- Email: [email protected]
pelvis fractures
G Model OTSR-1533; No. of Pages 4
ARTICLE IN PRESS Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original article
Predictors of 30-day mortality following hip/pelvis fractures A.C. Dodd , C. Bulka , A. Jahangir , H.R. Mir , W.T. Obremskey , M.K. Sethi ∗ The Vanderbilt Orthopaedic Institute Center for Health Policy, 1215 21st Avenue South, Suite 4200, Medical Center East, South Tower, Nashville, TN 37232, USA
a r t i c l e
i n f o
Article history: Received 18 November 2015 Accepted 30 May 2016 Keywords: 30-day mortality Complications Risk factors Hip fracture Pelvis fracture
a b s t r a c t Introduction: With the cost of healthcare in the United States reaching $2.9 trillion in 2013 and expected to increase with a growing geriatric population, the Center for Medicare and Medicaid Services (CMS) and Hospital Quality Alliance (HQA) began publicly reporting 30-day mortality rates so that hospitals and physicians may begin to confront clinical problems and promote high-quality and patient-centered care. Though the 30-day mortality is considered a highly effective tool in measuring hospital performance, little data actually exists that explores the rate and risk factors for trauma-related hip and pelvis fractures. Therefore, in this study, we sought to explore the risk factors associated with 30-day mortality in traumarelated hip and pelvic fractures. Materials and methods: Utilizing the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, 341,062 patients undergoing orthopaedic procedures from 2005 to 2013 were identified through a Current Procedural Terminology (CPT) code search. A second CPT code search identified 24,805 patients who sustained a hip/pelvis fracture. Patient demographics, preoperative comorbidities, operative characteristics and postoperative complications were collected and compared using Chi-squared test, Wilcoxon-Mann-Whitney test and multivariate logistic regression analysis. Results: Preoperative and postoperative risk factors for 30-day mortality following a hip/pelvis fracture were found: ASA classification, ascites, disseminated cancer, dyspnea, functional status, history of congestive heart failure (CHF), history of chronic obstructive pulmonary disease (COPD), a recent blood transfusion, and the postoperative complications: pneumonia, myocardial infarction, stroke, and septic shock. Discussion: Several preoperative patient risk factors and postoperative complications greatly increased the odds for patient mortality following 30-days after initial surgery. Orthopaedic surgeons can utilize these predictive risk factors to better improve patient care. Level of evidence: Retrospective study. Level IV. © 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction The cost of healthcare in the United States reached $2.9 trillion, or $9255 per person, in 2013 – equaling 17.4 percent of the Gross Domestic Product (GDP) [1]. As healthcare expenditures continue to rise, hospitals are seeking avenues to increase quality of care while decreasing costs [2]. In an effort to promote high-quality and patient-centered care, the Centers for Medicare & Medicaid Services (CMS) and Hospital Quality Alliance (HQA) recently began publicly reporting the 30-day mortality rates for patients with acute conditions [3]. By providing the 30-day mortality rates for hospitals, patients and physicians are better able to assess the essential clinical differences among hospitals [4]. Although 30-day mortality
∗ Corresponding author. E-mail address: [email protected] (M.K. Sethi).
is now being highly considered as an effective measure of hospital performance, minimal data exists exploring the rate and risk factors for orthopaedic trauma patients with a hip or pelvis fracture. In order to improve patient care and long-term outcomes, surgeons need to understand the risk factors that increase 30-day mortality in these patients. Previous studies investigating risk factors impacting mortality for orthopaedic trauma patients have: • focused on six-month and one year mortality for hip; • evaluated only specific fracture types, such as geriatric hip fractures; • included a limited number of risk factors [5–8]. For example, the study by Aharonoff et al. found that for geriatric hip fracture patients, age over 85 years, dependent functional status, history of cancer, ASA score of 3 or 4, and the development
http://dx.doi.org/10.1016/j.otsr.2016.05.016 1877-0568/© 2016 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model OTSR-1533; No. of Pages 4
ARTICLE IN PRESS A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
2
of one or more postoperative complications were significant risk factors for one year mortality [9]. Belmont et al. determined the inpatient mortality rate following a hip fracture to be 4.5% and that dialysis, cardiac disease, male sex, and injury severity score (ISS) were significant predictors of in-patient mortality; however, this study used a weighted sample and did not evaluate a 30-day mortality [6]. Khan et al., in 2014, determined that 30-day mortality for hip fractures was only related to abnormal creatinine, but this study had a relatively small cohort (n = 516) and did not evaluate preoperative comorbidities or postoperative complications as risk factors [10]. Our study aimed to identify the rate and risk factors impacting 30-day mortality for all orthopaedic trauma patients who had sustained a hip or pelvic fracture, allowing surgeons to better assess high-risk patients and ultimately improve patient outcomes and care. Our hypothesis here are that there are significant pre- and postoperative clinical risk factors associated with 30-day mortality following hip or pelvic fractures in the orthopaedic trauma population. 2. Patients and methods All procedures performed were in accordance with the ethical standards of the institution and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Utilizing the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, 341,062 patients undergoing orthopaedic procedures from 2005 to 2013 were identified through a Current Procedural Terminology (CPT) code search. A second CPT code search was conducted to identify all patients who underwent surgical fixation of a hip or pelvis fracture. Patients’ information was evaluated for complete data regarding 30-day mortality. A number of 24,805 hip/pelvis fracture patients were identified based on these criteria. Mortality for all patients was defined as death within 30-days following surgery. We investigated whether clinically appropriate patient factors had a significant impact on the rate of 30-day mortality. Patient demographics, preoperative, as well as operative characteristics were collected for each patient. Postoperative complications within 30-days following surgery were also assessed as
Table 1 Select patient characteristics associated with increased mortality following hip/pelvis fracture. Demographics Age, median* (IQR) ASA, median* (IQR) BMI, median* (IQR) Sex, N (%) Male Female *
All patients 83 (73–88) 3 (3–3) 24 (21–28) 7,534* (30.4) 17,271* (69.6)
Died (6.2%)
30-day survival (9.38%)
87 (81–91) 3 (3–4) 23 (20–27) 618* (40.2) 918* (59.8)
82 (73–88) 3 (3–3) 24 (21–28) 6,916* (29.7) 16,353* (70.3)
P < 0.05.
