economic advantages to a distraction decision tree-1.30

8/6/2019 Economic Advantages to a Distraction Decision Tree-1.30

1/13

PEDIATRIC/CRANIOFACIAL

Economic Advantages to a Distraction DecisionTree Model for Management of Neonatal

Upper Airway ObstructionEmil Kohan, M.D.

Salar Hazany, M.D.Jason Roostaeian, M.D.

Karam Allam, M.D.Christian Head, M.D.

Samuel Wald, M.D.Raj Vyas, M.D.

James P. Bradley, M.D.

Los Angeles, Calif.

Background: Neonatal upper airway obstruction demands urgent attention.Tracheostomy can prove to be lifesaving but has morbidities. Recently, theauthors found reduced morbidity/mortality when using a distraction decisiontree model compared with conventional case-by-case management. In thiscurrent study, the authors assess the long-term costs of (1) a decision tree model

versus conventional treatment and (2) tracheostomy versus distraction osteo-genesis.Methods: An inpatient cost-matrix analysis study on neonates with upper airwayobstruction and micrognathia was performed (n 149). In Part I, conven-tionally treated neonates managed on a case-by-case basis received home

monitoring or a tracheostomy. Decision tree modelmanaged newborns hadspecialist consultations and diagnostic testing to determine whether homemonitoring, tracheostomy, or distraction osteogenesis would be implemented.In Part II, tracheostomy treatment was compared directly to distraction osteo-genesis.Results: In Part I (conventional versus decision tree model), taking into accountthe costs of the distraction, tracheostomy, hospital stay, diagnostic studies,physician fees, and emergency department visits, the total per patient treatmentcost was 1.5 greater in the conventional treatment group ($332,673) compared

with the decision tree model ($225,998) (p 0.05). In Part II (tracheostomy versus distraction osteogenesis), the total per-patient treatment cost in thetracheostomy group was two times greater than in the distraction group($382,246 versus $193,128) (p 0.05).

Conclusions: In treating newborns with micrognathia and upper airway ob-struction, a decision tree model with mandibular distraction decreases long-term health care costs compared with conventional treatment. Furthermore,

when comparing distraction to tracheostomy, similar decreases in long-termhealth care costs occurred. (Plast. Reconstr. Surg. 126: 1652, 2010.)

Neonatal upper airway obstruction requiresimmediate management. A secure airwaymust be established immediately after

birth, often with use of a nasopharyngeal tube orwith endotracheal intubation.1 Subsequent plansregarding long-term management must be madeexpeditiously to minimize the morbidity associ-

ated with prolonged intubation. This includes vo-cal cord injury, respiratory infections, and nutri-tional deprivation. Although a tracheostomy maybe considered definitive treatment, it is also asso-ciated with complications such as pneumonia, la-ryngomalacia, subglottic stenosis, and death.14

From the Division of Plastic and Reconstructive Surgery,University of California, Los Angeles School of Medicine.Received for publication January 6, 2010; accepted April 16,2010.The first two authors should be considered cofirst authors.Copyright 2010 by the American Society of Plastic Surgeons

DOI: 10.1097/PRS.0b013e3181ef8e82

Disclosure: The authors have no financial interests,including products, devices, or drugs, associated withthis article. There are no commercial associations thatmight pose or create a conflict of interest with informa-tion presented. All sources of funds supporting thecompletion of this article are under the auspices of theUniversity of California, Los Angeles.

www.PRSJournal.com1652


2/13

Furthermore, given that removal of tracheosto-mies placed for upper airway obstruction in thenewborn often require multiple procedures andare usually not accomplished until 5 to 11 years ofage, patients may have future difficulties withspeech, growth, and development.15

Distraction osteogenesis offers an accepted al-ternative to tracheostomy for newborns with upperairwayobstruction andPierreRobin sequence,Stick-ler syndrome, Treacher-Collins syndrome, Nagersyndrome, and other craniofacial deformities.2,615

Several series on newborns with micrognathia havedocumented lower jaw lengthening using distrac-tion to advance the tongue base and epiglottis andimprove upper airway obstruction.610 Rapid resultsachieved using distraction have the added benefit ofavoiding the many long-term postoperative compli-cations associated with tracheostomy.16 In the prop-erly selected patient, neonatal distraction has beenshown to improve clinical outcomes; however, thereare costs involved with this treatment, includingthose for preoperative studies, multiple operations,distraction devices, and postoperative care.16

