effective treatment of posttraumatic and postoperative ...control group. the second group, the...

Effective Treatment of Posttraumatic andPostoperative Edema in Patients with

Ankle and Hindfoot FracturesA Randomized Controlled Trial Comparing Multilayer Compression

Therapy and Intermittent Impulse Compressionwith the Standard Treatment with Ice

Manuela Rohner-Spengler, MPTSc, Angela Frotzler, PhD, Philipp Honigmann, MD, and Reto Babst, MD, PhD

Investigation performed at the Departments of Rheumatology and Physiotherapy and Trauma Surgery,Lucerne Cantonal Hospital, Lucerne, Switzerland

Background: After ankle and hindfoot fractures, edema has a major impact on the time for surgical intervention and mayincrease the risk of wound complications and infection postoperatively. The aim of this study was to evaluate the efficacyof multilayer compression and intermittent impulse compression therapy in reducing ankle and hindfoot edema comparedwith the standard treatment with elevation and ice.

Methods: This was a randomized, controlled, single-blinded clinical trial using a repeated-measures design. Fifty-eightpatients with unilateral fractures of the ankle or hindfoot were randomized into the cold pack (control) group, the bandagegroup, or the impulse compression group and were analyzed according to the intention-to-treat principle. The primaryoutcome was the reduction of edema as measured with the figure-of-eight-20 method.

Results: Preoperatively and postoperatively, there were significant differences in edema reduction between the bandagegroup and the control group. After two days of intervention, the median preoperative edema reduction in the control groupwas 22.0 mm (25%) compared with 211.0 mm (223%) in the bandage group (p < 0.017), and 20.3 mm (0%) in theimpulse compression group (p > 0.017). Postoperatively, after two days, the median edema changes were 13.5 mm(17%) in the control group compared with 27.3 mm (222%) in the bandage group (p < 0.017) and 15.0 mm (146%) inthe impulse compression group (p > 0.017).

Conclusions: Multilayer compression therapy results in a faster reduction of ankle and hindfoot edema, although withless ankle dorsiflexion on postoperative day three than the control group, and can be recommended as an alternativetreatment. Intermittent impulse compression applied without any extra compression by stockinette or bandage andwithout elevation in off-session periods cannot be recommended as a superior alternative to the treatment with ice.

Level of Evidence: Therapeutic Level I. See Instructions to Authors for a complete description of levels of evidence.

Fractures of the ankle and hindfoot are common and fre-quently require surgical stabilization. The concomitantedema has a major impact on the time point of surgical

intervention and may increase the risk of wound complicationsand infection postoperatively. Prevention, control, and reductionof edema are of great interest, and different techniques targeted at

Disclosure:One ormore of the authors received paymentsor services, either directly or indirectly (i.e., via his or her institution), froma third party in support of anaspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, withany entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had anyother relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. Thecomplete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.

Peer Review: This article was reviewed by the Editor-in-Chief and one Deputy Editor, and it underwent blinded review by two or more outside experts. It was also reviewedby an expert in methodology and statistics. The Deputy Editor reviewed each revision of the article, and it underwent a final review by the Editor-in-Chief prior to publication.Final corrections and clarifications occurred during one or more exchanges between the author(s) and copyeditors.

1263

COPYRIGHT � 2014 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED

J Bone Joint Surg Am. 2014;96:1263-71 d http://dx.doi.org/10.2106/JBJS.K.00939

edema reduction are available. A common and well-knownmethod is cryotherapy. According to Knight et al.1, ice can di-minish pain, metabolism, and muscle spasm by decreasing tissuetemperature and therefore can minimize the inflammatoryprocess after soft-tissue trauma. Nevertheless, its proper appli-cation throughout the different stages after trauma remainspoorly understood1-5, and evidence is still controversial6-8. An-other treatment to reduce edema is compression therapy. Al-though there is evidence for its positive effects on edemareduction for various techniques9-13, its value for treating acutetrauma patients is not well established. Partsch11 postulated thatposttraumatic edema is a good indication for compressiontherapy, and different techniques of using several layers of woolaugmented by several layers of a compression bandage have beendescribed and promoted by various authors since the late 1940s10-12.More recently, multilayer bandaging has been effective in re-ducing edema in patients with lymphedema13-16 and in promotingwound-healing in patients with venous crural ulcers16,17. To ourknowledge, no randomized controlled trial investigated the ef-fects of multilayer compression therapy in patients with acutetrauma on edema reduction or on complications, such as woundinfections, or on patient satisfaction. Multilayer compression

bandaging has been used in our hospital with positive effects intreating edema in patients with trauma. Various publications18-23

also suggest the use of intermittent impulse compression devicesin reducing traumatic edema. These devices contain an inflatablepad that intermittently applies pressure to the plantar arch of thefoot to compress the concomitant veins and lymphatic vessels ofthe lateral plantar artery, thereby enhancing venous backflow24,25.