risk factors for mortality. Bivariate analyses using the chi-squared test and Wilcoxon-Mann-Whitney test were performed to compare risk factors between those who died and those who survived to postoperative day 30, as appropriate. Statistical significance was set at ␣ = 0.05. A multivariate logistic regression was then conducted to determine the risk factors that significantly predict 30-day mortality for orthopaedic trauma patients following a hip or pelvis fracture. All confounding variables, including patient demographics, preoperative comorbidities, operative characteristics and postoperative complications were included in the analysis. Patients with missing data were excluded in the analysis. 3. Results Patients with hip or pelvis injuries had a mortality rate of 6.2% (n = 1536; 6.2%). Table 1 lists mortality rates based on select patient demographics. The median age was 83 years (IQR: 73–88) for patients with hip and pelvis fractures, and patients were in the majority female (n = 17,271; IQR 69.6). The patients who died within 30-days were significantly older and had a lower median BMI than patients who survived. Of the patients who died, 59.8% were females. Table 2 provides the individual preoperative risk factors significantly associated with 30-day mortality. Based on the chi-squared analysis, patients with a history of ascites, bleeding disorder, disseminated cancer, dyspnea at rest, history of congestive heart failure (CHF), or a history of chronic obstructive pulmonary disease (COPD) were significantly more likely to die within 30-days following surgery (P < 0.001). Steroid use, weight loss greater than
Table 2 Preoperative risk factors for early mortality following hip/pelvis fracture. Risk factor
ASA class (median) Ascites, N (%) Bleeding disorder, N (%) Diabetic, N (%) Disseminated cancer, N (%) Dyspnea, N (%) None With moderate exertion At rest Functional status, N (%) Independent Partially dependent Totally dependent History of CHF, N (%) History of COPD, N (%) On dialysis, N (%) Use of steroids, N (%) Weight loss > 10% in last 6 months, N (%) Hypertension requiring medication, N (%) Recent blood transfusion, N (%)
Univariate analysis
Adjusted odds ratios
Died
30-day survival
P-value
OR (95% CI)
P-value
3 (3–4) 24 (1.6) 361 (23.5) 272 (6.2) 134 (8.7)
3 (3–3) 55 (0.2) 3,770 (16.2) 4,152 (17.8) 721 (3.1)
< 0.001 < 0.001 < 0.001 0.893 < 0.001 < 0.001
2.11 (1.91–2.34) 7.19 (4.18–12.36) 1.13 (0.99–1.300) 0.95 (0.82–1.11) 4.80 (3.82–6.04)
< 0.001 < 0.001 0.081 0.532 < 0.001
1295 (84.3) 168 (10.9) 73 (4.8)
21,365 (91.8) 1,591 (6.8) 313 (1.4)
1.00 (ref) 1.08 (0.89–1.32) 1.92 (1.42–2.61)
0.421 < 0.001
811 (52.8) 531 (34.6) 194 (12.6) 148 (9.6) 308 (20.1) 64 (4.2) 124 (8.1) 49 (3.2) 1,097 (71.4) 154 (10.0)
16,935 (72.8) 5,261 (22.6) 1,073 (4.6) 678 (2.9) 2,557 (11.0) 434 (1.9) 1,315 (5.7) 347 (1.5) 15,811 (68.0) 1,203 (5.2)
1.00 (ref) 1.58 (1.40–1.79) 2.81 (2.34–3.38) 1.72 (1.39–2.12) 1.34 (1.14–1.58) 1.82 (1.34–2.47) 1.07 (0.86–1.33) 1.26 (0.89–1.77) 0.83 (0.73–0.95) 1.36 (1.12–1.65)
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.547 0.187 0.005 0.002
< 0.001
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.005 < 0.001
CHF: congestive heart failure; COPD: chronic obstructive pulmonary disease.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model
ARTICLE IN PRESS
OTSR-1533; No. of Pages 4
A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
3
Table 3 Postoperative risk factors for early mortality following hip/pelvis fracture. Risk factor
Pneumonia, N (%) Urinary tract infection, N (%) Deep wound infection, N (%) Myocardial infarction, N (%) Pulmonary embolism, N (%) Deep vein thrombosis, N (%) Stroke, N (%) Sepsis, N (%) Septic shock, N (%)
Univariate analysis
Adjusted odds ratios
Died
30-day survival
P-value
OR (95% CI)
P-value
246 (16.0) 131 (8.5) 8 (0.5) 111 (7.2) 21 (1.4) 19 (1.2) 55 (3.6) 75 (4.9) 246 (16.0)
629 (2.7) 1303 (5.6) 103 (0.4) 255 (1.1) 148 (0.6) 264 (1.1) 91 (0.4) 324 (1.4) 629 (2.7)
< 0.001 < 0.001 0.657 < 0.001 0.001 0.714 < 0.001 < 0.001 < 0.001
3.36 (2.79–4.05) 1.06 (0.85–1.31) 1.10 (0.47–2.54) 4.42 (3.40–5.73) 2.22 (0.53–9.28) 1.25 (0.51–3.06) 6.78 (2.35–19.51) 1.53 (0.72–3.25) 9.66 (4.75–19.67)
< 0.001 0.622 0.827 < 0.001 0.275 0.623 < 0.001 0.273 < 0.001
10% within the last six months, dialysis, dependent functional status, or a recent blood transfusion was also significantly associated with mortality following hip/pelvis fractures (P < 0.001). After controlling for all individual patient comorbidities with a multivariable regression, it was determined that each one-unit increase in ASA score doubles the likelihood of death within 30-days (OR: 2.11; 95% CI: 1.91–2.34; P < 0.001), and a person with a history of ascites was also 7.19 times more at risk for 30-day mortality (95% CI: 4.18–12.36, P < 0.001). Other significant risk factors for 30-day mortality following multivariate analysis included: history of disseminated cancer and totally dependent functional status, which increased the likelihood of death 4.80 times (95%: 3.82–6.04, P < 0.001) and 2.81 times (2.34–3.38, P < 0.001), respectfully. To understand the risk factors for 30-day mortality, postoperative events were also evaluated. Table 3 summarizes the postoperative risk factors for death after controlling for all confounding variables. Pneumonia (OR: 3.36; 95% CI [2.79–4.05], P < 0.001), myocardial infarction (MI) (OR: 4.42, 95% CI [3.40–5.73], P < 0.001), and stroke (OR 6.78, 95% CI [2.35–19.51], P < 0.001) were all significant postoperative risk factors for 30-day mortality. A patient who developed septic shock was 9.66 times (95%: 4.75–19.67; P < 0.001) more likely to die in 30-days compared to a patient who did not develop the complication, representing the largest postoperative risk for mortality following hip/pelvis fractures. 