-0.60(moderate),k=0.61-0.80substantial,k=0.81-1.00almostperfect.Significantinterobserveragreementbeyondchancelevel(*p<0.05;†p<0.01).

BW_def.indd 56-57 07-02-11 22:25


The addition of 3D-CT and the 3D models to 2D-CT and radiographs led to significant improvements in sensitivity for diagnosis of fracture line separation of the entire articular surface from the radial neck, comminution of the radial neck, involvement of the articular surface, articular gap or step of 2 millimeters or greater, central impaction of the articular surface, presence of more than 3 articular frag-ments and to the presence of articular fragments too small to repair (all p < 0.01, Table II). There were no significant changes in specificity with more sophisticated imaging, which is not surprising given that improvements in the sensitivity of de-tecting fracture characteristics cannot be associated (statistically speaking) with in-creased specificity (Figure 1-3).

Figure 2: Agreement on treatment proposal stratified by 2D-CT and 2D/3D-CT.

Figure 1: Agreement on fracture characteristics stratified by 2D-CT and 2D/3D-CT.Table II. Sensitivity and Specificity Characteristics for 2D and 3D CT Methods

2D CT Alone 2D/3D CT 2D/3D CT with ModelVariable Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity

Fracture Line85

(91-99)

100

(90-100)

95*

(78-100)

94

(78-100)

95*

(81-100)

94

(78-100)

Comminution58

(31-83)

93

(79-99)

83*

(57-97)

96

(85-100)

92*

(68-100)

93

(79-99)

Articular Surface

94

(84-99)

67

(16-98)

97

(88-100)

67

(16-98)

97

(88-100)

67

(16-98)

Gap >2 mm93

(81-99)

78

(46-96)

97

(86-100)

78

(46-96)

97

(86-100)

89

(59-99)

Impaction57

(32-80)

88

(72-97)

86*

(62-97)

84

(67-95)

93*

(72-100)

96

(84-100)

>3 Fragments55

(34-74)

82

(60-95)

100*

(90-100)

85

(66-96)

91*

(74-98)

100

(89-100)

Small Fragments

74

(56-88)

83

(57-97)

100*

(93-100)

67

(39-88)

100*

(93-100)

75

(47-93)

Proposed Treatment

74

(56-88)

83

(57-97)

89*

(74-97)

75

(47-93)

100*

(93-100)

75

(47-93)

Valuesarepercentageswithresultsarebasedonattendingsurgeonforeachmethodcomparedtointraoperativedirectviewgoldstandard(N=39pairedcases).*Statisticallysignificantcomparedtosensitivityfor2DCTalone(allp<0.01).Therewerenosignificantdifferencesdetectedinspecificitybetweenthethreemethods.

BW_def.indd 58-59 07-02-11 22:25


DiscussionThe strengths of this investigation include the prospective design, the relatively large number of patients, and an intra-operative reference standard. The limitations of this investigation include the fact that images were usually rated after surgery (in part due to the inherent delay in receiving the physical 3D model), so that ratings of the radiological images were–in essence–retrospective; the injuries were relatively complex resulting in a spectrum bias in terms of all fractures of the radial head, al-though our work is representative of the types of fractures that would be studied with CT and operated on; two patients (one with addition of a capitellum/trochlea fracture and one anterior-transolecranon fracture dislocation) had non-nondisplaced fractures of the radial neck, which are relatively unusual–both fractures were seen only on operative exposure; and multiple physicians were involved in the ratings at two sites, which makes the results more generalizable, but less consistent. These data should also be interpreted in light of the fact that the first assistant was usually a resident or fellow, so that the observer variability may largely reflect differences in training and experience. It is not always feasible to have models available prior to operative treat-ment at this point, but 3D reconstructions of computed tomography images can be easily produced by surgeons using the DICOM (Digital Imaging and Communications in Medicine) files from the patient’s CT scan. Three-dimensional reconstructions are

made from CT-scans and therefore do not require additional scanning and do not expose the patient to additional radiation. It has been calculated at the investigators institution, that the cost for additional 3D reconstructions are an additional 20% of the cost of a CT-scan. Free software such as OsiriX 17 is available which makes it pos-sible for every orthopedic surgeon to quickly and easily create 3D reconstructions themselves with minimal training. This study found that increasing levels of sophistication in imaging/mod-eling: 1) improved the sensitivity for diagnosis of numerous fracture characteristics using the surgeon’s interpretation of the intraoperative findings as the reference standard; and 2) decreased observer variation between surgeon and first assistant. This is in concordance with prior studies that have demonstrated improved agree-ment in characterization and classification of fractures with 3D-CT compared to 2D-CT and radiographs alone 1, 3, 5, 8, 9, 11, 13, 19, 20. Prior studies that addressed the classification of radial head fractures specifically found substantial observer variation 4, 14, 15, 18 when fractures were evaluated by radiographs only. However, these studies differed in that they are based upon retrospective data in small groups of observers/patients and the reference standard was based upon surgeon recollection and the medical record (e.g. operative notes). We interpret this combination of findings to indicate that fracture clas-sification and characterization based on 3D imaging and models is more accurate and reliable, essentially helping to narrow the experience and training gap. While experienced surgeons sometimes suggest that little is added by more sophisticated imaging, science is establishing that more sophisticated imaging does improve our understanding of the injury. However, recommendations regarding the use of a new technology should be based on both diagnostic performance characteristics and clinical impact. The next steps are to investigate whether more sophisticated imag-ing leads to more effective treatment as measured by fewer complications with less functional impairment.

Figure 3: Sensitivity and specificity according to Broberg and Morrey modification of the Mason

classification stratified by 2D-CT, 2D/3D-CT and 2D/3D-CT with model.

BW_def.indd 60-61 07-02-11 22:25


References





Relat Res. 1987(216):109-19.








6. Harrell F. Regression modeling strategies with applications to linear models, logistic regres-

sion, and survival analysis. New York: Springer-Verlag; 2001. p. 331-42.

7. Jupiter JB, Ruder J, Roth DA. Computer-generated bone models in the planning of osteotomy of

multidirectional distal radius malunions. J Hand Surg Am. 1992;17(3):406-15.






10. Kocher MS, Zurakowski D. Clinical epidemiology and biostatistics: a primer for orthopaedic

surgeons. J Bone Joint Surg Am. 2004;86-A(3):607-20.



12. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics.

1977;33(1):159-74.



phy. J Hand Surg Am. 2009;34(9):1640-6.





16. Murase T, Oka K, Moritomo H, Goto A, Yoshikawa H, Sugamoto K. Three-dimensional corrective

osteotomy of malunited fractures of the upper extremity with use of a computer simulation

system. J Bone Joint Surg Am. 2008;90(11):2375-89.

17. Rosset A, Spadola L, Ratib O. OsiriX: An Open-Source Software for Navigating in Multidimen-

sional DICOM Images. Journal of Digital Imaging. 2004;17(3):205-16.

18. Sheps DM, Kiefer KR, Boorman RS, Donaghy J, Lalani A, Walker R, et al. The interobserver reli-

ability of classification systems for radial head fractures: the Hotchkiss modification of the

Mason classification and the AO classification systems. Can J Surg. 2009;52(4):277-82.




BW_def.indd 62-63 07-02-11 22:25

64 part ii | classification 65 reliability of radial head fracture classification | chapter 5

CHAPTER 5Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed Tomography


David Ring, MD PhD

Science of Variation Group*

InrevisionJBoneJointSurgAm

*ScienceofVariationGroup:

BrandonE.Earp,AmyL.Ladd,PeterJEvans,ChristinaE.Kuo,JanBiert,C.N.vanDijk,PhaniK.Dantuluri,DavidE.

Ruchelsman,K.J.Ponsen,MaximillianSoong,TimR.Davis,AshokK.Shyam,LauraS.Phieffer,C.M.LeCroy,Martin

Richardson,AndrewH.Schmidt,PeterL.Jebson,PaulE.Levin,GregoryJ.DellaRocca,CharlesA.Goldfarb,KyleJ.Jeray,

DavidM.Kalainov,GeorgeS.Dyer,NealT.Chen,ALeeOsterman,GeorgeS.Athwal,LukeP.Leenen,ThomasW.Wright,

MarcF.Swiontkowski,DavidJ.Slutsky,FredeFrihagen,ScottF.Duncan,RickF.Papandrea,KevinC.Chung,PhilipBla-

zar,RobertJ.Feibel,RobertD.Zura,HuubJ.vanderHeide,RobertZ.Tashijan,LeonElmans,JohnA.Jiuliano,Marco

M.Rizzo,SamirSodha,JohnA.McAuliffe,RandallW.Culp,JorgeOrbay,CharlesCassidy,RobertG.H.Albers,LeonidI.

Katolik,ReidAbrams,MarkE.Baratz,KennethA.Egol,JosephM.Conflitti,DougP.Hanel,JoseM.Nolla,Michaelhaus-

man,AndrewE.Caputo,RudolfW.Poolman,TerryS.Axelrod,MichaelD.McKee,J.C.Goslings,ParagK.Sancheti,Peter

R.Brink,CarrieR.Swigart,ThomasB.Hughes,KeithA.Segalman,P.V.vanEerten,BrettD.Crist,EdwardDiao,Richard

S.Page,LisaL.Lattanza,GeorgeThomas,JasonC.Fanuele,PeterKloen,TacoGosens,CharalamposZalavras,JohnS.

Taras,JeffreyA.Greenberg,EricM.Hammerberg,LouisW.Catalano3rd,RodrigoF.Pesantez,ArieB.vanVugt,SteveC.

Kronlage,MichaelA.Baskies,MartinI.Boyer,PeterV.Giannoudis,MichaelJ.Prayson,ElenaGrosso

BW_def.indd 64-65 07-02-11 22:25


Abstract

Background: The Broberg and Morrey modification of the Mason classification of ra-dial head fractures has substantial interobserver variation. This study used a large web-based collaborative of experienced orthopaedic surgeons to test the hypothesis that three-dimensional reconstructions of computed tomography scans (3D-CT) im-prove the interobserver reliability of the classification of radial head fractures ac-cording to the Broberg and Morrey modification of the Mason classification.Methods: Eighty-five orthopaedic surgeons evaluated twelve radial head fractures and were randomly assigned to review either radiographs and two-dimensional computed tomography scans (2D-CT) or radiographs, and 3D-CT images to determine the fracture classification, fracture characteristics and treatment plans. The kappa multirater measure (κ) was calculated to estimate agreement between observers.Results: 3D-CT had moderate and 2D-CT had fair agreement among observers for the Mason classification (κ3D = 0. 49 vs. κ2D = 0.37; p = < 0.001). Among seven fracture characteristics (fracture line, comminution, articular surface involvement, gap/step of 2mm or greater, central impaction, recognition of three articular fragments and articular fragments to small to repair) there was a significant difference in kappa value between 3D-CT and 2D-CT for three variables (articular fragments too small to repair [κ3D = 0.61 vs. κ2D = 0.47; p = < 0.001], recognition of three articular fragments [κ3D = 0.61 vs. κ2D = 0.38; p = < 0.001] and central impaction [κ3D = 0.15 vs. κ2D = 0.22; p = 0.006]). Among treatment recommendations there was fair agreement for both 3D-CT and 2D-CT (κ3D = 0.40 vs. κ2D = 0.26; p = < 0.001). Conclusion: Although 3D-CT led to some small but significant decreases in interob-server variation, there is still a notable degree of disagreement regarding classifica-tion and characterization of radial head fractures. Improvements in imaging may not be sufficient to optimize interobserver agreement. LevelofEvidence:DiagnosticLevelIII

IntroductionThe classification of radial head fractures according to Broberg and Morrey’s modi-fication of the Mason classification 1 has substantial interobserver variation 10, 12. As with classification and characterization of most fractures the interobserver variation is greater than the intraobserver variation. Evidence suggests that more sophisticat-ed images such as 3D-CT improve intraobserver reliability more than interobserver reliability 3, 4. A major limitation of most studies of observer variation is the use of only a few observers, most of them typically relatively junior surgeons. A new collaboration motivated to better understand interobserver varia-tion 11 consists of observers who have completed all training and are independently treating patients. This provides an opportunity to further investigate interobserver variability and how to reduce it. Treatment decisions for radial head fractures are often based on radiologi-cal criteria and measurements according to Broberg and Morrey’s modification of the Mason classification 5, 9. This investigation tested the hypothesis that 3D-CT im-ages improve the interobserver reliability of the classification and characterization of radial head fractures over 2D-CT and radiographs.

Materials and MethodsStudy DesignIndependent observers (all orthopaedic surgeons) from several countries were invit-ed to evaluate twelve cases from a convenience sample of radial head fractures (se-lected to represent a full spectrum of radial head fracture morphologies and overall injury patterns) in an online survey: they were randomly assigned to review either radiographs and 2D-CT or radiographs and 3D-CT and then to determine the fracture classification, fracture characteristics and treatment plans. The randomization se-quence was determined by a computer random number generator (Windows Excel; Microsoft, Redmond, WA). The study was performed under a protocol approved by the Institutional Research Board at the principal investigators hospital. This was the inaugural study from a nascent collaborative called the Science of Variation Group (SOVG). The objectives of the collaborative are to study variation in the definition, interpretation, and classification of injury and disease. The Sci-ence of Variation Group has created a web-based platform that facilitates large in-ternational interobserver studies. With multiple fully trained surgeons from diverse countries and institutions participating in studies, this approach should provide a powerful forum for studying, understanding, and ultimately reducing interobserver variation in aspects of patient care.

BW_def.indd 66-67 07-02-11 22:25


ObserversA total of 206 surgeons were invited via e-mail to join the Science of Variation Group. We used lists of various professional organizations as well as friends and acquain-tances (along with their friends and acquaintances) to identify surgeons to invite for participation. We welcome any interested surgeon to join. Other than an ac-knowledgement as part of the author collaborative in the paper, no incentives were provided. One-hundred surgeons were interested in participation and logged on to the website. Forty-eight surgeons were randomized to 2D-CT scans and radiographs and fifty-two to 3D-CT scans and radiographs. Four weekly reminders to complete the online survey were e-mailed. Eighty-eight surgeons completed the study, from which 3 observers were excluded because of inability to view the online study due to hospital restriction. This study presents an analysis of the eighty-five observers that completed the study; 39 in the 2D-CT group and 46 in the 3D-CT group.

FracturesRadiographs and computed tomography scans of radial head fractures were identi-fied from a list of all cases treated by the senior investigator between 2000 and 2006 at one level-1 trauma center. The scanning technique was evaluated to determine suitability for 3D reconstructions (slice thickness between 0.62 and 1.25 mm, no met-al implants). Inclusion criteria were: 1) Radial head fracture; 2) CT scan appropriate for 3D reconstruction; 3) Age 18 or older. Inadequate quality of the CT scan prompted exclusion from the study. Radiographs and CT scans of radial head fractures from 30 patients were blinded by an independent research fellow for use in this study. Two of the authors (one subspecialty trained upper extremity surgeon and one research fellow in upper extremity trauma) selected twelve cases that had radial head frac-tures of different size, morphology, and location; representing most of the different patterns of traumatic elbow instability with radial head fracture. Radiographs, 2D-CT scans, and 3D-CT reconstructions were uploaded to the research group’s website. The 3D-CT reconstructions were created with use of Vitrea imaging software (Vital Im-ages, Minnetonka, Minnesota). For each case, videos with 2D-CT and 3D-CT images along the sagittal, coronal and axial cuts were created. The 3D-CT videos included a reconstruction of the entire elbow and a reconstruction with the distal humerus subtracted. Observers could scroll through the videos or play them automatically.