Today, health care decision making takes intoaccount both clinical outcomes and cost-of-careanalysis. Considerations include both initial treat-ment cost and long-term care health care costs.Decision tree models are a way of providing op-timal care but may incur higher upfront costs withadditional consultations and tests. At our institu-tion, we have used a decision tree model at thetime of admission of newborns with micrognathiaand upper airway obstruction. We use a multidis-ciplinary team in evaluating each patient and in-terpreting findings of diagnostic testing dictatedby this decision tree model to properly select pa-tients for (1) home monitoring and positioning,(2) mandibular distraction osteogenesis, or (3)tracheostomy.16 Before the use of this decisiontree model, the decision to pursue tracheostomyversus other treatment strategies was made con-

ventionally on a case-by-case basis based on con-sultation advice from the subspecialists called bythe treating neonatologist.

A detailed cost comparison between tracheos-tomy and neonatal distraction as treatment mo-dalities for newborns with upper airway obstruc-tion has not yet been performed. The purpose ofour study was to evaluate the short and long-termcosts of treatment of neonatal upper airway ob-struction. In a two-part study, we used cost-matrixanalysis to compare (1) conventional treatmentversus a decision tree model and (2) mandibular

distraction versus tracheostomy.

PATIENTS AND METHODSNewborns with micrognathia and upper air-

way obstruction admitted to the neonatal or pe-diatric intensive care unit between January of 1994and December of 2005 were included in our study.Patients older than 2 months of age, those whohad already undergone tracheostomy placementat an outside hospital, and children with incom-plete records at final follow-up were excluded.

Part I: Conventional Treatment versus DecisionTree Model

For Part I, the conventional managementgroup was composed of those neonates born be-tween January of 1994 and December of 1999. Thiscohort received conventional case-by-case treat-ment with no distraction offered (see details ofmanagement below). The decision tree model

group consisted of patients born between Januaryof 2000 and December of 2005. The workup andmanagement were guided by the predeterminedalgorithm, with distraction available as a treatmentmodality (see details below). Comparative costanalysis was performed between these two groups.For completeness, patient demographics and pro-cedural complications are reported for the differ-ent groups.

Conventional Management GroupThe conventional management group was man-

aged on a case-by-case basis by the neonatologist or

pediatric intensivist. Consultants were requestedbased on their preference. Diagnostic and/or im-aging studies were ordered according to neonatol-ogist or consultant requests. For this group, man-agement typically consisted of tracheostomy orexpectant management with home monitoring. Al-though a nasopharyngeal airway was used at times,a tongue-lip adhesion or other surgical procedurewas not used. Decision for home monitoring, par-ticularly with Pierre Robin sequence patients, wasbased on physician comfort level with the infantsability to protect his or her airway by positional or

other means, not on relative size of the lower jaw.Part of the rationale for this care was that newbornswith Pierre Robin sequence were thought to exhibitcatch-up mandibular growth after birth, with sub-sequent improvement of airway obstruction.14 If theneonatologist or consultant felt that the upper air-way obstruction was significant enough to not advo-cate conservative measures such as positioning, tra-cheostomy was placed by the head and neck team.

Decision Tree Model Management(Study Group)

On admission of an infant with micrognathia

and upper airway obstruction to the pediatric or

Volume 126, Number 5 Neonatal Upper Airway Obstruction

1653


3/13

neonatal intensive care unit, pediatric specialists,including a plastic surgeon, head and neck sur-geon, anesthesiologist, geneticist, pulmonologist,and gastroenterologist, were consulted automati-cally (Fig. 1). Patients managed by our decisiontree model who were felt to have mild airway ob-struction were managed with home monitoring.Mild airway obstruction was defined as the abilityto tolerate side or oblique (on a unilateral bump)positioning with good oxygen saturation and with-out sternal retractions and signs of distress. (Notethat our study population included only infants

admitted to the intensive care unit and not themild cases evaluated in the office or craniofacialclinic.) A nasopharyngeal tube was placed in ap-proximately half of these infants with mild ob-struction for airway security before discharge. Thisnasopharyngeal airway was kept in place for sev-eral weeks to months and removed during a fol-low-up craniofacial team visit. Surgical tongue-lipadhesion was not part of our treatment algorithmbecause there has not been success with this tech-nique at our institution.

