Super elderly patients with intertrochanteric fractures do not predict worse outcomes and higher mortality than elderly patients: a propensity score matched analysis

Introduction

With the aging of the population, the prevalence and incidence of hip fractures have increased rapidly, particularly in developing countries. It has been reported that the number of elderly Chinese people aged over 60 had reached 249 million, accounting for 17.9% of the total population by 2018, and it has been estimated that by 2050, that number will exceed 450 million, comprising over 30% of the total population [1]. Wang et al. [2] also reported that the elderly population in China comprised nearly 20% of the elderly population worldwide. Intertrochanteric fractures, as one of the major orthopedic clinical problems in China that is a common cause of illness, disability and mortality in the elderly people, leads to a heavy socioeconomic pressure on society.

According to the literature, the postoperative recovery for these fragile patients with intertrochanteric fractures is not to be well expected due to preexisting severe comorbidities, delay of surgery, anemia and adverse reactions following anesthesia and blood transfusion, which are associated with a high risk of perioperative complications and mortality and thus can be a catastrophic event precipitating a steep decline in health and independence [3–10]. To the best of our knowledge, the possible role of age is still under discussion. Several previous publications [11–15] have shown that functional impairment increases with age, while others [16–20] have suggested that age is not associated with adverse outcomes. However, relevant research on whether super elderly patients aged over 90 years have significantly worse functional outcomes and survival rates than others is still limited.

The objective of this study was to evaluate the population of super elderly patients aged over 90 years with intertrochanteric fractures and treated with intramedullary fixation on the functional outcomes, perioperative complications, and mortality.

Results

From Jan 2014 to Jan 2019, a total of 3560 consecutive patients presenting with intertrochanteric fracture were screened and assessed for eligibility to participate in this study. A total of 1318 patients were eliminated by the exclusion criteria. Among these, 596 patients were under the age of 65 years (213 with age 60 to 64, 108 with age 55 to 59, 107 with age 50 to 54, 168 with age < 50); 468 patients (46 with age ≥ 90, 422 with age 65-89) received conservative treatment; 182 patients (22 with age ≥ 90, 160 with age 65-89) had an admission delay of more than 48 hours; 27 patients had multiple fractures or injuries, pathological or open hip fractures; and 45 patients were lost to follow-up. Finally, 2242 patients, including 2036 in group A with age < 90 years and 206 in group B with age ≥ 90 years, presenting with intertrochanteric fracture met our inclusion and exclusion criteria (Figure 1).

The baseline characteristics of patients in the two groups are summarized in Table 1. More than half of the fractures were stable and the majority of the patients were female in both groups. Characteristics including the American Society of Anesthesiologists (ASA) grade, body mass index (BMI), Hb level at admission, type of anesthesia, volume of intraoperative blood loss, transfusion requirement, and Elixhauser comorbidity scores [21, 22] were significantly different between the two groups. After propensity score weighting, there were 192 matched patients in each group, and the baseline characteristics were balanced between the two cohorts in terms of all covariates (Table 1).

Pre-matching and post-matching results, including perioperative complications, functional outcomes, and 30-day, 90-day, and 1-year mortality, are shown in Table 2. The statistical distribution showed that perioperative complications, including severe complications, cardiac complications, pulmonary complications, endocrine/metabolic complications and outcomes, were not significantly different between the two groups after propensity score matching (PSM), although the differences were significant before PSM (p< 0.001, p< 0.001, p< 0.001, p = 0.026 and p< 0.001, respectively). No significant difference was observed in other complications, such as neurological complications and hematological complications, before and after PSM. The crude mortality rate was 0.89% at 30 days, with no significant difference between the two groups before and after PSM. The mortality rates at 90 days and 1 year were higher in patients aged ≥ 90 years than in patients aged < 90 years group (p=0.015 and p<0.001). However, there was no significant difference between the two groups after PSM.

