ORIGINAL ARTICLE
Patient-Based and Surgical Risk Factors for 30-Day
Postoperative Complications and Mortality
After Ankle Fracture Fixation
Philip J. Belmont, Jr, MD,* Shaunette Davey, DO,* Nicholas Rensing, MD,*
Julia O. Bader, PhD, Brian R. Waterman, MD,* and Justin D. Orr, MD*
Objective: The purpose was to calculate the incidence rates and
determine risk factors for 30-day postoperative mortality and morbid-
ity after ankle fracture open reduction and internal xation (ORIF).
Methods: The NSQIP database was queried to identify patients
undergoing ankle fracture ORIF from 2006 to 2011, with extraction
patient-based or surgical variables and a 30-day clinical course.
Multivariable logistic regression analysis identied signicant
predictors on outcome measures.
Results: Mean age was 50.3 (618.2) years while diabetes mellitus
(12.8%) and body mass index $40 kg/m2 (9.2%) were documented
from a total of 3328 patients identied. The 30-day mortality rate
was 0.30%, and complications occurred in 5.1%. Chronic obstruc-
tive pulmonary disease [odds ratio (OR): 4.23, 95% condence inter-
val (CI): 1.1915.06] and a nonindependent functional status before
surgery (OR: 2.25, 95% CI: 1.134.51) were the sole independent
predictors of mortality and major local complications, respectively.
Major local complications occurred in 2.2% of patients, and signif-
icant predictors were peripheral vascular disease (OR: 6.14; 95% CI:
1.9519.35), open wound (OR: 5.04; 95% CI: 2.2511.27), nonclean
wound classication (OR: 3.02; 95% CI: 1.316.93), and smoking
(OR: 2.85; 95% CI: 1.425.70). Independent predictors of hospital
Accepted for publication March 2, 2015.
From the *Department of Orthopaedic Surgery, William Beaumont Army
Medical Center, Texas Tech University Health Sciences Center, El Paso,
TX; and Department of Mathematical Sciences, Statistical Consulting
Laboratory, University of Texas at El Paso, El Paso, TX.
The authors report no conict of interest.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions
of this article on the journals Web site (www.jorthotrauma.com).
P. J. Belmont, Jr, S. Davey, N. Rensing, B. R. Waterman, J. D. Orr are
employees of the US Federal Government and the United States Army. The
opinions or assertions contained herein are the private views of the authors
and are not to be construed as ofcial or reecting the views of William
Beaumont Army Medical Center, the Department of Defense, or United
States government. The American College of Surgeons National Surgical
Quality Improvement Program (ACS NSQIP) and the hospitals participat-
ing in the ACS NSQIP are the source of the data used herein; they have not
veried and are not responsible for the statistical validity of the data analysis
or the conclusions derived by the authors.
Presented in part at the Annual Meeting of the Society Of Military
Orthopaedic Surgeons, December 2014, Scottsdale, AZ.
This study has approval from the William Beaumont Army Medical Center IRB.
Reprints: Philip J. Belmont, Jr, MD, Department of Orthopaedic Surgery,
William Beaumont Army Medical Center, 5005 N. Piedras, El Paso, TX
79920 (e-mail: philip.j.belmont.mil@mail.mil).
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
e476 | www.jorthotrauma.com
Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
stay .3 days were cardiac disease, age 70 years or older, open
wound, partially/totally dependent functional status, American Soci-
ety of Anesthesiologists (ASA) classication $3, body mass index
$40 kg/m2, bimalleolar or trimalleolar ankle fracture pattern, female
sex, and diabetes.
Conclusions: Chronic obstructive pulmonary disease increased the
risk of mortality after ankle fracture ORIF. Risk factors for
postoperative complications included peripheral vascular disease,
open wound, nonclean wound classication, age 70 years or older,
and ASA classication $3.
Key Words: ankle fracture, surgery, mortality, complications, risk
factors, incidence rate
Level of Evidence: Prognostic Level II. See Instructions for
Authors for a complete description of levels of evidence.
(J Orthop Trauma 2015;29:e476e482)
INTRODUCTION
Ankle fractures are a common musculoskeletal injury
with an annual incidence rate of 187 per 100,000 person-
years in the general US population1 and as high as 830 per
1000 Medicare beneciaries.2 Among elderly patients, ankle
fractures are the third most common fracture, surpassed only
by hip and wrist fractures.3 Within the elderly population, an
increased ankle fracture incidence rate has been observed in
Western Europe.4,5 Ankle fracture surgical patients have been
reported to have signicant improvements in their health sta-
tus69 and health-related quality of life7,8 for until 2 years after
surgery. However, ankle fracture surgery has also been
associated with noteworthy perioperative complications and
mortality. Contemporary healthcare has increased life expec-
tancy, particularly those patients with chronic medical
conditions, and this has contributed to a corresponding
increase in patients with signicant medical comorbidities
undergoing open reduction and internal xation (ORIF) for
unstable ankle fractures.7,10
Well-designed, adequately powered studies evaluating
the effect of patient-based and surgical characteristics on
ankle fracture xation mortality and morbidity are limited.