The aim of this study was to compare the effects of mul-tilayer compression therapy and those of intermittent impulsecompression with the effects of the standard treatment with el-evation and ice on edema reduction in the treatment of patientswith ankle or hindfoot fractures. We hypothesized that both themultilayer compression therapy and the impulse compression aremore effective in reducing ankle edema than the standard treat-ment with elevation and ice. We hypothesized that these effectswould result in shorter preoperative and postoperative hospi-talization days and in better ankle motion postoperatively.

Materials and MethodsDesign

Aprospective, randomized, controlled, singled-blinded clinical trial, using arepeated-measures design, was carried out in the Department of

Fig. 1

CONSORT (Consolidated Standards of Reporting Trials) diagram. The asterisk indicates that the impulse compression (IC) group had fourteen patients in

the stand-alone treatment and six patients in the adapted protocol. The dagger indicates that the value was analyzed according to the intention-to-treat

principle.

1264

THE JOURNAL OF BONE & JOINT SURGERY d J B J S .ORG

VOLUME 96-A d NUMBER 15 d AUGUST 6, 2014TREAT ING POSTTRAUMATIC AND POSTOPERAT IVE EDEMA IN

PATIENTS WITH ANKLE AND HINDFOOT FRACTURES

Traumatology of the Lucerne Cantonal Hospital, Lucerne, Switzerland. The studywas approved by the Ethics Committee of the Canton of Lucerne, Switzerland(Approval-Nr.: 616), and the trial is registered with ClinicalTrials.gov(NCT01389089).

PatientsPatients with unilateral ankle or hindfoot fractures who were referred to theemergency unit from January 2007 to January 2009 were assessed for eligibility(Fig. 1). Patients meeting eligibility criteria (see Appendix) were informedabout the study. After obtaining written informed consent, patients were ran-domized into one of the three treatment groups using sequentially numberedopaque and sealed envelopes. Recruitment was conducted by the physicians ofthe emergency unit and treatment was assigned by the study coordinator orstudy assistant. The randomization sequence was generated by computerthrough an independent software specialist.

Primary and Secondary Outcome MeasuresOur primary outcome was the reduction of ankle edema during inpatientperiod. The figure-of-eight-20 method, as previously described and publishedby our group

26, was used to assess edema (see Fig. 2).

Several secondary outcomes were monitored and were followed as listedin Table I

27,28.

We also recorded the patients’ compliance with the general and group-specific intervention protocols, their activity during hospitalization, and thefollowing group-specific variables: number of ice gel packs applied per day, sub-bandage pressure, and the intermittent impulse system running time.

Edema and ankle motion were assessed by the examiner, who wasblinded to the treatment intervention. For details concerning assessment andblinding procedures, see the Appendix

29-34.

InterventionsTo control and reduce edema, ice gel packs were applied to the patients in thecontrol group. The second group, the bandage group, received multilayercompression bandages (two layers of wool and two or three layers of shortstretch bandage), and the third group, the impulse compression group, wastreated with the A-V Impulse System (Orthofix Vascular Novamedix,Andover, United Kingdom). For details on the interventions, see Table IIand the Appendix. For preoperative fracture stabilization and to standardizepostoperative care, a custom-made orthosis (VACOped vacuum orthosis;OPED, Steinhausen, Switzerland) was used with all patients without ex-ternal fixation treatment. Further co-interventions for all patients were

standardized according to the usual treatment regime of the hospital (seeAppendix).

Data AnalysisSample size calculation (a = 0.05, power = 0.8, sigma = 12) revealed thattwenty-eight patients per group would be necessary to detect a difference of8 mm in figure-of-eight-20 measurements

35. An 8-mm difference is of clinical

relevance, and it is above the minimal detectable change26. Thus, our targeted

sample size was 102 patients (thirty-four in each group), including a 20%dropout rate. A safety analysis was planned after the scheduled recruitmentperiod of two years

36. A check for normality was performed with the use of

box plots. Group medians were calculated for the description of patient de-mographic characteristics and baseline data. To evaluate if a specific treatmentintervention (e.g., multilayer compression therapy versus intermittent im-pulse compression versus standard treatment with ice) is important in edemareduction, mixed model analyses were used. The included factors in themodels were age, sex, body mass index, smoking, fracture type, number of

TABLE I Summary of Outcome Measures

Primary outcome measureEdema, figure-of-eight-20, measured in millimeters, measured daily for a maximum of five days preoperatively inpatient, daily formaximum of five days postoperatively inpatient, and at six weeks postoperatively

