Lower Complication Rate Following Ankle Fracture Fixation by Orthopaedic Surgeons Versus Podiatrists 2019;27: Copyright © the American Academy of Orthopaedic Surgeons

Home , Podiatrist

1,2 607 International , codes, and revision surgeries Among the Medicare patients, ankle fractures are the third most common juries, particularly in the elderly. on and débridement. Risk factors for retrospective cohort study — Level III Increased overlap in the scope of practice between Surgical treatment of ankle fractures by orthopaedic Using an insurance claims database, 11,745 patients who Overall, 11,115 patients were treated by orthopaedic urgical treatment oftures ankle has increased frac- as a result of surgeons was associated with lower rates of malunion/nonunion when compared with that by podiatrists. Theare reasons likely multifactorial for but these warrants differences further investigation.have important Our implications findings in patients who mustsurgically choose a manage surgeon their to ankle fracture, asdetermine well the as scope policymakers of who practice. orthopaedic surgeons and podiatrists has ledtreatment to of increased foot podiatric and ankle injuries.exists However, in the a literature paucity comparing of orthopaedic studies andfollowing podiatric ankle outcomes fracture fixation. Results: Discussion: Level of Evidence: Abstract Introduction: Methods: were identified by theComplications Current analyzed Procedural included Terminology malunion/nonunion, codes. vein infection, thrombosis, deep and rates of irrigati complications were compared using the Charlson Comorbidity Index. underwent ankle fracture fixation betweenretrospectively 2007 and evaluated. 2015 Patient were data wereprovider analyzed type. based Complications on were the identifiedClassification by of the Diseases, Ninth Revision surgeons and 630 patients were treated2015, by the podiatrists. percentage From of 2007 ankle to fracturespodiatrists surgically had treated increased, by whereas that treatedsurgeons by had orthopaedic decreased. Surgical treatment byassociated podiatrists with was higher malunion/nonunion rates amongankle all fractures. types No of differences in complicationspatients were with observed unimalleolar in fractures. In patientstrimalleolar with fractures, bimalleolar treatment or by a podiatristhigher was malunion/nonunion associated rates. with Patients treated bysurgeons orthopaedic versus podiatrists had similar comorbidity profiles. Research Article the increasing incidence of these in- S Lower Complication Rate Following Ankle Fracture Fixation by Orthopaedic Surgeons Versus Podiatrists 2019;27: Copyright © the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. Academy of Orthopaedic Surgeons. Correspondence to Dr. Chan: [email protected] J Am Acad Orthop Surg 607-612 DOI: 10.5435/JAAOS-D-18-00630 Copyright 2018 by the American From the Department ofSurgery, Orthopaedic Stanford Health Care, Stanford, CA. August 15, 2019, Vol 27, No 16 Jeremy N. Truntzer, MD Michael J. Gardner, MD Julius A. Bishop, MD Jeremy Y. Chan, MD