4. Discussion Our study is the first to examine the pre- and postoperative risk factors associated with 30-day mortality for orthopaedic trauma patients following a hip or pelvis fracture. While the overall rate of 30-day mortality was low for hip/pelvis fractures, there are significant factors, such as history of congestive heart failure and postoperative pneumonia that drive mortality within the first thirty days postsurgery. By identifying significant risk factors associated with 30-day mortality, orthopaedic surgeons will be better equipped to assess a patient’s risk for death prior to and following surgery. Previous studies have found varying rates of mortality for orthopaedic patients depending on location of injury and length of follow-up[11,12]. For orthopaedic trauma patients with hip and pelvis fractures within our study, we determined the 30-day mortality to be 6.2%, representing the lower range of mortality. Most studies, however, have focused mainly on geriatric patients within all orthopaedic subspecialties rather than on all orthopaedic trauma patients [9,13]. Older age significantly increased the risk for premature death following hip/pelvis fractures. Several studies evaluating patients undergoing hip fracture surgery determined that advanced age was a significant predictor of higher mortality [14–20]. However, a study by Streubel et al., did not find an effect of age on survival for distal femur fractures. Our study additionally showed that
preoperative ascites was a significant predictor of 30-day mortality. History of ascites increased the risk for 30-day mortality by 7.19 times. To our knowledge, previous studies have not included preoperative ascites as a risk factor for mortality, even though it has been noted in a study by Khan et al. that mortality in patients with ascites is high [21]. Our study also showed that a history of COPD and dialysis were significant predictors for 30-day mortality. One study found that COPD increased the risk of acute in-patient mortality for hip fractures 1.87 times (95%: 1.01–1.92) compared to a patient without COPD, which is slightly higher than our finding (OR: 1.34, 95%: 1.14–1.58) [22]. Postoperative risk factors were also clear predictors of 30-day mortality. Pneumonia increased the risk of 30-day mortality 3.36 times. Other studies have shown that postoperative pneumonia drastically increases the risk for inpatient mortality [22–24]. It is essential that surgeons are aware of the close association between mortality and pneumonia even within 30-days following surgery. Myocardial infarction and septic shock were also significantly associated with increased 30-day mortality after controlling for all confounding variables. There were some limitations in this study. First, the ACS-NSQIP database limits the risk factors available to be evaluated. Postoperative complications, such as osteomyelitis, could have been significant predictors of 30-day mortality. Due to limitations of the ACS-NSQIP database, the time before surgery was also not included in the analysis, which has been highly associated with increased early mortality for orthopaedic patients [25]. Nevertheless, our study helps identify several clinically relevant risk factors for 30-day mortality for orthopaedic trauma patients who sustain a hip or pelvis fracture. As our healthcare systems shifts towards containing costs while improving quality care, orthopaedists must better understand the risk factors for 30-day mortality in order to reduce costs while increasing patient quality care. Disclosure of interest Author WTO has previously consulted for Biometrics; done expert testimony in legal matters, and has a grant from the Department of Defense. ACD, CB, AJ, HM, and MKS declare that they have no competing interest. References [1] Centers for Medicare & Medicaid Services. National health expenditures 2013 highlights; 2013 [http://www.cms.gov/Research-StatisticsData-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/ downloads/highlights.pdf. Accessed 2 February 2015]. [2] Bundled payment: AHA research synthesis report. American Hospital Association Committee on Research. http://www.aha.org/research/cor/content/ BundledPayment.pdf. Accessed 2 February 2015. [3] Outcome Measures. Centers for Medicare & Medicaid Services. http://www. cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/ HospitalQualityInits/OutcomeMeasures.html. Accessed 2 February 2015.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model OTSR-1533; No. of Pages 4 4
ARTICLE IN PRESS A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