Secondary outcome measuresAnkle plantar flexion and dorsiflexion, measured in degrees, measured daily for a maximum of five days postoperatively inpatient,and at six weeks postoperativelyVisual analog scale for pain27,28 measured daily preoperatively and postoperatively inpatient and at six weeks postoperativelyVisual analog scale for patient satisfaction27,28 measured daily preoperatively and postoperatively inpatient and at twelve weeksand one year postoperativelyNumber of inpatient days measured daily preoperatively and postoperatively inpatientUse of medication measured daily preoperatively and postoperatively inpatientWound-healing measured daily postoperatively inpatient and at six weeks postoperativelyAdverse events measured daily preoperatively and postoperatively inpatient and at six weeks and one year postoperativelyFoot and Ankle Ability Measure measured at twelve weeks and one year postoperativelyShort Form-36 measured preoperatively inpatient according to the patient state of health before trauma, and at twelve weeksand one year postoperatively

Fig. 2

A photograph showing the assessment of ankle edema with the figure-of-

eight-20method. The intraclass correlation coefficients are >0.99 for both

intra-rater and interrater reliability, and the minimal detectable change is

7.3 mm when landmarks are used, as in this study26.

1265




intervention days (time), treatment intervention, edema at baseline, and se-verity of hematoma. The two models (one preoperative and one postopera-tive) with the smallest Akaike information criterion (AIC) were used todetermine the significance level of the tested variables. Kruskal-Wallis testsacross all three study groups, followed by post hoc analyses with t tests orMann-Whitney U tests, were then used to examine if there was a significantdifference between the groups in edema changes or in any of the secondaryoutcomes. The significance level, for multiple comparisons, was adjusted

according to the Bonferroni method. The effect sizes were calculated ac-cording to Cohen

37.

All analyses were performed according to the intention-to-treat prin-ciple

38,39. Significance was considered at a two-tailed level of p < 0.05.

Source of FundingThis study was partially funded by Orthofix and by a donation from the na-tional professional group of physiotherapists specializing in lymphatic therapy,

TABLE II Group-Specific Study Interventions

Group Study Interventions

Control Elevation for twenty-four hours with a Hess splint, four ice gel packs a day (minimum of twenty minutesper application), and no compression (no stockinette, no bandages)

Bandage Elevation for twenty-four hours with a Hess splint; multilayer compression bandage (twenty-two hoursof compression, one hour of bandage removal for blinding before measurements, and one hour ofreapplying the bandage): tube gauze, two layers of wool, two or three layers of a short-stretch bandage,extra-padding with wool behind both malleoli*; tightness: moderate compression as easily supported by thepatient, without having a feeling of discomfort; in patients with external fixation treatment: tube gauze or regulargauze for skin comfort (not mandatory), same bandage material as that for patients without external fixation;partial unrolling of bandages prior to application (for fitting underneath the fixators); no cold application

Impulse compression Intermittent impulse compression: 130 mm Hg, for one second, every twenty seconds; twenty-four hours ifpossible but at least a mean (and standard deviation) of 8 ± 2 hours a day; at least two consecutive hours persession; lower limb in the horizontal position or lower during the impulse compression session (foot positionedunderneath the level of the heart); lower limb in horizontal position in off-session periods (no elevation of thelimb leads to impulse compression as a stand-alone treatment); no cold application, no additional compression(no stockinette, no bandages)

*The materials used were Tricofix tube gauze, Artiflex 10-cm padding bandage, and Comprilan short stretch compression bandages (all BSNMedical, Hamburg, Germany).

TABLE III Changes in Preoperative Edema

Control Group Bandage Group Impulse Compression Group P Value*

At one day 0.01No. of analyzed patients 19 16 11Figure-of-eight-20† (mm) 0.3 (23.0; 3.3) 26.7 (212.4; 20.2)‡ 2.7 (23.3; 8.0)

At two days 0.03No. of analyzed patients 19 14 9Figure-of-eight-20† (mm) 22.0 (28.0; 4.7) 211.0 (216.0; 25.0)§ 20.3 (210.5; 5.8)

At three days 0.01No. of analyzed patients 13 14 8Figure-of-eight-20† (mm) 21.3 (29.7; 5.3) 215.2 (217.9; 29.8)‡§ 21.8 (211.3; 4.9)

At four days 0.05No. of analyzed patients 10 11 4Figure-of-eight-20† (mm) 27.1 (214.8; 22.8) 220.7 (222.7; 213.3)§ 213.5 (219.3; 23.0)

At five days 0.04No. of analyzed patients 6 4 3Figure-of-eight-20† (mm) 213.7 (217.3; 21.1) 221.3 (233.5; 219.4) 220.7 (229.7; 217.7)

*The p values are between-group differences across all three groups, based on Kruskal-Wallis tests.†The values are given as the median, with the25% and 75% quartiles in parentheses; negative numbers reflect edema reduction, and positive numbers reflect an augmentation of edema. ‡Onthe basis of the Mann-Whitney tests (adjusted alpha = 0.017), there were significant differences between the impulse compression group and thebandage group, but no significant difference between the impulse compression group and the control group. §On the basis of the Mann-Whitneytests (adjusted alpha = 0.017), there were significant differences between the control group and the bandage group.