Newborns were considered to have moderate

to severe obstruction if they were intubated at

birth and failed extubation, or demonstrated im-mediate oxygen desaturation with retractions andsigns of distress even in a side or oblique position.Of these patients, all underwent, according to ourdecision tree model, preoperative sleep study (todistinguish between central and obstructive ap-nea), milk scan with occasional 24-hour pH probe(to assess the severity of gastroesophageal reflux),and laryngobronchoscopy (to rule out obstructiveanatomical lesions, other than obstruction fromthe posterior tongue base or epiglottis, such as atracheal web, vascular malformation, or severe tra-

cheomalacia). Severe gastroesophageal reflux dis-ease was considered both a positive milk scan or24-hour pH probe (more than four episodes perhour) and evidence of significantly inflamed/ul-cerating supraglottic mucosa on endoscopy. Thissevere inflammation was considered a secondpoint of airway obstruction, in addition to thecollapsed tongue base, and reason for a tracheos-tomy. The above tests were performed to assesswhich patients would benefit most from an inter-nal mandibular distraction procedure. Patientsmanaged with our decision tree model were ex-

cluded for mandibular distraction and offered a

Fig.1. Adecisiontreemodelwasusedinourstudygroupfornewbornsadmittedtotheneonatalorpediatricintensivecareunitwith

micrognathia and upper airway obstruction. A multidisciplinary pediatric team followed this algorithm to determine appropriate

treatment,including mandible distraction(green rectangles, diagnoses;pinkellipses, studies;orangesquares, diagnostic tests;purple

hexagons, treatments).

Plastic and Reconstructive Surgery November 2010

1654


4/13

tracheostomy if they had (1) central apnea, (2)severe gastroesophageal reflux disease, or (3)other airway lesions. The remaining patients wereconsidered appropriate candidates for mandibu-lar distraction or tracheostomy, and were offeredthe option of having either procedure performed.

Patients who were appropriate candidates for themandibular distraction procedure according tothe decision tree model outlined above had aninternal microdistractor device (KLS-Martin, Jack-sonville, Fla.) placed through intraoral and Ris-don incisions after an inverted-L osteotomy.

Activation was begun the next day, and dis-traction continued twice a day for the next 7 to 14days, for a total of 15 to 30 mm of advancement(1.8 mm/day). Patients were generally left intu-bated and sedated in the intensive care unit untildistraction was finished. Nasopharyngeal tube

feeding or total parenteral nutrition was used toprovide caloric intake during this time. Patientswere returned to the operating room for turningarm removal and extubation, and then sent homeonce oral feeding was adequate. After 3 months ofconsolidation, patients were returned to the op-erating room for removal of their mandibular dis-traction devices. Imaging involved preoperativeand follow-up lateral head radiographs and com-puted tomographic scans.

Part II: Tracheostomy versus Neonatal Distraction

For Part II, all neonates with micrognathia andupper airway obstruction from 1994 to 2004 whoreceived tracheostomy were compared with thosewho received distraction osteogenesis. This compar-ison was independent of the above group associa-tion. Comparative cost analysis was performed be-tween these tracheostomy and distraction groups.

Follow-Up CareFollow-up visits for all patients were conducted

at the craniofacial clinic at least every 6 months forthe study period. Patients had comprehensive

evaluations performed by pediatricians, geneti-cists, head and neck surgeons, plastic surgeons,pediatric dentists, orthodontists, speech patholo-gists, audiologists, and social workers.

Cost Outcomes AnalysisCost data analysis was performed with cost ma-

trices based on fees using past and current item-ized bills at the University of California, Los An-geles Medical Center. Length of hospital stay(intensive care unit and/or floor bed), preoper-ative evaluation, physician and health care provider consultations, imaging studies, operations,

equipment and device use, and emergency de-

partment visits were documented by a thoroughsearch of each patients medical chart(s) and elec-tronic record(s), with careful review of inpatientrecords, emergency department visits, routine pe-diatric care, and all follow-up visits with pediatri-cians, specialists, surgeons, and the entire cranio-facial team.

