The elderly population of the United States is expected to increase exponentially in the coming decades, with an estimated 98 million individuals being older than 65 years by 2060.1 With this aging population, the surgical demand to address the associated increase in geriatric hip fractures will also increase, with 6.26 million hip fractures estimated to occur worldwide by 2050.2 Given the significant mortality risk following surgical fixation of hip fractures,3 there is a critical need to better understand the risk factors associated with increased mortality in these patients to appropriately identify high-risk patients and execute early intervention strategies to mitigate poor outcomes.
Previous studies have identified both injury- and patient-specific risk factors that are associated with increased morbidity and mortality after hip fracture. Although some studies have previously shown that patients with intertrochanteric femur fractures have higher mortality rates than those with intracapsular femoral neck fractures,4–6 other studies have shown no difference between these fracture types.7–9 Regardless of injury pattern, patient-specific risk factors that are associated with worse outcomes are often more likely to occur with older age, such as renal failure, cardiac disease, cancer, and diabetes mellitus.7,10,11 However, not all patient-specific risk factors are necessarily age related. For example, modifiable risk factors such as malnutrition have been associated with increased length of hospital stay and in-hospital mortality.10,12
To encompass a multidimensional state of weakness and decreased physiologic reserve that is associated with, but not necessarily dependent on, age, the term “frailty” has been more recently used in the surgical and orthopedic literature as a risk factor for mortality. Increased frailty index or geriatric mortality scores have been associated with increased mortality rates in patients with a hip fracture.13,14 However, despite the available evidence, there is a paucity of data regarding frailty and the risk of mortality following intertrochanteric femur fractures. Furthermore, there is a lack of consensus regarding which frailty index scores to use and at what associated thresholds scores become clinically relevant.
The aim of this study was to determine the association between frailty and 30-day morbidity and mortality in patients with intertrochanteric femur fractures using a previously validated 11-point modified frailty index (mFI) score. The authors hypothesized that higher mFI scores would be associated with increased morbidity and mortality in patients with intertrochanteric femur fractures. Furthermore, they sought to identify an mFI score threshold that may help to identify high-risk patients.
Materials and Methods
This retrospective observational study involved patients who presented to a single American College of Surgeons–verified level I trauma center for the treatment of an intertrochanteric femur fracture between January 1, 2013, and December 31, 2016. After institutional review board approval was obtained, medical records were queried using International Classification of Diseases, Ninth Revision, code 820 and Current Procedural Terminology codes 27244 and 27245. This resulted in 444 consecutive patients available for inclusion in this study. Exclusion criteria were age younger than 50 years and fractures from high-energy trauma. High-energy trauma was defined by mechanism and included injury secondary to motor vehicle or motorcycle accident, fall from height, or pedestrian vs automobile accidents. After exclusion criteria were applied, 229 patients remained available for chart review and statistical analyses.
The primary outcome variables included 30-day morbidity and mortality. Postoperative morbidity included acute kidney injury, delirium, unplanned intensive care unit admission, acute respiratory distress syndrome, myocardial infarction, urinary tract infection, central line infection, deep venous thrombosis, pulmonary embolism, stroke, and 30-day readmission. Patients discharged from the hospital directly to hospice for end-of-life care were considered a mortality.
Modified Frailty Index Calculation
The mFI is an adaptation of the thorough, but clinically cumbersome, Canadian Study on Health and Aging Frailty Index. This mFI is a previously validated 11-point mFI scale.15–17 The 11 frailty determinant variables include diabetes mellitus, functional status index of 2 (partially dependent) or higher, chronic obstructive pulmonary disease or pneumonia, congestive heart failure, myocardial infarction, percutaneous coronary intervention and/or stenting or angina, hypertension requiring medication, peripheral vascular disease or ischemic rest pain, impaired sensorium, transient ischemic attack or cerebrovascular accident, and cerebrovascular accident with deficit. Each patient's chart was reviewed for the presence or absence of each of these variables and these were noted. Then, mFI scores were calculated as follows: for each variable that was present, 1 point was added (possible outcomes 0, 1, 2, 3, … 11) and then the sum of positive variables was divided by 11 to calculate the final score (Table 1).