The majority of research detailing postoperative morbidity
have preferentially focused on wound complications after
ankle fracture xation and involve either single institutional
reports8,11 or narrow patient cohorts, such as the diabetic12 or
J Orthop Trauma  Volume 29, Number 12, December 2015
 J Orthop Trauma  Volume 29, Number 12, December 2015
elderly.7 A few investigations using state or national registries
have investigated postoperative mortality and complications
after ankle fracture xation.10,12,13 However, these broader
registry studies lack sufcient detailed quality data concern-
ing patient-based and surgical characteristics to have a robust
impact on outcomes analysis. The purpose of this study was
to assess the incidence rates and patient-based or surgical risk
factors for 30-day postoperative mortality, complications, and
length of hospital stay among a prospective patient cohort
derived from the National Surgical Quality Improvement Pro-
gram (NSQIP) with ankle fractures requiring ORIF.
MATERIALS AND METHODS
After institutional review board exemption, this inves-
tigation received approval from the NSQIP of the American
College of Surgeons and the 20062011 national dataset was
obtained. The civilian NSQIP methodology has been thor-
oughly described in previous publications,14,15 and it can also
be referenced in the programs participant user guide.16 For
enrolled orthopaedic surgical procedures, participating facili-
ties have risk-assessment nurses who undergo rigorous uni-
form training17 and prospectively collect patient-level data
during the designated surveillance window. Approximately
135 surgical variables are recorded, including preoperative
diagnosis, laboratory values, demographic information, med-
ical comorbidities, open versus closed ankle fracture, proce-
dure as dened by the Current Procedural Terminology (CPT)
code, operative characteristics, and mortality. All periopera-
tive complications and cases of mortality are monitored for 30
days postoperatively to include after the patient is discharged
from the surgical facility.18 Institutions contributing to the
multispecialty model are required to gather and submit at least
20% of their orthopaedic surgical caseload. Systematic sam-
pling limits the potential for selection bias, and the NSQIP
has been validated as a reliable and accurate prognostic tool.
Its methodology has been carefully vetted, with specic focus
on case collection and data management, resulting in an in-
terrater reliability agreement rate of 98.6%.19,20 The NSQIP
dataset has been shown to be a precise instrument for mod-
eling mortality and morbidity in general surgery17 and ortho-
paedic surgery.14,15 Secondary surgeries performed within the
30-day postoperative time frame are annotated accordingly
but are not separately counted to avoid duplicate entries.
A search of the NSQIP was performed to identify all
unique patients undergoing operative ankle fracture xation,
as designated by CPT codes 27766 (medial malleoli), 27792
(lateral malleoli), 27814 (bimalleoli), and 27822 or 27823
(trimalleoli). Demographic data and medical comorbidities
were exported, and surgical data such as type of anesthesia,
American Society of Anesthesiologists (ASA) classication,
open versus closed ankle fracture, and type of ankle fracture
type were recorded (Table 1). For this study, the types of
operative ankle fracture were organized as either unimalleolar
(CPT codes 27766 or 27792) or bimalleolar/trimalleolar
(27814/27822/27823) for subsequent analysis. Thirty-day
postoperative mortality, complications, and hospital length
of stay were the primary outcome measures. Systemic and
local complications were broadly classied as either major
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Postoperative Complications and Mortality
or minor within 2 separate groups in accordance with pre-
viously published methodologies (Tables 2 and 3). Major
systemic complications were dened as any complication
requiring complex medical intervention, whereas major local
complications encompassed all cases of deep wound infec-
tion, peripheral nerve injury, and/or surgical site reoperation.
Because there is no specic code for revision ankle fracture
procedures, subsequent cases of operation for ORIF identied
within the dataset were readmissions with a secondary pro-
cedure code consistent with ORIF.