For surgical procedures, the total cost in-cluded operating room facility fees, anesthesiafees, equipment costs (e.g., implantable devices,fixation screws, tracheostomy supplies), itemizedcosts for routine preoperative and postoperativelaboratory tests, and medicines associated witheach patients operation. As mentioned previ-ously, infants who underwent neonatal distractiontypically had three procedures: (1) osteotomy anddevice placement, (2) removal of turning armsand extubation, and (3) removal of distractiondevices after consolidation. Tracheostomy pa-tients typically had just one procedure for trache-ostomy placement; however, it was not uncommonfor tracheostomy patients to undergo multiple op-erations for endoscopic evaluation of airway andsmall procedures (laser for granulation tissue re-duction). In some cases, additional procedureswere necessary for subsequent tracheostomy re-moval (laser, laryngotracheal reconstruction, tra-cheostomy site closure). Operating facility feeswere based on a combination of time units (every15 minutes) in the operating room and the com-plexity of the case.

Hospital stay costs were calculated using theaverage per diem current cost or cost at the timeof admission for the pediatric or neonatal inten-sive care unit ($10,500 to $13,000) or step-down orfloor bed ($5500 to $7000). The number of daysspent in the pediatric or neonatal intensive careunit was multiplied by the average cost of a pedi-atric intensive care unit stay. The number of daysspent in a floor bed was multiplied by the averagefloor bed cost. For the decision tree model pa-

tients, the number of consultations, imaging stud-ies, and their associated fees were similar for eachpatient. For the conventional treatment patients,the number and type of consultations and studiesvaried for each patient and so cost matrices wereused to calculate and average these varied costs.Reimbursed surgeons and consultant physicianfees were estimated from Medicare fee schedules.

The average costs associated with treatment ofcomplications or illnesses such as pneumonia forour study populations included fees for routinelaboratory tests; bacterial, viral, and fungal cul-

tures; airway management including nebulizer


1655


5/13


6/13

cent) undergo tracheostomy, incurring a per-pa-tient cost of $5735, significantly greater than thecostin the decision tree model group (p 0.05), with a3.1-fold increase in cost per patient. The remaining

patients who did not undergo a tracheostomy ordistraction procedure in both groups were senthome with positioning instructions and monitoring.Hospital stay costs on the day of surgical procedureswere not included here to avoid duplicating thesecosts, which were already accounted for in the hos-

pital stay cost category.Hospital stay costs excluded all of the fees de-

scribed above. They were adjusted to reflect an av-erage cost taking account the number of intensivecare unit and nonintensive care unit bed stays. Theaverage intensive care unit/nonintensive care unitdays was similar in both groups accounting for anaverage of 65or 67 percent of the hospital stay. Thus,the average daily cost was approximately $7428 perday. The average hospital stay per patient in thedecision tree model group was 18 days, incurring aper-patient average cost of $133,704. The average

hospital stay per patient in the Conventional groupwas 33 days, incurring a per-patient average costof $245,124. This cost was significantly higher inthe conventional group (p 0.05), with the con-ventional group incurring a 1.8-fold increase incost per patient.

Table 3. Cost-Related Aspects of theDecision Tree Model Compared with ConventionalMethods (Control Group)

Average

Cost

DTM

Group (%)

Average Costper Patientin the DTM

Group (n

78)

ConventionalControl

Group (%)

Average Cost perPatient in the

Control Group

(n

71)Distraction osteogenesis $55,010 43 (55) $30,326 0 (0) $0Tracheostomy $7540 19 (24) $1837 54 (76) $5735Hospital stay $7428 per day 18 days* $133,704 33 days* $245,124Diagnostic tests/studies $23,200 $29,500Physician fees $29,900 $31,100Pneumonia $10,416 3 (4) $401 17 (24) $2494Emergency room visits $1950 3.4 visits* $6630 9.6 visits* $18,720Average total cost per patient

in their respective group $225,998 $332,673

DTM, decision tree model.*Mean.

Table 4. Cost-Related Aspects of Distraction Osteogenesis Compared with Tracheostomy

AverageCost

DOGroup (%)

Average Cost per Patient

in the DOGroup (n 43) Tracheostomy (%)

Average Cost perPatient in theTracheostomy

Group (n 73)

Distraction osteogenesis $55,010 43 (100) $55,010 0 (0) $0Tracheostomy $7540 0 (0) $0 73 (100) $7540Hospital stay $7428 per day 12 days* $89,136 38 days* $282,264Diagnostic tests/studies $22,100 $31,600Physician fees $24,300 $36,800Pneumonia $10,416 1 (2) $242 25 (34) $3567Emergency room visits $1950 1.2 visits* $2340 10.5 visits* $20,475Average total cost per patient

in their respective group $193,128 $382,246

DO, distraction osteogenesis.*Mean.