The 11-Item Modified Frailty Index
The distribution of continuous numerical data including mFI was examined in descriptive histograms and box plots, and a Kolmogorov–Smirnov test was used to confirm a normal distribution. For descriptive analysis, absolute mean values for mFI were expressed with standard deviation (SD). An independent samples t test was used to compare mean mFI values between patients who had postoperative morbidity or mortality with those who had an uncomplicated hospital course. Modified frailty index values were further characterized as a categorical variable with progressive cutoffs. Chi-square analyses and receiver operating characteristic (ROC) analyses were performed to determine the sensitivity and specificity of mFI in predicting 30-day morbidity and mortality. Receiver operating characteristic was reported with 95% confidence intervals (CIs). All statistical analyses were performed with Stata release 14 statistical software (StataCorp, College Station, Texas).
Of the 229 patients included in this study, 82 (36%) had a postoperative complication and there were 10 (4%) mortalities. The most common complications were delirium (n=40; 17%) and acute kidney injury (n=25; 11%). Of patients, 118 were men and 111 were women. Mean age was 73 years (SD, 13 years).
30-Day Morbidity Outcomes
Mean mFI score was 0.24 (SD, 0.18) for those who developed a postoperative complication compared with 0.14 (SD, 0.14) for those who did not (P<.001) (Table 2). Mean age was 75 years (SD, 13 years) for those who developed a postoperative complication compared with 73 years (SD, 13 years) for those who did not (P=.161) (Table 2).
30-Day Morbidity and Mortality by Age and Modified Frailty Index
Multivariate logistic regression revealed that an increased mFI was associated with the development of postoperative complications independent of age (P<.001). However, when controlling for mFI, older age was not an independent risk factor for the development of complications (P=.410). The sensitivity and specificity for mFI to predict 30-day morbidity and mortality for various cutoffs are listed in Table 3.
Sensitivity and Specificity Thresholds for 30-Day Morbidity and Mortality by Modified Frailty Index Scores
The most robust predictor of morbidity was an mFI score of 0.27 or more (P<.001). Postoperative morbidity increased from 21% with mFI 0 to 57% with mFI of 0.27 or more. Multivariate regression analysis revealed that the odds of developing a complication were 4 times higher in patients with an mFI score of 0.27 or more (odds ratio [OR], 4.0; 95% CI, 2.2–7.1; P<.001). Even after controlling for age, patients with an mFI of 0.27 or more were still almost 4 times as likely to have a postoperative complication compared with those with an mFI of less than 0.27 (OR, 3.8; 95% CI, 2.1–6.9; P<.001).
30-Day Mortality Outcomes
Mean mFI score was 0.31 (SD, 0.13) for the mortality group compared with 0.17 (SD, 0.16) for the nonmortality group (P=.010) (Table 2). Mean age was 79 years (SD, 14 years) for the mortality group compared with 73 years (SD, 13 years) in the nonmortality group (P=.209) (Table 2).
Multivariate logistic regression revealed that an increased mFI was associated with increased mortality independent of age (P=.020). However, when controlling for mFI, older age was not an independent risk factor for increased mortality (P=.304). Similar to the ROC analysis for morbidity, the most robust predictor of mortality was an mFI of 0.27 or more (P=.001) (Table 3).
The mortality rate increased from 0% for mFI of 0 to 11% for mFI of 0.27 or more. Patients with an mFI of 0.27 or more were more than 9 times as likely to have a mortality compared with patients with an mFI of less than 0.27 (OR, 9.1; 95% CI, 1.9–43.9; P=.006). Even after controlling for age, multivariate logistic regression revealed that patients with an mFI of 0.27 or more were still more than 8 times as likely to have a mortality (OR, 8.4; 95% CI, 1.7–41.1; P=.008).
This study demonstrated that an increased mFI was associated with increased 30-day morbidity and mortality rates in patients with intertrochanteric femur fractures. Notably, the authors showed that mFI was an age-independent risk factor for increased morbidity and mortality. In contrast, older age alone had no impact on 30-day outcomes.
Although age has been found to be a significant predictor of mortality in many studies,2,3,7,18,19 there is evidence that when patients are separated by general health status, the clinical and statistical significance of age is diminished.20 The current results support such findings and suggest that although age and age-associated morbidities may certainly contribute to in-hospital mortality, overall health status may provide fewer biased predictors of postoperative complications. Importantly, by solely emphasizing age as a risk factor for postoperative morbidity and mortality, relatively younger hip fracture patients (eg, age <65 years) with poor general health status may receive less frailty-specific care because of their age.
Although this study used mFI as the primary risk index, other indices have been used to assess risk factors for postoperative morbidity and mortality, including the American Society of Anesthesiologists classification, the Charleston Comorbidity Index, and the Orthopedic Multidimensional Prognostic Index. Although useful in some clinical scenarios, such indices are not without limitations, including variability in interobserver consistency and limited generalizability due to their focus on particular patient subpopulations.21–25 Given that frailty is not a static parameter and the mFI calculation can change as a patient's functional status changes, it can be a dynamic and reliable tool to assess postoperative complications and mortality, as the current study has demonstrated.
