Background

Due to the aging population, the worldwide prevalence of osteoporosis is rising, and correspondingly, the incidene of fragility fractures are increasing [1]. National Osteoporosis Foundation estimates that over 53 million individuals have or are at a high risk of osteoporosis in the United States, and by 2040, the total yearly expenses of caring for fragility fractures will increase to over $95 billion [2]. While spine and wrist fractures are also common, proximal femoral fractures are the most common fragility fractures, accounting for up to 40% of all osteoporotic fractures [3]. Thus, proximal femoral fractures represent a substantial proportion of fracture-related admissions and impose a significant burden to healthcare systems worldwide, accounting for up to 72% of all fracture-related costs in the United States [4, 5]. Similar data have been reported from Europe, in which the cost of treating fragility fracture was €37.5 billion in 2017 [6].

Proximal femoral fractures significantly contribute to health deterioration and long-term morbidity and mortality. The arduous rehabilitation, functional decline, and reduced quality of life affect patients’ independence and livelihood [7, 8]. Additionally, proximal femoral fractures are associated with a significant mortality risk during a hospital stay and following discharge. Hip “fracture fatality rate,“ or the proportion of patients who died following a hip fracture in a year, is an important health indicator in patients with this type of injury [9]. This index could be considered and monitored as the health system’s efficiency in dealing with elderly patients and geriatric care. The reported mortality rate of proximal femoral fractures in the literature ranges from 11% to more than 30% [10].

In line with the improvements in the prevention and medical care of proximal femoral fractures, orthopedic treatment has also undergone a dramatic shift in the past decades [11]. As our understanding of the biomechanics of these fractures has evolved, implants have been redesigned to ultimately improve the standards of care [12]. The appropriate implant selection is based on the access to the device, fracture characteristics, the patient’s bone quality and underlying medical conditions, and the surgeon’s preferences [12].

Although several large-scale studies have been performed in developed countries, there is a paucity in the literature regarding the outcomes of proximal femoral fractures in develo** countries like Iran [13, 14]. Particularly, the incidence and risk factors of in-hospital and post-discharge mortality and functional outcomes are areas with minimal data available. Therefore, the goal of this study was to evaluate patients with a proximal femoral fracture at a level I tertiary referral trauma center in Iran, to determine the early and one-year mortality and its associated risk factors. Furthermore, we sought to evaluate functional outcomes of proximal femoral fractures at one year and the correlations with patients, injury, and treatment characteristics. The findings of this study could be implemented to improve the care and outcomes of elderly patients with this type of injury in future.

Materials and methods

Study design and population

In an IRB-approved retrospective cohort study, our institutional hospital information system (HIS) was queried for all patients admitted with a femoral neck or intertrochanteric fracture diagnosis from January 1, 2016, through December 31, 2018 to one referral orthopedic surgery center affialited with Kermanshah University of Medical Sciences, Kermanshah, Iran. Medical records were reviewed to include patients over 60 years of age who had a proximal femoral fracture and had a complete medical record and radiographs. The 60 years age cut-off for inclusion of the elderly population with proximal femoral fracture in this study was chosen based on the prevalence of this type of fracture in Iranian elederly population. Exclusion criteria included patients with pathological fractures, cancer under active treatment, follow-up loss, and patient access loss.

Study variables and data collection

Basic patient information including sex, mechanism of injury, past medical history, American Society of Anesthesiologists (ASA) score status, the delay from fracture to surgery, type of anesthesia, surgical technique and the utilized hardware, the duration of hospital stay, and discharge status of patients were recorded. The admission radiographs were reviewed on the local picture archiving and communication system (PACS), and each fracture was classified according to the 2018 AO fracture and dislocation classification [15]. Associated fractures and follow-up radiographs were also extracted and reviewed. Anesthesia records were pre-operatively reviewed to extract the American Society of Anesthesiologists (ASA) score assigned to each patient by the attending anesthesiologist [16].

In-hospital and post-discharge mortalities at one and 12 months due to included types of fractures were identified and reviewed as the primary outcome in this study. Additionally, follow-up visits were reviewed to determine whether the patient recovered or died. The two time points of one-month and one-year for patient follow-up were chosen based on a literature review and similar studies in the field as most of the adverse outcomes following surgery of this condition in elderly population happen in the first month after surgery and also the first-year assessment is a crucial time-point in surgical outcomes investigations. Modified Harris Hip Scores (mHHS), collected at the latest clinic visit, were also recorded [17]. Furthermore, patients who did not have a post-operative visit at or after January 1, 2020, were called by phone between November 2020 and January 2021 to ascertain their health status. The mHHS form was completed over the phone by a trained investigator for these patients. The assessors who extracted data from charts and called patients were not involved in patient care and surgery. The treating surgeons were also not involved in data analysis.