EvaluationObservers logged in independently on the website. Upon login to the website, they were asked to provide demographic and professional information: 1) location of prac-tice; 2) years in independent practice; 3) training of surgical trainees; 4) number of radial head fractures treated per year, and; 5) clinical specialty. Subsequently, observ-

ers were asked to classify the fractures according to Broberg and Morrey’s modifica-tion of the Mason classification 5, 9. Type 4 fractures were excluded, since we were interested in the radial head fracture independent of associated injuries. Observers were provided with the original description and corresponding images of the clas-sification system. The observers were also asked 7 questions regarding fracture characteristics: 1) Does the fracture line separate the entire articular surface from the radial neck? 2) Is there any comminution of the radial neck? 3) Does the fracture involve the articular surface? 4) Is there an articular step or gap of greater than 2 millimeters? 5) Is there any central impaction of the articular surface, 6) Are there more than 3 articular frag-ments? 7) Are any of the fragments too small to repair? They were also asked their preferred management: 1) Non-operative management; 2) Open reduction and internal fixation with screws, wires or pins; 3) Open reduction and internal fixation with plate and screws; 4) radial head excision; 5) radial head replacement (arthroplasty). Obser-vers were blinded to clinical information. Observers could comment on each case and all questions had to be completed in order to continue with the next case. The obser-vers completed the study at their own time and pace.

Statistical AnalysisThe kappa multirater measure (κ) was used to estimate agreement among surgeons with respect to fracture classification, fracture characteristics and treatment ap-proach. It is a commonly used statistical method to describe chance-corrected agree-ment in a variety of intra-observer and interobserver studies 2, 7, 13. Agreement among observers was calculated with use of the multirater kappa measure described by Sie-gel and Castellan 15. Kappa values were interpreted using the guidelines proposed by Landis and Koch 7: values of 0.01 to 0.20 indicate slight agreement; 0.21 to 0.40, fair agreement; 0.41 to 0.60, moderate agreement; 0.61 to 0.80, substantial agreement; and more than 0.81, almost perfect agreement. Zero indicates no agreement beyond that expected due to chance alone, – 1.00 means total disagreement, and + 1.00 rep-resents perfect agreement 7, 13. Two-sample independent Z-tests were performed for each variable to compare the kappa for 2 dimensional CT with that of 3 dimensional CT. Since the samples compared in this study were not independent (the same set of fractures were rated by the 2D-CT and 3D-CT group), this method produced conserva-tive estimates of the p-values. A post-hoc power analysis was performed using nQue-ry Advisor (version 7.0, nQuery Advisor, Statistical Solutions, Saugus, MA) to identify the power of each comparison and the sample size necessary to achieve 80% given both effect size and rater ratio remain constant at each iteration (Table V).

BW_def.indd 68-69 07-02-11 22:25


Sources of FundingNo funding was received in direct support of this study.

ResultsObserver Demographics A total of 85 observers participated in this investigation. The observer demographics are summarized in table I. Among the others surgeons there were 3 hand surgeons (no wrist), 2 trauma surgeons and 3 upper extremity surgeons (hand, wrist, elbow and shoulder).

ClassificationThe interobserver variation for classification of the fractures according to Broberg and Morrey’s modification of the Mason classification 1 was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.37, SE 0.010, and κ3D = 0.49, SE 0.023; p = < 0.001) (Table IIA-B).

Fracture Characteristics Agreement on central impaction of the articular surface was fair with use of 2D-CT scans and slight with use of 3D-CT reconstructions (κ2D = 0.22, SE 0.027, and κ3D = 0.15, SE 0.010; p = 0.006). Interobserver agreement on presence of more than 3 ar-

Table IIA. Interobserver Agreement 2-Dimensional CT

2-Dimensional CTVariable Categorical κ 95% CI P Value

Broberg and Morrey Modified Mason Classification Fair 0.37 (0.35, 0.39) <0.001*

Fracture CharacteristicsFracture line Moderate 0.41 (0.31, 0.51) 0.139Comminution radial neck Fair 0.33 (0.25, 0.41) 0.328Articular surface involvement Slight 0.02 (-0.46, 0.49) 0.961Step/gap >2 millimeters Substantial 0.65 (0.58, 0.71) 0.152Central impaction Fair 0.22 (0.17, 0.28) 0.006*Three articular fragments Fair 0.38 (0.35, 0.40) <0.001*Too small to repair Moderate 0.47 (0.44, 0.49) <0.001*

Proposed Treatment Fair 0.26 (0.24, 0.29) <0.001*

*Statisticallysignificantcomparedto3-Dimensional.CI=confidenceinterval.

Table IIB. Interobserver Agreement 3-Dimensional CT

3-Dimensional CTVariable Categorical κ 95% CI P Value

Broberg and Morrey Modified Mason Classification Moderate 0.49 (0.45, 0.54) <0.001*

Fracture CharacteristicsFracture line Moderate 0.50 (0.43, 0.58) 0.139Comminution radial neck Fair 0.38 (0.33, 0.42) 0.328Articular surface involvement Poor 0.00 (-0.81, 0.80) 0.961Step/gap >2 millimeters Substantial 0.73 (0.63, 0.82) 0.152Central impaction Slight 0.15 (0.13, 0.16) 0.006*Three articular fragments Substantial 0.61 (0.59, 0.63) <0.001*Too small to repair Substantial 0.61 (0.59, 0.63) <0.001*

Proposed Treatment Fair 0.40 (0.37, 0.43) <0.001*

*Statisticallysignificantcomparedto2-DimensionalCT.CI=confidenceinterval.

Table I. Observer demographics

2-Dimensional CT

(N = 39)3-Dimensional

CT (N = 46)Total

(N = 85) N % N % N %Practice

Asia 1 3 2 4 3 4Australia 1 3 1 2 2 2Canada 2 5 2 4 3 4Europe 6 15 9 20 15 18United Kingdom 1 3 0 0 1 1United States 28 72 31 67 59 69Other 0 0 1 2 1 1

Years In practice0-5 11 28 7 15 18 216-10 6 15 15 33 21 2511-20 18 46 17 37 35 4121-30 4 10 7 15 11 13

Supervise ResidentsYes 35 90 40 87 75 88No 4 10 6 13 10 12

Fractures per year0-5 10 26 3 7 13 156-10 10 26 15 33 25 2911-20 13 33 11 24 22 26>20 6 15 17 37 23 27

SpecializationGeneral orthopaedics 2 5 1 2 3 4Orthopaedic traumatology 12 31 13 28 25 29Shoulder and elbow 7 18 4 9 11 13Hand and wrist 14 36 24 52 38 45Other 4 10 4 9 8 9

BW_def.indd 70-71 07-02-11 22:25


ticular fragments was fair with use of 2D-CT scans and substantial with use of 3D-CT reconstructions (κ2D = 0.38, SE 0.011, and κ3D = 0.61, SE 0.010; p = < 0.001). Agreement on presence of fragments too small to repair was moderate with use of 2D-CT scans and substantial with use of 3D-CT reconstructions (κ2D = 0.47, SE 0.013, and κ3D = 0.61, SE 0.010; p = < 0.001) (Table IIA-B).

Treatment Interobserver agreement on treatment was fair with both 2D-CT scans and 3D-CT reconstructions (κ2D = 0.26, SE 0.012, and κ3D = 0.40, SE 0.013; p = < 0.001) (Table II).

Observer demographics and Mason classification When classifying fractures according to the Mason classification agreement among United States observers was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.32, SE 0.01, and κ3D = 0.52, SE 0.03; p = < 0.001) (Table IIIA-B). Agreement among observers who were in practice 5 or fewer years was moderate with use of 2D-CT scans and substantial with use of 3D-CT reconstructions (κ2D = 0.44, SE 0.03, and κ3D = 0.62, SE 0.18; p = 0.039). Agreement among observers

who were in practice from 6 to 10 years was fair with use of 2D-CT scans and moder-ate with use of 3D-CT reconstructions (κ2D = 0.32, SE 0.05, and κ3D = 0.53, SE 0.05; p = 0.002). Agreement among observers who were in practice from 11 to 20 years was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.35, SE 0.02, and κ3D = 0.45, SE 0.04; p = 0.011). Agreement among observers who treated 5 or fewer radial head fractures per year was fair with both 2D-CT scans and 3D-CT reconstructions (κ2D = 0.27, SE 0.03, and κ3D = 0.32, SE 0.14; p = 0.76). Agreement among observers who treated 6 to 10 radial head fractures per year was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.39, SE 0.04, and κ3D = 0.48, SE 0.04; p = 0.063). Agreement among observers who treated 11 to 20 radial head fractures per year was moderate with both 2D-CT scans and 3D-CT reconstructions (κ2D = 0.44, SE 0.03, and κ3D = 0.46, SE 0.05; p = 0.66). Agreement among observers that treated more than 20 radial head fractures per year was moderate with both 2D-CT scans and 3D-CT recon-structions (κ2D = 0.46, SE 0.06, and κ3D = 0.52, SE 0.05; p = 0.44). Agreement among orthopaedic traumatology specialist observers was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.37, SE 0.03, and κ3D = 0.47, SE 0.04; p = < 0.05). Agreement among hand and wrist spe-

Table IIIA. Differences in Agreement on Broberg and Morrey modification of the Mason classification between Observer demographics and Imaging Modality

2-Dimensional CT P-Value Variable N Catagorical κ SE Practice

Europe 6 Moderate 0.50 0.05 0.275United States 28 Fair 0.32 0.01 < 0.001Other 5 Moderate 0.44 0.07 0.561

Years In practice0-5 11 Moderate 0.44 0.03 0.0396-10 6 Fair 0.32 0.05 0.00211-20 18 Fair 0.35 0.02 0.01121-30 4 Moderate 0.47 0.10 0.449

Fractures per year0-5 10 Fair 0.27 0.03 0.7576-10 10 Fair 0.39 0.04 0.06311-20 13 Moderate 0.44 0.03 0.663>20 6 Moderate 0.46 0.06 0.411

SpecializationGeneral orthopaedics 2 Moderate 0.49 0.21 .Orthopaedic traumatology 12 Fair 0.37 0.03 < 0.05Shoulder and elbow 7 Moderate 0.50 0.05 0.107Hand and wrist 14 Fair 0.31 0.03 < 0.001Other 4 Moderate 0.47 0.09 0.396

N=NumberofObservers,SE=standarderror;P-value<0.05=Significantcomparedto3-DimensionalCT

Table IIIB. Differences in Agreement on Broberg and Morrey modification of the Mason classification between Observer demographics and Imaging Modality

3-Dimensional CT P-Value Variable N Catagorical κ SE Practice

Europe 9 Moderate 0.43 0.04 0.275United States 31 Moderate 0.52 0.03 < 0.001Other 6 Moderate 0.51 0.10 0.561

Years In practice0-5 7 Substantial 0.62 0.08 0.0396-10 15 Moderate 0.53 0.05 0.00211-20 17 Moderate 0.45 0.04 0.01121-30 7 Fair 0.38 0.08 0.449

Fractures per year0-5 3 Fair 0.32 0.14 0.7576-10 15 Moderate 0.48 0.04 0.06311-20 11 Moderate 0.46 0.05 0.663>20 17 Moderate 0.52 0.05 0.411

SpecializationGeneral orthopaedics 1 . . . .Orthopaedic traumatology 13 Moderate 0.47 0.04 < 0.05Shoulder and elbow 4 Fair 0.30 0.11 0.107Hand and wrist 24 Moderate 0.54 0.03 < 0.001Other 4 Fair 0.35 0.11 0.396

N=NumberofObservers,SE=standarderror;P-value<0.05=Significantcomparedto2-DimensionalCT

BW_def.indd 72-73 07-02-11 22:25


cialist observers was fair with use of 2D-CT scans and moderate with use of 3D-CT reconstructions (κ2D = 0.31, SE 0.03, and κ3D = 0.54, SE 0.03; p = < 0.001). There were significant differences in agreement among the Broberg and Morrey modification of the Mason classification on 2D-CT between surgeons in Eu-rope and the United States (p = 0.001), surgeons who treat 0-5 and >20 fractures per year (p = 0.006) and between orthopedic traumatology and shoulder and elbow spe-cialists (p = 0.017). With 3D-CT the only differences were between surgeons with 0-5 and 21-30 years in practice (p = 0.037) (Table IV).

Table IV. Differences in Agreement between Observers for Broberg and Morrey modification of the Mason Classification Among Imaging Modality

Variable 2-Dimensional CT 3-Dimensional CTPractice

Europe vs. United States 0.001* 0.06Years In practice

0-5 vs. 21-30 0.715 0.037*Fractures per year

0-5 vs. >20 0.006* 0.175Specialization

Ortho Trauma vs. Shoulder & elbow 0.017* 0.175

*=Significant(p<0.05)

DiscussionThe collaborative, internet based approach has facilitated large, international stud-ies of inter-rater variation 6, 8. Additionally, only fully trained surgeons, many with substantial clinical experience participated. Inclusion of surgeons from multiple countries and continents should increase the generalizability of the results. Using high-speed internet connections and improved compression techniques, we were able to provide high quality reproduction images and movies via the internet. There are many weaknesses in this study. First, the quality of the radio-graphs was limited to what had been obtained at the time of injury, which reflects usual practice, but not what might be achieved with specific protocols. In addition, we provided limited information about the patient and the injury. There was also a spectrum bias by selecting cases to represent the known variety of injuries, with the result that less common complex fractures were over represented compared to the more common minimally or slightly displaced fractures. Our study reflects what would be expected with relatively complex fractures of the radial head–the reli-ability would be expected to be higher if we included more of the non-displaced or minimally displaced fractures that makeup the majority of radial head fractures. Another shortcoming is the fact that a small number of observers either uncom-monly or never treat radial head fractures, but we did not plan for exclusions on this basis and therefore did not do so after the fact to avoid introducing bias. The power is based on the total number of observations allowing us to use a smaller number of cases and thereby decrease burden and increase participation of observers. Given the small kappa differences between 2D and 3D for certain questions (e.g. articular sur-face involvement) and the large variabilities, power was low, and huge sample sizes would be required for 80% power for detecting differences between 2D and 3D given the levels of agreement observed. For other questions (e.g. three articular fragments), the observed power was very high with the numbers of surgeons participating and even fewer would have attained the traditional 80% power. Finally, this is an artifi-cial research situation given that in clinical practice patients would have both the two and three-dimensional reconstructions available to them. We speculate that the very poor agreement regarding articular surface in-volvement might reflects misunderstanding of the question-based on comments re-ceived as part of the survey some observers probably thought we were referring to involvement of the part of the radial head that articulates with the lesser sigmoid notch of the ulna. The poor agreement regarding central impaction likely reflects the lack of a precise or consistent definition of this term. The findings of this study are otherwise consistent with prior studies on the distal humerus, distal radius and the coronoid 3, 4, 8. Three-dimensional reconstructions are made from CT-scans and there-

Table V. Post-hoc Power Analysis for Comparing 2D and 3D CT Images

VariablePower(1 – b)*

Minimum N per Group for 80% Power*

Broberg-Morrey Modified

Mason Classification.99 20

Fracture CharacteristicsFracture line .29 179Comminution radial neck .19 289Articular surface involvement .05 500Step/gap >2 mm .25 209Central impaction .72 57Three articular fragments .99 5Too small to repair .99 8

Proposed Treatment .99 10

*PowerisbasedoncomparingkappainterobserveragreementusingindependentgroupsZ-test(39surgeons2D-CT,46surgeons3D-CT).

BW_def.indd 74-75 07-02-11 22:25


fore do not require additional scanning or expose the patient to additional radia-tion. It has been calculated at the investigators institution, the cost for additional 3D reconstructions are an additional 20% of the costs of a CT-scan. Free software such as OsiriX 14 is available which makes it possible for every orthopedic surgeon to quickly and easily create 3D reconstructions themselves with minimal training. Three-dimensional CT images led to small but significant decreases in vari-ation between observers for fracture classification and some fracture characteristics compared to 2D-CT, but a notable amount of variation remains even with more so-phisticated imaging. Our belief that 3D-CT images are easier for surgeons to interpret is supported by the observation that 3D-CT produced a higher agreement for Broberg and Morrey’s modification of the Mason classification than previously reported in the literature 10, 12 and 3D-CT was associated with less disagreement in classification than 2D-CT across various cultures, training, subspecialty and levels of experience. Nonetheless agreement was only fair or moderate at best even with 3D-CT. Further-more, some might interpret this data as showing much less influence one interob-server variation than one might guess. Other potential sources of interobserver variation include unfamiliar or unclear definitions, and differences in culture, training, and exposure. In our opin-ion, the fact that well-trained, experienced observers disagree indicates that there are variations in these factors that lead different experts to see different things in sophisticated images. In other words, reducing interobserver variation seems to de-pend on something more than better imaging. Additional research to identify and reduce sources of observer variation in the interpretation of diagnostic images is merited.