In a previous study that examined the relationship between hip fractures and frailty, Dwyer et al13 used 2 geriatric mortality calculators to show that frailty could differentiate survivors vs nonsurvivors following hip fracture. In another study of the association between patients with a hip fracture and mortality, Krishnan et al14 calculated a frailty index score that encompassed motivation, sleep, emotional state, feeding status, walking, dressing, polypharmacy, appetite, weight change, and medical history to find that higher frailty index scores were associated with increased 30-day mortality. A significant drawback to a frailty index score such as that used by Krishnan et al14 is how difficult and cumbersome it would be to calculate for patients with a hip fracture in the acute setting, such as the emergency department. In contrast, an advantage of the current study was the use of validated frailty index score that is quick and easy to calculate.
Although previous studies have examined both intra- and extra-capsular hip fractures and compared the risk of mortality between intertrochanteric and femoral hip fractures,2,4,6,7,9 the current study was unique in that the authors isolated the risk of morbidity and mortality in intertrochanteric femur fractures alone. Furthermore, they defined a specific threshold mFI score of 0.27 or more, at which patients have a significantly increased risk of morbidity and mortality. The 30-day mortality rate for the current cohort was 9 times higher in patients with an mFI of 0.27 or more. The importance of such findings cannot be overstated because such information may be used both to communicate risk to patients and families and to preoperatively identify patients who may benefit from a more structured, interdisciplinary care pathway.
Although the current results do not allow for determination of causality, it has been postulated that frailty contributes to complications and mortality as it is an “accumulation of deficits.”26 Given this deficit in physiological reserve, it follows that the physiological stress of trauma and surgical intervention may be less well tolerated in the frail patient.
This study had several limitations. First, although the authors' use of the 11-point mFI scale was a strength in its practicality for ease of everyday clinical use and its application to large patient databases, it could also be considered a weakness because the authors could not exclude the possibility that other factors may also contribute to the frailty of patients that are currently not accounted for by mFI calculations. In addition, there are inherent weaknesses in the retrospective study design, such as selection and misclassification bias. Finally, this study was limited by lack of long-term follow-up.
This study demonstrated that mFI is associated with 30-day morbidity and mortality in patients aged 50 years or older with intertrochanteric femur fractures. Whereas previous works have focused on the association between frailty and mortality in elderly patients with intracapsular femoral neck fractures, this study offered new evidence that frailty is a significant predictor of short-term morbidity and mortality in patients with intertrochanteric femur fractures. Importantly, the authors identified an mFI score of 0.27 or more as the most robust predictor of increased 30-day morbidity and mortality following surgical fixation of intertrochanteric femur fractures.
- Mather M, Jacobsen LA, Pollard KM. Aging in the United States. Washington, DC: Population Reference Bureau; 2015.
- Kesmezacar H, Ayhan E, Unlu MC, Seker A, Karaca S. Predictors of mortality in elderly patients with an intertrochanteric or a femoral neck fracture. J Trauma. 2010;68(1):153–158. doi:10.1097/TA.0b013e31819adc50 [CrossRef]
- Abrahamsen B, van Staa T, Ariely R, Olson M, Cooper C. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos Int. 2009;20(10):1633–1650. doi:10.1007/s00198-009-0920-3 [CrossRef]19421703
- Fox KM, Magaziner J, Hebel JR, Kenzora JE, Kashner TM. Intertrochanteric versus femoral neck hip fractures: differential characteristics, treatment, and sequelae. J Gerontol A Biol Sci Med Sci. 1999;54(12):M635–M640. doi:10.1093/gerona/54.12.M635 [CrossRef]
- Haentjens P, Autier P, Barette M, Venken K, Vanderschueren D, Boonen SHip Fracture Study Group. Survival and functional outcome according to hip fracture type: a one-year prospective cohort study in elderly women with an intertrochanteric or femoral neck fracture. Bone. 2007;41(6):958–964. doi:10.1016/j.bone.2007.08.026 [CrossRef]17913614
- Keene GS, Parker MJ, Pryor GA. Mortality and morbidity after hip fractures. BMJ. 1993;307(6914):1248–1250. doi:10.1136/bmj.307.6914.1248 [CrossRef]8166806
- Aharonoff GB, Koval KJ, Skovron ML, Zuckerman JD. Hip fractures in the elderly: predictors of one year mortality. J Orthop Trauma. 1997;11(3):162–165. doi:10.1097/00005131-199704000-00004 [CrossRef]9181497
- Kenzora JE, McCarthy RE, Lowell JD, Sledge CB. Hip fracture mortality: relation to age, treatment, preoperative illness, time of surgery, and complications. Clin Orthop Relat Res. 1984;(186):45–56.