Table 2. Treatment Decisions and Outcomes

DTMConventional

Treatment

Distraction osteogenesis 43 (55%) 0 (0%)Tracheostomy 19 (24%) 54 (76%)Home monitoring 16 (21%) 17 (24%)

Outcome of distractionosteogenesis 1 tracheostomy/1 redistraction Outcome of

tracheostomy 1 deathOutcome of home

monitoring2 tracheostomy,

2 distraction

Deaths 0* 3 (all airway)Pneumonia

(% of patients) 4%* 24%*GE reflux

(% of patients) 12%* 69%*Wound infections as

a result of distractionosteogenesis 4%

DTM, decision tree model; GE, gastroesophageal.*Significant difference between DTM and control groups (p 0.05).


1657


7/13

Costs for diagnostic studies excluded the fees

above and included all initial work-up tests, suchas computed tomographic scans, lateral head ra-diographs, 24-hour pH probe, sleep study, endos-copy, and subsequent studies during later admis-sions. The average per patient decision tree modelinitial study costs were higher initially at $21,600compared with the initial conventional costs of$15,300. However, because of subsequent testsfrom hospital visits during the 4-year study period,the overall cost in the conventional groups wasslightly higher (decision tree model, $23,200; con-

ventional, $29,500) but not significant.

Physician and health care provider costs ex-

cluded the costs above and included all consultingand surgeons fees. Although the physician feesduring the first few weeks were higher in thedecision tree model compared with the conven-tional group, the costs were similar by the endof the study period. For the decision tree modelpatients, initial physician fees were $28,200; forthe conventional treatment patients, physicianfees were $19,300. By the end of the study pe-riod, the decision tree model physician fees were$29,900 and the conventional treatment physi-

cian fees were $31,100.

Fig. 2. (Above, left) Frontal preoperative view of a newborn with micrognathia and upper airway obstruction

treated with distraction osteogenesis. (Above, right) Frontal postoperative view 2 months after the distraction

procedure. (Below, left) Right lateral preoperative view. (Below, right) Right lateral postoperative view.


1658


8/13

Costs associated with treatment of pneumoniaexcluded the fees described above and had a totalcost matrix of $10,416. Hospital stay costs were notincluded here to avoid duplicating these costs,which were already accounted for in the hospitalstay cost category. Three patients in the decisiontree model group (4 percent) suffered from pneu-monia, resulting in a per-patient cost of $401. Sev-

enteen patients in the control group (24 percent)

suffered from pneumonia, incurring a per-patientcost of $2494, a significantly higher per-patientcost than in the decision tree model group (p0.05) and a 6.2-fold increase.

Emergency department visit costs included allof the fees described above and totaled on average$1950 per visit. The decision tree model groupaveraged 3.4 visits per patient, incurring an aver-

age cost of $6630 per patient. The control group

Fig. 3. Bar graphs demonstrating costs associated with the respective groups. DO, distraction osteogenesis; Trach, tracheostomy;

PNA, pneumonia; ER, emergency room. (Above) Costs associated with decision tree model versus conventional treatment. (Below)

Costs associated with distraction osteogenesis versus tracheostomy.


1659


9/13

averaged 9.6 emergency department visits, incur-ring an average cost of $18,720 per patient. Thisis a significantly higher cost (p 0.05), with a2.8-fold increase.

The total per patient treatment cost in thedecision tree model group, taking into accountthe costs of the distraction osteogenesis, trache-ostomy, hospital stay, diagnostic studies, physi-

cian fees, treatment of pneumonia, and emer-

gency department visits, was $225,998. This issignificantly less than the total per-patient treat-ment cost of the control group $332,673 (p0.05), and a 1.5-fold increase.

Part II: Cost-Related Outcomes of DistractionOsteogenesis versus Tracheostomy

The 149 neonates were then studied indepen-

dent of group association. Specifically, we com-

Fig.4. Bargraphs demonstratingfold differencein costsassociated withthe respectivegroups.DO, distraction osteogenesis;Trach,

tracheostomy; PNA, pneumonia; ER, emergency room. (Above) Fold difference in costs associated with decision tree model versus

conventional treatment. (Below) Fold difference in costs associated with distraction osteogenesis versus tracheostomy.


1660


10/13

pared the costs associated with distraction osteogen-esis with the costs associated with tracheostomy formanagement of upper airway obstruction, as we didin Part I.