Statistical analysis

Statistical calculations and analyses were performed using SPSS Statistics for Windows, Version 25.0 (IBM). Statistically significant p for type I error was set at < 0.05. Descriptive statistics were used to calculate mean and standard deviation (± SD). Categorical variables were reported as frequency and percentage. Student’s t or Mann-Whitney U tests were used to compare means depending on the normality of data distribution. Pearson’s chi-square or Fischer’s exact test compared categorical variables. Kaplan-Meier survival analysis was performed to estimate the probability of survival at a specific time after the fracture (1-month and one-year periods), and the log-rank test was used to compare survival between groups. The Cox proportional hazard models were used to identify the predictors of mortality. Independent variables in the recruited models included age, sex, AO type of fracture, time from admission to surgery, orthopedic hardware utilized, and ASA score. Crude and adjusted hazard ratios with 95% confidence intervals (CI) were calculated for each variable.

Ethical considerations

The study protocol was reviewed and approved by the Research Ethics Committee of Kermanshah University of Medical Sciences (IR.KUMS.REC.1398.1175). This study was performed following the Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans, and the human subjects’ privacy rights were respected.

Results

General findings

A total of 1004 cases with a proximal femoral fracture were identified during the study period. Of these, 889 patients satisfied the inclusion criteria. After excluding 101 (11.4%) patients who were lost to follow-up or had incomplete records/radiographs, a total of 788 patients including 412 females (52.3%) and 376 males (47.7%) with a female to male ratio of 1.1 were included in the final analysis (Fig. 1). The mean age of patients was 76.05 ± 10.01 (range, 60–111) years, and 341 patients (43.3%) were over 80 years old. No statistically significant difference was detected in the mean age between male and female patients (p = 0.24). Right-sided fractures were slightly more common, occurring in 409 patients (51.9%).

Fig. 1
figure 1

Flowchart of study population selection process

Full size image

The majority of our patients (94%) sustained a fracture after a simple fall, and only 6% of fractures resulted from high-energy trauma. Concomitant fractures were diagnosed in 117 patients (15.0%), most commonly a distal radius or a humerus fracture. A history of previous hip fracture was present in 74 patients (9.5%). In 462 patients (59.4%), at least two comorbidities were present, the most common of which were dyslipidemia, diabetes mellitus, ischemic heart disease, and Alzheimer’s disease (Table 1).

Table 1 Basic demographics, injury, and outcome characteristics of included sample of populations in this study
Full size table

Fracture type

Among patients, 573 (72.7%) had an intertrochanteric fracture, while 215 (27.3%) had a femoral neck fracture. According to the AO classification, 180 fractures (22.8%) were considered type 31A1, 233 (29.6%) were 31A2, 160 (20.3%) were 31A3, and 215 (27.3%) were 31B. Patients with a femoral neck fracture were significantly younger than those with an intertrochanteric fracture (p = 0.01).

Overall, 23 patients (2.9%) did not undergo surgery and were treated non-operatively, with the remaining 765 patients (97.1%) received surgical treatment in one of the forms of screws, dynamic hip screws (DHS), open reduction and internal fixation (ORIF) with a proximal femoral plate, or bipolar hemiarthroplasty (Table 2).

Table 2 Treatment characteristics, broken down by fracture type
Full size table

Mortality and survival

The mean follow-up period in this study was 33.31 months (range, 24–48), and survival assessment was determined based on health status on 1/1/2020. Overall mortality rate was 33.1% (262 patients), of which 6 deaths (2.3%) were in-hospital, 45 (17.2%) occurred during the first post-fracture month, and 159 (60.7%) happened during the first year following fracture. Although the mean overall survival was slightly higher in male patients (34.3 vs. 32.9 months, p = 0.009), men’s survival was lower in the ≥ 80-years age group (30.8 vs. 33.5 months).

Analysis of 1-month mortality showed a significant mortality difference in patients operated after 48 h of fracture (p = 0.01) and in patients with ASA scores of 3–4 compared to ASA scores of 1–2 (p = 0.001). One-year mortality data showed that the mortality rate in femoral neck fractures was lower compared to other types of fracture. Surgical delay of > 48 h, ASA scores of 3–4, and treatment by proximal femoral plate were associated with shorter survival (Table 3).

Table 3 Kaplan-Meier survival analysis for 1-month and 1-year mortality
Full size table

Univariate Cox regression survival analysis revealed that femoral neck fractures had a significantly lower mortality risk compared to the intertrochanteric fractures (p = 0.001). Additionally, delayed surgery (p = 0.02) and an ASA score of 3–4 (p = 0.04) significantly increased the mortality risk, while treatment with bipolar hemiarthroplasty decreased the risk of mortality (p = 0.001). On multivariable Cox regression analysis, age ≥ 80 years, ORIF with a proximal femoral plate, delayed surgery, and ASA scores of 3–4 were risk factors of mortality, while bipolar hemiarthroplasty reduced the risk of mortality (Fig. 2). Table 4 summarizes the results of Cox regression analysis.