- Cipitria JA, Sosa MM, Pezzotto SM, Puche RC, Bocanera R. Outcome of hip fractures among elderly subjects. Medicina (B Aires). 1997;57(5):530–534.
- Hu F, Jiang C, Shen J, Tang P, Wang Y. Pre-operative predictors for mortality following hip fracture surgery: a systematic review and meta-analysis. Injury. 2012;43(6):676–685. doi:10.1016/j.injury.2011.05.017 [CrossRef]
- Jiang HX, Majumdar SR, Dick DA, et al. Development and initial validation of a risk score for predicting in-hospital and 1-year mortality in patients with hip fractures. J Bone Miner Res. 2005;20(3):494–500. doi:10.1359/JBMR.041133 [CrossRef]15746995
- Koval KJ, Maurer SG, Su ET, Aharonoff GB, Zuckerman JD. The effects of nutritional status on outcome after hip fracture. J Orthop Trauma. 1999;13(3):164–169. doi:10.1097/00005131-199903000-00003 [CrossRef]10206247
- Dwyer JG, Reynoso JF, Seevers GA, et al. Assessing preoperative frailty utilizing validated geriatric mortality calculators and their association with postoperative hip fracture mortality risk. Geriatr Orthop Surg Rehabil. 2014;5(3):109–115. doi:10.1177/2151458514537272 [CrossRef]25360340
- Krishnan M, Beck S, Havelock W, Eeles E, Hubbard RE, Johansen A. Predicting outcome after hip fracture: using a frailty index to integrate comprehensive geriatric assessment results. Age Ageing. 2014;43(1):122–126. doi:10.1093/ageing/aft084 [CrossRef]
- Karam J, Tsiouris A, Shepard A, Velanovich V, Rubinfeld I. Simplified frailty index to predict adverse outcomes and mortality in vascular surgery patients. Ann Vasc Surg. 2013;27(7):904–908. doi:10.1016/j.avsg.2012.09.015 [CrossRef]23711971
- Farhat JS, Velanovich V, Falvo AJ, et al. Are the frail destined to fail? Frailty index as predictor of surgical morbidity and mortality in the elderly. J Trauma Acute Care Surg. 2012;72(6):1526–1531. doi:10.1097/TA.0b013e3182542fab [CrossRef]22695416
- Vu CCL, Runner RP, Reisman WM, Schenker ML. The frail fail: increased mortality and postoperative complications in orthopaedic trauma patients. Injury. 2017;48(11):2443–2450. doi:10.1016/j.injury.2017.08.026 [CrossRef]28888718
- Kannegaard PN, van der Mark S, Eiken P, Abrahamsen B. Excess mortality in men compared with women following a hip fracture: national analysis of comedications, comorbidity and survival. Age Ageing. 2010;39(2):203–209. doi:10.1093/ageing/afp221 [CrossRef]20075035
- Richmond J, Aharonoff GB, Zuckerman JD, Ko val KJ. Mortality risk after hip fracture. J Orthop Trauma. 2003;17(1):53–56. doi:10.1097/00005131-200301000-00008 [CrossRef]
- Mullen JO, Mullen NL. Hip fracture mortality: a prospective, multifactorial study to predict and minimize death risk. Clin Orthop Relat Res. 1992;(280):214–222.1611747
- Arnson WL, McAuliffe MS, Miller K. Variability in the American Society of Anesthesiologists physical status classification scale. AANA J. 2003;71(4):265–274.