Bivariate logistic regression analysis and chi-square
analysis were performed to assess the effect of itemized
patient- and surgery-based risk factors on the aforementioned
outcome measures. The patient-based variables encompassed
age (classied as ,60, 6069, and $70 years), sex, body
mass index (BMI) (#29.9, 30.039.9, and $40 kg/m2),
ASA classication (12 compared with 35), specied
medical comorbidities, functional status (independent com-
pared with partially/totally dependent on additional care-
givers), and wound classication (clean compared with
clean-contaminated/contaminated/dirty or infected). The sur-
gical variables included type of anesthesia (general compared
with spinal/epidural) and ankle fracture type (unimalleolar
compared with bimalleolar/trimalleolar). Patients were classi-
ed as having cardiac disease if they met one or more of the
following criteria: new diagnosis or exacerbation of preexist-
ing, chronic congestive heart failure within 30 days of index
surgery, history of angina within 30 days of the surgery,
history of myocardial infarction within 6 months of surgery,
and/or any percutaneous cardiac intervention or history of
other major cardiac surgery. Renal insufciency was classi-
ed as a preoperative serum creatinine $2.0 mg/dL. Factors
with P , 0.2 after initial univariate/chi-square analysis were
then inputted into multivariable logistic regression analysis to
objectively assess for the inuence of given risk factor on the
primary outcome measure. Odds ratios (ORs) and 95% con-
dence intervals (CIs) were reported during both bivariate
and multivariable logistic regression analysis. A P , 0.05
and 95% CI excluding 1.0 after multivariable testing were
necessary to recognize a signicant independent risk factor.
RESULTS
Between 2006 and 2011, a total of 3328 patients
undergoing operative ankle fracture xation were identied
in the NSQIP database. The mean age of the patient cohort
was 50.3 (618.2) years. The majority of patients were func-
tionally independent (88.5%), nonsmoking (75.7%), and
female (59.4%), and surgery more commonly entailed unim-
alleolar operative ankle fracture xation (38.3%) and general
anesthesia (84.7%). Notable medical comorbidities among the
patient cohort included diabetes (12.8%) and morbid obesity
(BMI $40 kg/m2) (9.2%) (Table 1). The average length of
stay for all patients was 2.2 days, and 81.9% of patients had
a length of stay #3 days.
In this 30-day period, a total of 229 complications
were noted among 170 patients (5.1%) (Tables 2 and 3).
There were 50 major systemic complications in 41 individ-
uals (1.2%) and 59 minor systemic complications among 55
www.jorthotrauma.com | e477
 Belmont et al J Orthop Trauma  Volume 29, Number 12, December 2015

TABLE 1. Patient Demographic and Preoperative TABLE 1. (Continued ) Patient Demographic and Preoperative
Characteristics Characteristics
Patients for Whom Patients for Whom
Characteristic Was Characteristic Was
Characteristic Value Determined (N) Characteristic Value Determined (N)
Age [average age (SD)], yrs 50.3 (18.2) 3328 Dirty or infected 30 (0.90)
#59, n (%) 2248 (67.6) Type of anesthesia 3328
6069, n (%) 561 (16.9) General 2818 (84.7)
$70, n (%) 519 (15.6) Spinal/epidural 448 (13.5)
Sex, n (%) 3321 Other 62 (1.9)
Male 1382 (41.6)
*EtOH more than 2 drinks per day in the 2 weeks before admission.
Female 1939 (59.4) CHF within 30 days before surgery/chronic CHF with new signs or symptoms in
BMI, mean 6 SD, n (%), 3104 30 days before surgery; history of myocardial infarction within the past 6 months before
kg/m2 surgery; history of percutaneous cardiac stent placement.
Noted preoperatively or intraoperatively.
#29.9 kg/m2 1731 (55.8) CHF, congestive heart failure.
30.039.9 1088 (35.1)
$40.0 285 (9.2)
CPT codes 3328
27766; 27792 (medial 1273 (38.3) patients (1.7%). Among major systemic complications, post-
malleoli; lateral operative sepsis/septic shock (34%) and pulmonary embolism
malleoli) (30%) were the most common causes. Among minor systemic
27814 (bimalleoli) 1213 (36.5) complications, urinary tract infection (49%) and deep venous
27822; 27823 (trimalleoli) 842 (25.3) thrombosis (31%) were the most common diagnoses. A total of
ASA classication 3321 72 major local complications occurred in 66 patients (2.2%)
12: no or mild 2371 (71.4) and 48 minor local in 45 individuals (1.4%). Reoperation
disturbance, n (%) (1.7%) was the most frequent major local complication,
35: severe or life- 950 (28.6) whereas supercial wound infection (1.0%) was the most
threatening disturbance,
n (%) common minor local complication. When compared with un-
Medical comorbidities, imalleolar xation, bimalleolar/trimalleolar fracture xation
n (%) was a signicant predictor for developing a major systemic
Diabetes 426 (12.8) 3328 complication (OR: 2.22, 95% CI: 1.064.67) in the univariate
Diabetes requiring insulin 202 (6.1) 3328 analysis, although this failed to achieve statistical signicance
Smoking (current smoker 807 (24.3) 3328 in the multivariate analysis.
within 1 year) The overall 30-day mortality rate was 0.30% (Table 4).