1266




Fachgruppe Lymphologische Physiotherapie Schweiz. Smith & Nephew do-nated the bandage material. Orthofix served mainly as a consultant for themodification of the A-V Impulse System application used in the study and alsomonitored the conduct of the study. The Department of Clinical Affairs atCovidien reviewed and commented on the publication draft. Neither sponsorhad influence on the analysis or interpretation of the study data. The funds wereused to pay for partial additional physiotherapist salaries generated because ofthe conduction of the study.

ResultsParticipant Flow

After the scheduled recruitment period of two years, whensixty-seven patients were enrolled, safety analysis showed

that significant differences in the main outcome had occurredand recruitment was halted.

Three of the recruited patients were directly excludedafter randomization because of fracture instability. Six patientswho originally were randomized into the impulse compressiongroup are not included in the present analysis because of aprotocol change during the conduct of the study. Therefore,

fifty-eight patients were analyzed according to the intention-to-treat principle (Fig. 1).

Baseline CharacteristicsPatient demographic characteristics and baseline variables arepresented in the Appendix40. Not all of the fifty-eight patientshad preoperative and postoperative data available as some wereincluded preoperatively and postoperatively and some wereincluded only postoperatively.

Edema ChangesThe medians and upper and lower quartiles of the changes inedema, as measured with the figure-of-eight-20 in millimeters,for each group, over a period of five days preoperatively and fivedays postoperatively, are presented in Tables III and IV and inFigures 3 and 4.

Preoperatively, the mixed model analysis revealed thatthe variables of time (p < 0.001), treatment intervention (p =0.005), and fracture type (p = 0.016) are significant factors in

Fig. 3

A bar graph showing preoperative edema changes in each group over a

period of five intervention days. The horizontal black line in the bar rep-

resents the median change in millimeters as measured with the figure-of-

eight-20 method. Negative values represent the amount of edema

reduction from baseline. Positive values represent an augmentation of

edema. The asterisks indicate a significant difference as compared with

the control group (p < 0.017). The daggers indicate a significant difference

as compared with the impulse compression (IC) group (p < 0.017). The

error bars indicate the upper and lower quartiles. The circular symbols

represent outliers.

Fig. 4

A bar graph showing postoperative edema changes in each group over a

period of five intervention days. The horizontal black line in the bar

represents the median change in millimeters as measured with the figure-

of-eight-20 method. Negative values represent the amount of edema

reduction from baseline. Positive values represent an augmentation of

edema. The asterisks indicate significant differences as compared with

the control group (p < 0.017). The error bars indicate the upper and

lower quartiles. The circular symbols represent outliers. IC = impulse

compression.

1267




preoperative edema reduction. The other factors in the model,smoking (p = 0.532), edema at baseline (p = 0.301), severity ofhematoma (p = 0.258), and sex (p = 0.156), were not signifi-cant. Postoperatively, the mixed model included the variablesof time (p < 0.001), treatment intervention (p = 0.019), edemaat baseline (p = 0.098), and severity of hematoma (p = 0.478) asfactors. The first group to reach a clinically meaningful medianedema reduction (more than28 mm) was the bandage group,after two days of intervention.

The relative mean preoperative edema reduction aftertwo days was 25% in the control group (effect size, 20.1),223% in the bandage group (effect size,20.67), and 0% in theimpulse compression group. Postoperatively, after two days,the relative mean edema reduction was 222% in the bandagegroup, whereas there was an increase in edema of 7% in thecontrol group and 46% in the impulse compression group.After three days of intervention, the effect sizes were large in thebandage group (preoperatively, 20.94, and postoperatively,20.83) and small in the two other groups (less than 20.3).

Secondary Outcomes and Follow-upThe bandage group showed significantly less median ankledorsiflexion (p = 0.03) on the third postoperative day than thecontrol group. For further secondary outcome measures andfollow-up, see the Appendix.