Downloaded from https://journals.lww.com/jaaos by k38zdtHxkv6xg2UYa2wsBqVVcyoCw+D3OHfTep5dUHN1DdlN7IOIME5DMgTrHulUgu30se/Xi/mVP1xJ+4f7LKVpRwtAn8DAzwNtKAqe1cq0R1T2zzZIi3472bylXFgmfO12LGB+GS0= on 01/12/2020 Downloaded from https://journals.lww.com/jaaos by k38zdtHxkv6xg2UYa2wsBqVVcyoCw+D3OHfTep5dUHN1DdlN7IOIME5DMgTrHulUgu30se/Xi/mVP1xJ+4f7LKVpRwtAn8DAzwNtKAqe1cq0R1T2zzZIi3472bylXFgmfO12LGB+GS0= on 01/12/2020 Ankle Fracture Outcomes by Surgeon Type extremity fracture, costing more ilar complication rates as compared Healthcare Provider Taxonomy than half a billion dollars per year.3,4 with podiatrists. Code Set for orthopaedic surgeons Although most ankle fracture care (207X0004X, 207XX0801X) and continues to be provided by ortho- podiatrists (213E00000X, 213EG0000X, paedic surgeons, the growing pres- Methods 213ES0103X, 213ES0131X). A ence and expanding scope of practice sample of our search algorithm for of podiatrists has resulted in changes A retrospective cohort study was unimalleolar fractures is illustrated in the management of foot and ankle conducted using data from all patients (Figure 1). injuries. For example, following the between the ages of 20 and 80 years Among the identified patients who introduction of podiatric staff priv- within the Humana subset of the underwent ankle fracture fixation, ileges at a level-I trauma center, the PearlDiver Patient Record Database subsequent complication rates were overall proportion of foot and ankle (Pearl-Diver Technologies) who identified using a combination of ICD-9 consults seen by podiatrists increased underwent ankle fracture fixation and CPT codes. Within 90 days of sixfold from 9% to 58% within 5 between 2007 and 2015. The research surgical treatment, the following com- years.5 Similarly, the proportion of and compliance office at our institu- plications were identified: new diag- foot and ankle injuries surgically tion deemed the study exempt from noses of infection (ICD-9 996.67, treated by podiatrists increased from human studies review because the 996.69, 998.51, 998.59), deep vein 8% to 41% in the same time frame.5 data extracted for this study was thrombosis (DVT) (ICD-9 453.40, Despite the greater involvement of from a publicly available source and 453.41 and 453.42), and revision sur- podiatrists in the surgical manage- all information received was de- gery for irrigation and débridement ment of foot and ankle injuries, identified. Patients who were diag- (CPT 20000, 20005, 10140, 10180, currently, no studies exist in the lit- nosed with an ankle fracture were 11042, 11043, 11044, 27603, 27607, erature that directly compare ortho- identified using codes from the Inter- 27610). These revision surgery codes paedic and podiatric outcomes national Classification of Diseases, were chosen as indicators of postop- following ankle fracture fixation. Ninth Revision (ICD-9), list for un- erative wound dehiscence or infection In addition, no published studies imalleolar (824.0, 824.1, 824.2, that could potentially be attributable to have compared longitudinal trends 824.3), bimalleolar (824.4, 824.5), the treatment provider. Within 1 year in the proportion of ankle fractures and trimalleolar (824.6, 824.7) frac- of surgical treatment, the following treated by orthopaedic surgeons tures. Inclusion criteria for this study complications were identified: new versus podiatrists. required patients to have undergone diagnoses of malunion or nonunion Therefore, this study evaluated subsequent surgical treatment within (ICD-9 733.81, 733.82). We did not short-term complication rates fol- 30 days of their primary ICD-9 code analyze complications such as cardiac lowing ankle fracture fixation based to limit the analysis to the treatment or respiratory events because of the on the provider type. We also sought of acute ankle fractures alone. Sur- likelihood of confounding factors that to identify changes in the proportion gical treatment was identified using were not attributable to the surgical of ankle fractures treated by ortho- the Current Procedural Terminology treatment directly. paedic surgeons versus podiatrists. (CPT) codes for unimalleolar (27766, Demographic data collected in- Our hypotheses were that (1) podia- 27792), bimalleolar (27814), and cluded sex, age, and Charlson Co- trists would have increasing involve- trimalleolar (27822, 27823) fracture morbidity Index (CCI). Statistical ment in surgically treated ankle fixation with or without concomi- analyses to compare demographic data fractures and (2) surgical treatment of tant syndesmotic fixation (27829). and complication rates were per- ankle fractures by orthopaedic sur- Patients were then separated based formed using chi-squared test for geons would be associated with sim- on their provider type using the proportions and two-sample t-test

Dr. Gardner or an immediate family member has received IP royalties from Synthes; is a member of a speakers’ bureau or has made paid presentations on behalf of KCI; serves as a paid consultant to Biocomposites, BoneSupport AB, Conventus, Globus Medical, KCI, Pacira Pharmaceuticals, SI-Bone, StabilizOrtho, and Synthes; has stock or stock options held in Conventus, Imagen Technologies; has received research or institutional support from Medtronic, SmartDevices, SMV Medical, Synthes, and Zimmer Biomet; and serves as a board member, owner, officer, or committee member of the American Orthopaedic Association, Orthopaedic Research Society, and Orthopaedic Trauma Association. Dr. Bishop or an immediate family member has received IP royalties from Globus Medical and Innomed; serves as a paid consultant to DePuy, Globus Medical, and KCI; has received research or institutional support from Conventus; and serves as a board member, owner, officer, or committee member of the Western Orthopaedic Association. Neither of the following authors nor any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Chan and Dr. Truntzer.