For relationships between the Elixhauser comorbidity score and survival state/functional outcomes, the data shown in Table 3 demonstrated progressive increases in mortality and poor functional outcomes with increasing Elixhauser comorbidity score, which represented the severity index of patients preoperatively (p<0.05). By using the Spearman method, Table 4 showed the weak correlation between patient age and the factors of ASA grade, Elixhauser comorbidity score, Hb level at admission, intraoperative blood loss, and transfusion requirement.

The average follow-up was 37 months. The overall mortality rate of all patients was 20.38% at the end of this study, and deaths occurred predominantly during the first 4 years. The cumulative survival rates of the two groups of patients were gradually decreased within the first three years after injury and stabilized thereafter. Patients older than 90 years had a risk of death at 1 year that was 0.98 times as high, a risk of death at 2 years that was 1.6 times as high, a risk of death at 3 years that was 1.5 times as high and a risk of death at 4 years that was 1.3 times as high as the risk of the patients less than 90 years. However, the Kaplan-Meier survival curve showed no significant difference between the two groups of patients in cumulative survival rate. (Figure 2, p=0.081, log-rank).

Discussion

In the current study involving 2242 patients, we observed that in general, there was a higher ASA grade, Elixhauser comorbidity scores, and transfusion rates and a lower Hb level at admission, BMI, volume of intraoperative blood loss and percentage of general anesthesia in super elderly patients (with age ≥ 90 years) than in younger patients (with age < 90 years). However, after PSM and McNemar’s tests, we confirmed that super elderly patients do not predict a worse clinical outcomes in terms of functional outcomes, perioperative complications, and mortality than younger patients. Additionally, the discrepancy of all outcomes can likely be attributed to the differences in the baseline characteristics of the study populations. In other words, age is not a well-studied prognostic factor for patients’ functional outcomes, perioperative complications, and mortality, especially in super elderly patients based on the results of this study.

Functional outcomes after hip fracture surgeries are likely to be multifactorial. Numerous studies [11–14, 16] have shown that age, gender, ASA-grade, transfusion requirement/volume and comorbidities were predictors of morbidity and mortality after surgery for hip fractures. In addition, the impact of other factors, such as surgery delay, type of anesthesia, duration of operation, and volume of intraoperative blood loss, have not been consistently demonstrated. For example, some studies [13, 16, 23, 24] suggested better survival states and lower morbidity rates with early surgery while other studies [25, 26] suggested no difference due to surgery delay. Another example is that the choice of anesthesia for hip fracture surgery in elderly patients is still controversial [10, 27–30]. However, it should be noted that while the age factor has been shown to be associated with mortality after hip fracture surgery, it is not always adjusted for other covariables that may confound survival analyses.

In this study, we used PSM to minimize the effects of possible confounding variables. The differences represented the effect of age on ASA grade, BMI, Hb level at admission, type of anesthesia, volume of intraoperative blood loss, transfusion requirement, and Elixhauser comorbidity scores before PSM were eliminated in the present study. Our result implies that it is not advanced age itself but other factors associated with age that appear to be responsible for adverse outcomes and mortality in elderly patients, which is consistent with Sophia et al. [17]. In a review of twelve studies based on large databases that addressed the influence of age on outcome, Sophia et al. [17] revealed that it is not age itself but concomitant factors, such as severity of illness and premorbid functional status, that might influence prognosis. Similarly, several previous studies [18–20] using multivariate analyses also concluded that age was not an independent risk factor for mortality. A cohort study performed by Somme et al. [20] also found that advanced age is not directly associated with mortality. Compared to these former studies, the strength of our study lies in the more recent data included from 2014 to 2019, with a relatively large sample size.