During the inpatient period, one patient in the controlgroup and one patient in the bandage group had delayedwound-healing. Three patients in the bandage group reported once eachthat the compression applied was too tight and the bandage had

to be adjusted. Of the control group, two patients reported painwhen the ice gel pack was applied, and this treatment wastemporarily stopped. Of the impulse compression group, sevenpatients discontinued their allocated treatment. Four perceivedmore pain when the impulse compression system was on, twodeveloped a wound complication, and one both perceivedmore pain when the impulse compression system was on anddeveloped a wound complication.

Use of medication, activity level, and adherence to the studyprotocols were comparable, with no significant differences amongthe groups.

In the follow-up period, in the impulse compressiongroup, one patient had a superficial wound infection withStaphylococcus aureus and another patient had skin necrosis.The two patients were treated with antibiotics and local de-bridement. In both the control group and the bandage group,one patient in each group had a delay in wound-healing. Alladverse events were unrelated to the study treatments, and allfractures were healed at the latest follow-up.

We found no significant differences in the Foot andAnkle Ability Measure or in the Short Form-36 between thethree groups after three months or one year postoperativelyafter adjusting for Bonferroni.

Considering the interventions, patients in the controlgroup received approximately 4.5 ice gel packs a day preoper-atively and approximately four ice gel packs postoperatively.

The mean sub-bandage pressure (and standard devia-tion) was 31.5 ± 8.9 mmHg preoperatively and 27.3 ± 9.1 mmHgpostoperatively. Patients in the impulse compression group

TABLE IV Changes in Postoperative Edema

Control Group Bandage Group Impulse Compression Group P Value*

At one day 0.03No. of analyzed patients 22 20 13Figure-of-eight-20† (mm) 3.0 (22.3; 6.3) 21.0 (27.7; 3.0)‡ 4.6 (23.5; 12.8)

At two days 0.01No. of analyzed patients 18 15 11Figure-of-eight-20† (mm) 3.5 (24.4; 7.4) 27.3 (214.7; 0.3)‡ 5.0 (29.7; 13.0)

At three days 0.08No. of analyzed patients 16 12 6Figure-of-eight-20† (mm) 22.2 (29.3; 4.4) 29.0 (217.9; 26.5)‡ 25.0 (216.8; 11.1)

At four days 0.38No. of analyzed patients 9 7 6Figure-of-eight-20† (mm) 26.3 (218.8; 5.2) 217.0 (226.0; 22.3) 26.5 (218.2; 6.8)

At five days 0.55No. of analyzed patients 5 5 4Figure-of-eight-20† (mm) 25.0 (232.0; 21.7) 221.0 (231.0; 28.3) 23.3 (228.7; 8.8)

*The p values are for between-group differences across all three groups, based on Kruskal-Wallis tests.†The values are given as the median, withthe 25% and 75% quartiles in parentheses; negative numbers reflect edema reduction, and positive numbers reflect an augmentation of edema.‡On the basis of the Mann-Whitney tests (adjusted alpha = 0.017), there were significant differences between the control group and the bandagegroup, but no significant difference between the impulse compression group and the control group or between the bandage group and the impulsecompression group.

1268




had a mean system running time (and standard deviation)of 16.6 ± 5.9 hours preoperatively and 13.8 ± 5.9 hourspostoperatively.

Discussion

The results of this study show that multilayer compressiontherapy is a very effective edema-resolving method in pa-

tients with ankle and hindfoot fractures, which may reduce thelength of hospital stay. Considering the impulse compressiontreatment, there seems to be a tendency toward a better edemareduction as compared with ice, when applied preoperativelyfor more than three days. In contradiction to the findings ofprior studies18-23, we could not find significantly different effectsof the impulse compression treatment as compared with thestandard treatment with ice, when the impulse compressionwas applied as a stand-alone treatment (with no elevation ofthe affected limb and no additional compression by bandageor stockinette). Stockle et al.21 found that the impulse com-pression showed the fastest swelling reduction in all measuredareas both before and after surgery as compared with coolpacks (applied four times daily). They reported an averagetwenty-four-hour edema reduction of 53% preoperatively inthe impulse compression group, whereas the cool pack grouphad a reduction of 10%. Caschman et al.18 found that theimpulse compression in-cast system significantly reduced thetime taken for ankle swelling to diminish prior to surgerywhen compared with elevation only. They also found thatwound and skin complications could be reduced. We believethat the use of different modes of impulse compression and itsco-interventions throughout prior studies contribute to ourcontroversial findings. We suspect that positioning of the limb(with or without elevation) during impulse compression ses-sions and in off-session periods and the addition of a com-pressive force, as applied, for example, by stockinette, play animportant role. There is a lack of precise description of theimpulse compression application and its co-interventions inprior studies18-23, making comparisons of the results difficult.We believe that limb positioning and any sort of compressionare potential confounders in clinical investigations. In ourstudy, we tried to avoid having compression be a potentialconfounder. Therefore, the patients in the control and impulsecompression groups did not receive any sort of preoperativecompression therapy. Postoperatively, all patients used theVACOped orthotic support system fully (vacuum on) for mo-bilization and removed it afterwards, thereby allowing com-pressional forces only while the limbs were kept in a verticalposition during walking.

