Clinical and Demographic Characteristics of Centenarians versus Other Age Groups Over 75 Years with Hip Fractures

Introduction

The number of individuals aged 100 years or more is estimated to quadruple by 2033 in Spain. According to the Spanish National Statistics Institute, the number of centenarians will increase from 11,248 in 2018 to 46,366 in 2033. Nevertheless, the literature on centenarian patients is scarce.¹

The incidence of hip fractures in centenarians is seven-fold higher than in younger age groups.² Several studies in Spain and elsewhere have shown that hip fractures are the most common cause of hospital admission in centenarians, ahead of heart failure and pneumonia.^3,4

Studies on hip fractures in centenarians are limited in number of patients and show contradictory results.^5–24 Differences have been described in relation to the type of fracture,^2–8 the relationship between time to surgery and outcomes⁵ and the risk of complications during hospitalisation.⁹ Discrepancies have also been published regarding short and long-term outcomes.^6–8,11,12 A recent study including more than 4000 centenarians with hip fracture focused exclusively on complications during hospitalization, but did not include outcomes related to mortality or functional status in the first month after the fracture.¹⁰

Considering the above, the primary objective of this study was to describe the clinical characteristics, management, and outcomes of centenarians admitted with fragility hip fracture and to compare them with those observed in patients aged 75 to 79, 80 to 89 (octogenarians) and 90 to 99 years old (nonagenarians) using data from the Spanish National Hip Fracture Registry (Registro Nacional de Fracturas de Cadera, RNFC). The secondary objective was to determine the variables associated with length of stay, and both in-hospital and 30-day mortality in older patients including the centenarian group.

Materials and Methods

The RNFC registry is a multicentric, prospective and continuous registry of patients treated for hip fractures in participating hospitals. The RNFC has been approved by the Research Ethics Committees of the participating centres and complies with the Declaration of Helsinki.

The RNFC includes all patients 75 years or older admitted for fragility hip fractures (due to falls from standing height) who have signed an informed consent (signed by the patient or their next of kin) in 86 participating hospitals from Spain between January 2017 and June 2019. Hip fractures occurring as a result of high-energy trauma and fractures in patients with known disseminated cancer are excluded.

The design and organization of the RNFC have been described in greater detail in a previous article.²⁵ Baseline variables were recorded such as age, sex, place of residence, pre-fracture mobility (using Functional Ambulation Categories, FAC), cognitive status (using Pfeiffer’s SPMSQ) and the ASA (American Society of Anaesthesiology) score, as well as whether the patient had received bone protection medication before the fracture. The clinical variables associated with hospital admission were type of fracture, surgical delay, type of surgery performed, type of anaesthesia, use of peripheral nerve blocks, development of pressure ulcers, and whether the patient had been mobilized on the first postoperative day. Variables included at discharge were destination at discharge, length of hospital stay, anti-osteoporotic medication prescribed at discharge and in-hospital mortality. Data at 30 days was collected by contacting the patient or caregivers directly and included readmission, repeat surgery, use of anti-osteoporotic medication, functional status and mortality.

Statistical Analysis

Patients were divided into four age groups: 75–79 years old, 80–89 years old, 90–99 years old and ≥100 years old. These groups were characterized by means of a descriptive analysis for categorical variables using absolute and relative frequencies; numerical variables were characterized using the mean and standard deviation or median and interquartile ranges, according to the normality of the data. The χ² test was used to compare categorical variables and the Kruskal–Wallis test to compare numerical variables. Linear regression was carried out to study variables related to the length of stay (in days) and logistic regression to assess in-hospital and 30-day mortality. Correlation coefficients and OR with their corresponding 95% confidence intervals are shown, respectively. The differences were considered as statistically significant if the p value was <0.05. The statistical package used was Stata/IC v.15.1. (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC).

Results

Of 25,938 patients included, 253 (1.0%) were centenarians, 2,888 (11.1%) were between 75 and 79 years of age, 14,762 (56.9%) were octogenarians, and 8,035 (31.0%) were nonagenarians.

In the centenarian group, 83.3% were women, 32.8% had severe dementia, 8.8% had severe dependency and 36.0% lived in residential care homes. The most frequent fracture was intertrochanteric (59.6%) and the most common procedure used was cephalomedullary nailing (63.7%). Peripheral nerve blocks were used in 17.1% of patients. Time to surgery ≤48 hours occurred in 52.8% of the centenarians. Six out of ten patients were mobilized the first day after surgery and 12.2% developed pressure ulcers in hospital. The median length of hospital stay was 8.6 days, and in-hospital mortality was 10.3%. At discharge, 14.1% of them received anti-osteoporotic medication, 33.6% calcium and 55.9% vitamin D supplements. Of the patients, 42.3% were transferred to nursing homes and 11.9% to geriatric rehabilitation units. Regarding outcomes at 30 days, 20.9% died, 1.8% were readmitted and 4.9% underwent repeat surgery.