[4] Krumholz HM, Normand SLT. Public reporting of 30-day mortality for patients hospitalized with acute myocardial infarction and heart failure. Circulation 2008;13:1394–7. [5] Hietala P, Strandberg M, Kiviniemi T, et al. Usefulness of troponin T to predict short-term and long-term mortality in patients after hip fracture. Am J Cardiol 2014;114:193–7. [6] Belmont Jr PJ, Garcia EJ, et al. Risk factors for complications and in-hospital mortality following hip fractures: a study using the National Trauma Data Bank. Arch Orthop Trauma Surg 2014;134:597–604. [7] Clement ND, Duckworth AD, McQueen MM, et al. The outcome of proximal humeral fractures in the elderly predictors of mortality and function. Bone Joint J 2014;96:970–7. [8] Copeland CE, Mitchell KA, Brumback RJ, et al. Mortality in patients with bilateral femoral fractures. J Orthop Trauma 1998;12:315–9. [9] Aharonoff GB, Koval KJ, Skovron ML, et al. Hip fractures in elderly: predictors of one-year mortality. J Orthop Trauma 1997;11:162–5. [10] Khan SK, Shirley M, Glennie C, et al. Achieving best practice tariff may not reflect improved survival after hip fracture treatment. Clin Interv Aging 2014;9:2097–102. [11] Richmond J, Aharonoff, Zuckerman JD, et al. Mortality risk after hip fracture. J Orthop Trauma 2003;17:53–6. [12] Hu F, Jiang C, Shen J, et al. Preoperative predictors for mortality following hip fracture surgery: a systematic review and meta-analysis. Injury 2012;43:676–85. [13] White BL, Fisher WD, Laurin CA. Rate of mortality for elderly patients after fracture of the hip in the 1980s. J Bone Joint Surg Am 1987;69:1335–40. [14] Elliott J, Beringer T, Kee F, et al. Predicting survival after treatment for fracture of the proximal femur and the effect of delays to surgery. J Clin Epidemiol 2003;56:788–95.
[15] Calder SJ, Anderson GH, Jagger C, et al. Unipolar or bipolar prosthesis for displaced intracapsular hip fracture in octogenarians: a randomised prospective study. J Bone Joint Surg Br 1996;78:391–4. [16] Hommel A, Ulander K, Bjorkelund KB, et al. Influence of optimised treatment of people with hip fracture on time to operation, length of hospital stay, reoperations and mortality within 1 year. Injury 2008;39:1164–74. [17] Oztürk A, Ozkan Y, Akgöz S, et al. The risk factors for mortality in elderly patients with hip fractures: postoperative one-year results. Singapore Med J 2010;51:137–43. [18] Paksima N, Koval KJ, Aharanoff G, et al. Predictors of mortality after hip fracture: a 10-year prospective study. Bull NYU Hosp Jt Dis 2008;66:111–7. [19] Rosencher N, Vielpeau C, Emmerich J, et al. Venous thromboembolism and mortality after hip fracture surgery: the ESCORTE study. J Thromb Haemost 2005;3:2006–14. [20] Trombetti A, Herrmann F, Hoffmeyer P, et al. Survival and potential years of life lost after hip fracture in men and age-matched women. Osteoporos Int 2002;13:731–7. [21] Khan J, Pikkarainen P, Karvonen AL, et al. Ascites: aetiology, mortality and the prevalence of spontaneous bacterial peritonitis. Scand J Gastroenterol 2009;44:970–4. [22] Myers AH, Robinson EG, Van Natta ML, et al. Hip fractures among the elderly: factors associated with in-hospital mortality. Am J Epidemiol 1991;134:1128–37. [23] Streubel PN, Ricci WM, Wong A, et al. Mortality after distal femur fractures in elderly patients. Clin Orthop Relat Res 2011;469:1188–96. [24] Bhattacharyya T, Iorio R, Healy WL. Rate of and risk factors for acute inpatient mortality after orthopaedic surgery. J Bone Joint Surg Br 2002;84:562–72. [25] Moran CG, Wenn RT, Sikand M, Taylor AM. Early mortality after hip fracture: is delay before surgery important? J Bone Joint Surg 2005;87:483–9.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
ARTICLE IN PRESS Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original article
Predictors of 30-day mortality following hip/pelvis fractures A.C. Dodd , C. Bulka , A. Jahangir , H.R. Mir , W.T. Obremskey , M.K. Sethi ∗ The Vanderbilt Orthopaedic Institute Center for Health Policy, 1215 21st Avenue South, Suite 4200, Medical Center East, South Tower, Nashville, TN 37232, USA
a r t i c l e
i n f o
Article history: Received 18 November 2015 Accepted 30 May 2016 Keywords: 30-day mortality Complications Risk factors Hip fracture Pelvis fracture
a b s t r a c t Introduction: With the cost of healthcare in the United States reaching $2.9 trillion in 2013 and expected to increase with a growing geriatric population, the Center for Medicare and Medicaid Services (CMS) and Hospital Quality Alliance (HQA) began publicly reporting 30-day mortality rates so that hospitals and physicians may begin to confront clinical problems and promote high-quality and patient-centered care. Though the 30-day mortality is considered a highly effective tool in measuring hospital performance, little data actually exists that explores the rate and risk factors for trauma-related hip and pelvis fractures. Therefore, in this study, we sought to explore the risk factors associated with 30-day mortality in traumarelated hip and pelvic fractures. Materials and methods: Utilizing the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, 341,062 patients undergoing orthopaedic procedures from 2005 to 2013 were identified through a Current Procedural Terminology (CPT) code search. A second CPT code search identified 24,805 patients who sustained a hip/pelvis fracture. Patient demographics, preoperative comorbidities, operative characteristics and postoperative complications were collected and compared using Chi-squared test, Wilcoxon-Mann-Whitney test and