Regular alcohol use* 158 (5.8) 2742 Six of 10 patient deaths (60%) sustained a total of 10 com-
COPD 107 (3.2) 3328 plications other than mortality. The most frequent complica-
Cardiac issues (congestive 26 (0.78) 3328 tions among those who died were cardiac arrest requiring
heart failure/myocardial
infarction)
Hypertension 1194 (35.9) 3328
PVD 24 (0.72) 3328 TABLE 2. Total Number of Major/Minor Systemic
Renal insufciency 65 (3.0) 2138 Complications
(Cr $ 2.0) Characteristic N (%)
Dialysis 27 (0.81) 3328 Major systemic complications
Steroid use 59 (1.8) 3328 Pulmonary embolism 15 (0.45)
Open wound or wound 156 (4.7) 3328 Other systemic complication 9 (0.27)
infection
Postoperative sepsis 13 (0.39)
Functional status 3305
Septic shock 4 (0.12)
Independent 2924 (88.5)
Cerebrovascular accident 1 (0.03)
Partially dependent 366 (11.1)
Acute renal failure 2 (0.06)
Totally dependent 15 (0.45)
Cardiac arrest requiring CPR 4 (0.12)
Preoperative laboratory 873
values Myocardial infarction 2 (0.06)
Prealbumin #3.5 (g/dL) 254 (29.1) Minor systemic complications
Wound classication 3328 Urinary tract infection 29 (0.87)
Clean 3164 (95.1) Deep venous thrombosis 18 (0.54)
Clean-contaminated 69 (2.1) Pneumonia 9 (0.27)
Contaminated 65 (2.0) Renal insufciency 3 (0.09)
CPR, cardiopulmonary resuscitation.

e478 | www.jorthotrauma.com Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 J Orthop Trauma  Volume 29, Number 12, December 2015 Postoperative Complications and Mortality

TABLE 3. Total Number of Major/Minor Local Complications by Fracture Type

CPT 27814; 27822; 28823
CPT 27766; 27792 Unimalleolar, Bimalleolar/Trimalleolar, Odds Ratio
Characteristic Overall, N (%) N = 1273, n (%) N = 2055, n (%) (95% CI) P
Major local complications 66 (2.0) 24 (1.9) 42 (2.0) 1.09 (0.651.80) 0.75
Deep wound infection 16 (0.48) 4 (0.31) 12 (0.58) 0.54 (0.171.67) 0.28
Reoperation 56 (1.7) 21 (1.7) 35 (1.7) 1.03 (0.601.78) 0.90
Minor local complications 45 (1.4) 11 (0.86) 34 (1.7) 1.93 (0.973.82) 0.06
Supercial wound 34 (1.0) 10 (0.79) 24 (1.2) 1.49 (0.713.13) 0.29
infection
Wound dehiscence 14 (0.42) 1 (0.08) 13 (0.63) 8.09 (1.0661.90) 0.04

cardiopulmonary resuscitation, septic shock, and renal insuf-
ciency. A total of 106 patients (3.2%) died or experienced
a major complication in the rst 30 days after operative ankle
fracture xation.
Univariate analysis identied multiple risk factors for
mortality, any complication, major systemic complication,
minor systemic complication, major local complication,
minor local complication, and length of stay $3 days (see
Table, Supplemental Digital Content 1, http://links.lww.
com/BOT/A337, which demonstrates univariate analysis of
risk factors for complications). Signicant risk factors for
mortality, complications, and length of stay as determined
by multivariable logistic regression analysis are categorized
in Table 5. Multivariate regression logistic analysis deter-
mined that chronic obstructive pulmonary disease (COPD)
(OR: 4.23, 95% CI: 1.1915.06) was the sole independent
predictor of mortality. Signicant independent predictors of
developing any postoperative complication were peripheral
vascular disease (PVD) (OR: 3.67, 95% CI: 1.3010.39),
open wound or wound infection (OR: 2.73, 95% CI: 1.52
4.91), nonclean wound classication (OR: 2.34, 95% CI:
1.284.29), age 70 years or older (OR: 2.25, 95% CI: 1.38
3.67), and ASA classication $3 (OR: 1.64; 95% CI: 1.05
2.57). A nonindependent functional status before surgery
(OR: 2.25, 95% CI: 1.134.51) was the only signicant risk
factor for developing a major systemic complication. The
most important factors associated with major local complica-
tions, in the descending order of magnitude, were PVD, open
wound, nonclean wound classication, and smoking. Risk
factors with ORs exceeding 2.0 for increasing hospital length
of stay, in the descending order of magnitude, were cardiac
disease, age 70 years or older, open wound, and partially/
totally dependent functional status.