References


Relat Res. 1987(216):109-19.

2. Cohen J. A coefficient of agreement for nominal scales. Edoc Psychol Meas. 1960;20(1):37–46.

3. Doornberg J, Lindenhovius A, Kloen P, van Dijk CN, Zurakowski D, Ring D. Two and three-dimen-

sional computed tomography for the classification and management of distal humeral frac-

tures. Evaluation of reliability and diagnostic accuracy. J Bone Joint Surg Am. 2006;88(8):1795-

801.

4. Harness NG, Ring D, Zurakowski D, Harris GJ, Jupiter JB. The influence of three-dimensional

computed tomography reconstructions on the characterization and treatment of distal radial

fractures. J Bone Joint Surg Am. 2006;88(6):1315-23.



6. Karanicolas PJ, Bhandari M, Kreder H, Moroni A, Richardson M, Walter SD, et al. Evaluating

agreement: conducting a reliability study. J Bone Joint Surg Am. 2009;91 Suppl 3:99-106.

7. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics.

1977;33(1):159-74.



phy. J Hand Surg Am. 2009;34(9):1640-6.






1999;87(6):2007-15.



13. Posner KL, Sampson PD, Caplan RA, Ward RJ, Cheney FW. Measuring interrater reliability among

multiple raters: an example of methods for nominal data. Stat Med. 1990;9(9):1103-15.

14. Rosset A, Spadola L, Ratib O. OsiriX: An Open-Source Software for Navigating in Multidimen-

sional DICOM Images. Journal of Digital Imaging. 2004;17(3):205-16.

15. Siegel S, Castellan NJ. Nonparametric Statistics for the Behavioral Sciences. 2nd ed. New York:

McGraw-Hill; 1988.

BW_def.indd 76-77 07-02-11 22:25

78 part iii | treatment 79 stretch pain after radial head fracture | chapter 6

CHAPTER 6Attitude Towards Stretch Pain of the Elbow After Radial Head Fracture


Ana-Maria Vranceanu, PhD

David Ring, MD PhD

SubmittedtoClinOrthopRelatRes

Part III: Treatment

BW_def.indd 78-79 07-02-11 22:25


Abstract

Hypothesis: This study was designed to test the hypothesis that agreement with the idea that “stretching of the elbow beyond the point were it becomes painful is im-portant in recovery” leads to greater elbow range of motion one month after injury.Methods: Seventy-one patients with an isolated Broberg and Morrey modified Ma-son Type 1 or Type 2 radial head fracture seen within 14 days after injury were en-rolled prospectively. They completed the Pain Catastrophizing Scale (PCS), Center for Epidemiologic Studies Depression Scale (CES-D) and were asked to rate their agree-ment with a statement regarding pain and recovery from their injury on a 5-point Likert scale, which were collapsed into 3 categories (disagree, neutral, agree) to fa-cilitate statistical power. One-month later, patients completed the Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire and elbow and forearm motion were measured with a goniometer.Results: Nine patients (12.6%) disagreed with the role of pain in recovery, 6 (8%) were neutral, and 56 (78.9%) agreed. Patients that disagreed with the role of stretch pain in recovery were older (p = 0.031), had more depressive symptoms (CES-D; p = 0.047), and achieved less elbow extension (p = 0.050) and forearm rotation (p = 0.017) ap-proximately one month after injury. Conclusions: A negative attitude towards stretch pain during recovery from fracture of the radial head is associated with less elbow motion one month after injury. Fu-ture studies should address the ability to improve recovery by encouraging a change in pain paradigm. LevelofEvidence:PrognosticLevel1

IntroductionIsolated, stable, minimally displaced fractures of the radial head (Types 1 and 2 of the Broberg and Morrey modification of the Mason Classification 3) are common frac-tures that are usually treated non-operatively. The most common sequel of these fractures is elbow stiffness 8-10, 17, 21, 24, 25. In our experience, the elbow stiffness may be a result of excessive immobilization or ineffective stretching exercises. It can be counter-intuitive to intentionally cause pain in the setting of an injury. Vulnerability and protectiveness are enhanced by automatic thoughts such as “pain indicates harm”, “the pain is permanent”, or other aspects of a maladaptive response to nociception that psychologists have termed pain catastrophizing 1, 12. Re-search suggests that fear of pain, thinking the worst in response to nociception (pain catastrophizing) and pain anxiety may be important determinants of recovery af-ter an acute fracture 1, 12. Similarly, depression hinders recovery after fracture 16, 22. We tested the hypothesis that agreement that painful stretches are an important part in recovery leads to greater motion one month after injury. Secondarily, we tested the hypothesis that depression and pain catastrophizing correlate with lack of agree-ment that painful stretches are important, as well as less motion and more disability one month after injury.

Materials and MethodsInclusion and Exclusion CriteriaUnder an IRB (Institutional Review Board) approved protocol, we prospectively in-clude patients with a radial head fracture seen at one Level 1 trauma center. Inclu-sion criteria were 1) A non-operatively treated fracture of the radial head; 2) Type 1 or Type 2 according to the Broberg and Morrey modification of Mason’s classification; 3) Seen within 14 days of injury, 4) Skeletal maturity; 5) Cognitive and physical ability to follow exercise instructions; 6) Isolated injury and 7) No clinical or radiographic evidence of injury to the medial elbow ligaments. Exclusion criteria were pregnant women and patients unable to give informed consent. Patients with prior elbow injury, disease, or arthritis were not excluded. Eighty-five patients satisfied the in-clusion and exclusion criteria. Fourteen patients did not return 1 month after injury resulting in a final cohort of 71 patients. Among the 71 patients, the average age was 44.4 years (range, 19 to 72 years). There were 18 men (25.4%) and 53 (74.6%) women. Sixty-five fractures were Mason type 1 (91.5%) and six were Mason type 2 (8.5%). Thirty-four patients (47.9%) injured their right elbow and 37 (52.1%) their left elbow. The dominant side was injured in 38 cases (53.5%) and the non-dominant side in 33 cases (46.5%). Fifty-seven patients (80.3%) fractured their elbow in a fall from a standing height, 3 (4.2%) from a greater

BW_def.indd 80-81 07-02-11 22:25


height and 11 (15.5%) patients suffered from a multi-vehicle accident (MVA). Forty-eight patients (67.6%) were white-collar workers. Six doctors participated in the care of these patients, from which one doctor treated 56 (78.9%) patients. The average follow-up was 33.6 days (range, 13 to 70 days). EvaluationPatients were approached during the initial outpatient visit to an orthopaedic surgeon and informed consent was obtained. The patients completed the Pain Catastrophiz-ing Scale (PCS) and the Center for the Epidemiological Study of Depression Instrument (CES-D). The Pain Catastrophizing Scale (PCS) is a reliable and valid 13-item question-naire, developed by Sullivan et al. 23 to measure the extent to which people think the worst in response to pain in clinical and nonclinical populations Each PCS item is rated on a 4-point scale: 1 (not at all) to 4 (all the time). The 13 items are summed to create a total score. The CES-D scale is a reliable and valid 19 item self-report scale designed to measure depressive symptoms in the general population 18. Subjects were also asked to rate their agreement with the following state-ment regarding recovery from their injury: “Do you think that you should perform the exercise beyond the point were it gets painful to reach the best possible recovery following your fracture?” with use of a 5-point Likert scale (5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, and 1 = strongly disagree). During the routine one-month follow-up with their treating surgeon, all enrolled subjects completed the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire to measure upper extremity specific disability with a lower score in-dicating less disability 11. In addition, the elbow flexion arc and range of forearm rota-tion was measured with a goniometer by a research fellow that was not involved with the treatment of that patient.

RehabilitationDuring the initial outpatient visit, patients were advised to actively move the el-bow into as much flexion, extension, supination and pronation as possible and to self-assist by pushing with the uninjured hand (or a wall or desk as a fulcrum) in order to stretch the elbow to achieve greater motion. Additionally, patients were in-structed that weight bearing on the elbow, lifting, and grasping and that pain con-trol either physically e.g. cold application or using pain medication was allowed. They were advised to return to deskwork, but limit forceful activities or risk of an-other fall. Additionally, patients were instructed that the pain might make them feel protective and cautious, but that it was a false alarm as no harm could come from these stretches and that they were an important and helpful part of recovery. The instructing doctor (blinded) was a hand fellowship trained orthopedic surgeon.

Statistical AnalysisPower analysis indicated that a minimum sample size of 18 subjects (6 in each group) would provide 80% statistical power to detect a significant difference in elbow flex-ion-extension arc, assuming an effect size of 2.0 or greater (mean difference of 20 degrees, standard deviation of 10 degrees) (alpha = 0.05, beta = 0.20) using one-way ANOVA. Two-tailed p values of <0.05 were considered to be significant. Frequencies were used to describe demographics, fracture characteristics and outcome variables. Two main sets of analyses were conducted. In the first set, we used analyses of variance to look for differences in depression, pain catastrophizing, disability, elbow flexion and forearm rotation by category of response to the state-ment about pain. Due to a small number of responses in some of the categories, and in order to achieve statistical power (at least 6 participants per group), we combined the strongly agree and agree categories, as well as the strongly disagree and disagree categories. In the second set, we used Spearman correlations to test the relationship between ranked level of agreement with the pain statement and the outcome vari-ables.

ResultsNine patients (12.6%) disagreed with the role of pain in recovery, 6 (8%) were neutral, and 56 (78.9%) agreed. Demographics, fracture characteristics and outcome variables were compared between patients who agreed, were neutral or disagreed that pain was useful for recovery. Patients who disagreed with the role of pain in recovery were older than those who were neutral or agreed with the role of pain in recovery (F = 3.79, p = 0.031). Patients who were neutral regarding the role of pain had more depressive symptoms than those who agreed or disagreed with the role of pain in recovery (F = 3.12, p = 0.047). Patients who disagreed that pain is important in recovery had less one-month post-injury pronation (F = 4.29, p = 0.016), elbow extension (F = 3.12, p = 0.050) and combined forearm motion (F = 4.27; p = 0.017). There was no statistical difference between groups in PCS (p = 0.17), DASH (p = 0.20), elbow flexion (p = 0.34), and flexion-extension arc (p = 0.13). There was also no statistical difference in time from injury to final evaluation (p = 0.67) (Table I). There was a significant correlation between 5-point Likert ranked agree-ment with the role of pain in recovery and PCS (rho = -0. 26, p = 0. 031), a trend to-ward significance for CES-D total score (rho = -0.22, p = 0.070), and no correlation with DASH (rho = -0.10, p = 0.39), one month post-injury elbow flexion (rho = -0.027, p = 0. 82), extension (rho = 0.13, p = 0.28), pronation (rho = 0.12, p = 0.30), supination (rho = -0.053, p = 0.66), flexion-extension arc (rho = -0.12, p = 0. 32) or arc of forearm rotation

BW_def.indd 82-83 07-02-11 22:25


(rho = 0.001, p = 0.99) (Table II). There was a significant correlation between CES-D score and PCS score (rho = 0.47, p = <0.001).

DiscussionIn our experience, many health care providers-occupational and physical therapists in particular-advise patients to work to pain, but not beyond 6. They often further admonish that painful activities may cause “inflammation”, which they feel would be counterproductive. There does not appear to be a scientific basis for these recom-mendations and they amount, more or less, to a culture or tradition 5, 14, 15. Our obser-vation has been that patients recover greater motion and do so more rapidly after injury when they are confident and feel good about stretching their arm. In our opinion, according to this paradigm, stretches to gain motion are the same as ath-letic stretches–both are an intentional tearing of tissue. We believe pain indicates that one is doing the exercise correctly. Maladaptive responses to nociception are associated with greater pain intensity and arm-specific disability 4, 7, 13, 19, 20. Advising patients that pain during stretching exercises may be harmful risks reinforcing these maladaptive coping strategies. The current study was designed to evaluate the association of attitude towards stretch pain with recovery of elbow motion after a minimally displaced frac-ture of the radial head. The strengths of this study include the prospective design and enrollment of patients from 6 different surgeons’ practices. Limitations include: 1) Enrollment after diagnosis, reassurance, and coaching with motion exercises, all of which are therapeutic interventions that may affect attitudes towards pain (and may explain why the majority of patients agreed with the importance of pain for recovery); 2) Ceiling effects since most patients with minimally displaced radial head fractures regain near normal motion regardless of their paradigm; 3) Meaningless variation (or “noise”) added to the data by virtue of the fact that the measurement error of a hand-held goniometer is comparable to the small differences in elbow flexion con-tractures observed; 4) The measure of agreement with a statement regarding the role of pain in recovery is an indirect measure of confidence with exercises-a more direct and objective measure would be preferable; patients may state agreement on a questionnaire, but still have a hard time performing stretches beyond pain; 5) Wide range in the time that people returned for the “one-month” follow-up (although this did not correlate with any of the outcome measures); and 6) Limited power due to unequal distribution of patients among agreement groups.

Table I. Bivariate Analysis of Factors Associated with Agreement on Stretch Pain

Variable Agree (N = 56) Neutral (N = 6) Disagree (N = 9) P-ValueAge, years 48 48 63 0.031*Sex 0.15

Male 17 1 0Female 39 8 6

Mason Type 0.421 50 9 62 6 0

Side 0.060Right 30 1 3Left 26 8 3

Dominance 0.22Right 44 9 4Left 12 0 2

Dominant Injured 0.12Yes 32 2 4No 24 7 2

Mechanism 0.68Fall lower height 44 7 6Greater Height 3 0 0MVA 9 2 0

Occupation 0.17White collar 40 6 2Other 16 3 4

Doctor 0.621 45 6 5Other 11 3 1

Follow-up, Days 30 29 31 0.67CES-D 10 19 16 0.047*PCS 16 23 18 0.17DASH 15 10 26 0.20Range of Motion, degrees Pronation 90 90 90 0.016* Supination 90 90 90 0.083* Elbow Flexion 135 135 133 0.34 Elbow Extension 0 0 7.5 0.050* FE Arc 135 135 133 0.13 PS Arc 180 180 180 0.017*

*P<0.05=Statisticallysignificant,ContinuousVariablesasMedian

Table II. Spearman Rho Correlations for Agreement with Stretch Pain

CES-D PCS DASH Flexion Extension Pronation Supination FE Arc PS ArcRho -0.216 -0.256 -0.104 -0.027 0.129 0.125 -0.053 -0.120 0.001P-Value 0.070 0.031* 0.389 0.824 0.282 0.300 0.658 0.319 0.993

*P<0.05=Statisticallysignificant

BW_def.indd 84-85 07-02-11 22:25


In spite of these shortcomings, we did find that a patient’s paradigm with respect to the role of pain in recovery predicted motion one month after injury and that a patient’s paradigm had small but significant correlation with pain catastroph-izing. The lack of correlation between attitude and disability might be due to the small number of patients in the neutral (9) and disagree (6) categories versus the agree category (55). There is a nearly 11 point difference in the means DASH score in patients that agree (17.4) or are neutral (17.3) about the role of pain and those that disagree (28.0), which seems clinically important. Of note is that the mean DASH for patients that disagree is higher than what has been reported in patients with frac-tures, while the mean DASH for the other categories is lower (21). A small subset of patients get stiff after radial head fractures and we prob-ably need a larger study to adequately study these issues. If additional studies cor-roborate the role of automatic thoughts and beliefs (intuition, “gut feelings”) in re-covery from injury, as well as the correlation of these thoughts with maladaptive responses to nociception or depression, then there is room for improvement in our teaching and coaching of post-injury exercises. As revered hand surgery pioneer Paul Brand noted in his book “The Gift of Pain”, nociception exists for our protection 2. It’s no surprise that pain after injury can make us feel vulnerable and protective. The key may be to help our patients change their mindset from vulnerability to recovery, seeing a painful exercise more as a useful stretch exercise and the post-exercise pain more as that rewarding ache after a great work out.

ConclusionThis study found that a negative attitude towards stretch pain during recovery from fracture of the radial head is associated with less elbow motion one month after in-jury. Future studies should address the ability to improve recovery by encouraging a change in pain paradigm.