All 43 patients in the distraction osteogenesisgroup underwent the procedure, incurring a per-patient cost for the distraction procedures of$55,010. The tracheostomy group had no patientundergo distraction osteogenesis, thus incurring aper-patient cost of $0 for the distraction proce-dures, significantly less than the distraction osteo-

genesis group (p

0.05) (Table 2 and Figs. 3

through 5). As in Part I, hospital stay cost on theday of the surgical procedures was not includedhere to avoid duplicating these costs, which werealready accounted for in the hospital stay costcategory.

All 73 patients in the tracheostomy group un-derwent the procedure, incurring a per-patientcost of $7540. The distraction osteogenesis grouphad no patient undergo a tracheostomy proce-dure, thus incurring a per-patient cost of $0, sig-nificantly less than the tracheostomy group (p

0.05). Again, hospital stay cost on the day of the

Fig.5. Lineargraphsdemonstratingcostsassociatedwiththerespectivegroupsovertimefrominitialtreatment

to 48-month follow-up. (Above) Total costs over time for decision tree model versus conventional treatment.

(Below) Total costs over time for distraction osteogenesis versus tracheostomy.


1661


11/13

surgical procedures was not included here toavoid duplicating these costs, which were alreadyaccounted for in the hospital stay cost category.

The average hospital stay per patient in thedistraction osteogenesis group was 12 days, incur-ring a per-patient average cost of $89,136. The

average hospital stay per patient in the tracheos-tomy group was 38 days, incurring a per-patientaverage cost of $282,264. This cost was signifi-cantly higher in the tracheostomy group (p 0.05), with the tracheostomy group incurring a3.2-fold increase in cost per patient.

The total cost for diagnostic studies in thedistraction osteogenesis group ordered based onthe treatment algorithm and any follow-up testswas $22,100. The cost of studies obtained in thetracheostomy group over the 4-year study periodwas slightly higher at $31,600. However, this was

only a 1.4-fold increase. Physician consultationand surgical fees for the distraction osteogenesispatients were $24,300, whereas the physician feesfor the tracheostomy group were $36,800. This wasa 1.5-fold increase.

One patient in the distraction osteogenesisgroup (2 percent) suffered from pneumonia, re-sulting in a per-patient cost of $242. Twenty-fivepatients in the tracheostomy group (34 percent)suffered from pneumonia, incurring a per-patientcost of $3567, a significantly higher per patientcost than in the decision tree model group (p

0.05) and a 14.7-fold increase. The distractionosteogenesis group averaged 1.2 emergencyroom visits per patient, incurring an averagecost of $2340 per patient. The control groupaveraged 10.5 emergency department visits, in-curring an average cost of $20,475 per patient.This is a significantly higher cost (p 0.05), withan 8.8-fold increase.

The total per-patient treatment cost in the dis-traction osteogenesis group taking into accountthe costs of distraction osteogenesis, tracheos-tomy, hospital stay, diagnostic studies, physician

fees, treatment of pneumonia, and emergencydepartment visits was $193,128. This is signifi-cantly less than the total per-patient treatmentcost of the control group of $382,246 (p 0.05),and a 2-fold increase.

DISCUSSIONSevere micrognathia leading to upper airway ob-

struction in newborns requires organized and effi-cient management to prevent immediate and long-term morbidity and mortality.1,2 We instituted a

decision tree model at admission to our neonatal or

pediatric intensive care unit to allow for a systematicapproach for diagnosis and treatment of upper air-way obstruction. In a previous study, using this de-cision tree model with mandibular distraction, wedemonstrated better outcomes for neonatal upperairway obstruction patients compared with conven-tional treatment without distraction.16 However, theinitial diagnostic studies, subspecialty consultants,and distraction procedures may be costly. Becauseno large cost-comparative studies existed, in this cur-rent follow-up study, we looked at the cost of carebetween our decision tree model and conventionaltreatment, and between distraction and tracheos-tomy patients. Some alternative treatment optionssuch as tongue-lip adhesion and subperiosteal man-dibular stripping were notincluded in ouralgorithmbecause they are not performed at our institution. A

multi-institutional study would be necessary to in-clude these modalities in a cost analysis.Our data showed that despite the initial up-