According to the baseline characteristics of patients in the two groups, a greater percentage of patients had an ASA grade 4 or 5 and an Elixhauser comorbidity score of 6 or higher in the super elderly patient group than in the elderly patient group, indicating that the older the patients, the more severe the comorbidities they had. In view of this, early identification of super elderly individuals prone to postoperative morbidity and mortality and accelerated care, consisting of rapid medical optimization, are of vital importance. Previous studies have shown that good postoperative recovery for elderly patients with hip fractures was not expected because of poor surgical tolerance and tendency for anemia, which were associated with significant adverse clinical outcomes [3, 4, 31, 32]. Gruson et al. [31] found that 45.6% of emergency surgical patients had anemia, and unexplained anemia accounted for approximately one-third of these patients [3]. In the current study, there were significantly more patients with an Hb level of 10 g/dL or less at admission in the super elderly group than in the elderly group (51.0% vs. 27.4%). Among surgery-related factors, patients aged 90 years or older received more regional anesthesia and had more transfusion requirements but less intraoperative blood loss than the younger patients. After using the Spearman method, we further demonstrated that the aforementioned factors of ASA grade, Elixhauser comorbidity score, Hb level at admission, intraoperative blood loss, and transfusion requirement, even though there was only a weak correlation, were related to the patient age. However, we did not find any significant relationship between anesthesia type and age, despite regional anesthesia being most common in the 90 years or older age group.

Although many published articles [6, 7, 33, 34] have previously performed survival analyses and studied the short-term outcomes of elderly patients with hip fractures, longer-term survival and functional outcomes are less frequently reported. In our seven-year follow-up, 457 of 2242 total patients died at the end of the study, and the mortality rates of super elderly patients and elderly patients were 30.1% and 19.4% before PSM, respectively. Our data suggest that the mortality rate is lower than previous findings [6, 7, 33]. This can be ascribed to the patients who receive conservative treatment being excluded due to the severe comorbidities, and we restricted the study population to surgical patients. After PSM, the survival analysis showed that there was no significant difference between the two groups of patients in terms of cumulative survival rate. (p=0.081, log-rank). Marginally significant results (p-value close to 0.05) should be interpreted with caution; however, there was also a marginally significant trend towards a difference in survival rate between the two groups. There is possible selection bias, and caution should therefore be taken when generalizing related results and conclusions.

Surprisingly, we found that most patients could walk independently or with the help of walking aids, while only a relatively small percentage of patients were restricted to wheelchairs or beds. Ekstrom et al. [34] concluded that only approximately 55% of patients regain their previous walking ability, and approximately 34% of patients lose their previous level of daily living function. Based on our data, 27.2% of super elderly patients and 41.4% of elderly patients walked independently before PSM, while the percentages were 27.1% and 37.0%, respectively, after PSM. In addition, a notable proportion of individuals are able to walk independently but still require help with walking aids to avoid falling again, which leads to a decrease in the independent walking status ratio.

Many previous studies have assessed the factors that influence the functional outcomes and mortality rates of elderly hip fracture patients. However, this is the first study to assess the effect of age on functional outcomes, perioperative complications, and mortality in super elderly patients with intertrochanteric hip fracture. Such quantitative analyses might increase the orthopedist’s confidence in treatment management for super elderly patients and be beneficial for clinicians looking to establish probabilities for adverse results in the future and making and establishing rational goals of medical care for this patient population. The strength of this study includes an appropriate comprehensive preoperative evaluation of functional status using Elixhauser comorbidity scores, a validated and simple scoring system that accounts for 30 comorbidities. Other strengths are the PSM method we used and the specific cohort of patients who received surgery by single internal fixation, which eliminated the effects of possible confounding variables. Finally, the cohort comprises a relatively large number of patients and has a long-term follow-up. The limitations of this study include its retrospective design and the data being collected in a single center. In addition, we did not control all risk factors, such as smoking and other unknown confounders, including perioperative laboratory values and surgeon practice, for analysis, which may potentially influence the findings. However, this is the first study including multiple relative contributions of patient, fracture, surgical, anesthetic and transfusion factors, which have not been studied together.

In conclusion, super elderly patients aged more than 90 years do not predict worse clinical outcomes in terms of functional outcomes, perioperative complications, and mortality than elderly patients aged 65-90 years treated surgically with proximal femoral nail anti-rotation (PFNA) for intertrochanteric fractures.

Materials and Methods