DISCUSSION
With the increasing prevalence of medical comorbid-
ities and prolonged life expectancy, the identication of
relevant patient-based and surgical variables is essential for
stratifying individual surgical outcomes.7,10 The prevalence of
ASA classication 35 (28.6%), age 70 years or older
(15.6%), diabetes (12.8%), and BMI $40 kg/m2 (9.2%) sub-
stantiates the cumulative medical concerns encountered
within 21st century patients. Osteoporosis, unlike distal
radius, hip and spine fractures, does not seem to be a signif-
icant risk factor for ankle fractures.21,22 However, obesity, as
dened by BMI, has been consistently reported as a risk fac-
tor for ankle fracture2124 and further displacement.25 The
prevalence of obesity (BMI $30 kg/m2) among the US adults
has increased 270% over the past 50 years to its current rate
of 36.1%, and morbid obesity (BMI $40 kg/m2) is estimated
at 6.6%.26 Similarly, diabetes is diagnosed in at least 11.3% of
individuals aged 20 years or older.27 With the existent trend
toward increasing comorbid disease burden, orthopaedic sur-
geons can anticipate a corresponding rise in the incidence of
patients with these complicating conditions that require oper-
ative xation of ankle fractures.23,24
The current investigation is the rst to conrm COPD
as an independent risk factor for early postoperative mortality
specically after ankle fracture xation (see Table, Supple-
mental Digital Content 1, http://links.lww.com/BOT/A337,
which demonstrates univariate analysis of risk factors for

TABLE 4. Thirty-Day Mortality, Any Complication, Major/Minor Systemic Complication, and Major/Minor Local Complication
Rates by Individual Patient (N) and Fracture Type
Value Unimalleolar Bimalleolar/Trimalleolar Odds Ratio
Characteristic Overall, N % N % N % (95% CI) P
Mortality 10 0.30 3 0.24 7 0.34 1.45 (0.375.60) 0.59
Any complication 170 5.1 54 4.2 116 5.6 1.35 (0.971.88) 0.07
Major systemic 41 1.2 9 0.71 32 1.6 2.22 (1.064.67) 0.0352
Minor systemic 55 1.7 18 1.4 37 1.8 1.28 (0.732.26) 0.40
Major local 66 2.0 24 1.9 42 2.0 1.09 (0.651.80) 0.75
Minor local 45 1.4 11 0.86 34 1.7 1.93 (0.973.82) 0.06
Mortality or major complication 106 3.2 31 2.4 75 3.7 1.52 (0.992.32) 0.06

Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved. www.jorthotrauma.com | e479

Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Belmont et al J Orthop Trauma  Volume 29, Number 12, December 2015

TABLE 5. Significant Risk Factors for Mortality, Any Complication, Major and Minor Systemic Complications, Major/Minor Local
Complications, and Length of Stay as Determined by Multivariate Logistic Regression Analysis
Dependent Variable Risk Factor P OR (95% CI)
Mortality COPD 0.0260 4.23 (1.1915.06)
Any complication PVD 0.0141 3.67 (1.3010.39)
Open wound or wound infection 0.0008 2.73 (1.524.91)
Wound classication 0.0060 2.34 (1.284.29)
Age $70 vs. #59 yrs 0.0069 2.25 (1.383.67)
ASA classication $3 0.0299 1.64 (1.052.57)
Major systemic complication Functional status 0.0218 2.25 (1.134.51)
Minor systemic complication Age $70 vs. #59 yrs 0.0250 2.70 (1.226.00)
Open wound or wound infection 0.0371 2.67 (1.066.73)
Male 0.0469 0.48 (0.230.99)
Major local complication PVD 0.0019 6.14 (1.9519.35)
Open wound or wound infection ,0.0001 5.04 (2.2511.27)
Wound classication 0.0093 3.02 (1.316.93)
Smoking 0.0031 2.85 (1.425.70)
Minor local complication Open wound or wound infection 0.0236 2.79 (1.156.76)
Hypertension 0.0079 2.67 (1.295.50)
Wound classication 0.0382 2.60 (1.056.42)
Length of stay Cardiac disease 0.0358 4.61 (1.1119.21)
Age $70 vs. #59 yrs ,0.0001 3.39 (2.434.73)
Open wound or wound infection ,0.0001 3.11 (1.825.32)
Functional status ,0.0001 2.03 (1.492.76)
ASA classication $3 ,0.0001 1.82 (1.392.38)
BMI $40.0 vs. #29.9 kg/m2 0.0038 1.67 (1.122.48)
Multimalleolar vs. unimalleolar 0.0003 1.60 (1.242.07)
Diabetes 0.0345 1.42 (1.031.97)
Male 0.0035 0.69 (0.530.88)

complications). The effects of COPD on postoperative out-
comes have not been extensively investigated. In a recent
report of over 465,000 surgical patients and over 24,000
orthopaedic patients, multivariate logistic regression model-
ing revealed that COPD was independently associated with an
increased risk of postoperative mortality (OR: 1.29; 95% CI:
1.191.39).28 The 0.30% 30-day mortality rate after ankle
fracture ORIF serves as a yardstick among a diverse national
patient sample within the United States. Using an integrated
state discharge database, the 90-day mortality rate after ankle
fracture xation in over 57,000 patients was 1.07%.13 Koval
et al10 reviewed the US Medicare database and found that
ankle fracture patients treated nonoperatively had a signi-
cantly higher mortality than those treated operatively at 6
months, 1 year, and 2 years, even after adjusting for the other
variables such as age and medical comorbidities. These data
not only underscore the relative merits of ankle fracture x-
ation, even in an aging demographic with comorbid disease.