Lack of blinding21, or the use of unvalidated outcomesmeasurements, as seen in previous studies18-23, might also havecontributed to the controversial findings of our study ascompared with others. In our study, blinding was successful inapproximately 91% of the measurements. Unblinding mostlyoccurred in the later postoperative phase and was random. Thesurgeons were not blinded to the interventions. Therefore, ourresults regarding the number of days until possible operationshould be interpreted with caution. Nevertheless, because the

mean edema (and standard deviation) at the time of readinessfor surgery, as measured by the blinded physiotherapists, was27.1 ± 12.4 mm in the control group, 21.2 ± 12.4 mm in thebandage group, and 25.1 ± 14.4 mm in the impulse compres-sion group, we believe that the surgeons were not influenced infavor of the two intervention groups. On the contrary, we be-lieve that the effect of the two intervention groups in reducingdays until possible operation might even be underestimatedsomewhat, as the control group was considered ready for op-eration with a larger mean preoperative edema than the twoother groups. Despite the finding that the patients in thebandage group tended to perceive somewhat more pain pre-operatively than the patients treated with ice, the interventionwas well tolerated and no patient wanted to discontinue thebandage. Conversely, despite the finding that the patients in theimpulse compression group had pain levels comparable withthose of the control group, a subset of these patients did nottolerate the intervention well and requested to be changed tothe intervention of the control group. Other authors reportedsimilar negative effects of impulse compression18,21. Accordingto the findings of many authors who concluded that ice iseffective in reducing pain41,42, we assumed that the patients inthe control group probably would have had less pain as com-pared with the patients in the two other groups. After surgery,the patients in the bandage group had a tendency toward lesspain than the patients in the control group. At the expense ofankle motion, the bandage has a stabilizing and protectiveeffect.

There were several limitations in this study. We choseelevation and ice as treatment for the control group. One ar-gument to choose ice for the treatment in the control group isthat its use is still widespread and clinicians and patients con-sider ice as the standard treatment. Nevertheless, the applica-tion of ice in the control group could have contributed to anoverestimation or underestimation of the effects of the twointervention groups.

During recruitment of the patients, we observed that theeffect of the impulse compression intervention was not as ex-pected. Despite our attempts to keep the medical staff neutraltoward the patients’ interventions, impartiality could not beguaranteed throughout the study. It is possible that their ob-servations induced a decreasing tolerance toward the impulsecompression device. As a consequence of these problems, webelieved that there was a need to modify the original impulsecompression intervention protocol. The last six patients whowere randomized into the impulse compression group werethus treated according to an adapted modified protocol, withelevation in off-session periods. To avoid flawed results andconclusions, the data of these six patients were not included inthe current intention-to-treat analyses.

After the scheduled study period, before reaching thetargeted sample size (n = 84), because of financial constraintsand for feasibility reasons, we stopped patient recruitment.Even though the statistical analyses revealed some significanteffects, the study overall remained underpowered (somewhatunderpowered considering the results of the bandage group

1269




compared with the control group, and strongly underpoweredconsidering the comparisons between the impulse compres-sion group and the two other groups).

Although the results of this study were obtained in in-patient care, we believe that they can also be applied to out-patient care. Postoperatively, patients can easily be trained toapply the bandage themselves. Before fracture stabilization, themultilayer bandage can be applied by specialists in the emer-gency unit and afterwards by trained home care providers. Inthe outpatient setting, we believe that patients with soft-tissueinjury only, such as patients with ankle sprains, could alsobenefit from multilayer bandaging.

In conclusion, in patients with acute trauma, multilayercompression therapy leads more quickly to a clinically relevantand significant reduction of ankle and hindfoot edema than thestandard treatment with ice. We cannot recommend inter-mittent impulse compression as a superior alternative to thetreatment with ice, when it is applied as a stand-alone treat-ment (e.g., without elevation in off-session periods and with-out any form of external compression).

AppendixText describing the details of the assessment and blindingprocedures, a figure demonstrating postoperative appli-

cations for each group, and tables showing eligibility criteria,

patient demographic characteristics and baseline data for pre-operatively and postoperatively included patients, secondaryoutcomes, and standardized co-interventions for all patients areavailable with the online version of this article as a data sup-plement at jbjs.org. nNOTE: We thank Dorothee Rhein Staub, MD, and her colleagues for their contributions to theenrollment of participants in this study. We furthermore thank all the physiotherapists, nurses, andmedical doctors who were involved in the study for their great contribution to this interdisciplinarywork. We especially thank Beate Walter (physiotherapist), Petra Hofstetter (nurse), and AlexanderBrunner, MD, PD.