front costs, the decision tree model comparedwith conventional treatment without distractionhad substantial economic advantages over a 4-yearperiod (1.5-fold less cost). In addition, the costover 4 years with the mandibular distraction pa-tients was 2.0times less than with the tracheostomypatients. The tracheostomy patients had costs in-volved with management of tracheostomy and itscomplications or sequelae, even years after the

initial tracheostomy procedure. By contrast, afterthe mandibular distraction procedure, subse-quent costs were relatively minimal. A smallerstudy that supports our findings showed that theinitial hospitalization costs for procedures andsupplies were nearly double for the tracheostomygroup (n 5) as compared with the distractionosteogenesis group (n 5).6,9 Our study confirmsthis trend with a substantially larger sample sizebut also assesses the cost-to-benefit ratio using adecision tree model.

The upfront expenses of distraction are attrib-utable to the cost of additional procedures and thecost of the implantable devices. One cost studylooked at older patients undergoing orthog-nathic surgery to correct malocclusion and com-pared mandibular distraction to sagittal splitosteotomy.17 For this group of patients, distrac-tion was found to be expensive and time con-suming. By contrast, our study showed that themandibular distraction procedure offered to ne-onates for micrognathia and upper airway ob-struction showed better cost effectiveness in the

long run compared with a tracheostomy.


1662


12/13

Our cost comparison study did not include thecost of outpatient care. Additional outpatient costsare known for tracheostomy patients.18,19 Theseoutpatient costs include items such as home airwaymonitoring systems, suctioningapparatus, and tra-cheostomy supplies (e.g., tracheostomy tubes, ties,and cleaning equipment). Home health care nurs-ing is sometimes needed, and 14 of our tracheos-tomy patients lived in specialized 24-hour nursingcare facilities at a cost of $600 per day for the entirestudy period. Most of these additional outpatientcosts were not necessary for distraction patients.Other costs for tracheostomy patients includedthose associated with the increased incidence ofgastroesophageal reflux disease, feeding diffi-culties, and speech abnormality. All of thesewere significantly higher in prevalence and withlonger duration in the tracheostomy patientscompared with the distraction patients. All ofthe variables studied demonstrated significantincreases in the conventional treatment and tra-cheostomy groups, compared with our decisiontree model group and distraction osteogenesisgroup, respectively.

Our decision tree model was designed to act asan efficient algorithm using multidisciplinary col-laboration to obtain the necessary tests and toarrive at an informed decision regarding long-term airway management of these affected new-borns with micrognathia and upper airway ob-struction. Even allowing for the possibility thatmore severe cases underwent tracheostomy (al-though the decision tree model and conventionalgroups were demographically similar), if given thechoice between mandibular distraction or trache-ostomy using the decision tree model for the less-severe cases, distraction has proven to be lesscostly in the long term and to have significantlyless associated morbidity/mortality and improvedaesthetic results. In addition, the use of an algo-rithm with multidisciplinary consultations may

have itself contributed to improved outcomes,fewer complications, and decreased costs.

SUMMARYAlthough tracheostomy might be the only mo-

dality appropriate for certain newborns with up-per airway obstruction, for others, mandibular dis-traction osteogenesis offers a more cost-effectivealternative with better outcomes. Our data showedthat newborns with micrognathia and upper air-way obstruction will have a 54 percent reduction

in costs over long-term follow-up when a decision

tree model with multidisciplinary specialists,timely studies, and an option for mandibular dis-traction osteogenesis is implemented. Not onlydid the decision tree model group a have signif-icant reduction in morbidity and mortality andsignificantly lower costs at follow-up, but when

selecting patients within the decision tree modelgroup who underwent distraction osteogenesisand comparing them directly to those who under-went tracheostomy, there was a similar benefit inmorbidity and cost savings.

James P. Bradley, M.D.Division of Plastic and Reconstructive Surgery

University of California, Los Angeles200 Medical Plaza, Suite 465

Los Angeles, Calif. [email protected]

PATIENT CONSENT

Parents or guardians provided written consent forthe use of the patients image.

ACKNOWLEDGMENTThe authors thank David McArthur, M.P.H., for

assistance in statistical analysis and revisions of this article.

REFERENCES

1. Liechty KW,Crombleholme TM.Management of fetal airwayobstruction. Semin Perinatol. 1999;23:496506.2. Izadi K, Yellon R, Mandell D, et al. Correction of upper

airway obstruction in the newborn with internal mandib-ular distraction osteogenesis. J Craniofac Surg. 2003;14:493499.