The 30-day postoperative complication rate of 5.1% after
ankle fracture xation provides a benchmark using an
established complication classication system (Tables 2
and 3). This study identied PVD, open fracture, nonclean
wound, age 70 years or older, and ASA classication $3 as
signicant predictors for the development of postoperative
complications, which are novel contributions. Comparisons
with previous reports citing overall complication rates are
difcult given the variable denitions of major or minor
complications, length of follow-up, narrow demographic
focus,12,29,30 failure to account for systemic complications,7
and fundamental differences in incidence reporting, which
may fail to account for patients with multiple complications.
A report of 160,000 adult ankle fracture surgery patients eval-
uating the role of diabetes as a risk factor established a 3.3% in-
hospital complication rate; furthermore, multivariate analysis
found that diabetes and trimalleolar fractures were signicant
risk factors for complications.12 In contrast, in this study,
bimalleolar/trimalleolar fractures undergoing ankle fracture x-
ation had a trend toward increased complications (P = 0.07)
only in the univariate analysis.
Major systemic complications occurred in 1.2% of
ankle fracture xation patients, and its only signicant
predictor was partially dependent/totally dependent functional
status (Tables 2 and 5). There have been few ankle fracture
xation studies examining the rate of major complications,
and no known investigation has identied risk factors.29,31
Pulmonary embolism (0.45%) was the most common major
complication and occurred more frequently than the 0.32%
0.34% rate previously cited at a minimum of 90 days post-
operatively.13,32 Minor systemic complications occurred in
1.4% of ankle fracture xation patients; urinary tract infection
and deep venous thrombosis were the most common minor
complications.
Smoking, PVD, and open fracture were signicant risk
factors for major local complications according to multivariate

e480 | www.jorthotrauma.com Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 J Orthop Trauma  Volume 29, Number 12, December 2015
regression analysis (Table 5). Although studies reporting the
effects of smoking on local wound complications specically
after ankle fracture xation have been varied,11,31 a recent
investigation of over 390,000 surgical patients found that cur-
rent smokers had signicantly higher rates of surgical site
infection.33 Our study corroborates that smoking is a signicant
modiable risk factor for major local complications after ankle
fracture xation. Previous authors have studied the correlation
between PVD and complications after ankle fracture xa-
tion.11,13,30,31,34 SooHoo et al13 noted a 6.87% rate of wound
infection in 582 ankle fracture xation patients with PVD.
Although the study did not isolate the local surgical site
complications, it found a signicantly higher rate of short-
term complications in PVD patients.13 Shivarathre et al31 re-
ported that both smoking and PVD were independent risk
factors for wound infection in elderly ankle fracture xation
patients after univariate analysis. This study is the rst to iden-
tify PVD as a signicant independent predictor for developing
major wound complications (OR: 6.14, 95% CI: 1.9519.35)
in the general population while adjusting for covariates.
Finally, in accordance with previous reports, open fracture
was a signicant risk factor for both minor (OR: 2.79, 95%
CI: 1.156.76) and major (OR: 5.04, 95% CI: 2.2511.27)
local complications.1113,34
Considerable literature details the inherent local compli-
cations associated with diabetic ankle fracture xation.1113,31,34
Recent studies have focused on the effects of complicated (e.g.
end-organ effects) versus uncomplicated diabetes on ankle frac-
ture xation outcomes.12,34 A limitation of the NSQIP registry
is the inability to determine the severity of an individual
patients comorbidity (i.e., complicated vs. uncomplicated
diabetic patient). Wukich et al34 found that complicated
diabetic patients undergoing ankle fracture xation were
3.8 times more likely to experience a postoperative compli-
cation, and there was a similar trend toward increased wound
infections in these patients. Multivariate logistic regression
analysis in this study did not identify either diabetes or
insulin-dependent diabetes as a risk factor for mortality or
any postoperative complications other than increased length
of hospital stay (Table 5). In this study, PVD was also a risk
factor for major local complications. Wukich et al34 cautioned
that many diabetic patients may be unaware that they have
PVD. Notably, PVD, and not diabetes, was a signicant inde-
pendent risk factor for major local complications in this
study. We contend that peripheral vasculopathy, which rep-
resents a stage of complicated diabetes, may be the more
specic determinant of postoperative local complications
rather than diabetes alone.