Manuela Rohner-Spengler, MPTScPhilipp Honigmann, MDReto Babst, MD, PhDDepartments of Rheumatology and Physiotherapy (M.R.-S.)and Trauma Surgery (M.R.-S., P.H., and R.B.),Lucerne Cantonal Hospital,Spitalstrasse,6000 Lucerne 16, Switzerland.E-mail address for M. Rohner-Spengler: [email protected]

Angela Frotzler, PhDClinical Trial Unit,Swiss Paraplegic Centre,Guido A. Zach Strasse 1,6207 Nottwil, Switzerland

References

1. Knight KL, Brucker JB, Stoneman PD, RubleyMD.Muscle injury management withcryotherapy. Athl Ther Today. 2000 Jul;5:26-30.2. MacAuley D. Do textbooks agree on their advice on ice? Clin J Sport Med. 2001Apr;11(2):67-72.3. Mac Auley DC. Ice therapy: how good is the evidence? Int J Sports Med. 2001Jul;22(5):379-84.4. Bleakley C, McDonough S, MacAuley D. The use of ice in the treatment of acutesoft-tissue injury: a systematic review of randomized controlled trials. Am J SportsMed. 2004 Jan-Feb;32(1):251-61.5. Hubbard TJ, Denegar CR. Does cryotherapy improve outcomes with soft tissueinjury? J Athl Train. 2004 Sep;39(3):278-9.6. Meeusen R, Lievens P. The use of cryotherapy in sports injuries. Sports Med.1986 Nov-Dec;3(6):398-414.7. Rivenburgh DW. Physical modalities in the treatment of tendon injuries. ClinSports Med. 1992 Jul;11(3):645-59.8. Thorsson O. [Cold therapy of athletic injuries. Current literature review].Lakartidningen. 2001 Mar 28;98(13):1512-3. Swedish.9. Matsen FA 3rd, Krugmire RB Jr. The effect of externally applied pressure on post-fracture swelling. J Bone Joint Surg Am. 1974 Dec;56(8):1586-91.10. Brodell JD, Axon DL, Evarts CM. The Robert Jones bandage. J Bone Joint SurgBr. 1986 Nov;68(5):776-9.11. Partsch H. Compression therapy of the legs. A review. J Dermatol Surg Oncol.1991 Oct;17(10):799-805.12. Gibbons CE, Solan MC, Ricketts DM, Patterson M. Cryotherapy compared withRobert Jones bandage after total knee replacement: a prospective randomized trial.Int Orthop. 2001;25(4):250-2.13. McNeely ML, Magee DJ, Lees AW, Bagnall KM, Haykowsky M, Hanson J. Theaddition of manual lymph drainage to compression therapy for breast cancer relatedlymphedema: a randomized controlled trial. Breast Cancer Res Treat. 2004Jul;86(2):95-106.14. Damstra RJ, Partsch H. Compression therapy in breast cancer-related lymphe-dema: A randomized, controlled comparative study of relation between volume andinterface pressure changes. J Vasc Surg. 2009 May;49(5):1256-63.15. Badger C, Preston N, Seers K, Mortimer P. Physical therapies for reducing andcontrolling lymphoedema of the limbs. Cochrane Database Syst Rev. 2004 Oct18;(4):CD003141.16. Foldi M, Foldi E. Foeldi’s textbook of lymphology: for physicians and lymphe-dema therapists. 2nd ed. Munich: Elsevier GmbH, Urban & Fischer; 2006.