3. Conway WA, Victor LD, Magilligan DJ, Fujita S, Zorick FJ,Roth T. Adverse effects of tracheostomy for sleep apnea.

JAMA. 1981;246:347350.4. Sasaki CT, Masatoshi H, Koss N. Tracheostomy related sub-

glottic stenosis: Bacteriologic pathogenesis. Laryngoscope1979;89:857865.

5. Hill BP, Singer LT. Speech and language development afterinfant tracheostomy. J Speech Hear Disord. 1990;55:1520.

6. Judge B, Hamlar D, Rimell FL. Mandibular distraction os-

teogenesis in a neonate. Arch Otolaryngol Head Neck Surg.1999;125:10291032.

7. Denny AD, Kalantarian B. Mandibular distraction in neo-nates: A strategy to avoid tracheostomy. Plast Reconstr Surg.2002;109:896904.

8. Schaefer R,Stradler J, Gosain A.To distract or not to distract: An algorithm for airway management in isolated PierreRobin sequence. Plast Reconstr Surg. 2004;113:11131125.

9. Perlyn CA, Schmelzer RE, Sutera SP, Kane AA, Govier D,Marsh JL. Effect of distraction osteogenesis of the mandibleon upper airway volume and resistance in children withmicrognathia. Plast Reconstr Surg. 2002;109:18091818.

10. Denny AD, Talisman R, Hanson PR, Recinos RF. Mandibulardistraction osteogenesis in very young patients to correct

airway obstruction. Plast Reconstr Surg. 2001;108:302311.


1663


13/13

11. Fritz MA, Sidman JD. Distraction osteogenesis of the mandible.Curr Opin Otolaryngol Head Neck Surg. 2004;12:513518.

12. Gosain AK. Distraction osteogenesis of the craniofacial skel-eton. Plast Reconstr Surg. 2001;107:278280.

13. McCarthy JG, Stelnicki EJ, Mehrara BJ, Longaker MT. Dis-traction osteogenesis of the craniofacial skeleton. Plast Re-constr Surg. 2001;107:18121827.

14. Diner PA, Kollar EM, Martinez H, Vazquez MP. Intraoraldistraction for mandibular lengthening: A technical inno-vation. J Craniomaxillofac Surg. 1996;24:9295.

15. Greenwald JA, Luchs JS, Mehrara BJ, et al. Pumping the re-generate: An evaluation of oscillating distraction osteogenesisin the rodent mandible. Ann Plast Surg. 2000;44:516521.

16. Vyas R, Dipple K, Head C, et al. Management of neonatalupper airway obstruction: Decreased morbidity by utilizing adecision tree model with mandibular distraction osteogen-esis. J Neonatal Perinatal Med. 2008;1:2129.

17. van Strijen PJ, Perdijk FB, Breuning KH. Distraction osteo-genesis: A new development in craniofacial surgery (inDutch). Ned Tijdschr Tandheelkd. 1998;105:129131.

18. Newacheck PW, Inkelas M, Kim SE. Health services useand health careexpenditures for children withdisabilities.Pediatrics 2004;114:7985.

19. Harnick C, Bissell C, Parsons A. The impact of pediatrictracheotomy on parental caregiver burden and health status.Arch Otolaryngol Head Neck Surg. 2003;129:10651069.

Recent Supplements in Plastic and Reconstructive Surgery Body Contouring Surgery after Massive Weight Loss (January 2006)

Consensus Recommendations for Soft-Tissue Augmentation with Nonanimal Stabilized Hyaluronic Acid(Restylane) (March 2006)

Current Concepts in Wound Healing (June 2006)

Semipermanent and Permanent Dermal/Subdermal Fillers (September 1, 2006) Advances in Breast Augmentation (December 2006)

Temporary Dermal and Soft-Tissue Fillers (November 2007)

Craniofacial Trauma (December 2007)

Silicone Breast Implants: Outcomes and Safety (December 2007)

Advances in Facial Rejuvenation: Botulinum Toxin Type A, Hyaluronic Acid Dermal Fillers, andCombination TherapiesConsensus Recommendations (May 2008)

Advances in Pediatric Plastic Surgery (online, July 2009)

Evidence-Based Patient Safety Advisory for Ambulatory Surgery (October 2009)

Advances in Reconstructive Microsurgery (online, December 2009)


1664

economic advantages to a distraction decision tree-1.30

Documents