References

1. Arinzon Z, Gepstein R, Shabat S, Berner Y. Pain perception during the rehabilitation phase fol-

lowing traumatic hip fracture in the elderly is an important prognostic factor and treatment

tool. Disability and Rehabilitation. 2007;29(8):651-8.

2. Brand PW, Yancey P. The gift of pain. Grand Rapids, Michigan: Zondervan; 1997. p. 1-352.


Relat Res. 1987(216):109-19.

4. Ciechanowski P, Sullivan M, Jensen M, Romano J, Summers H. The relationship of attachment

style to depression, catastrophizing and health care utilization in patients with chronic pain.

Pain. 2003;104(3):627-37.

5. Clarkson PM, Hubal MJ. Exercise-induced muscle damage in humans. Am J Phys Med Rehabil.

2002;81(11 Suppl):S52-69.

6. Davila S, Evelyn M. Therapist’s management of fractures and dislocations of the elbow. In:

Mackin E, editor. Rehabilitation of the hand and upper extremity. St. Louis: Mosby; 2002. p.

1-2180.

7. Geisser ME, Robinson ME, Keefe FJ, Weiner ML. Catastrophizing, depression and the sensory,

affective and evaluative aspects of chronic pain. Pain. 1994;59(1):79-83.

8. Hammacher ER, van der Werken C. Radial head fractures: operative or conservative treatment?

The Greek temple model. Acta Orthop Belg. 1996;62 Suppl 1:112-5.




10. Hotchkiss RN. Fractures of the Radial Head and Related Instability and Contracture of the

forearm. Instructional Course Lectures; 1998. p. 173-7.

11. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome measure:

the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper Extremity Col-

laborative Group (UECG). Am J IndMed. 1996;29(6):602-8.

12. Linton SJ, Buer N, Samuelsson L, Harms-Ringdahl K. Pain-related fear, catastrophizing and pain

in the recovery from a fracture. Scandinavian Journal of Pain. 2010;1(1):38-42.

13. Main CJ, Wood PL, Hollis S, Spanswick CC, Waddell G. The Distress and Risk Assessment Method.

A simple patient classification to identify distress and evaluate the risk of poor outcome. Spine

(Phila Pa 1976). 1992;17(1):42-52.


1999;87(6):2007-15.

15. Morgan DL, Proske U. Popping sarcomere hypothesis explains stretch-induced muscle damage.

Clin Exp Pharmacol Physiol. 2004;31(8):541-5.

16. Mossey JM, Mutran E, Knott K, Craik R. Determinants of recovery 12 months after hip fracture:

the importance of psychosocial factors. Am J Public Health. 1989;79(3):279-86.

BW_def.indd 86-87 07-02-11 22:25


17. Radin EL, Riseborough EJ. Fractures of the radial head. A review of eighty-eight cases and

analysis of the indications for excision of the radial head and non-operative treatment. J Bone

Joint Surg Am. 1966;48(6):1055-64.

18. Radloff LS. The CES-D scale: A self report depression scale for research in the general popula-

tion. Applied Psychological Measurement. 1977;1:385-401.

19. Ring D, Kadzielski J, Fabian L, Zurakowski D, Malhotra LR, Jupiter JB. Self-reported upper extrem-

ity health status correlates with depression. J Bone Joint Surg Am. 2006;88(9):1983-8.

20. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back pain patients: relation-

ship to patient characteristics and current adjustment. Pain. 1983;17(1):33-44.

21. Schwarz N. [Conservative therapy of isolated radius head fracture--immobilization or early

functional movement treatment?]. Aktuelle Traumatol. 1983;13(3):97-102.

22. Shyu YI, Chen MC, Cheng HS, Deng HC, Liang J, Wu CC, et al. Severity of depression risk predicts

health outcomes and recovery following surgery for hip-fractured elders. Osteoporosis Inter-

national. 2008;19(11):1541-7.

23. Sullivan MJL, Bishop SR, Pivik J. The Pain Catastrophizing Scale: Development and validation.

Psychol Assess. [Article]. 1995;7(4):524-32.

24. Unsworth-White J, Koka R, Churchill M, D’Arcy JC, James SE. The non-operative management of

radial head fractures: a randomized trial of three treatments. Injury. 1994;25(3):165-7.

25. Weseley MS, Barenfeld PA, Eisenstein AL. Closed treatment of isolated radial head fractures. J

Trauma. 1983;23(1):36-9.

BW_def.indd 88-89 07-02-11 22:25

90 part iV | outcome 91 radiographic arthrosis after traumatic elbow injuries | chapter 7

CHAPTER 7Incidence and Risk Factors for the Development of Radiographic Arthrosis After Traumatic Elbow Injuries


David Zurakowski, PhD

C. Niek van Dijk, MD PhD

David Ring, MD PhD

JHandSurgAm.2010;35(12):1976-80.

Part IV: Outcome

BW_def.indd 90-91 07-02-11 22:25


Abstract

Purpose: Radiographic arthrosis is a common sequela of elbow trauma. Few studies have addressed risk factors for radiographic arthrosis after elbow injury, especially in the long term. Data from multiple long-term follow-up studies of patients with sur-gically treated elbow fractures provided us with an opportunity to assess risk factors for long-term radiographic arthrosis after elbow injury.Methods: During a 5-year period, we obtained radiographs during a research-specific evaluation of 139 patients (81 men and 58 women) 10 or more years (median, 19.5 y; range, 10–34 y) after surgical treatment of an elbow fracture as part of multiple retrospective studies. Radiographic arthrosis was graded according to the system of Broberg and Morrey. Bivariate and multivariable analyses evaluated risk factors for radiographic arthrosis.Results: Of 139 patients, 75 had radiographic evidence of arthrosis at final evalua-tion and 32 had moderate or severe radiographic arthrosis. Mechanism of injury, age, gender, follow-up time, occupation, and limb dominance were not associated with radiographic arthrosis. Multiple logistic regression analysis identified the type of in-jury as the only independent predictor of moderate to severe radiographic arthrosis. Patients with a bicolumnar fracture of the distal humerus, a capitellum/trochlear fracture, or an elbow fracture–dislocation were 8.0, 7.3, and 5.2 times more likely (odds ratio), respectively, to develop radiographic evidence of moderate or severe ra-diographic arthrosis than the average patient in this cohort.Conclusions: Distal humerus fractures (both columnar and capitellum/trochlea) and elbow fracture–dislocations are more likely than fractures of the olecranon and ra-dial head to develop moderate or severe radiographic arthrosis in the long term.LevelofEvidence:IV,prognosticstudy

IntroductionRadiographic arthrosis is a common sequela of elbow trauma resulting from direct cartilage injury, instability, and articular incongruity 5, 6, 17, 18. It is understood that over the long term, many patients develop radiographic signs of osteoarthrosis after elbow trauma, although symptoms vary and few patients present for treatment 9, 11. There is a limited relationship between radiographic evidence of arthrosis and impairment or disability 7, 12. Not much has been published regarding risk factors for radiographic arthrosis after elbow injury, especially in the long term. Data from multiple long-term follow-up studies of injured elbows provided the opportunity to assess the risk factors for posttraumatic elbow arthrosis on radiographs. Our null hypothesis was that different types of elbow injuries have rates of radiographic arthrosis (indepen-dent of function or outcome) that are comparable at equivalent follow-up times 7, 12, 13.

Patients and MethodsPatientsDuring a 5-year period (2005–2010), we collected radiographs and clinical data from a research-specific, long-term evaluation of patients who had surgical treatment of displaced and unstable elbow fractures between 1975 and 1998 as part of 7 retrospec-tive studies 3, 7, 8, 12, 14. All 7 studies were performed at one institution in The Nether-lands and were approved by the institutional review board. Inclusion criteria for the present study were: (1) surgically treated elbow fracture, (2) age greater than 18 years at the time of injury, and (3) date of evaluation more than 10 years after the initial injury or surgery. Among the 235 patients in the prior studies, 139 satisfied the inclu-sion and exclusion criteria and made up the study cohort. There were 81 men and 58 women, with a mean age of 38 years at the time of injury (range, 18–73 years). The dominant arm was involved in 69 patients. Twenty-nine patients were employed as laborers at the time of injury; 110 were nonlaborers. Injuries included fracture–dislocation of the elbow in 29 patients, distal hu-merus columnar fracture in 29, isolated radial head fracture in 20, proximal ulna/olecranon fracture in 54, and capitellum/trochlear fracture in 7. According to the AO comprehensive classification of fractures 15, 2 distal humerus columnar fractures were classified as type B and 27 as type C. Among the capitellum fractures, there were 7 type B. Among the ulna/olecranon fractures, there were 49 type B and 5 type C, according to the AO classification. All isolated radial head fractures were type 2 (partial articular, displaced more than 2 mm), according to the Broberg and Morrey modification of Mason’s classification 2. Among fracture dis-locations of the elbow, there were 13 posterior olecranon, 6 anterior olecranon, and 10 dislocations with associated fracture of the radial head (2 with concomitant fracture of the coronoid-the so-called terrible triad).

BW_def.indd 92-93 07-02-11 22:25


Patients were injured in a fall from a standing height in 52 cases, in a fall from a greater height in 37 cases, in a bicycle accident in 16 cases, from a direct blow in 8 cases, and in a motor vehicle collision in 26 cases. A total of 129 patients had open reduction and plate and screw fixation and 10 had excision of a radial head fracture.

EvaluationInvestigators who were not involved in patients’ care evaluated each radiograph at a mean of 19 years (range, 10–34 y) after the injury. Two independent observers (both orthopedic surgeons) evaluated anteroposterior and lateral radiographs of the in-volved elbow once for radiographic arthrosis, according to the system of Broberg and Morrey: grade 0 indicates no radiographic arthrosis, grade 1 indicates slight joint-space narrowing with minimum osteophyte formation, grade 2 indicates moderate joint-space narrowing with moderate osteophyte formation, and grade 3 indicates severe degenerative change with gross destruction of the joint 1. The senior author reviewed all discrepancies in a blinded manner and made a final determination. We did not measure interobserver variability as part of this study.

Statistical AnalysisContinuous data are presented as the mean when they are normally distributed; oth-erwise, the median and interquartile range are reported. The number of patients in the severe group was too small (n = 9) and we thought that pooling moderate or se-vere and mild or no grades was a more reasonable approach for analysis of factors as-sociated with the development of radiographic arthrosis. We compared independent variables such as demographics and fracture characteristics one at a time between patients with moderate or severe and mild or no radiographic arthrosis by bivariate analysis to examine associations, including unpaired Student’s t-test for age, Mann-Whitney U test for follow-up time (owing to skewness), and Fisher’s exact test or Pearson chi-square for categorical data such as injury type, and injury mechanism. We entered variables into multivariable analysis using stepwise logistic regression (backward selection) to identify factors that were independently associated with moderate or severe arthrosis. We used the Wald test (distributed as chi-square) to as-sess significance of predictors and the Hosmer-Lemeshow test to evaluate regression model fit to the data 10. Odds ratios were calculated with the 95% confidence interval for significant predictors. Power analysis indicated that a minimum sample size of 80 patients with radiographic arthrosis provided 80% power (α = 0.05, β = 0.20) to identify significant predictors based on the odds ratio and for evaluating 8 variables using multivariable logistic regression 4. Two-tailed values of p<.05 were considered statistically significant.

ResultsAmong the 139 patients, 32 had moderate or severe radiographic arthrosis and 107 had mild or no radiographic arthrosis on final radiographs. Median follow-up was not significantly different among patients with moderate or severe radiographic arthrosis and with mild or no radiographic arthrosis (18 vs 19 y; p=.23). Duration of follow-up was not significantly different between injury types (p>.10). Bivariate analysis indicated an association between the presence of moderate or severe radio-graphic arthrosis and injury type (p<.001). Mechanism of injury, age, gender, follow-up time, occupation, and limb dominance were not associated with severe or moder-ate radiographic arthrosis (Table I).

Table I. Bivariate Analysis: Comparison Between Patients With Moderate or Severe Arthrosis Versus Mild or No Arthrosis

Variable

Moderate or Severe Arthrosis(N = 32)

Mild or No Arthrosis

(N = 107)P-Value

Age, (y) 40.2 ± 15.5 37.1 ± 14.1 0.27Gender 0.88 Male 19 (59) 62 (58) Female 13 (41) 45 (42)Dominance, (n [%]) Dominant 18 (56) 51 (48) 0.43 Nondominant 14 (44) 56 (52)Occupation, (n [%]) 0.32 Nonlaborer 23 (72) 87 (81) Laborer 9 (28) 20 (19)Injury Type, (n [%]) <0.001* Proximal ulna/olecranon 5 (16) 49 (45) Distal humerus fracture 13 (41) 16 (15) Elbow dislocation with fracture 10 (31) 19 (18) Capitellum/trochlea fracture 3 (9) 4 (4) Radial head fracture 1 (3) 19 (18)Mechanism, (n [%]) 0.25 Fall–standing height 11 (35) 41 (38) Fall–greater height 10 (31) 27 (25) Motor vehicle accident 9 (28) 15 (14) Bicycle accident 1 (3) 17 (16) Direct blow injury 1 (3) 7 (7)Follow-up, (y) Median 18 19 0.23 Interquartile range 14 – 22 15 – 24 Full range 10 – 31 10 – 34*Statisticallysignificant.

BW_def.indd 94-95 07-02-11 22:25


Figure 1 illustrates the number of patients with moderate or severe radio-graphic arthrosis and the injury type. Multiple logistic regression confirmed that in-jury type was the only factor independently associated with radiographic arthrosis (p<.001). The odds of moderate or severe radiographic arthrosis were greater for dis-tal humerus fractures, capitellum trochlear fractures, and fracture–dislocations than for isolated fractures of the radial head or olecranon (Table II).

DiscussionInjury type was the only significant independent predictor of moderate or severe radiographic arthrosis. Different injury types may lead to more severe degrees of articular surface injury and realignment. Distal humerus fractures and capitellum/trochlea fractures create the greatest articular injury 8 and were associated with the greatest risk of radiographic arthrosis. Radial head fractures and proximal ulna/olec-ranon fractures were associated with lower incidence of radiographic arthrosis. Un-fortunately, we could not accurately or reliably measure intra-articular displacement and comminution from records available decades after treatment of these injuries. These findings are all consistent with the previous literature on posttraumatic elbow radiographic arthrosis, which notes that intra-articular distal humerus fractures are a common source of posttraumatic elbow radiographic arthrosis 16. Consistent with prior data, none of these patients requested surgery specifi-cally to address radiographic arthrosis. Most patients develop radiographic signs of posttraumatic radiographic arthrosis after elbow trauma, but only a few present for treatment 9, 11. Radiographic arthrosis was not related to follow-up time, age, hand dom-inance, occupation, gender, or mechanism of injury. This suggests that postinjury activities and occupation are not important risk factors for the development or ad-vancement of radiographic arthrosis. This finding is reassuring and enabling, al-though counterintuitive. The limitations of this report include the fact that we have only cross-sec-tional (rather than longitudinal) data and that we did not address symptoms and dysfunction. We based assessment of arthrosis on radiographs alone, and we ac-knowledge the known limited relationships among radiographic evidence of arthro-sis and symptoms, impairment, and disability. Interobserver variability for the radio-graphic arthrosis rating was not measured. Unfortunately, available reproductions of the initial injury radiographs and initial postoperative radiographs were inadequate to quantify articular incongruity, intra-articular comminution, or fracture severity, and we do not have sufficient numbers to analyze the influence of subclassification of each injury type (eg, most of the injuries in each category were of a single AO type). For instance, nearly all of the columnar distal humerus fractures were type C according to the AO classification, and we did not feel confident about measures of articular incongruity. In addition, there were a large number of surgeons involved and fixation was often performed with older techniques that would be considered nonstandard at this time. Consequently, this study looks broadly at general types of injuries rather than at the influence of articular incongruity or fracture pattern and concludes that isolated fractures of the radial head and olecranon are less prone to moderate or se-

Table II. Multivariate Analysis: Type of Injury is Predictive of Moderate or Severe Arthrosis

Injury Type Odds Ratio 95% CI p Value

Proximal ulna/olecranon (reference category) —

Distal humerus fracture 8.0 2.5 – 25.8 <0.001*

Elbow dislocation with fracture 5.2 1.8 – 17.0 <0.01*

Capitellum/trochlea fracture 7.3 1.5 – 42.6 0.02*

Radial head fracture 0.5 0.1 – 4.7 0.56

*Statisticallyhigherriskofmoderateorseverearthrosisthanproximalulna/olecranonreferencegroup.Analysisisadjustedforage,gender,dominance,occupation,mechanismofinjury,treatment,andlengthoffollow-up(allp>0.10).