multivariate logistic regression analysis. Results: Preoperative and postoperative risk factors for 30-day mortality following a hip/pelvis fracture were found: ASA classification, ascites, disseminated cancer, dyspnea, functional status, history of congestive heart failure (CHF), history of chronic obstructive pulmonary disease (COPD), a recent blood transfusion, and the postoperative complications: pneumonia, myocardial infarction, stroke, and septic shock. Discussion: Several preoperative patient risk factors and postoperative complications greatly increased the odds for patient mortality following 30-days after initial surgery. Orthopaedic surgeons can utilize these predictive risk factors to better improve patient care. Level of evidence: Retrospective study. Level IV. © 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction The cost of healthcare in the United States reached $2.9 trillion, or $9255 per person, in 2013 – equaling 17.4 percent of the Gross Domestic Product (GDP) [1]. As healthcare expenditures continue to rise, hospitals are seeking avenues to increase quality of care while decreasing costs [2]. In an effort to promote high-quality and patient-centered care, the Centers for Medicare & Medicaid Services (CMS) and Hospital Quality Alliance (HQA) recently began publicly reporting the 30-day mortality rates for patients with acute conditions [3]. By providing the 30-day mortality rates for hospitals, patients and physicians are better able to assess the essential clinical differences among hospitals [4]. Although 30-day mortality
∗ Corresponding author. E-mail address: [email protected] (M.K. Sethi).
is now being highly considered as an effective measure of hospital performance, minimal data exists exploring the rate and risk factors for orthopaedic trauma patients with a hip or pelvis fracture. In order to improve patient care and long-term outcomes, surgeons need to understand the risk factors that increase 30-day mortality in these patients. Previous studies investigating risk factors impacting mortality for orthopaedic trauma patients have: • focused on six-month and one year mortality for hip; • evaluated only specific fracture types, such as geriatric hip fractures; • included a limited number of risk factors [5–8]. For example, the study by Aharonoff et al. found that for geriatric hip fracture patients, age over 85 years, dependent functional status, history of cancer, ASA score of 3 or 4, and the development
http://dx.doi.org/10.1016/j.otsr.2016.05.016 1877-0568/© 2016 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model OTSR-1533; No. of Pages 4
ARTICLE IN PRESS A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
2
of one or more postoperative complications were significant risk factors for one year mortality [9]. Belmont et al. determined the inpatient mortality rate following a hip fracture to be 4.5% and that dialysis, cardiac disease, male sex, and injury severity score (ISS) were significant predictors of in-patient mortality; however, this study used a weighted sample and did not evaluate a 30-day mortality [6]. Khan et al., in 2014, determined that 30-day mortality for hip fractures was only related to abnormal creatinine, but this study had a relatively small cohort (n = 516) and did not evaluate preoperative comorbidities or postoperative complications as risk factors [10]. Our study aimed to identify the rate and risk factors impacting 30-day mortality for all orthopaedic trauma patients who had sustained a hip or pelvic fracture, allowing surgeons to better assess high-risk patients and ultimately improve patient outcomes and care. Our hypothesis here are that there are significant pre- and postoperative clinical risk factors associated with 30-day mortality following hip or pelvic fractures in the orthopaedic trauma population. 2. Patients and methods All procedures performed were in accordance with the ethical standards of the institution and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Utilizing the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, 341,062 patients undergoing orthopaedic procedures from 2005 to 2013 were identified through a Current Procedural Terminology (CPT) code search. A second CPT code search was conducted to identify all patients who underwent surgical fixation of a hip or pelvis fracture. Patients’ information was evaluated for complete data regarding 30-day mortality. A number of 24,805 hip/pelvis fracture patients were identified based on these criteria. Mortality for all patients was defined as death within 30-days following surgery. We investigated whether clinically appropriate patient factors had a significant impact on the rate of 30-day mortality. Patient demographics, preoperative, as well as operative characteristics were collected for each patient. Postoperative complications within 30-days following surgery were also assessed as
Table 1 Select patient characteristics associated with increased mortality following hip/pelvis fracture. Demographics Age, median* (IQR) ASA, median* (IQR) BMI, median* (IQR) Sex, N (%) Male Female *
All patients 83 (73–88) 3 (3–3) 24 (21–28) 7,534* (30.4) 17,271* (69.6)
Died (6.2%)
30-day survival (9.38%)
87 (81–91) 3 (3–4) 23 (20–27) 618* (40.2) 918* (59.8)
82 (73–88) 3 (3–3) 24 (21–28) 6,916* (29.7) 16,353* (70.3)
P < 0.05.