The group of centenarians was compared to the other age groups, showing that the frequency of women (p<0.001), those living in care facilities (p<0.001), and those with severe dementia (p<0.001), severe functional impairment (p<0.001), and intertrochanteric fracture (p<0.001), increased in the older age groups (Table 1 and Table 2). Non pathological fractures in the surgical pieces were diagnosed in centenarian patients (1.6%, 0.7%, 0.5% and 0%, respectively per age group) and three centenarians had atypical fractures (0.5%, 0.5%, 0.5% and 1.2%; p<0.001).

Table 1 Demographic and Baseline Characteristics of the Patients Included for Analysis

Table 2 Fracture and in-Hospital Management Characteristics of the Patients Included for Analysis

Time to surgery was less than 48 hours in a greater proportion of centenarians than in the younger age groups (p=0.015). The older age groups had lower percentages of mobilization on the first day after surgery (p<0.001) and had higher frequencies of pressure ulcers developed in hospital (p<0.001) (Table 2).

Regarding hospital outcomes, in-hospital mortality was higher as the age of the groups increased (p<0.001) (Table 2). The median length of stay was also shorter among centenarians (8.8 vs 9.0 vs 9.0 vs 8.6 days, respectively per age group; p=0.004). At discharge, a greater proportion of patients in the older age groups were sent to residential care facilities (19.8%, 30.0%, 38.7% and 42.3%). Centenarians were also less likely to be referred to geriatric recovery units (19.5%, 22.1%, 19.6% and 11.7%). Anti-osteoporotic medication, calcium and vitamin D supplements were much less likely to be prescribed in the older age groups (Figure 1A–C). Figure 2 shows functional status at baseline and 30 days after the fracture using Functional Ambulation Categories.

Figure 1 Bone health prescriptions before the fracture, at discharge and at one-month follow-up, by age group. (A) Anti-osteoporotic medication (antiresorptive treatment and bone-forming agents); (B) Calcium supplements; (C) Vitamin D supplements.

Figure 2 Functional status at baseline and 30 days after the fracture using Functional Ambulation Categories.

Table 3 shows characteristics at 30 days after hip fracture by age group. Centenarians had a higher risk of 30-day mortality than the other age groups (p<0.001).

Table 3 Characteristics at One-Month Follow-Up After Hip Fracture, by Age Group

In-Hospital Mortality

Multivariate analysis found that the following variables were associated with in-hospital mortality: female sex (OR 0.58, 95% CI: 0.48–0.70; p<0.001), time to surgery >48 hours (OR 1.30, 95% CI: 1.08–1.55; p=0.005]), baseline ambulatory capacity with FAC 0 as a reference (FAC 2 [OR: 0.59, 95% CI: 0.37–0.92; p=0.022], FAC 3 [OR: 0.59, 95% CI: 0.37–0.95; p=0.032], FAC 4 [OR: 0.49 [95% CI: 0.36–0.67; p<0.001] and FAC 5 [OR: 0.29, 95% CI: 0.21–0.40; p<0.001]), anaesthetic risk with ASA score I as a reference (ASA score IV [OR 9.29, 95% CI: 1.27–68.1; p=0.028] and ASA score V [OR 15.4 [95% CI: 1.27–187]]; p=0.032)), peripheral nerve block (OR 0.67, 95% CI: 0.51–0.87; p=0.003]), neuraxial anaesthesia (OR 0.66, 95% CI: 0.49–0.90; p=0.008]), and age group, taking <80 years as a reference (80–89 years [OR 1.44, 95% CI: 1.00–2.07; p=0.048]; 90–99 years [OR 2.41 [95% CI: 1.67–3.48; p<0.001]]; ≥100 years [OR 4.55, 95% CI: 2.26–9.13; p<0.001]). In centenarians, multivariate analysis showed no association with in-hospital survival for any of these variables.

Factors Related to Length of Stay

Multivariate analysis found that the variables associated with length of stay were time to surgery >48 hours (coefficient: 3.86, 95% CI: 3.59–4.13; p<0.001), baseline ambulatory capacity, taking FAC 0 as reference (FAC 2 [coefficient: 2.01, 95% CI: 1.18–2.84; p<0.001], FAC 3 [coefficient: 1.75, 95% CI: 0.87–2.63; p<0.001], FAC 4 [coefficient: 1.62, 95% CI: 0.99–2.25; p<0.001] and FAC 5 [coefficient: 1.67, 95% CI: 1.04–2.29; p<0.001]), anaesthetic risk, taking ASA score I as reference (ASA score IV [coefficient 2.11, 95% CI: 0.36–3.85; p<0.001]), and type of fracture, taking as reference non-displaced intracapsular fracture (subtrochanteric fracture [coefficient 0.84, 95% CI: 0.17–1.51; p=0.014] and other fracture [coefficient 1.86, 95% CI: 0.12–3.61; p=0.036]).

In centenarians, only a time to surgery over 48 hours was related to length of stay (coefficient 3.99 [95% CI: 2.35–5.64; p<0.001]).