Figure 1: Histogram of percent moderate or severe radiographic arthrosis for each injury type.

BW_def.indd 96-97 07-02-11 22:25


vere radiographic arthrosis in the long term than fractures of the distal humerus and fracture–dislocations. This finding is expected; nevertheless, the data objectively document and quantify the differences using long-term evaluations that are hard to come by.

References

1. Broberg MA, Morrey BF. Results of delayed excision of the radial head after fracture. J Bone

Joint Surg Am. 1986;68(5):669-74.


Relat Res. 1987(216):109-19.

3. Brouwer KM, Guitton TG, Doornberg JN, Kloen P, Jupiter JB, Ring D. Fractures of the medial

column of the distal humerus in adults. J Hand Surg Am. 2009;34(3):439-45.

4. Demidenko E. Sample size determination for logistic regression revisited. Stat Med.

2007;26:3385-97.

5. Doornberg J, Ring D, Jupiter JB. Effective treatment of fracture-dislocations of the olecranon

requires a stable trochlear notch. Clin Orthop Relat Res. 2004(429):292-300.

6. Doornberg JN, Ring DC. Fracture of the anteromedial facet of the coronoid process. J Bone Joint

Surg Am. 2006;88(10):2216-24.

7. Doornberg JN, van Duijn PJ, Linzel D, Ring DC, Zurakowski D, Marti RK, et al. Surgical treatment

of intra-articular fractures of the distal part of the humerus. Functional outcome after twelve

to thirty years. J Bone Joint Surg Am. 2007;89(7):1524-32.

8. Guitton TG, Doornberg JN, Raaymakers EL, Ring D, Kloen P. Fractures of the capitellum and

trochlea. J Bone Joint Surg Am. 2009;91(2):390-7.




10. Hosmer D, Lemeshow S. Applied logistic regression. In: Wiley J, editor. 2nd ed. New York; 2000.

p. 34-69.

11. Konrad GG, Kundel K, Kreuz PC, Oberst M, Sudkamp NP. Monteggia fractures in adults: long-

term results and prognostic factors. J Bone Joint Surg Br. 2007;89(3):354-60.

12. Lindenhovius AL, Brouwer KM, Doornberg JN, Ring DC, Kloen P. Long-term outcome of opera-

tively treated fracture-dislocations of the olecranon. J Orthop Trauma. 2008;22(5):325-31.

13. Lindenhovius AL, Buijze GA, Kloen P, Ring DC. Correspondence between perceived disability and

objective physical impairment after elbow trauma. J Bone Joint Surg Am. 2008;90(10):2090-7.

14. Lindenhovius AL, Felsch Q, Ring D, Kloen P. The long-term outcome of open reduction and inter-

nal fixation of stable displaced isolated partial articular fractures of the radial head. J Trauma.

2009;67(1):143-6.

15. Müller ME, Nazarian S, Koch P, Chatzker J. The comprehensive classification of fractures of long

bones. Berlin: Springer-Verlag; 1990.

16. O’Driscoll SW. Elbow Arthritis: Treatment Options. J Am Acad Orthop Surg. 1993;1(2):106-16.

17. Regel G, Weinberg AM, Seekamp A, Blauth M, Tscherne H. [Complex trauma of the elbow].

Orthopade. 1997;26(12):1020-9.

18. Ring D. Elbow stiffness associated with malunion or nonunion. In: JB J, editor. The stiff elbow.

1st ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2006. p. 41-9.

BW_def.indd 98-99 07-02-11 22:25

100 part V | general discussion 101 discussion | chapter 8

CHAPTER 8Discussion

Thierry G. Guitton

Part V: General discussion

BW_def.indd 100-101 07-02-11 22:25


DiscussionAdvancements in technology, further insight into patient related factors, and the availability of long-term functional outcome data potentially allowed us to improve the classification, treatment and outcome of radial head fractures. We applied these recent advancements to radial head fracture treatment: 1) to gain further insight in radial head fractures and 2) to function as a model for general improvements in or-thopedic trauma surgery. This discussion will be continued with a more detailed address of each chapter separately and finally summarized in a conclusion:

chAPTer 2

This chapter was the successful validation of our Q3D-CT modeling technique. Addi-tionally, we derived linear regression models capable of estimating the volume and proximal articular surface area of the radial head prior to fracture. The limitations of this investigation included the fact that the accuracy of this Q3D-CT modeling tech-nique depended on the quality of the CT scan. Because CT scans do not account for articular cartilage, our measurements will differ from those based on MRI 1 or direct measurements of fresh cadaveric bone 12, 13, 30. We did not thoroughly evaluate inter- and intra-observer variability in creation of the models because our method was time and resource intensive and, based on experience with 2 observers doing several models during training; the method leaves limited room for bias. When one person created the same model five times, we found very little variation in the measures of volume and surface area. The differences between surface area calculated using our formula and that using simple geometry probably reflect the ovoid shape of the ra-dial head. In the final multivariable models height, weight and radial neck diameter did not significantly contribute to the fit of the model (as gauged by adjusted R2) and were therefore not used. The strong points of this investigation included the fact that we used: 1) a relatively large number of CT scans 1, 6, 12, 17, 18, 26, 30, 31; 2) a consistent algorithm for bone identification (on CT slides); and 3) automated curve and polygon mesh creation, which left limited room for judgment or bias on the part of the individual creating the model. The relatively small standard deviations of the measured volumes and surface areas, the relatively narrow 95% confidence intervals of the predictive linear models, and the fact that our multivariable models account for over 70% of the vari-ability of volume and surface area, all indicated that we could make reasonable and useful estimation of these parameters in fractured radial heads. To our knowledge, measurement of proximal articular surface area and ra-dial head volume has not been attempted. We produced equations capable of esti-mating the volume and proximal articular surface area of the intact radial head–on

the basis of parameters usually available in fractured radial heads–with an average relative percent difference of 0.5%. The ability to estimate the volume and surface area of the bone prior to fracture provides useful information when we analyze a fractured radial head. For instance, it allowed us to measure the percentage of the surface area involved in the fracture, which is one criterion in Broberg and Morrey’s modification of Mason’s classification. Keeping in mind the many shortcomings of our approach, we believe that it will, nonetheless improve our analysis and charac-terization of radial head fracture patterns. These Q3D-CT methods are, at least initially, more important for clinical re-search. We will be using this technique to study fracture fragment size and injury pattern, and the ability to estimate percentage involvement helps make the results more intuitive for clinicians. For instance, in radial head fracture classifications and management, decisions often refer to 30% of the surface area 3, but it’s not clear that this is an important cutoff, that we can make this measurement accurately from radiographs, or that it is representative of the fracture patterns that actually occur. More detailed analysis with these sophisticated techniques may help to clarify these issues. This Q3D-CT modeling technique can be applied to any intact or fractured bone in the human body and therefore has a huge potential in orthopedic trauma surgery, but additional work is needed to better define the accuracy and reliability of our method and determine how sensitive it is to the quality of the CT scan and the person doing the analysis.

chAPTer 3

In this chapter, quantitative analysis of CT scans provided measurements of the volume and articular surface area of radial head fracture fragments. The strengths of this investigation include the fact that we developed this technique with widely used software. A consistent algorithm was used for bone identification (on CT slides) and automated curve and polygon mesh creation, which left limited room for judg-ment or bias on the part of the individual creating the model. The limitations of this paper include the fact that we could not use the op-posite radial head for volume and surface area estimates, and the estimates of bone loss based on formulae are less precise. Additionally, we used both volume and ar-ticular surface area measurement. The volume measurements are straightforward whereas the articular surface area measurements may be less reliable. Finally, our definition of a small fragment was a mathematical and arbitrary categorization that, from our experience seems reasonable, but might not reflect the actual clinical situ-ation. We found that partial head (Mason 2) fractures are usually multi-fragment-ed (73%) and often have small fragments that are difficult to repair by volume and

BW_def.indd 102-103 07-02-11 22:25


surface area criteria (23 fractures [88%]) particularly when the fracture is displaced and unstable. We conclude that small fragments are most common with partial head fractures, at least among the unstable fractures associated with elbow dislocation or fracture of the ulna. Interestingly, small partial head fractures have not been ad-dressed much in the literature 5, 16, 23-25, and it may be commonly assumed that these are more straightforward to repair simply because a large part of the articular sur-face is not fractured. Another important finding is that, according to the surface area criterion, many of the Mason 2 fractures involved less than a third of the radial head (12 out of 26). This means that about half of these fractures would not satisfy the Broberg and Morrey criterion of greater than 30% of the articular surface area to be considered Type 2 fractures 3, 19. Quantitative analysis of 3D-CT scans is a useful technique for analyzing articular fracture pattern and morphology. Using this technology we identified: 1) that partial head (Mason 2) fractures frequently involve less than a third of the ra-dial head surface area; 2) that partial head fractures have more small and difficult to repair fragments than whole head fractures (Mason 3); and 3) that whole head fractures with more than 3 fragments are relatively uncommon (4 patients, 23% of Mason type 3 and 8.7% of total), but some 3 fragment-fractures have small fragments. These findings may influence our conception and classification of radial head frac-tures. For now, this technique is primarily designed for research purposes and not for patient care, but with further development Q3D-CT might prove useful in manage-ment decisions for individual patients.

chAPTer 4

This chapter investigated if classification and characterization of fractures of the ra-dial head is more accurate with 3D-CT images and 3D models than 2D-CT images and radiographs, using a prospective study design with intraoperative inspection as the reference standard. The limitations of this investigation include the fact that images were usu-ally rated after surgery (in part due to the inherent delay in receiving the physical 3D model), so that ratings of the radiological images were–in essence–retrospective; the injuries were relatively complex resulting in a spectrum bias in terms of all fractures of the radial head, although our work is representative of the types of fractures that would be studied with CT and operated on; two patients (one with addition of a capi-tellum/trochlea fracture and one anterior-transolecranon fracture dislocation) had non-nondisplaced fractures of the radial neck, which are relatively unusual–both fractures were seen only on operative exposure; and multiple physicians were in-volved in the ratings at two sites, which makes the results more generalizable, but

less consistent. These data should also be interpreted in light of the fact that the first assistant was usually a resident or fellow, so that the observer variability may largely reflect differences in training and experience. The strengths of this investigation in-clude the prospective design, the relatively large number of patients, and an intra-operative reference standard. This study found that increasing levels of sophistication in imaging/model-ing: 1) improved the sensitivity for diagnosis of numerous fracture characteristics us-ing the surgeon’s interpretation of the intraoperative findings as the reference stan-dard; and 2) decreased observer variation between surgeon and first assistant. This is in concordance with prior studies that have demonstrated improved agreement in characterization and classification of fractures with 3D-CT compared to 2D-CT and radiographs alone 2, 4, 7, 10, 11, 14, 15, 28, 29. Prior studies that addressed the classification of radial head fractures specifically found substantial observer variation 5, 20, 21, 27 when fractures were evaluated by radiographs only. However, these studies differed in that they are based upon retrospective data in small groups of observers/patients and the reference standard was based upon surgeon recollection and the medical record (e.g. operative notes). We interpret this combination of findings to indicate that fracture clas-sification and characterization based on 3D imaging and models is more accurate and reliable, essentially helping to narrow the experience and training gap. While experienced surgeons sometimes suggest that little is added by more sophisticated imaging, science is establishing that more sophisticated imaging does improve our understanding of the injury. However, recommendations regarding the use of a new technology should be based on both diagnostic performance characteristics and clinical impact. The next steps are to investigate whether more sophisticated ima-ging leads to more effective treatment as measured by fewer complications with less functional impairment.

chAPTer 5

This chapter investigated in a large web-based collaborative of experienced ortho-paedic surgeons if 3D-CT improve the interobserver reliability of the classification of radial head fractures according to the Broberg and Morrey modification of the Mason classification. The collaborative, web-based approach has facilitated large international studies of inter-rater variation 9, 15. Additionally, only fully trained surgeons, many with substantial clinical experience participated. Inclusion of surgeons from mul-tiple countries and continents should increase the generalizability of the results. Us-ing high-speed Internet connections and improved compression techniques, we were able to provide sophisticated reproduction images and movies via the Internet.

BW_def.indd 104-105 07-02-11 22:25


There are some weaknesses in this study. First, the quality of the radiographs was limited to what had been obtained at the time of injury, which reflects usual prac-tice, but not what might be achieved with specific protocols. In addition, we provided limited information about the patient and the injury. There was also a spectrum bias by selecting cases to represent the known variety of injuries, with the result that less common complex fractures were over represented compared to the more common minimally or slightly displaced fractures. Our study reflects what would be expected with relatively complex fractures of the radial head–the reliability would be expect-ed to be higher if we included more of the non-displaced or minimally displaced frac-tures that make up the majority of radial head fractures. Another shortcoming is the fact that a small number of observers either uncommonly or never treat radial head fractures, but we did not plan for exclusions on this basis and therefore did not do so after the fact to avoid introducing bias. Finally, this is an artificial research situation given that in clinical practice clinicians would have both the 2D and 3D reconstruc-tions available to them. Three-dimensional CT images led to small but significant decreases in vari-ation between observers for fracture classification and some fracture characteristics compared to 2D-CT, but a notable amount of variation remains even with more so-phisticated imaging. Our belief that 3D-CT images are easier for surgeons to interpret is supported by the observation that 3D-CT produced a higher agreement for Broberg and Morrey’s modification of the Mason classification than previously reported in the literature 20, 21 and 3D-CT was associated with less disagreement in classification than 2D-CT across various cultures, training, subspecialty and levels of experience. Nonetheless agreement was only fair or moderate at best even with 3D-CT. Further-more, some might interpret this data as showing much less influence on interob-server variation than one might guess. Other potential sources of interobserver variation include unfamiliar or un-clear definitions, and differences in culture, training, and exposure. In our opinion, the fact that well-trained, experienced observers disagree indicates that there are variations in these factors that lead different experts to see different things in so-phisticated images. In other words, reducing interobserver variation seems to depend on something more than better imaging. Additional research to identify and reduce sources of observer variation in the interpretation of diagnostic images is merited.

chAPTer 6

This chapter investigated the psychosocial aspects of radial head fractures. More spe-cifically, if agreement with the idea that “stretching of the elbow beyond the point were it becomes painful is important in recovery” leads to greater elbow range of motion one month after injury.

The strengths of this study included the prospective design and enrollment of patients from 6 different surgeons’ practices. Limitations include: 1) Enrollment after diagnosis, reassurance, and coaching with motion exercises, all of which were therapeutic interventions that may have affected attitudes towards pain (and may explain why the majority of patients agreed with the importance of pain for recov-ery); 2) Ceiling effects since most patients with minimally displaced radial head fractures regained near normal motion regardless of their paradigm; 3) Meaning-less variation (or “noise”) added to the data by virtue of the fact that the measure-ment error of a hand-held goniometer was comparable to the small differences in elbow flexion contractures observed; 4) The measure of agreement with a statement regarding the role of pain in recovery was an indirect measure of confidence with exercises—a more direct and objective measure would be preferable; patients may state agreement on a questionnaire, but still have a hard time performing stretches beyond pain; 5) Wide range in the time that people returned for the “one-month” follow-up (although this did not correlate with any of the outcome measures); and 6) Limited power due to unequal distribution of patients among agreement groups. In spite of these shortcomings, we did find that a patient’s paradigm with respect to the role of pain in recovery predicted motion one month after injury and that a patient’s paradigm had small but significant correlation with pain catastro-phizing. The lack of correlation between attitude and disability might be due to the small number of patients in the neutral (9) and disagree (6) categories versus the agree category (55). There was a nearly 11 point difference in the mean DASH score in patients that agree (17.4) or were neutral (17.3) about the role of pain and those that disagree (28.0), which seems clinically important. Of note is that the mean DASH for patients that disagree was higher than what had been reported in patients with frac-tures, while the mean DASH for the other categories was lower (21). This line of research should be pursued. If additional studies corroborate the role of automatic thoughts and beliefs (intuition, “gut feelings”) in recovery from injury, as well as the correlation of these thoughts with depressive symptoms and maladaptive responses to nociception, then there is room for improvement in our teaching and coaching of post-injury exercises. As Paul Brand noted in his book “The Gift of Pain”, nociception exists for our protection. It is no surprise that pain after in-jury may make us feel vulnerable and protective. The key may be to help our patients change their mindset from vulnerability to recovery, seeing a painful exercise more as a useful stretch exercise and the post-exercise pain more as that rewarding ache after a great work out.