risk factors for mortality. Bivariate analyses using the chi-squared test and Wilcoxon-Mann-Whitney test were performed to compare risk factors between those who died and those who survived to postoperative day 30, as appropriate. Statistical significance was set at ␣ = 0.05. A multivariate logistic regression was then conducted to determine the risk factors that significantly predict 30-day mortality for orthopaedic trauma patients following a hip or pelvis fracture. All confounding variables, including patient demographics, preoperative comorbidities, operative characteristics and postoperative complications were included in the analysis. Patients with missing data were excluded in the analysis. 3. Results Patients with hip or pelvis injuries had a mortality rate of 6.2% (n = 1536; 6.2%). Table 1 lists mortality rates based on select patient demographics. The median age was 83 years (IQR: 73–88) for patients with hip and pelvis fractures, and patients were in the majority female (n = 17,271; IQR 69.6). The patients who died within 30-days were significantly older and had a lower median BMI than patients who survived. Of the patients who died, 59.8% were females. Table 2 provides the individual preoperative risk factors significantly associated with 30-day mortality. Based on the chi-squared analysis, patients with a history of ascites, bleeding disorder, disseminated cancer, dyspnea at rest, history of congestive heart failure (CHF), or a history of chronic obstructive pulmonary disease (COPD) were significantly more likely to die within 30-days following surgery (P < 0.001). Steroid use, weight loss greater than
Table 2 Preoperative risk factors for early mortality following hip/pelvis fracture. Risk factor
ASA class (median) Ascites, N (%) Bleeding disorder, N (%) Diabetic, N (%) Disseminated cancer, N (%) Dyspnea, N (%) None With moderate exertion At rest Functional status, N (%) Independent Partially dependent Totally dependent History of CHF, N (%) History of COPD, N (%) On dialysis, N (%) Use of steroids, N (%) Weight loss > 10% in last 6 months, N (%) Hypertension requiring medication, N (%) Recent blood transfusion, N (%)
Univariate analysis
Adjusted odds ratios
Died
30-day survival
P-value
OR (95% CI)
P-value
3 (3–4) 24 (1.6) 361 (23.5) 272 (6.2) 134 (8.7)
3 (3–3) 55 (0.2) 3,770 (16.2) 4,152 (17.8) 721 (3.1)
< 0.001 < 0.001 < 0.001 0.893 < 0.001 < 0.001
2.11 (1.91–2.34) 7.19 (4.18–12.36) 1.13 (0.99–1.300) 0.95 (0.82–1.11) 4.80 (3.82–6.04)
< 0.001 < 0.001 0.081 0.532 < 0.001
1295 (84.3) 168 (10.9) 73 (4.8)
21,365 (91.8) 1,591 (6.8) 313 (1.4)
1.00 (ref) 1.08 (0.89–1.32) 1.92 (1.42–2.61)
0.421 < 0.001
811 (52.8) 531 (34.6) 194 (12.6) 148 (9.6) 308 (20.1) 64 (4.2) 124 (8.1) 49 (3.2) 1,097 (71.4) 154 (10.0)
16,935 (72.8) 5,261 (22.6) 1,073 (4.6) 678 (2.9) 2,557 (11.0) 434 (1.9) 1,315 (5.7) 347 (1.5) 15,811 (68.0) 1,203 (5.2)
1.00 (ref) 1.58 (1.40–1.79) 2.81 (2.34–3.38) 1.72 (1.39–2.12) 1.34 (1.14–1.58) 1.82 (1.34–2.47) 1.07 (0.86–1.33) 1.26 (0.89–1.77) 0.83 (0.73–0.95) 1.36 (1.12–1.65)
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.547 0.187 0.005 0.002
< 0.001
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.005 < 0.001
CHF: congestive heart failure; COPD: chronic obstructive pulmonary disease.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model
ARTICLE IN PRESS
OTSR-1533; No. of Pages 4
A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
3
Table 3 Postoperative risk factors for early mortality following hip/pelvis fracture. Risk factor
Pneumonia, N (%) Urinary tract infection, N (%) Deep wound infection, N (%) Myocardial infarction, N (%) Pulmonary embolism, N (%) Deep vein thrombosis, N (%) Stroke, N (%) Sepsis, N (%) Septic shock, N (%)
Univariate analysis
Adjusted odds ratios
Died
30-day survival
P-value
OR (95% CI)
P-value
246 (16.0) 131 (8.5) 8 (0.5) 111 (7.2) 21 (1.4) 19 (1.2) 55 (3.6) 75 (4.9) 246 (16.0)
629 (2.7) 1303 (5.6) 103 (0.4) 255 (1.1) 148 (0.6) 264 (1.1) 91 (0.4) 324 (1.4) 629 (2.7)
< 0.001 < 0.001 0.657 < 0.001 0.001 0.714 < 0.001 < 0.001 < 0.001
3.36 (2.79–4.05) 1.06 (0.85–1.31) 1.10 (0.47–2.54) 4.42 (3.40–5.73) 2.22 (0.53–9.28) 1.25 (0.51–3.06) 6.78 (2.35–19.51) 1.53 (0.72–3.25) 9.66 (4.75–19.67)
< 0.001 0.622 0.827 < 0.001 0.275 0.623 < 0.001 0.273 < 0.001