608 Journal of the American Academy of Orthopaedic Surgeons

Figure 1

Flow chart showing the data search algorithm used to identify complication and revision surgery rates following unimalleolar ankle fractures from the PearlDiver claims database. for means (Med Calc Software treated by orthopaedic surgeons and Version15.1).Withbothofthese Results 630 ankle fracture patients who were tests, P values less than or equal to treated by podiatrists. When com- 0.05 were considered statistically We identified a total of 11,115 ankle paring the overall demographics of significant. fracture patients who were surgically the two patient cohorts, no difference

August 15, 2019, Vol 27, No 16 609

Table 1 4.9%). The calculated relative risk for malunion/nonunion in patients Patient Demographics with a bimalleolar or trimalleolar Factor Ortho Podiatry P Value ankle fracture was 1.7 (95% confi- Patients 11115 630 dence interval, 1.2 to 2.4; P = 0.006) when surgical treatment was per- Age (%) formed by a podiatrist. ,60 yr 34.5 36.1 0.41 .60 yr 65.5 63.9 Sex (%) Discussion Male 32.5 35.6 0.11 Female 67.5 64.4 Although the amount of overlap in CCI (mean 6 SD) 2.7 6 3.4 2.8 6 3.3 0.47 the scope of practice between podiatrists and orthopaedic surgeons var- CCI = Charlson Comorbidity Index ies from state to state, surgical treatment of unstable ankle fractures is a common procedure performed by was found in sex or the proportion of interval, 1.2 to 2.1; P = 0.002) when both provider types. However, cur- patients who were younger than or treated by podiatrists as compared rently no studies exist in the literature older than 60 years (Table 1). In with orthopaedic surgeons. No sig- that directly compare patient out- addition, the mean CCI score was nificant differences were reported in comes following ankle fracture comparable between the orthopaedic the observed rate of infection, DVT, or treatment by orthopaedic surgeons cohort (mean, 2.7 6 3.4) and the irrigation and débridement (Table 2). and podiatrists. Our study evaluated podiatric cohort (mean, 2.8 6 3.3). Among the subgroup of patients short-term complication rates fol- From 2007 to 2015, the proportion with unimalleolar ankle fractures, lowing ankle fracture fixation by of ankle fractures treated by podia- a trend toward a higher rate of orthopaedic surgeons and podiatrists doubled from 3.5% to 7.0%. malunion/nonunion in ankle frac- trists, and revealed higher rates of The proportion of ankle fractures tures was observed in those treated malunion or nonunion following treated by orthopaedic surgeons by podiatrists compared with those podiatric surgery among all types of decreased over the period from treated by orthopaedic surgeons ankle fractures. This difference in 96.5% to 93.0%. Across all types of (6.2% versus 4.0%), although this malunion and nonunion rate was ankle fractures, a statistically signifi- was not statistically significant. driven primarily by higher complex- cantly higher rate of malunion/nonunion Among the subgroup of patients with ity bimalleolar and trimalleolar ankle was found in the patient cohort treated bimalleolar or trimalleolar ankle frac- fractures. No significant differences by podiatrists compared with the cohort tures, surgical treatment by podiatrists were found in the rates of postoper- treated by orthopaedic surgeons (7.3% was associated with a significantly ative infection, DVT, or irrigation versus 4.6%). The relative risk for higher rate of malunion/nonunion and débridement. In addition, our malunion/nonunion across all ankle compared with surgical treatment by study demonstrated a longitudinal fractures was 1.6 (95% confidence orthopaedic surgeons (8.2% versus trend toward increasing involvement