Table 4 Risk factors of mortality according to the univariate and multivariable-adjusted Cox regression analysis
Full size table
Fig. 2
figure 2

Cox regression Cumulative Hazard Ratio (CHR) curves in patients with proximal femoral fracture for each variable. A: CHR for all variable, B: CHR for Sex, C: CHR for each age category (1=60-70 Years, 2= 70-80, 3= >80), D: CHR for each AO fracture type, E: CHR based on surgical device, F: CHR based on time of admission to surgery (1= <48-hour, 2= 2–7-day, 3= >1 week), G: CHR for ASA score (1 = score 1 and 2, 3 = score 3,4)

Full size image

A separate multivariable Cox regression analysis was performed to evaluate the risk factors of mortality among intertrochanteric fractures only (AO type 31B cases were excluded). The results indicated that a bipolar hemiarthroplasty decreased the risk of mortality with a hazard ratio of 0.61 (95% CI: 0.39–0.95, p = 0.02), while age ≥ 80 years, delayed surgery, ORIF by a proximal femoral plate, and ASA scores of 3–4 were significant risk factors of mortality (Table 5).

Table 5 Risk factors of mortality in intertrochanteric fractures according to multivariable-adjusted Cox regression analysis
Full size table

Functional outcomes

The results of the mHHS scores at follow-up are listed in Table 6. The overall mean mHHS score was 53.80 ± 20.78. Patients with a femoral neck fracture (56.14 ± 21.34, p = 0.005), treatment with bipolar hemiarthroplasty (56.46 ± 21.01, p = 0.001), and DHS (54.68 ± 19.80, p = 0.001) had significantly higher mHHS scores, while patients with a type 31A3 fracture (50.06 ± 1.57, p = 0.005) and those who underwent ORIF with a proximal femoral plate (50.91 ± 19.92, p = 0.001) had a significantly lower mean mHHS. No statistically significant difference was detected among age groups (p = 0.651) and between two sexes (p = 0.412).

Table 6 Functional outcomes, according to HHS, are broken down by subscales and treatment
Full size table

Discussion

As the global life expectancy rises and the populations are aging, the incidence of fragility fractures, particularly hip fractures, is increasing worldwide. While preventive measures are being implemented, the evidence clearly shows that hip fractures are a global health challenge. This study was performed to establish the mortality rates for hip fractures in the elderly in Iran, besides the patient-, injury-, and treatment-related risk factors of early and mid-term mortality. Finally, we aimed to evaluate the functional outcomes of hip fractures in elderly population residing in Iran. The main findings of this study were the significant contribution of patients’ age, type of fracture, time of receiving surgical treatment, and type of surgical repair to the mortality following proximal femoral fractures.

We collected data from 788 patients aged over 60 years old who were treated for a hip fracture at a referral trauma center in Iran. Although several studies have been performed in the develo** world on hip fracture mortality, the common limitation of the majority is the high loss to follow-up rate. This common shortage might be largely the result of the limited access to healthcare and insurance [18, 19]. We attempted to mitigate this limitation by calling the patients with less than a year of follow-up to ascertain the patient’s health status and determine their functional outcomes according to the mHHS. After this study, we have had a largely positive experience with virtual clinic visits implemented in response to the COVID-19 pandemic and have continued the practice [20, 21]. With this approach, only about 11% of the total patient population was unavailable at one year, increasing our results’ validity.

It is well-established that women have a higher risk of hip fractures, with a female to male ratio of 1.7 to 2.5 in the literature [22, 23]. Interestingly, studies from the Middle East region have reported a much closer incidence between males and females, ranging from 0.9 to 1.4 [24, 25], which has been replicated in previous studies from Iran, at 1.1 [26, 27]. We also found a female to male ratio of 1.1 in our patients. Although this was not an epidemiologic study, our findings in line with the findings of previous studies, call the need for epidemiologic studies to determine the underlying determinants of these results.

The 1-month mortality has been reported about 3–14% in the literature. The large mortality range is partly explained by the baseline patient and injury characteristics. Regardless, we found a 1-month mortality rate of 5.7%, which is in line with the literature. Of note, only six in-hospital mortalities were recorded. We also found a 1-year mortality rate of 20.2%.

Interestingly, older studies have reported higher mortality rates of about 21–39% [28, 29], while more recent studies have reported a mortality rate of about 2.5–14.6% [30,

Conclusion

To conclude, in the study of 788 Iranian elderly patients with a hip fracture, we found 1-month and 1-year mortality rates 5.7% and 20.2%, respectively. We found several risk factors of mortality, including age ≥ 80 years, a > 48-hour delay to surgery, and pre-operative ASA scores of 3–4. Furthermore, the use of a proximal femoral plate was a significant risk factor for mortality and lower mHHS scores, while a bipolar hemiarthroplasty was associated with a decreased risk of mortality and higher mHHS scores.