In this study, length of hospital stay was signicantly
increased by several factors (Table 5). The average length of
stay in this study was 2.2 days, with only 18.1% of patients
staying .3 days. There is sparse literature regarding what
constitutes or predicts increased postoperative hospital stay
after ankle fracture xation. Davidovitch et al7 dened .4
days as increased hospital stay and found no trend toward
increased hospital stay in patients older than 60 years who
underwent ankle fracture xation. Conversely, this study
demonstrated that age $70 years signicantly increased
length of hospital stay (OR: 3.39, 95% CI: 2.434.73).
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Postoperative Complications and Mortality
Ganesh et al12 using multivariate analysis determined that
the presence of diabetes alone, regardless of fracture severity,
presence of dislocation, or open fracture wound, predicted
longer hospital stay and increased health care costs. In con-
trast, this study determined that diabetes along with both
bimalleolar/trimalleolar fracture patterns and open fractures
were signicant predictors of hospital stay .3 days. In addi-
tion to the aforementioned risk factors, this study is the rst to
report that cardiac disease, partially dependent/totally depen-
dent functional status, ASA classication $3, female sex,
and morbid obesity were independent predictors of hospital
stay .3 days after ankle fracture xation.
The present investigations strengths lie in the prospec-
tive data collection of the patient-based and surgical charac-
teristics of greater than 3300 patients who underwent ankle
fracture xation. The primary limitations of this study stem
from its reliance on data from a national patient registry. The
NSQIP uses a regimented oversight program to avoid under/
over-reporting of complications1720; however, patient comor-
bidity and complications data are contingent on accurate
input of data into the NSQIP database. Similarly, this dataset
does not include those deaths and complications occurring
outside the 30-day postoperative period. Perhaps most impor-
tantly, the authors are unable to determine the severity of
patients comorbidities. This may limit the external validity
of these data to the individual readers patient population. The
authors acknowledge this and offer that the large volume of
patients drawn from the NSQIP database reduces this poten-
tial limitation. Finally, the authors cannot draw conclusions
regarding functional outcomes after ankle fracture xation,
because the NSQIP database does not specically assess for
changes in patients functional status or use instrumented out-
come measures.
This study reects the most comprehensive study that
denes the risk factors for morbidity and mortality in a large
volume of patients undergoing ankle fracture xation.
COPD increased the risk of mortality after ankle fracture
xation. Predictive factors increasing the risk of postoper-
ative complications included PVD, open wound, nonclean
wound classication, age 70 years or older, and ASA
classication $3.
REFERENCES
1. Daly PJ, Fitxgeral RH Jr, Melton LJ, et al. Epidemiology of ankle frac-
tures in Rochester, Minnesota. Acta Orthop Scand. 1987;58:539544.
2. Koval KJ, Lurie J, Zhou W, et al. Ankle fractures in the elderly: what you
get depends on where you live and who you see. J Orthop Trauma. 2005;
19:635639.
3. Sporer SM, Weinstein JM, Koval KJ. The geographic incidence and
treatment variation among common fractures of elderly patients. J Am
Acad Orthop Surg. 2006;14:246255.
4. Kannus P, Parkkari J, Niemi S, et al. Epidemiology of osteoporotic
ankle fractures in elderly persons in Finland. Ann Intern Med. 1996;
125:975978.
5. Thur CK, Edgren G, Jansson KA, et al. Epidemiology of adult ankle frac-
tures in Sweden between 1987 and 2004. Acta Orthop. 2012;83:276281.
6. Bhandari M, Sprague S, Hanson B, et al. Health-related quality of life
following operative treatment of unstable ankle fractures: a prospective
observational study. J Orthop Trauma. 2004;18:338345.
7. Davidovitch RI, Walsh M, Spitzer A, et al. Functional outcomes after
operatively treated ankle fractures in the elderly. Foot Ankle Int. 2009;30:
728733.
www.jorthotrauma.com | e481
 Belmont et al
8. Egol KA, Tejwani NC, Walsh MG, et al. Predictors of short-term func-
tional outcome following ankle fracture surgery. J Bone Joint Surg Am.
2006;88:974979.
9. Obremskey WT, Dirschl DR, Crowther JD, et al. Change over time of the
SF-36 functional outcomes for operatively treated unstable ankle frac-
tures. J Ortho Trauma. 2002;16:3033.
10. Koval KJ, Zhou W, Sparks MJ, et al. Complications after ankle fracture
in elderly patients. Foot Ankle Int. 2007;28:12491255.
11. Miller AG, Margules A, Raikin SM. Risk factors for wound complications
after ankle fracture surgery. J Bone Joint Surg Am. 2012;94:20472052.