Table 2 Complication Rates by Provider Type Following Ankle Fracture Fixation All Ankle Fractures Single Malleolar Fractures Bimalleolar or Trimalleolar Fractures P P P Factor Orthopaedic Rate Podiatry Rate Value Orthopaedic Rate Podiatry Rate Value Orthopaedic Rate Podiatry Rate Value

Total patients 11,115 630 3,638 275 7,477 355 Malunion or 512 4.6% 46 7.3% 0.002 144 4.0% 17 6.2% 0.07 368 4.9% 29 8.2% 0.006 nonunion Infection 476 4.3% 21 3.3% 0.23 149 4.1% 8 2.9% 0.33 327 4.4% 13 3.7% 0.53 Irrigation and 347 3.1% 20 3.2% 0.89 110 3.0% 9 3.3% 0.78 237 3.2% 11 3.1% 0.92 débridement Deep vein 190 1.7% 14 2.2% 0.35 53 1.5% 7 2.5% 0.20 137 1.8% 7 2.0% 0.78 thrombosis

610 Journal of the American Academy of Orthopaedic Surgeons

Copyright © the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. Jeremy Y. Chan, MD, et al of podiatry in the surgical treatment observed difference in malunion and Orthopaedic surgeons are required to of ankle fractures. nonunion rates. In our study, a large complete 4 years of medical school, 5 The disparity in malunion and discrepancy was observed in the years of residency, and often another nonunion rates between podiatrists number of ankle fractures treated by year of subspecialized fellowship and orthopaedic surgeons is impor- orthopaedic surgeons versus podia- training. In contrast, podiatry train- tant because these complications trists. Previous studies have impli- ing was recently standardized in 2011 have the potential to impact patient cated surgical case volume as a factor to 4 years of podiatry school followed outcome. Fibular and medial mal- that can affect patient outcomes. In by a 3-year residency. Over the course leolar malunion resulting in tibiotalar lumbar spine surgery, surgeons and of residency, orthopaedic surgeons are malalignment has been shown to hospitals with surgical volume in the afforded a longer surgical training significantly alter tibiotalar contact top 25% of the National Inpatient time and extensive exposure to frac- pressures.6-10 Moreover, complica- Sample were found to have signifi- ture care in all areas of the body. This tions such as delayed union or non- cantly lower rates of mortality and increased exposure likely improves union following ankle fracture perioperative complications.15 In total competency and may help to optimize fixation has been associated with hip arthroplasty, patients treated patient outcome.18,19 Furthermore, a increased rates and decreased latency by surgeons who performed less previous study revealed shortcomings time to the development of post- than 35 cases per year were found to in general musculoskeletal knowledge traumatic ankle osteoarthritis.11 have higher rates of dislocations as among podiatric residents compared The underlying cause for the dif- well as revision surgeries.16 Simi- with orthopaedic residents,20 which ference in malunion and nonunion larly, in total ankle arthroplasty, may be relevant to patient outcomes rate remains unclear but is likely surgeons with case volume greater following ankle fracture surgery. The multifactorial. Risk factors for post- than the 90th percentile were found only other studies that have been operative complication rates follow- to have decreased rates of complica- published, which have directly com- ing ankle fracture fixation can tions and intraoperative fractures.17 pared surgical outcomes based on generally be categorized into patient- However, the association between surgeon specialty, are from the spine specific factors and surgeon-related case volume and complications has literature. These studies used patients variables. In regard to patient- not been demonstrated in studies on from the National Surgical Quality specific factors, no observed differ- ankle fractures to date.13,14 Although Improvement Program database and ences was observed in the age and sex the claims database that our data found no difference in postoperative proportions between the two cohorts were extracted from precluded an complication rates following certain that would account for the difference analysis of the average volume of procedures between neurosurgeons in malunion or nonunion. Patient ankle fractures treated by individual and orthopaedic surgeons.21-23 How- comorbidities, such as diabetes and surgeons, more than 90% of all sur- ever, the intensity and length of peripheral vascular disease, have gically treated ankle fractures identi- training between neurosurgery and previously been demonstrated to be a fied were treated by an orthopaedic orthopaedic surgery are more com- risk factor for short-term complica- surgeon. Notably, the ratio of ankle parable, which may minimize any tions following ankle fracture fixa- fractures treated by orthopaedic sur- observed differences in outcomes. tion.12-14 However, similar CCI geons compared with those treated Limitations of the present study scores were reported in our podiatric by podiatrists was even higher in include those associated with the use and orthopaedic cohorts, suggesting a the bimalleolar and trimalleolar of an administrative claims data set comparable level of comorbidities. subgroup than in the unimalleolar such as PearlDiver. Patients who re- Our study did not specifically exclude subgroup. The lower volume of located or changed insurance pro- patients with multiple injuries or open bimalleolar and trimalleolar ankle viders postoperatively would not be fractures, which could potentially fractures treated by podiatrists could captured by this study. Similarly, the affect the observed complication rates. certainly be a contributing factor to rate of complications such as infec- These surgeries are typically performed the increased rates of malunion and tion, malunion, and nonunion is by orthopaedic surgeons on an inpa- nonunion observed with these higher reliant on accurate coding and diag- tient basis though, which would be complexity patterns. nosis by the treatment providers. We more likely to bias complications Finally, significant differences exist did not have objective radiographic toward the orthopaedic cohort. in the training that orthopaedic sur- measures to corroborate the diagnoses Surgeon-related variables, such as geons and podiatrists receive, which of malunion or nonunion. In addi- surgical technique, case volume, and could potentially affect outcomes tion, we did not have access to training, may also contribute to the following ankle fracture fixation. patient-reported outcomes to assess