BW_def.indd 106-107 07-02-11 22:25


chAPTer 7

This chapter assessed the risk factors for posttraumatic elbow arthrosis on radio-graphs after elbow injury in the long term. The limitations of this investigation included the fact that we have only cross-sectional (rather than longitudinal) data and that we did not address symptoms and dysfunction. We based assessment of ar-throsis on radiographs alone, and we acknowledge the known limited relationships among radiographic evidence of arthrosis and symptoms, impairment, and disabil-ity. Interobserver variability for the radiographic arthrosis rating was not measured. Unfortunately, available reproductions of the initial injury radiographs and initial postoperative radiographs were inadequate to quantify articular incongruity, intra-articular comminution, or fracture severity, and we did not have sufficient numbers to analyze the influence of subclassification of each injury type (eg, most of the in-juries in each category were of a single AO type). For instance, nearly all of the co-lumnar distal humerus fractures were type C according to the AO classification, and we did not feel confident about measures of articular incongruity. In addition, there were a large number of surgeons involved and fixation was often performed with older techniques that would be considered nonstandard at this time. Injury type was the only significant independent predictor of moderate or severe radiographic arthrosis. Different injury types may lead to more severe degrees of articular surface injury and realignment. Distal humerus fractures and capitellum/trochlea fractures created the greatest articular injury 8 and were associ-ated with the greatest risk of radiographic arthrosis. Radial head fractures and prox-imal ulna/olecranon fractures were associated with lower incidence of radiographic arthrosis. Unfortunately, we could not accurately or reliably measure intra-articular displacement and comminution from records available decades after treatment of these injuries. These findings are all consistent with the previous literature on post-traumatic elbow radiographic arthrosis, which notes that intra-articular distal hu-merus fractures are a common source of posttraumatic elbow radiographic arthro-sis 22. Radiographic arthrosis was not related to follow-up time, age, hand domi-nance, occupation, gender, or mechanism of injury. This suggests that post-injury activities and occupation are not important risk factors for the development or ad-vancement of radiographic arthrosis. This finding is reassuring and enabling, al-though counterintuitive. Consequently, this study looked broadly at general types of injuries rather than at the influence of articular incongruity or fracture pattern and concludes that isolated fractures of the radial head and olecranon are less prone to moderate or se-vere radiographic arthrosis in the long term than fractures of the distal humerus and fracture–dislocations. This finding is expected; nevertheless, the data objectively

document and quantify the differences using long-term evaluations that are hard to come by.

ConclusionThroughout the various chapters in this thesis, our results showed that advance-ments in technological, imaging, psychosocial and long-term outcome can help im-prove classification, treatment and outcome after radial head fracture. First, we vali-dated our new Q3D-CT modeling technique and successfully applied this technique to radial head fractures. Measurement of proximal articular surface area and radial head volume has not been attempted before and it will improve our analysis and characterization of radial head fracture patterns. We demonstrated prospectively and in a multi-rater study that 3D images led to increased agreement in classification of radial head fractures. We also demonstrated that patients who agree that pain is a necessary part of recovery have improved outcomes after radial head fractures. Additionally, we identified predictors for arthrosis after elbow trauma in the long term. In general, these advancements can help gain more insight in the classifica-tion, treatment and outcome of fractures in orthopedic trauma surgery. As shown in this thesis, advancements in technical analysis, imaging mo-dalities, increased interest in psychosocial aspects of treatment and the availability of long-term outcome data can help improve classification, treatment and outcome in fractures of the radial head. Implementation and application of new technologies as they emerge in medicine are needed to allow further improvement of our cur-rent treatments. Dogmas in orthopedics on the psychosocial aspects of treatment should be set aside. With all these new technologies and advancements widely avail-able, it is our duty to carefully evaluate them and–if proven beneficial–to use them in the treatments of our patients. It is science that created these advancements and through adequate scientific evaluation of these advancements we can continue cre-ating more effective treatments for our patients.

BW_def.indd 108-109 07-02-11 22:25


References







Relat Res. 1987(216):109-19.







J Hand Surg Am. 2006;31(1):53-7.



8. Guitton TG, Doornberg JN, Raaymakers EL, Ring D, Kloen P. Fractures of the capitellum and

trochlea. J Bone Joint Surg Am. 2009;91(2):390-7.

9. Karanicolas PJ, Bhandari M, Kreder H, Moroni A, Richardson M, Walter SD, et al. Evaluating

agreement: conducting a reliability study. J Bone Joint Surg Am. 2009;91 Suppl 3:99-106.














phy. J Hand Surg Am. 2009;34(9):1640-6.

16. Lindenhovius AL, Felsch Q, Doornberg JN, Ring D, Kloen P. Open reduction and internal fixation

compared with excision for unstable displaced fractures of the radial head. J Hand Surg Am.

2007;32(5):630-6.












22. O’Driscoll SW. Elbow Arthritis: Treatment Options. J Am Acad Orthop Surg. 1993;1(2):106-16.

23. Ring D, Quintero J, Jupiter JB. Open reduction and internal fixation of fractures of the radial

head. J Bone Joint Surg Am. 2002;84-A(10):1811-5.

24. Ring D. Open reduction and internal fixation of fractures of the radial head. Hand Clin.

2004;20(4):415-27, vi.

25. Ring D. Displaced, unstable fractures of the radial head: fixation vs. replacement--what is the

evidence? Injury. 2008;39(12):1329-37.



27. Sheps DM, Kiefer KR, Boorman RS, Donaghy J, Lalani A, Walker R, et al. The interobserver reli-

ability of classification systems for radial head fractures: the Hotchkiss modification of the

Mason classification and the AO classification systems. Can J Surg. 2009;52(4):277-82.






31. van Riet RP, van Glabbeek F, Neale PG, Bimmel R, Bortier H, Morrey BF, et al. Anatomical consid-


BW_def.indd 110-111 07-02-11 22:25

112 part Vi | summary 113 summary / samenvatting | chapter 8

SummaryPart VI: Summary

BW_def.indd 112-113 07-02-11 22:25


inTroDucTion

The advancements in analyzing techniques, 3D imaging and modeling, increased interest in psychosocial aspects of treatment and the recent availability of multiple long-term outcome studies gives us the opportunity to further investigate the clas-sification, treatment and outcome of radial head fractures. This all could lead to im-proved treatment and possible better outcomes for patients. The aim of this thesis was to apply these advancements to the radial head in order to: 1) gain further in-sight in classification, treatment and outcome of radial head fractures and 2) func-tion as a model for general improvements in orthopedic trauma surgery.

chAPTer 2

QuAnTiTATive meAsuremenTs of The volume AnD surfAce AreA of The rADiAl

heAD

This chapter investigated if quantitative Q3D-CT modeling technique based on ana-tomical and demographic data that can measure size, shape, and proximal articular surface area could be used to develop formulas that could predict the volume and proximal surface area of the intact radial head in patients with fractures of the radial head. To our knowledge, measurement of proximal articular surface area and radial head volume has not been attempted. We produced equations capable of estimating the volume and proximal articular surface area of the intact radial head–on the basis of parameters usually available in fractured radial heads–with an average relative percent difference of 0.5%. The ability to estimate the volume and surface area of the bone prior to fracture provides useful information when we analyze a fractured radial head. For instance, it allows us to measure the percentage of the surface area involved in the fracture, which is one criterion in Broberg and Morrey’s modification of Mason’s classification.We will be using this technique to study fracture fragment size and injury pattern. chAPTer 3

QuAnTiTATive Three-DimensionAl comPuTeD TomogrAPhy meAsuremenT of

rADiAl heAD frAcTures

This chapter analyzed radial head fracture fragment morphology on Q3D-CT imag-es in terms of size, shape, and articular surface area. Quantitative analysis of 3D-CT scans proved to be a useful technique for analyzing articular fracture pattern and morphology. Using this technology we identified that partial head (Mason 2) frac-tures frequently involve less than a third of the radial head surface area; that partial head fractures have more small and difficult to repair fragments than whole head fractures (Mason 3); and that whole head fractures with more than 3 fragments are relatively uncommon, but some 3 fragment-fractures have small fragments. These findings may influence our conception and classification of radial head fractures.

chAPTer 4

DiAgnosTic AccurAcy of Two-DimensionAl AnD Three-DimensionAl imAging

AnD moDeling of rADiAl heAD frAcTures

This chapter investigated if classification and characterization of fractures of the ra-dial head is more accurate with 3D-CT and 3D models than 2D-CT and radiographs, using a prospective study design with intraoperative inspection as the reference standard. We found that increasing levels of sophistication in imaging/modeling: 1) improved the sensitivity for diagnosis of numerous fracture characteristics; and 2) decreased observer variation between surgeon and first assistant. We found that fracture classification and characterization based on three-dimensional imaging and models is more accurate and reliable, essentially helping to narrow the experi-ence and training gap.

chAPTer 5

inTerobserver reliAbiliTy of rADiAl heAD frAcTure clAssificATion:

Two-DimensionAl vs. Three-DimensionAl comPuTeD TomogrAPhy

This chapter investigated in a large web-based collaborative of experienced ortho-paedic surgeons if 3D reconstructions of CT scans improved the interobserver reliabil-ity of the classification of radial head fractures according to the Broberg and Morrey modification of the Mason classification. Three-dimensional CT images led to small but significant decreases in variation between observers for fracture classification and some fracture characteristics compared to 2D-CT, but a notable amount of varia-tion remained even with more sophisticated imaging. We believe that 3D-CT images are easier for surgeons to interpret. Nonetheless agreement was only fair or moder-ate at best even with 3D-CT. Furthermore, some might interpret this data as show-ing much less influence on interobserver variation than one might guess. Reducing interobserver variation seems to depend on something more than better imaging.

chAPTer 6

ATTiTuDe TowArDs sTreTch PAin of The elbow AfTer rADiAl heAD

This chapter investigated if agreement with the idea that “stretching of the elbow beyond the point were it becomes painful is important in recovery” leads to greater elbow range of motion one month after injury. We found that a negative attitude towards stretch pain during recovery from fracture of the radial head is associated with less elbow motion one month after injury.

BW_def.indd 114-115 07-02-11 22:25


chAPTer 7

inciDence AnD risk fAcTors for The DeveloPmenT of

rADiogrAPhic ArThrosis AfTer TrAumATic elbow injuries

This chapter assessed the risk factors for posttraumatic elbow arthrosis on radio-graphs after elbow injury in the long-term. We found that injury type was the only significant independent predictor of moderate or severe radiographic arthrosis. Ra-diographic arthrosis was not related to follow-up time, age, hand dominance, occu-pation, gender, or mechanism of injury. This suggests that post-injury activities and occupation are not important risk factors for the development or advancement of radiographic arthrosis. Fractures of the radial head and olecranon are less prone to moderate or severe radiographic arthrosis in the long term than fractures of the dis-tal humerus (both columnar and capitellum/trochlea) and elbow fracture–disloca-tions.

ConclusionsThis thesis shows that advancements in technical analysis, imaging modalities, in-creased interest in psychosocial aspects of treatment and the availability of long-term outcome data can help improve classification, treatment and outcome in frac-tures of the radial head. It is science that created these advancements and through adequate scientific evaluation of these advancements we can continue creating more effective treatments for patients.

BW_def.indd 116-117 07-02-11 22:25

118 part Vi | summary 119samenvatting | chapter 8

Samenvatting

BW_def.indd 118-119 07-02-11 22:25


IntroductieDe nieuwe ontwikkelingen in analyse technieken, driedimensionale (3D) beeldvor-ming en modellering, toegenomen belangstelling voor psychosociale aspecten van de behandeling en de recente beschikbaarheid van meerdere lange termijn studies biedt ons de gelegenheid om verder onderzoek te doen naar de classificatie, de be-handeling en uitkomst van radiuskop fracturen. Dit alles kan leiden tot een betere behandeling en eventueel betere resultaten voor patiënten. Het doel van dit proef-schrift was om deze nieuwe ontwikkelingen toe te passen op de radiuskop om: 1) meer inzicht te verkrijgen in de classificatie, de behandeling en de uitkomst van ra-diuskop fracturen en 2) te functioneren als een model voor algemene verbeteringen in de orthopedische traumatologie.

hoofDsTuk 2

kwAnTiTATieve meTingen vAn De inhouD en De oPPervlAkTe vAn De

rADius koP

In dit hoofdstuk werd onderzocht of de kwantitatieve 3D computertomografie (Q3D-CT) techniek, die grootte, vorm en proximale articulaire oppervlakte meet, kan wor-den toegepast op de intacte radiuskop om zo formules, gebaseerd op anatomische en demografische gegevens, te creëren die het volume en de proximale oppervlakte kunnen voorspellen bij patiënten met fracturen van de radiuskop. Voor zover onze kennis reikt, is de meting van proximale articulaire oppervlakte en radiuskop volume niet eerder uitgevoerd. We produceerde vergelijkingen die kunnen berekenen wat de omvang van het proximale articulaire oppervlakte en het volume is van een intacte radiuskop, op basis van parameters die meestal beschikbaar zijn in de gebroken radi-us kop met een gemiddeld relatief procentueel verschil van 0,5%. De mogelijkheid om het volume en het proximale articulaire oppervlak van het bot in te schatten vooraf-gaand aan de breuk, is waardevolle informatie wanneer we een gebroken radiuskop analyseren. Bijvoorbeeld, het stelt ons in staat om exact te meten welk percentage van de oppervlakte betrokken is bij de breuk. Dit is een criterium in de Broberg en Morrey’s classificatie volgens Mason. We zullen deze techniek gebruiken om fractuur fragment grootte en fractuur patronen te bestuderen. hoofDsTuk 3

kwAnTiTATieve DrieDimensionAle comPuTer TomogrAfie AnAlyse vAn

rADiuskoP frAcTuren

In dit hoofdstuk werden radiuskop fracturen geanalyseerd op fragment morfologie met driedimensionale CT-beelden (Q3DCT) in termen van grootte, vorm en articu-laire oppervlakte. Kwantitatieve analyse van 3D-CT scans bleek een nuttige techniek voor het analyseren van patronen en morfologie van articulaire fracturen. Met be-

hulp van deze technologie hebben we vastgesteld dat: 1) partiële radiuskop fracturen (Mason type 2) frequent breken met minder dan een derde van de radiuskop opper-vlakte; 2) partiële radiuskop fracteren meer kleine en moeilijk te herstellen fragmen-ten hebben dan hele radiuskop fracturen (Mason 3); en 3) hele radiuskop fracturen met meer dan 3 fragmenten relatief zeldzaam zijn, en dat enkele 3-fragment-radius-kop fracturen kleine fragmenten hebben. Deze bevindingen kunnen onze opvatting en indeling van radiuskop fracturen beïnvloeden.

hoofDsTuk 4

DiAgnosTische nAuwkeurigheiD vAn TweeDimensionAle en DrieDimensio-

nAle beelDvorming en moDellen vAn rADiuskoP frAcTuren

In dit hoofdstuk werd onderzocht of de classificatie en karakterisering van fractu-ren van de radiuskop nauwkeuriger is met 3D-CT en 3D modellen dan met 2D-CT en röntgenfoto’s, met behulp van een prospectieve studie opzet met intraoperatieve in-spectie als de referentiestandaard. We vonden dat toename van de geavanceerdheid in de beeldvorming/modellen leidde tot: 1) verbeterde sensitiviteit voor de diagnose van een groot aantal fractuur kenmerken, en 2) gedaalde waarnemer variatie tussen de chirurg en de eerste assistent. We vonden dat fractuur classificatie en karakterise-ring op basis van drie-dimensionale beeldvorming en modellen nauwkeuriger en be-trouwbaarder is, en zodoende helpen om de ervaring- en opleidings-kloof te dichten.

hoofDsTuk 5

inTer-beoorDelAArs beTrouwbAArheiD vAn rADiuskoP frAcTuur clAssificA-

Tie: TweeDimensionAAl vs. DrieDimensionAAl comPuTer TomogrAfie

In dit hoofdstuk werd in een groot online samenwerkingsverband van ervaren or-thopedisch chirurgen onderzocht of 3D reconstructies van CT-scans verbeterde inter-beoordelaars betrouwbaarheid van de classificatie van radiuskop fracturen gaven met de Broberg en Morrey classificatie volgens Mason. Driedimensionale CT beelden leidde tot een kleine, maar significante afname van de variatie tussen waarnemers voor fractuur indeling en een aantal fractuur kenmerken in vergelijking met 2D-CT, maar een opmerkelijke hoeveelheid variatie bleef aanwezig, zelfs met meer geavan-ceerde beeldvorming. Wij geloven dat 3D-CT beelden gemakkelijker voor chirurgen te interpreteren zijn. Toch was de overeenstemming redelijk of op zijn best schap-pelijk, zelfs met 3D-CT. Bovendien zullen sommigen deze gegevens interpreteren als hetgeen waaruit blijkt dat 3D-CT veel minder invloed op de interobserver variatie heeft dan men zou verwachten. Het verminderen van interobserver variatie lijkt dus op meer te berusten dan alleen betere beeldvorming.