10% within the last six months, dialysis, dependent functional status, or a recent blood transfusion was also significantly associated with mortality following hip/pelvis fractures (P < 0.001). After controlling for all individual patient comorbidities with a multivariable regression, it was determined that each one-unit increase in ASA score doubles the likelihood of death within 30-days (OR: 2.11; 95% CI: 1.91–2.34; P < 0.001), and a person with a history of ascites was also 7.19 times more at risk for 30-day mortality (95% CI: 4.18–12.36, P < 0.001). Other significant risk factors for 30-day mortality following multivariate analysis included: history of disseminated cancer and totally dependent functional status, which increased the likelihood of death 4.80 times (95%: 3.82–6.04, P < 0.001) and 2.81 times (2.34–3.38, P < 0.001), respectfully. To understand the risk factors for 30-day mortality, postoperative events were also evaluated. Table 3 summarizes the postoperative risk factors for death after controlling for all confounding variables. Pneumonia (OR: 3.36; 95% CI [2.79–4.05], P < 0.001), myocardial infarction (MI) (OR: 4.42, 95% CI [3.40–5.73], P < 0.001), and stroke (OR 6.78, 95% CI [2.35–19.51], P < 0.001) were all significant postoperative risk factors for 30-day mortality. A patient who developed septic shock was 9.66 times (95%: 4.75–19.67; P < 0.001) more likely to die in 30-days compared to a patient who did not develop the complication, representing the largest postoperative risk for mortality following hip/pelvis fractures. 4. Discussion Our study is the first to examine the pre- and postoperative risk factors associated with 30-day mortality for orthopaedic trauma patients following a hip or pelvis fracture. While the overall rate of 30-day mortality was low for hip/pelvis fractures, there are significant factors, such as history of congestive heart failure and postoperative pneumonia that drive mortality within the first thirty days postsurgery. By identifying significant risk factors associated with 30-day mortality, orthopaedic surgeons will be better equipped to assess a patient’s risk for death prior to and following surgery. Previous studies have found varying rates of mortality for orthopaedic patients depending on location of injury and length of follow-up[11,12]. For orthopaedic trauma patients with hip and pelvis fractures within our study, we determined the 30-day mortality to be 6.2%, representing the lower range of mortality. Most studies, however, have focused mainly on geriatric patients within all orthopaedic subspecialties rather than on all orthopaedic trauma patients [9,13]. Older age significantly increased the risk for premature death following hip/pelvis fractures. Several studies evaluating patients undergoing hip fracture surgery determined that advanced age was a significant predictor of higher mortality [14–20]. However, a study by Streubel et al., did not find an effect of age on survival for distal femur fractures. Our study additionally showed that
preoperative ascites was a significant predictor of 30-day mortality. History of ascites increased the risk for 30-day mortality by 7.19 times. To our knowledge, previous studies have not included preoperative ascites as a risk factor for mortality, even though it has been noted in a study by Khan et al. that mortality in patients with ascites is high [21]. Our study also showed that a history of COPD and dialysis were significant predictors for 30-day mortality. One study found that COPD increased the risk of acute in-patient mortality for hip fractures 1.87 times (95%: 1.01–1.92) compared to a patient without COPD, which is slightly higher than our finding (OR: 1.34, 95%: 1.14–1.58) [22]. Postoperative risk factors were also clear predictors of 30-day mortality. Pneumonia increased the risk of 30-day mortality 3.36 times. Other studies have shown that postoperative pneumonia drastically increases the risk for inpatient mortality [22–24]. It is essential that surgeons are aware of the close association between mortality and pneumonia even within 30-days following surgery. Myocardial infarction and septic shock were also significantly associated with increased 30-day mortality after controlling for all confounding variables. There were some limitations in this study. First, the ACS-NSQIP database limits the risk factors available to be evaluated. Postoperative complications, such as osteomyelitis, could have been significant predictors of 30-day mortality. Due to limitations of the ACS-NSQIP database, the time before surgery was also not included in the analysis, which has been highly associated with increased early mortality for orthopaedic patients [25]. Nevertheless, our study helps identify several clinically relevant risk factors for 30-day mortality for orthopaedic trauma patients who sustain a hip or pelvis fracture. As our healthcare systems shifts towards containing costs while improving quality care, orthopaedists must better understand the risk factors for 30-day mortality in order to reduce costs while increasing patient quality care. Disclosure of interest Author WTO has previously consulted for Biometrics; done expert testimony in legal matters, and has a grant from the Department of