August 15, 2019, Vol 27, No 16 611

Copyright © the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited. Ankle Fracture Outcomes by Surgeon Type the clinical importance of the observed Finnish statistics in 1970-2006 and 13. Pincus D, Veljkovic A, Zochowski T, prediction for the future. Bone 2008;43: Mahomed N, Ogilivie-Harris D, complication rates. Furthermore, we 340-342. Wasserstein D: Rate of and risk factors for intentionally limited our analysis to intermediate-term reoperation after ankle 2. Thur CK, Edgren G, Jansson KA, short-term complications in an at- fracture fixation: A population-based Wretenberg P: Epidemiology of adult ankle cohort study. J Orthop Trauma 2017;31: tempt to only capture complications fractures in Sweden between 1987 and e315-e320. that could be directly related to the 2004: A population-based study of 91,410 Swedish inpatients. Acta Orthop 2012;83: 14. SooHoo NF, Krenek L, Eagan MJ, Gurbani treatment provider. As such, we were 276-281. B, Ko CY, Zingmond DS: Complication unable to assess long-term complica- rates following open reduction and internal 3. Belatti DA, Phisitkul P: Economic burden of fixation of ankle fractures. J Bone Joint tions such as the development of foot and ankle surgery in the US medicare Surg Am 2009;91:1042-1049. posttraumatic ankle arthritis or sub- population. Foot Ankle Int 2014;35: 334-340. 15. Farjoodi P, Skolasky RL, Riley LH: The sequent conversion to ankle arthro- effects of hospital and surgeon volume on desis or arthroplasty. Finally, our 4. Sporer SM, Weinstein JN, Koval KJ: The postoperative complications after geographic incidence and treatment study was designed only to identify the LumbarSpine surgery. Spine (Phila Pa variation of common fractures of elderly 1976) 2011;36:2069-2075. rates of complication and revision patients. J Am Acad Orthop Surg 2006;14: 246-255. 16. Ravi B, Jenkinson R, Austin PC, et al: surgery based on provider type rather Relation between surgeon volume and risk than the underlying cause for these 5. Jakoi AM, Old AB, O’Neill CA, et al: of complications after total hip differences. Further studies will need Influence of podiatry on orthopedic surgery arthroplasty: Propensity score matched at a level I trauma center. Orthopedics cohort study. BMJ 2014;348:g3284. to be performed to address the reasons 2014;37:e571-e575. behind the observed difference in 17. Basques BA, Bitterman A, Campbell KJ, 6. Bariteau JT, Hsu RY, Mor V, Lee Y, Haughom BD, Lin J, Lee S: Influence of complication rates. DiGiovanni CW, Hayda R: Operative surgeon volume on inpatient complications, In conclusion, our study found that versus nonoperative treatment of geriatric cost, and length of stay following total ankle fractures: A medicare part A claims ankle arthroplasty. Foot Ankle Int 2016;37: surgical treatment of ankle fractures by database analysis. Foot Ankle Int 2015;36: 1046-1051. podiatrists was associated with higher 648-655. 