BW_def.indd 120-121 07-02-11 22:25


hoofDsTuk 6

houDing Ten oPzichTe vAn rek Pijn in De elleboog nA rADiuskoP frAcTuur

Dit hoofdstuk onderzocht of instemming met het idee dat “rekken van de elleboog voorbij het punt dat het pijnlijk wordt belangrijk is in het herstel” leidt tot een gro-ter bereik van de elleboog beweging een maand na het letsel. Wij vonden dat een negatieve houding ten opzichte van rek pijn tijdens het herstel van een fractuur van de radiuskop was geassocieerd met minder elleboog beweging een maand na letsel.

hooDsTuk 7

inciDenTie en risicofAcToren voor De onTwikkeling vAn rADiogrAfische

ArTrose nA TrAumATische leTsels vAn De elleboog

In dit hoofdstuk werden de risicofactoren voor posttraumatische elleboog arthrose op röntgenfoto’s na het doormaken van een elleboog fractuur in de lange termijn on-derzocht. We vonden dat het type fractuur de enige significante onafhankelijke voor-speller van een matige of ernstige radiografische artrose is. Radiografische artrose was niet gerelateerd aan de tijd sinds letsel, leeftijd, dominantie, beroep, geslacht of mechanisme van de verwonding. Dit suggereert dat post-traumatische activiteiten en beroep geen belangrijke risicofactoren voor de ontwikkeling of voortgang van ra-diografische artrose zijn. Fracturen van de radiuskop en olecranon zijn minder gevoe-lig voor matige of ernstige radiografische artrose op de lange termijn dan fracturen van de distale humerus (zowel collum en capitellum/trochlea) en elleboog fractuur-dislocaties.

ConclusieDit proefschrift toont aan dat nieuwe ontwikkelingen in technische analyse, beeld-vorming, interesse in de psychosociale aspecten van de behandeling en de be-schikbaarheid van lange termijn studies, kunnen helpen bij het verbeteren van de classificatie, de behandeling en uitkomst van fracturen van de radiuskop. Het is de wetenschap die deze nieuwe ontwikkelingen heeft gecreëerd en door middel van adequate wetenschappelijke evaluatie van deze nieuwe ontwikkelingen kunnen we verder met het vervaardigen van meer effectieve behandelingen voor patiënten.

BW_def.indd 122-123 07-02-11 22:25

124 part Vi | summary 125glossery | chapter 8

Glossery

BW_def.indd 124-125 07-02-11 22:25

126 127 glossery | chapter 8part Vi | summary

Glossery2D Two-dimensional

3D Three-dimensional

AO Arbeitsgemeinschaft für Osteosynthesefragen

AP Anteroposterior

B&M Broberg and Morrey Score

CAD Computer-Aided Design

CES-D Center for Epidemiologic Studies Depression Scale

CI Confidence interval

CMM Coordinator measuring machine

CT Computed Tomography

CV Coefficient of variation

DASH Disabilities of the Arm, Shoulder, and Hand questionnaire

DICOM Digital Imaging and Communications in Medicine

GEE Generalized estimating equations

IRB Institutional Review Board

Matlab Numerical computing environment

MRI Magnetic Resonance Imaging

MVA Multi-vehicle accident

MVC Motor vehicle collision

NURBS Non-Uniform Rational B-Spline

ORIF Open reduction and internal fixation

PCS Pain Catastrophizing Scale

POFD Posterior-olecranon-fracture dislocation

Q3D-CT Quantitative Three-dimensional CT

Rhinoceros Three-dimensional modeling tool based on NURBS

ROM Range of Motion

SD Standard Deviation

Vitrea Visualization solution that creates Three-dimensional reconstructions

X-ray Röntgen radiation

The Broberg & Morrey Modification of the Mason Classification

The Broberg-Morrey modification of the Mason classification of radial head and neck fractures

Type I = fx of the radial head or neck displaced <2 mm

Type II = fx of the radial head or neck displaced >=2 mm and involving >=30% of the joint surface

Type III = comminuted fx of the radial head or neck

Type IV = elbow dislocation with any of the above fx types

Type I

Type III

Type II

Type IV

BW_def.indd 126-127 07-02-11 22:25

128 appendices 129bibliographyappendices

Appendices

Bibliography

BW_def.indd 128-129 07-02-11 22:25


Original Articles

1. Guitton TG, Ring D. Three-dimensional computed tomographic imaging and modeling in the upper extremity. Hand Clin; 2010. p. 447-53, viii.

2. Bachoura A, Guitton TG, Vrahas M, Smith M, Zurakowski D, Ring D. Surgical Site Infections in Orthopaedic Trauma. In press CORR. 2010.

3. Brouwer KM, Guitton TG, Doornberg JN, Kloen P, Jupiter JB, Ring D. Fractures of the medial column of the distal humerus in adults. J Hand Surg Am. 2009;34(3):439-45.

4. Guitton TG, Albers RG, Ring D. Anterior olecranon fracture-dislocations of the el-bow in children. A report of four cases. J Bone Joint Surg Am. 2009;91(6):1487-90.

5. Guitton TG, Doornberg JN, Raaymakers EL, Ring D, Kloen P. Fractures of the capi-tellum and trochlea. J Bone Joint Surg Am. 2009;91(2):390-7.

6. Guitton TG, Duckworth AB, McQueen MM, Kloen P, Ring D. Elbow subluxation and articular fracture of the distal humerus. Shoulder & Elbow. 2009;2(2):111 - 7.

7. Guitton TG, Jupiter JB. 90-90 versus parallel plating of the distal humerus; What is the evidence? AO Dialogue. 2009(2):28-32.

8. Guitton TG, Ring D. Elbow arthritis. Current Orthopaedic Practice. 2008;19(6):626-32.

9. Guitton TG, Ring D. Nonsurgically treated terrible triad injuries of the elbow: re-port of four cases. J Hand Surg Am. 2010;35(3):464-7.

10. Guitton TG, Ring D, Kloen P. Long-Term Evaluation of Surgically Treated An-terior Monteggia Fractures in Skeletally Mature Patients. J Hand Surg Am. 2009;34(9):1618-24.

11. Guitton TG, van der Werf HJ, Ring D. Quantitative measurements of the volume and surface area of the radial head. J Hand Surg Am. 2010;35(3):457-63.

12. Guitton TG, van der Werf HJ, Ring D. Quantitative three-dimensional comput-ed tomography measurement of radial head fractures. J Shoulder Elbow Surg. 2010;19(7):973-7.

13. Guitton TG, Van Der Werf HJ, Ring D. Quantitative measurements of the coronoid in healthy adult patients. J Hand Surg Am. 2011;36(2):232-7.

14. Guitton TG, Van Dijk NC, Raaymakers EL, Ring D. Isolated Diaphyseal Fractures of the Radius in Skeletally Immature Patients. Hand (N Y). 2009.

15. Guitton TG, van Leerdam RH, Ring D. Necessity of routine pathological examina-tion after surgical excision of wrist ganglions. J Hand Surg Am. 2010;35(6):905-8.

16. Guitton TG, Zurakowski D, van Dijk NC, Ring D. Incidence and risk factors for the development of radiographic arthrosis after traumatic elbow injuries. J Hand Surg Am. 2010;35(12):1976-80.

17. Lu HT, Guitton TG, Capo JT, Ring D. Elbow instability associated with bico-

lumnar fracture of the distal humerus: report of three cases. J Hand Surg Am. 2010;35(7):1126-9.

18. Rineer CA, Guitton TG, Ring D. Radial head fractures: loss of cortical contact is associated with concomitant fracture or dislocation. J Shoulder Elbow Surg. 2009;19(1):21-5.

19. Soong M, van Leerdam R, Guitton TG, Got C, Katarincic J, Ring D. Fracture of the distal radius: risk factors for complications after locked volar plate fixation. J Hand Surg Am. 2011;36(1):3-9.

20. van der Werf HJ, Guitton TG, Ring D. Non-operatively Treated Fractures of the An-teromedial Facet of the Coronoid. Shoulder & Elbow. 2010;2(1):40 - 2.

21. Wiggers J, Guitton TG, Vrahas M, Smith M, Ring D. The Complexity of Transfers Versus Direct Admissions for Hip Fracture at a Regional Referral Center. In press JOT. 2010.

BW_def.indd 130-131 07-02-11 22:25


Conference Proceedings

1. Bachoura A, Guitton TG, Zurakowski D, Vrahas M, Smith M, Ring D. Surgical Site Infections in Orthopaedic Trauma. Annual Meeting American Academy of Or-thopaedic Surgeons; 2010 March 9-13; New Orleans, LA. 2010.

2. Brouwer KM, Guitton TG, Doornberg JN, Jupiter JB, Kloen P, Ring D. Fractures of the Medial Column of the Distal Humerus in Adults. New England Hand Society Annual Meeting; 2007; Sturbridge, MA, USA. 2007.

3. Brouwer KM, Guitton TG, Doornberg JN, Jupiter JB, Ring D, Kloen P. Fractures of the Medial Column of the Distal Humerus in Adults. 54th Nordic Orthopaedic Federation Congress; 2008; Amsterdam, the Netherlands. 2008.

4. Buijze GA, Guitton TG, Dijk van CN, Ring D, SVG;. The Effect of Training on Inter-Observer Reliability for the Diagnosis of Scaphoid Fracture Displacement. 65th Annual Meeting American Society for Surgery of the Hand; 2010 October 7-9; Bos-ton, Massachusetts. 2010.

5. Guitton TG, Bachoura A, Zurakowski D, Vrahas M, Smith M, Ring D. Surgical Site Infections in Orthopaedic Trauma. 11th European Congress for Trauma & Emer-gency Surgery; 2010 March 9-13; Brussels/Belgium. 2010.

6. Guitton TG, Bachoura A, Zurakowski D, Vrahas M, Smith M, Ring D. Surgical Site Infections in Orthopaedic Trauma. 55th Nordic Orthopaedic Federation Congress; 2010 5 – 7 May; Aarhus, Denmark. 2010.

7. Guitton TG, Bachoura A, Zurakowski D, Vrahas M, Smith M, Ring D. Surgical Site Infections in Orthopaedic Trauma. Deutschen Kongresses für Orthopädie und Unfallchirurgie; 2010 26 - 29 Oct; Messe/ICC, Berlin. 2010.

8. Guitton TG, Brouwer K, Dyer G, Zurakowski D, Mudgal C, Ring D. Comparison of the Accuracy of 2-D and 3-D Imaging and Modeling of Radial Head Fractures. An-nual Meeting American Academy of Orthopaedic Surgeons; 2011; San Diego. 2011.

9. Guitton TG, Dijk van CN, Raaymakers EL, Ring D. Isolated Diaphyseal Fractures of the Radius in Skeletally Immature Patients. Massachusetts General Hospital Clinical Research Day; 2008; Boston, MA. 2008.

10. Guitton TG, Doornberg JN, Raaymakers EL, Kloen P, Ring D. Capitellum and Troch-lear fractures of the distal humerus in adults. Massachusetts General Hospital Clinical Research Day; 2008; Boston, MA. 2008.

11. Guitton TG, Doornberg JN, Raaymakers EL, Ring D, Kloen P. Fractures of the capi-tellum and trochlea. 54th Nordic Orthopaedic Federation Congress; 2008; Am-sterdam, the Netherlands. 2008.

12. Guitton TG, Doornberg JN, Raaymakers EL, Ring D, Kloen P. Fractures of the cap-itellum and trochlea. 19th Annual Richard J Smith Lecture; 2008 May; Boston. 2008.

13. Guitton TG, Duijvenbode D, Raaymakers EL, Ring D, Kloen P. Long Term Evaluation of Isolated Radius Radius Fractures in Adults. Vereniging Orthopaedisch Chirur-gische Assistenten Annual Meeting; 2009; Leiden, the Netherlands. 2009.

14. Guitton TG, Lu H, Capo JT, Ring D. Instability after Open Reduction and Internal Fixation of the Distal Humerus. 11th European Congress of Trauma & Emergency Surgery; 2010; Brussels/Belgium. 2010.

15. Guitton TG, Rineer CA, Ring D. Radial head fractures: loss of cortical contact is as-sociated with concomitant fracture or dislocation. Combined Annual Meeting of the ASSH and ASHT; 2009; San Francisco, USA. 2009.

16. Guitton TG, Rineer CA, Ring D. Radial head fractures: loss of cortical contact is as-sociated with concomitant fracture or dislocation. 20th Annual Richard J Smith Lecture; 2009 May; Boston. 2009.

17. Guitton TG, Rineer CA, Ring D. Radial head fractures: loss of cortical contact is associated with concomitant fracture or dislocation. 7th Biennial AAOS/ASES Shoulder and Elbow; 2010; Aventura (North Miami Beach), FL. 2010.

18. Guitton TG, Rineer CA, Ring D. Radial head fractures: loss of cortical contact is associated with concomitant fracture or dislocation. Open Meeting of the Ameri-can Shoulder and Elbow Surgeons; 2010 Saturday, March 13; New Orleans, LA. 2010.

19. Guitton TG, Ring D. Quantitative Measurements of the Radial Head in Healthy Adult Patients Massachusetts General Hospital Clinical Research Day; 2009; Bos-ton, MA. 2009.

20. Guitton TG, Ring D. Nonsurgically treated terrible triad injuries of the elbow: re-port of four cases. 55th Nordic Orthopaedic Federation Congress; 2010 5 – 7 May; Aarhus, Denmark. 2010.

21. Guitton TG, Ring D, Science of Variation Group. Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed Tomography. 65th Annual Meeting American Society for Surgery of the Hand; 2010 October 7-9; Boston, Massachusetts. 2010.

22. Guitton TG, Ring D, Science of Variation Group. Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed Tomography. 28th Annual Adrian E Flatt Residents and Fellows Conference in Hand Surgery 2010 Oct 6; Boston, MA. 2010.

23. Guitton TG, Ring D, Science of Variation Group. Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed Tomography. 65th Annual Meeting of the ASSH; 2010 October 7-9; Boston, Mas-sachusetts. 2010.

24. Guitton TG, van der Werf HJ, Ring D. Quantitative Evaluation of Fracture Frag-ments of Radial Head and Coronoid Fractures. New Egnland Hand Society An-

BW_def.indd 132-133 07-02-11 22:25

134 appendices 135bibliography

nual Meeting; 2008; Sturbridge, MA, USA. 2008.25. Guitton TG, van der Werf HJ, Ring D. Quantitative Evaluation of Fracture Frag-

ments of Radial Head and Coronoid Fractures. 25th Annual Meeting of the Ortho-paedic Trauma Association; 2009; San Diego, CA. 2009.

26. Guitton TG, van der Werf HJ, Ring D. Quantitative Three-Dimensional Computed Tomography Measurement of Radial Head Fractures. Massachusetts General Hospital Clinical Research Day; 2009; Boston, MA. 2009.