12. Ganesh SP, Pitrobon R, Ceclio WAC, et al. The impact of diabetes on
patient outcomes after ankle fracture. J Bone Joint Surg Am. 2005;87:
17121718.
13. SooHoo NF, Krenek L, Eagan MJ, et al. Complication rates following
open reduction and internal xation of ankle fractures. J Bone Joint Surg
Am. 2009;91:10421049.
14. Belmont PJ Jr, Goodman GP, Waterman BR, et al. Thirty-day postoper-
ative complications and mortality following total knee arthroplasty: inci-
dence and risk factors among a national sample of 15,321 patients.
J Bone Joint Surg Am. 2014;96:2026.
15. Schoenfeld AJ, Ochoa LM, Bader JO, et al. Risk factors for immediate
postoperative complications and mortality following spine surgery:
a study of 3475 patients from the National Surgical Quality Improvement
Program. J Bone Joint Surg Am. 2011;93:15771582.
16. American College of Surgeons National Surgical Quality Improvement
Program. User Guide for the 2011 Participant Use Data File. 2012.
Available at: http://site.acsnsqip.org/wp-content/uploads/2012. Accessed
March 4, 2014.
17. Procter LD, Davenport DL, Bernard AC, et al. General surgical operative
duration is associated with increased risk-adjusted infectious complication
rates and length of hospital stay. J Am Coll Surg. 2010;210:6065.e1-2.
18. Ingraham AM, Richards KE, Hall BL, et al. Quality improvement in
surgery: the american college of surgeons national surgical quality
improvement program approach. Adv Surg. 2010;44:251.
19. Dimick JB, Staiger DO, Hall BL, et al. Composite measures for proling
hospitals on surgical morbidity. Ann Surg. 2013;257:6772.
20. Shiloach M, Frencher SK, Steeger JE, et al. Toward robust information;
data quality and inter-rater reliability in the American College of Sur-
geons National Surgical Quality Improvement Program. J Am Coll Surg.
2010;210:616.
e482 | www.jorthotrauma.com
Copyright 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Orthop Trauma  Volume 29, Number 12, December 2015
21. Hasselman CT, Vogt MT, Stone KL, et al. Foot and ankle fractures in
elderly white women: incidence and risk factors. J Bone Joint Surg Am.
2003;85:820825.
22. Valtola A, Honkanen R, Krger Tuppurainen M, et al. Lifestyle and other
factors predict ankle fractures in perimenopausal women: a population-
based prospective cohort study. Bone. 2002;30:238242.
23. Bstman OM. Body mass index of patients with elbow and ankle frac-
tures requiring surgical treatment. J Trauma. 1994;37:6265.
24. Strauss EJ, Frank JB, Walsh M, et al. Does obesity inuence the outcome
after the operative treatment of ankle fractures? J Bone Joint Surg Br.
2007;89:794798.
25. Spaine LA, Bollen SR. The bigger they come: the relationship
between body mass index and severity of ankle fractures. Injury.
1996;27:687689.
26. National Center of Health Statistics Health E-Stats. Prevalence of over-
weight, obesity, and extreme obesity among adults: United States, trends
1960-1962 through 2009-2010. Available at: http://www.cdc.gov/nchs/
data/hestat/obesity_adult_09_10/obesity_adult_09_10.pdf. Accessed
May 2, 2014.
27. Center for Disease Control and Prevention. 2011 National Diabetes Fact
Sheet. Available at: http://www.cdc.gov.diabetes/pubs/factsheet11.htm.
Accessed May 2, 2014.
28. Gupta H, Ramanan B, Gupta PK, et al. Impact of COPD on postoperative
outcomes: results from a national database. Chest. 2013;143:15991606.
29. Anderson SA, Li X, Franklin P, et al. Ankle fractures in the elderly:
initial and long-term outcomes. Foot Ankle Int. 2008;29:11841188.
30. Zaghloul A, Haddad B, Barkseld R, et al. Early complications of sur-
gery in operative treatment of ankle fractures in those over 60: a review
of 186 cases. Injury. 2014;45:780783.
31. Shivarathre DG, Chandran P, Platt SR. Operative xation of unstable
ankle fractures in pateints aged over 80 years. Foot Ankle Int. 2011;32:
599602.
32. Pelet S, Roger ME, Belzile EL, et al. The incidence of thromboembolic
events in surgically treated ankle fracture. J Bone Joint Surg Am. 2012;
94:502506.
33. Hawn MT, Houston TK, Campagna EJ, et al. The attributable risk of
smoking on surgical complications. Ann Surg. 2011;254:914920.
34. Wukich DK, Joseph A, Ryan M, et al. Outcomes of ankle fractures in
patients with uncomplicated versus complicated diabetes. Foot Ankle Int.
2011;32:11841188.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.