18. DeFroda SF, Gil JA, Blankenhorn BD, rates of malunion and nonunion com- 7. Curtis MJ, Michelson JD, Urquhart MW, Daniels AH: Variability in foot and ankle pared with treatment provided by Byank RP, Jinnah RH: Tibiotalar contact case volume in orthopaedic residency orthopaedic surgeons. This observa- and fibular malunion in ankle fractures: A training. Foot Ankle Spec 2017;10: cadaver study. Acta Orthop Scand 1992; 531-537. tion is particularly relevant given that 63:326-329. 19. Gil JA, Daniels AH, Weiss AP: Variability our study also identified an increasing 8. Lareau CR, Bariteau JT, Paller DJ, in surgical case volume of orthopaedic involvement of podiatrists in the sur- Koruprolu SC, DiGiovanni CW: surgery residents: 2007 to 2013. J Am Acad gical management of ankle fractures. Contribution of the medial malleolus to Orthop Surg 2016;24:207-212. tibiotalar joint contact characteristics. Foot Although the specific reasons for the Ankle Spec 2015;8:23-28. 20. Creech CL, Pettineo SJ, Meyr AJ: Podiatric difference in malunion and nonunion resident performance on a basic 9. Moody ML, Koeneman J, Hettinger E, competency examination in rate is likely multifactorial, our find- Karpman RR: The effects of fibular and musculoskeletal medicine. J Foot Ankle ings have important implications in talar displacement on joint contact areas Surg 2016;55:45-48. about the ankle. Orthop Rev 1992;21: patients who must choose a surgeon to 741-744. 21. Kim BD, Edelstein AI, Hsu WK, Lim S, Kim surgically manage their ankle fracture, JY: Spine surgeon specialty is not a risk 10. Thordarson DB, Motamed S, Hedman T, factor for 30-day complication rates in as well as policymakers who determine Ebramzadeh E, Bakshian S: The effect of single-level lumbar fusion: A propensity the scope of practice. fibular malreduction on contact pressures score-matched study of 2528 patients. in an ankle fracture malunion model. J Spine (Phila Pa 1976) 2014;39:E919- Bone Joint Surg Am 1997;79:1809-1815. E927.

References 11. Horisberger M, Valderrabano V, 22. Minhas SV, Chow I, Patel AA, Kim JY: Hintermann B: Posttraumatic ankle Surgeon specialty differences in single-level References printed in bold type are osteoarthritis after ankle-related fractures. J anterior cervical discectomy and fusion. Orthop Trauma 2009;23:60-67. Spine (Phila Pa 1976) 2014;39:1648-1655. those published within the past 5 years. 12. Koval KJ, Zhou W, Sparks MJ, Cantu RV, 23. Seicean A, Alan N, Seicean S, Neuhauser D, 1. Kannus P, Palvanen M, Niemi S, Parkkari J, Hecht P, Lurie J: Complications after ankle Benzel EC, Weil RJ: Surgeon specialty and Jarvinen M: Stabilizing incidence of low- fracture in elderly patients. Foot Ankle Int outcomes after elective spine surgery. Spine trauma ankle fractures in elderly people 2007;28:1249-1255. (Phila Pa 1976) 2014;39:1605-1613.

612 Journal of the American Academy of Orthopaedic Surgeons