17. O’Meara S, Cullum NA, Nelson EA. Compression for venous leg ulcers. Co-chrane Database Syst Rev. 2009 Jan 21;(1):CD000265.18. Caschman J, Blagg S, Bishay M. The efficacy of the A-V Impulse system in thetreatment of posttraumatic swelling following ankle fracture: a prospective ran-domized controlled study. J Orthop Trauma. 2004 Oct;18(9):596-601.19. Myerson MS, Juliano PJ, Koman JD. The use of a pneumatic intermittent im-pulse compression device in the treatment of calcaneus fractures. Mil Med. 2000Oct;165(10):721-5.20. Thordarson DB, Greene N, Shepherd L, Perlman M. Facilitating edema resolutionwith a foot pump after calcaneus fracture. J Orthop Trauma. 1999 Jan;13(1):43-6.21. Stockle U, Hoffmann R, Schutz M, von Fournier C, SudkampNP, Haas N. Fastestreduction of posttraumatic edema: continuous cryotherapy or intermittent impulsecompression? Foot Ankle Int. 1997 Jul;18(7):432-8.22. Myerson MS, Henderson MR. Clinical applications of a pneumatic intermittentimpulse compression device after trauma and major surgery to the foot and ankle.Foot Ankle. 1993 May;14(4):198-203.23. Gardner AM, Fox RH, Lawrence C, Bunker TD, Ling RS, MacEachern AG. Re-duction of post-traumatic swelling and compartment pressure by impulse com-pression of the foot. J Bone Joint Surg Br. 1990 Sep;72(5):810-5.24. McMullin G, Scott HJ, Colderidge-Smith PD, Scurr JH. An assessment of theeffect of the foot pump on venous emptying in chronic venous insufficiency. In: DavyA, Stremmer R, editors. Phlebologie. London: John Libbey; 1989. p 69-71.25. Gardner AM, Fox RH. The venous pump of the human foot—preliminary report.Bristol Med Chir J. 1983 Jul;98(367):109-12.26. Rohner-Spengler M, Mannion AF, Babst R. Reliability and minimal detectablechange for the figure-of-eight-20 method of measurement of ankle edema. J OrthopSports Phys Ther. 2007 Apr;37(4):199-205.27. Carlsson AM. Assessment of chronic pain. I. Aspects of the reliability and va-lidity of the visual analogue scale. Pain. 1983 May;16(1):87-101.28. Coll AM, Ameen JR, Mead D. Postoperative pain assessment tools in day sur-gery: literature review. J Adv Nurs. 2004 Apr;46(2):124-33.29. Moholkar K, Fenelon G. Diurnal variations in volume of the foot and ankle. J FootAnkle Surg. 2001 Sep-Oct;40(5):302-4.30. Croft PR, Nahit ES, Macfarlane GJ, Silman AJ. Interobserver reliability in mea-suring flexion, internal rotation, and external rotation of the hip using a plurimeter.Ann Rheum Dis. 1996 May;55(5):320-3.31. Attinger C, Cooper P. Soft tissue reconstruction for calcaneal fractures orosteomyelitis. Orthop Clin North Am. 2001 Jan;32(1):135-70.

1270




32. Bullinger M. German translation and psychometric testing of the SF-36 HealthSurvey: preliminary results from the IQOLA Project. International Quality of Life As-sessment. Soc Sci Med. 1995 Nov;41(10):1359-66.33. Martin RL, Irrgang JJ, Burdett RG, Conti SF, Van Swearingen JM. Evidence ofvalidity for the Foot and Ankle Ability Measure (FAAM). Foot Ankle Int. 2005Nov;26(11):968-83.34. Carcia CR, Martin RL, Drouin JM. Validity of the Foot and Ankle Ability Mea-sure in athletes with chronic ankle instability. J Athl Train. 2008 Apr-Jun;43(2):179-83.35. Rosner B. Fundamentals of biostatistics. 5th ed. Belmont, CA: Duxbury Press;1999.36. Pintadosi S. Clinical trials: a methodologic perspective. New York: John Wiley &Sons; 1997.37. Cohen J. Statistical power analysis for the behavioral sciences. New York: Ac-ademic Press; 1977. p 8.

38. Newell DJ. Intention-to-treat analysis: implications for quantitative and qualita-tive research. Int J Epidemiol. 1992 Oct;21(5):837-41.39. Gravel J, Opatrny L, Shapiro S. The intention-to-treat approach in randomizedcontrolled trials: are authors saying what they do and doing what they say? ClinTrials. 2007;4(4):350-6.40. Marsh JL, Slongo TF, Agel J, Broderick JS, Creevey W, DeCoster TA, Prokuski L,Sirkin MS, Ziran B, Henley B, Audige L. Fracture and dislocation classificationcompendium - 2007: Orthopaedic Trauma Association classification, database andoutcomes committee. J Orthop Trauma. 2007 Nov-Dec;21(10 Suppl):S1-133.41. Airaksinen OV, Kyrklund N, Latvala K, Kouri JP, Gronblad M, Kolari P. Efficacy ofcold gel for soft tissue injuries: a prospective randomized double-blinded trial. Am JSports Med. 2003 Sep-Oct;31(5):680-4.42. Bleakley CM, McDonough SM, MacAuley DC, Bjordal J. Cryotherapy for acuteankle sprains: a randomised controlled study of two different icing protocols. Br JSports Med. 2006 Aug;40(8):700-5; discussion 705. Epub 2006 Apr 12.

1271




effective treatment of posttraumatic and postoperative ...control group. the second group, the...

Documents