Defense. ACD, CB, AJ, HM, and MKS declare that they have no competing interest. References [1] Centers for Medicare & Medicaid Services. National health expenditures 2013 highlights; 2013 [http://www.cms.gov/Research-StatisticsData-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/ downloads/highlights.pdf. Accessed 2 February 2015]. [2] Bundled payment: AHA research synthesis report. American Hospital Association Committee on Research. http://www.aha.org/research/cor/content/ BundledPayment.pdf. Accessed 2 February 2015. [3] Outcome Measures. Centers for Medicare & Medicaid Services. http://www. cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/ HospitalQualityInits/OutcomeMeasures.html. Accessed 2 February 2015.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016
G Model OTSR-1533; No. of Pages 4 4
ARTICLE IN PRESS A.C. Dodd et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2016) xxx–xxx
[4] Krumholz HM, Normand SLT. Public reporting of 30-day mortality for patients hospitalized with acute myocardial infarction and heart failure. Circulation 2008;13:1394–7. [5] Hietala P, Strandberg M, Kiviniemi T, et al. Usefulness of troponin T to predict short-term and long-term mortality in patients after hip fracture. Am J Cardiol 2014;114:193–7. [6] Belmont Jr PJ, Garcia EJ, et al. Risk factors for complications and in-hospital mortality following hip fractures: a study using the National Trauma Data Bank. Arch Orthop Trauma Surg 2014;134:597–604. [7] Clement ND, Duckworth AD, McQueen MM, et al. The outcome of proximal humeral fractures in the elderly predictors of mortality and function. Bone Joint J 2014;96:970–7. [8] Copeland CE, Mitchell KA, Brumback RJ, et al. Mortality in patients with bilateral femoral fractures. J Orthop Trauma 1998;12:315–9. [9] Aharonoff GB, Koval KJ, Skovron ML, et al. Hip fractures in elderly: predictors of one-year mortality. J Orthop Trauma 1997;11:162–5. [10] Khan SK, Shirley M, Glennie C, et al. Achieving best practice tariff may not reflect improved survival after hip fracture treatment. Clin Interv Aging 2014;9:2097–102. [11] Richmond J, Aharonoff, Zuckerman JD, et al. Mortality risk after hip fracture. J Orthop Trauma 2003;17:53–6. [12] Hu F, Jiang C, Shen J, et al. Preoperative predictors for mortality following hip fracture surgery: a systematic review and meta-analysis. Injury 2012;43:676–85. [13] White BL, Fisher WD, Laurin CA. Rate of mortality for elderly patients after fracture of the hip in the 1980s. J Bone Joint Surg Am 1987;69:1335–40. [14] Elliott J, Beringer T, Kee F, et al. Predicting survival after treatment for fracture of the proximal femur and the effect of delays to surgery. J Clin Epidemiol 2003;56:788–95.
[15] Calder SJ, Anderson GH, Jagger C, et al. Unipolar or bipolar prosthesis for displaced intracapsular hip fracture in octogenarians: a randomised prospective study. J Bone Joint Surg Br 1996;78:391–4. [16] Hommel A, Ulander K, Bjorkelund KB, et al. Influence of optimised treatment of people with hip fracture on time to operation, length of hospital stay, reoperations and mortality within 1 year. Injury 2008;39:1164–74. [17] Oztürk A, Ozkan Y, Akgöz S, et al. The risk factors for mortality in elderly patients with hip fractures: postoperative one-year results. Singapore Med J 2010;51:137–43. [18] Paksima N, Koval KJ, Aharanoff G, et al. Predictors of mortality after hip fracture: a 10-year prospective study. Bull NYU Hosp Jt Dis 2008;66:111–7. [19] Rosencher N, Vielpeau C, Emmerich J, et al. Venous thromboembolism and mortality after hip fracture surgery: the ESCORTE study. J Thromb Haemost 2005;3:2006–14. [20] Trombetti A, Herrmann F, Hoffmeyer P, et al. Survival and potential years of life lost after hip fracture in men and age-matched women. Osteoporos Int 2002;13:731–7. [21] Khan J, Pikkarainen P, Karvonen AL, et al. Ascites: aetiology, mortality and the prevalence of spontaneous bacterial peritonitis. Scand J Gastroenterol 2009;44:970–4. [22] Myers AH, Robinson EG, Van Natta ML, et al. Hip fractures among the elderly: factors associated with in-hospital mortality. Am J Epidemiol 1991;134:1128–37. [23] Streubel PN, Ricci WM, Wong A, et al. Mortality after distal femur fractures in elderly patients. Clin Orthop Relat Res 2011;469:1188–96. [24] Bhattacharyya T, Iorio R, Healy WL. Rate of and risk factors for acute inpatient mortality after orthopaedic surgery. J Bone Joint Surg Br 2002;84:562–72. [25] Moran CG, Wenn RT, Sikand M, Taylor AM. Early mortality after hip fracture: is delay before surgery important? J Bone Joint Surg 2005;87:483–9.
Please cite this article in press as: Dodd AC, et al. Predictors of 30-day mortality following hip/pelvis fractures. Orthop Traumatol Surg Res (2016), http://dx.doi.org/10.1016/j.otsr.2016.05.016