- Mak PH, Campbell RC, Irwin MGAmerican Society of Anesthesiologists. The ASA physical status classification: inter-observer consistency. Anaesth Intensive Care. 2002;30(5):633–640. doi:10.1177/0310057X0203000516 [CrossRef]12413266
- Harse JD, Holman CD. Charlson's Index was a poor predictor of quality of life outcomes in a study of patients following joint replacement surgery. J Clin Epidemiol. 2005;58(11):1142–1149. doi:10.1016/j.jclinepi.2005.02.017 [CrossRef]
- Giusti A, Kammerlander C. Comprehensive geriatric assessment in orthogeriatrics. In: Pilotto A, Martin FC, eds. Comprehensive Geriatric Assessment. Cham, Switzerland: Springer; 2018:123–132. doi:10.1007/978-3-319-62503-4_11 [CrossRef]
- Vitale E, Notarnicola A, Tafuri S, Vicenti G, Cassano M, Moretti B. Orthopedic Multidimensional Prognostic Index (Ortho-MPI) in the elderly with hip or neck femur fracture: a pilot study. Arch Gerontol Geriatr. 2014;58(1):101–104. doi:10.1016/j.archger.2013.08.006 [CrossRef]
- Velanovich V, Antoine H, Swartz A, Peters D, Rubinfeld I. Accumulating deficits model of frailty and postoperative mortality and morbidity: its application to a national database. J Surg Res. 2013;183(1):104–110. doi:10.1016/j.jss.2013.01.021 [CrossRef]23415494
The 11-Item Modified Frailty Indexa
|Variable, if Present in Patient History||Modified Frailty Index Variable|
|Diabetes mellitus—insulin and noninsulin dependent||1|
|Congestive heart failure||2|
|Hypertension requiring medication||3|
|History of myocardial infarction||4|
|Previous percutaneous coronary intervention or angina||5|
|History of transient ischemic attack or cerebrovascular accident without neurological deficit||6|
|Cerebrovascular accident with neurological deficit||7|
|History of chronic obstructive pulmonary disease or pneumonia||9|
|History of peripheral vascular disease or rest pain||10|
|Functional health status before surgery—partially or totally dependent for activities of daily living||11|
30-Day Morbidity and Mortality by Age and Modified Frailty Index
|Factor, Mean (SD)||30-Day Morbidity||30-Day Morality|
|Age||73 (13)||75 (13)||.161||73 (13)||79 (14)||.209|
|Modified Frailty Index||0.14 (0.14)||0.24 (0.18)||<.001||0.17 (0.16)||0.31 (0.13)||.010|
Sensitivity and Specificity Thresholds for 30-Day Morbidity and Mortality by Modified Frailty Index Scoresa
|mFI Cutoff||Chi-square (P)||AUC||Sensitivity||Specificity||PPV||NPV||Likelihood Ratio|
|Outcome: 30-day postoperative complications|
| ≥0.09||.005||0.58 (0.53–0.64)||85.4% (75.8%–92.2%)||31.3% (23%.9–39.5%)||40.9% (33.5%–48.7%)||79.3% (66.6%–88.8%)||1.24 (1.08–1.43)||0.47 (0.26–0.83)|
| ≥0.18||.001||0.61 (0.55–0.68)||64.6% (53.3%–74.9%)||57.8% (49.4%–65.9%)||46.1% (36.8%–55.6%)||75.6% (65.6%–82.3%)||1.53 (1.20–1.96)||0.61 (0.44–0.85)|
| ≥0.27||<.001||0.65 (0.59–0.72)||52.4% (41.1%–63.3%)||78.2% (70.7%–84.6%)||57.3% (45.4%–68.7%)||74.7% (67.0%–81.3%)||2.41 (1.67–3.49)||0.61 (0.48–0.78)|
| ≥0.36||<.001||0.63 (0.57–0.69)||36.6% (26.2%–48.0%)||89.1% (82.9%–93.6%)||65.2% (49.8%–78.6%)||71.6% (64.5%–78.0%)||3.36 (1.95–5.79)||0.71 (0.60–0.85)|
|Outcome: 30-day mortality|
| ≥0.09||.060||0.63 (0.60–0.66)||100.0% (69.2%–100.0%)||26.5% (20.8%–32.9%)||5.9% (2.8%–10.5%)||100.0% (93.8%–100.0%)||1.36 (1.26–1.47)||0.0 (0.0–0.0)|
| ≥0.18||.010||0.71 (0.61–0.81)||90.0% (55.5%–99.7%)||51.6% (44.8%–58.4%)||7.8% (3.6%–14.3%)||99.1% (95.2%–100.0%)||1.86 (1.45–2.38)||0.19 (0.03–1.25)|
| ≥0.27||.001||0.75 (0.61–0.88)||80.0% (44.4%–97.5%)||69.4% (62.8%–75.4%)||10.7% (4.7%–19.9%)||98.7% (95.4%–99.8%)||2.61 (1.81–3.78)||0.29 (0.08–1.00)|
| ≥0.36||.108||0.60 (0.44–0.77)||40.0% (12.2%–73.8%)||80.8% (75.0%–85.8%)||8.7% (2.4%–20.8%)||96.7% (93.0%–98.8%)||2.09 (0.93–4.67)||0.75 (0.45–1.24)|