27. Guitton TG, van der Werf HJ, Ring D. Non-operatively Treated Fractures of the An-teromedial Facet of the Coronoid. New Egnland Hand Society Annual Meeting; 2009; Sturbridge, MA, USA. 2009.

28. Guitton TG, van der Werf HJ, Ring D. Quantitative Evaluation of Fracture Frag-ments of Radial Head Fractures. Annual Meeting American Academy of Ortho-paedic Surgeons; 2010 March 9-13,; New Orleans. 2010.

29. Guitton TG, van der Werf HJ, Ring D. Non-operatively Treated Fractures of the An-teromedial Facet of the Coronoid. 11th European Congress for Trauma & Emer-gency Surgery; 2010; Brussels/Belgium. 2010.

30. Guitton TG, van der Werf HJ, Ring D. Quantitative Measurements of the Radial Head in Healthy Adult Patients. 55th Nordic Orthopaedic Federation Congress; 2010 5 – 7 May; Aarhus, Denmark. 2010.

31. Guitton TG, van der Werf HJ, Ring D. Non-operatively Treated Fractures of the An-teromedial Facet of the Coronoid. 55th Nordic Orthopaedic Federation Congress; 2010 5 - 7 May; Aarhus, Denmark. 2010.

32. Guitton TG, van der Werf HJ, Ring D. Quantitative Evaluation of Fracture Frag-ments of Radial Head Fractures. 55th Nordic Orthopaedic Federation Congress; 2010 5 - 7 May; Aarhus, Denmark. 2010.

33. Guitton TG, van der Werf HJ, Ring D. Quantitative three-dimensional computed tomography measurement of radial head fractures. Deutschen Kongresses für Orthopädie und Unfallchirurgie; 2010 26 - 29 Oct; Messe/ICC, Berlin. 2010.

34. Guitton TG, van Leerdam R, Ring D. Necessity of Routine Pathological Examina-tion Following Surgical Excision of Wrist Ganglions. 65th Annual Meeting Amer-ican Society for Surgery of the Hand; 2010 October 7-9; Boston, Massachusetts. 2010.

35. Guitton TG, van Leerdam RH, Ring D. Necessity of routine pathological examina-tion after surgical excision of wrist ganglions. 55th Nordic Orthopaedic Federa-tion Congress; 2010 5 – 7 May; Aarhus, Denmark. 2010.

36. Guitton TG, van Leerdam RH, Ring D. Necessity of routine pathological exami-nation following surgical excision of wrist ganglions. Deutschen Kongresses für Orthopädie und Unfallchirurgie; 2010 26 - 29 Oct; Messe/ICC, Berlin. 2010.

37. Guitton TG, van Leerdam RH, Ring D. Necessity of Routine Pathological Examina-tion Following Surgical Excision of Wrist Ganglions. 28th Annual Adrian E Flatt Residents and Fellows Conference in Hand Surgery 2010 Oct 6; Boston, MA. 2010.

38. Guitton TG, Vranceanu A, Ring D. Attitude Toward Exercising Through Pain After Radial Head Fracture. 65th Annual Meeting of the ASSH; 2010 October 7-9; Bos-ton, Massachusetts. 2010.

39. Guitton TG, Vranceanu AM, Ring D. Attitude toward exercising through pain af-ter radial head fracture. 65th Annual Meeting American Society for Surgery of the Hand; 2010 October 7-9; Boston, Massachusetts. 2010.

40. Guitton TG, Vranceanu AM, Ring D. Attitude Toward Exercising Through Pain Af-ter Radial Head Fracture. 11th Triennial Congress of the International Federation of Societies for Surgery of the Hand; 2010 31 Oct - 4 Nov; Seol, Korea. 2010.

41. Guitton TG, Vranceanu AM, Ring D. Attitude Toward Exercising Through Pain Af-ter Radial Head Fracture. 28th Annual Adrian E Flatt Residents and Fellows Con-ference in Hand Surgery 2010 Oct 6; Boston, MA. 2010.

42. Guitton TG, Wiggers J, Vrahas M, Smith M, Ring D. The Complexity of Transfers Versus Direct Admissions for Hip Fracture at a Regional Referral Center. 55th Nor-dic Orthopaedic Federation Congress; 2010 5 - 7 May; Aarhus, Denmark. 2010.

43. Guitton TG, Wiggers J, Vrahas M, Smith M, Ring D. Transfer patients have worse observed and expected outcomes compared to non-transfer patients after treat-ment for hip fracture at a regional referral center. Deutschen Kongresses für Or-thopädie und Unfallchirurgie; 2010 26 - 29 Oct; Messe/ICC, Berlin. 2010.

44. Guitton TG, Zurakowski D, van Dijk CN, Ring D. Incidence and Risk Factors for the Development of Arthrosis After Traumatic Elbow Injuries. American Shoulder and Elbow Surgeons 2009 Closed Meeting; 2009 October 24-27; New York. 2009.

45. Guitton TG, Zurakowski D, van Dijk CN, Ring D. Incidence and Risk Factors for the Development of Arthrosis After Traumatic Elbow Injuries. 11th European Con-gress for Trauma & Emergency Surgery; 2010 October 24-27; Brussels/Belgium. 2010.

46. Guitton TG, Zurakowski D, van Dijk CN, Ring D. Incidence and risk factors for the development of arthrosis after traumatic elbow injuries. Deutschen Kongresses für Orthopädie und Unfallchirurgie; 2010 26 - 29 Oct; Messe/ICC, Berlin. 2010.

47. Huangling LT, Guitton TG, Capo JT, Ring D. Elbow Instability after Open Reduction and Internal Fixation of a Fracture of the Distal Humerus. 65th Annual Meeting of the ASSH; 2010 October 7-9; Boston, Massachusetts. 2010.

48. Lu H, Guitton TG, Capo JT, Ring D. Elbow Instability after Open Reduction and In-ternal Fixation of a Fracture of the Distal Humerus. 65th Annual Meeting Amer-ican Society for Surgery of the Hand; 2010 October 7-9; Boston, Massachusetts. 2010.

BW_def.indd 134-135 07-02-11 22:25

136 appendices 137bibliography

49. Lu H, Guitton TG, Capo JT, Ring D. Elbow Instability after Open Reduction and In-ternal Fixation of a Fracture of the Distal Humerus. 28th Annual Adrian E Flatt Residents and Fellows Conference in Hand Surgery 2010 Oct 6; Boston, MA. 2010.

50. Shore B, Guitton TG, Ring D. Fracture Patterns and Characteristics of Posterior Monteggia Fractures With and With-out Ulnohumeral Dislocation. 65th Annual Meeting of the ASSH; 2010 October 7-9; Boston, Massachusetts. 2010.

51. Shore BI, Guitton TG, Ring D. Fracture Patterns and Characteristics of Posterior Monteggia Fractures With and With-out Ulnohumeral Dislocation. 65th Annu-al Meeting American Society for Surgery of the Hand; 2010 October 7-9; Boston, Massachusetts. 2010.

52. Shore BI, Guitton TG, Ring D. Fracture Patterns and Characteristics of Posterior Monteggia Fractures With and Without ulnohumeral Dislocation. 11th Triennial Congress of the International Federation of Societies for Surgery of the Hand; 2010 31 Oct - 4 Nov; Seol, Korea. 2010.

53. Shore BI, Guitton TG, Ring D. Fracture Patterns and Characteristics of Posterior Monteggia Fractures With and With-out Ulnohumeral Dislocation. 28th Annual Adrian E Flatt Residents and Fellows Conference in Hand Surgery 2010 Oct 6; Bos-ton, MA. 2010.

54. Song M, van Leerdan R, Guitton TG, Ring D. Complications of volar plate fixation for distal radius fractures. 65th Annual Meeting American Society for Surgery of the Hand; 2010 October 7-9; Boston, MA. 2010.

55. Song M, van Leerdan R, Guitton TG, Ring D. Risk Factors for Complications after Locked Volar Plate Fixation of Distal Radius Fractures. 11th Triennial Congress of the International Federation of Societies for Surgery of the Hand; 2010 31 Oct - 4 Nov; Seol, Korea. 2010.

56. van der Werf HJ, Guitton TG, Ring D. Non-operatively Treated Fractures of the An-teromedial Facet of the Coronoid. Combined Annual Meeting of the ASSH and ASHT; 2009; San Francisco, California. 2009.

57. Wiggers J, Guitton TG, Vrahas M, Smith M, Ring D. Transfer patients have worse observed and expected outcomes compared to non-transfer patients after treat-ment for hip fracture at a regional referral center. 26th Annual Meeting of the Orthopaedic Trauma Association; 2010 Oct 13 - 16; Baltimore, Maryland, USA. 2010.

58. Guitton TG, Ring D, Variation So, editors. Interobserver Reliability of Radial Head Fracture Classification: Two-Dimensional vs. Three-Dimensional Computed To-mography. Jaarvergadering, Nederlandse Orthopaedische Vereniging; 2011; Gron-ingen, the Netherlands.

BW_def.indd 136-137 07-02-11 22:25

138 appendices 139acknowledgements

Acknowledgements

BW_def.indd 138-139 07-02-11 22:25

140 appendices 141acknowledgements

Acknowledgements

I would like to thank all patients that participated voluntarily in the studies of this thesis.

Prof. van Dijk, you received my idea to perform a PhD in Boston with great enthusiasm. You made me feel as if it was completely normal and realistic to start such a huge endeavor. Your comments were always very encouraging and our meet-ings productive. You guided my thoughts in the direction that lead to this PhD. It was wonderful to be able to show you around in Boston. I will never forget the lobster and the Peter Michael! Thank you for the confidence you gave me and I am looking forward to work with you as a resident in the near future! Prof. Jupiter, dr. J, although the soccer games at your house are famous, playing volleyball with “Europe” against “the world” was fun too! Your knowledge of trauma surgery during journal clubs was intriguing! It is an honor to have you as a promotor of my thesis. David, since my first introduction to you we started emailing back and forth almost everyday. You mentioned at your home in Beacon Hill that “we are good on email”. I think in the 21st century that does not sound even awkward anymore. It formed the basis of our extremely efficient collaboration and techni-cally the basis of this thesis. You taught me how to craft ideas from the clinic to a research study, ethics and of course psychology. I am thankful for your guidance and the opportunities you gave me and I hope to keep collaborating with you in the future. Peter, my research endeavor started with you; you explained to me how you went across the pond to start your PhD and ended up staying even more years to perform your residency and fellowship in the US. You opened my eyes to start clinical research. After the great start with our projects in Amsterdam, you were the first to suggest to do a PhD. It was wonderful to visit you and your family twice in Nantucket for a “research meeting” at the beautiful beach house! Thank you for inspiring me and for your mentorship throughout my PhD! Job, at our first meeting in the Blaffende Vis you showed me one of the in-famous “AO cards”. It was the start of a fruitful collaboration. I could always count on you as a colleague, thank you for coaching me! Thanks to all my colleagues from “Team Research”: Sebas, Anneluuk, Geert, Kim, Fiesky, Abdo, Huub, Roderick, Jim, Maarten Niekel, Maarten Koenis, Paul, Jean-Paul, Diederik, you all made my time at the research factory even more fun. I still remember the Thai Fridays, CCRP, Quiznos tuna melt, and of course the collaboration on exciting projects. It was a productive time for all of us and I hope to continue this

collaboration in our careers. Research coordinators, Shawn, Phil and Megan, thank you for letting me appreciate “the best country in the world” and feeding me with patriotism. Marga, Rosalie and Christine, thank you for all help with the paperwork. Thanks to my roomies at 238 Prospect street. Stan, jouw (economische) inzichten waren waardevol en jouw gezelligheid zeer gewaardeerd! Hopelijk weet je de weg naar het pluche snel te vinden opdat jij weer in ons midden bent! Monique, Lucas, Marjolein, Boaz, Sophie, Darinka, Cillian and Melody you guys made me feel home abroad! I keep remembering our nights at the phoenix landing, DOTC, Thanksgiving turkeys, grilling in the backyard, Morning Joe, brunches at S und S, and Shahi Korma’s at Punjabi Dhaba. It was perfect! All the Bostonians, Abdenour, Oliver, Lana and Mary; thanks for taking me around to kayak and fishing trips at the cape, the movie nights, Arguileh get togeth-ers and ski weekends. Thank you for your love and friendship. Carlos, thank you for the great escapes! We were living the dream! I hope to see you guys soon in Amster-dam, Morocco, Mexico, Lebanon, the USA, or some other place in the world! Lieve Bianca, ontzettend bedankt voor je hulp bij het ontwerpen van mijn proefschrift, ik ben erg trots op het resultaat! Ted en Sasja; mijn paranimfen, jullie support is van onschatbare waarde op deze belangrijke dag in mijn leven! Sasja, je bent een geweldige roomie geweest in Boston, bij jou kon ik altijd mijn verhaal kwijt. Ted, amice, om jou te vragen als para-nimf was vanzelfsprekend, jij bent eigenlijk al mijn paranimf sinds September 2003. Mamma, Pappa, Els, Thibault en Sébastien; al 27 jaar voel ik mij gesteund door jullie onvoorwaardelijke liefde. Mijn plan om 2 jaar naar Amerika te gaan was voor jullie geen grote verrassing, wat dat betreft waren jullie natuurlijk al heel wat gewend! Dankzij jullie ben ik wie ik ben en sta ik nu hier! Lieve Lou, jouw liefde en steun waren onmisbaar. Het is “onze” tijd in Ame-rika geworden; een van de mooiste periodes uit mijn leven die ik met jou heb kunnen delen. Ik hoop dat er nog veel zullen volgen; je t’aime ma cherie!

BW_def.indd 140-141 07-02-11 22:25

142 appendices 143curriculum vitae

Curriculum Vitae

BW_def.indd 142-143 07-02-11 22:25

144 appendices

Curriculum vitaeThierry Guillaume Guitton was born on October 16th 1983 in Delft, the Netherlands. Son of a Dutch mother and French father, he was the middle child of three boys. He grew up in Delft and attended the Vrije School Den Haag where he graduated in 2003. Right after graduating, Thierry was accepted at the medical school of the Uni-versity of Amsterdam. That same year he joined a student Society, the Amsterdamsche Studenten Corps en Amsterdamsche Vrouwelijke Studenten Vereniging (ASC/AVSV). He imme-diately became involved with organising committees and participated on several boards and was elected president of his fraternity later on. After finishing his nursery internship in Martinique in the French Carib-bean, he completed his first year and received his Bachelor’s degree in 2004. During his second year in medical school he started to work for Bio Implant Services (BIS) as part of the bone explantation team. In his fourth year of medical school, he followed courses from Dr. Peter Kloen. The synergies in their interests resulted in several re-search projects that he performed at the Orthopedic Research Centre Amsterdam (ORCA) from the department of orthopedics at the Academical Medical Centre (AMC) in Amsterdam under the guidance of Professor C. Niek van Dijk. After completing his Masters degree in November 2007, the driving ambition to explore the opportuni-ties on offer made him move, with support from several scholarships including the VSB, to Boston in the United States for a PhD Research Fellowship at Harvard Medi-cal School. Thierry continued the tradition of the MGH-AMC collaboration, which al-ready led to three PhD degrees under supervision of Professor C. Niek van Dijk at the University of Amsterdam. Under supervision of Professor Jesse B. Jupiter and Associate Professor Da-vid Ring from the Orthopeadic Hand and Upper Extremity Service at the Massachu-setts General Hospital from the Harvard University, Thierry investigated complex trauma of the upper extremity with a focus on the radial head. Thierry presented his work at multiple national and international meetings and was rewarded with honors. Besides his PhD research projects, he became the founder and director of the Science of Variation Group and the Science of Variation Foundation. In March 2010 he returned to Amsterdam to finish and defend his thesis in the beginning of 2011. Currently, Thierry is enrolled in his clinical rotations at the AMC to finish his medical degree in early 2012. Later that year he will start his orthopedic training at the University of Amsterdam Orthopaedic Residency Program under supervision of Professor C. Niek van Dijk.

BW_def.indd 144 07-02-11 22:25

uva-dare (digital academic repository) advancements in … · 2 3 ph.d thesis of thierry g....

Documents