30-Day Mortality

Multivariate analysis found that the following variables were associated with 30 day-mortality: female sex (OR 0.47, 95% CI: 0.41–0.54; p<0.001), time to surgery >48 hours (OR 1.24, 95% CI: 1.08–1.42; p=0.002]), baseline ambulatory capacity, taking FAC 0 as a reference (FAC 3 [OR: 0.54, 95% CI: 0.38–0.79; p=0.001], FAC 4 [OR: 0.48 [95% CI: 0.37–0.61; p<0.001] and FAC 5 [OR: 0.31, 95% CI: 0.24–0.40; p<0.001]), baseline cognitive function, taking as a reference no cognitive impairment (moderate dementia [OR: 1.34, 95% CI: 1.11–1.63; p=0.002] and severe dementia [OR: 1.53, 95% CI: 1.27–1.84; p<0.001]), anaesthetic risk, taking ASA score I as reference (ASA score III [OR 8.01, 95% CI: 1.11–58.0; p=0.039], ASA score IV [OR 16.3, 95% CI: 2.24–118; p=0.006] and V [OR 54.5, 95% CI: 5.61–530; p=0.001]), peripheral nerve block (OR 0.77, 95% CI: 0.63–0.93; p=0.007), neuraxial anaesthesia (OR 0.64, 95% CI: 0.51–0.80; p<0.001), and age group, taking <80 years as a reference (80–89 years [OR 1.72, 95% CI: 1.29–2.29; p<0.001]; 90–99 years [OR 3.09, 95% CI: 2.31–4.13; p<0.001]; ≥100 years [OR 6.43, 95% CI: 3.76–11.0; p<0.001]).

In centenarians, after multivariate analysis only the anaesthetic risk, taking ASA score V as reference (ASA score II [OR 0.25, 95% CI: 0.09–0.70; p=0.009] and III [OR 0.43, 95% CI: 0.19–0.96; p=0.039]), remained significantly associated with 30-day mortality.

Discussion

The present study provides three main results. Firstly, centenarians were more frequently women, had a high prevalence of dementia and dependency and nearly four out of ten were institutionalised. The most frequent fracture was intertrochanteric. The length of hospital stay was 8.6 days, in-hospital mortality was 10.3% and 30-day mortality was 20.9%.

Secondly, older age groups had greater percentages of women, severe functional dependency, severe dementia, intertrochanteric fracture, living in care facilities, and being discharged to nursing care. Early mobilisation, treatment for osteoporosis and being discharged to rehabilitation units were less frequent. In-hospital and 30-day mortality were higher with increasing age of groups.

Thirdly, time to surgery was independently associated with length of stay and anaesthetic risk was related to 30-day mortality in the centenarian group.

Our study showed a cohort of centenarians which were commonly women and had a high frequency of functional dependency, cognitive impairment and institutionalisation, in contrast to some previous studies.²⁶ The prevalence of cognitive impairment in patients with hip fractures varies widely among studies, ranging from values similar to those in our group of younger patients (40%) to others more similar to those in our group of centenarians (80%).²⁷

Intertrochanteric fractures were the most frequent in centenarians and also increased with ageing as observed in other studies from Spain and the United States.^4–7 In contrast, intracapsular fractures were reported to be more frequent in Denmark.⁸

One out of two centenarians underwent surgery in less than 48 hours. This frequency was even higher than that of some of the younger age groups. In our study, time to surgery >48 hours was associated with length of hospital stay but not with survival in centenarians. Our finding was consistent with that of other authors,^5,28 who concluded that surgery in centenarians in less than 48 hours did not reduce mortality. Some authors showed similar rates of surgery ≤48 hours without shorter length of hospital stay compared to other age groups,⁹ while others reported longer surgical delay (1.9 vs 3.6 days) and length of stay (14 vs 20 days).²⁸

We would like to highlight the lower percentage of mobilisation on the first day and the higher frequency of pressure ulcer development the older the age group. In the centenarian group, six out of ten were mobilized the first day after surgery and nearly one out of ten developed pressure ulcers during their hospital stay. This may be due to the higher baseline degree of cognitive impairment and dependency or the higher degree of in-hospital complications in this age group.^9,11,12

Centenarians received anti-osteoporotic medication (14.1%), calcium (33.6%) and vitamin D supplements (55.9%) at discharge. Bone-protective medication was much less likely to be prescribed in older age groups, despite a seven-fold higher incidence of hip fracture in centenarians than in younger patients² and despite hip fractures being the most common cause of hospitalization in centenarians.³ In view of the current evidence, the low rate of primary and secondary fracture prevention in centenarians does not seem justified.

The median length of hospital stay was 8.6 days, lower than that of some younger age groups. Length of stay was shorter in centenarians, as observed in other studies.²¹ As mentioned above, the length of hospital stay was associated with time to surgery >48 hours, baseline ambulatory capacity, anaesthetic risk and type of fracture in our entire sample. Only time to surgery was an independent factor in centenarians.

Institutionalization at discharge increased with age, but referral to geriatric rehabilitation units at discharge was lower in centenarians. Functional impairment was greater in centenarians compared to the younger age groups. It has been previously described that centenarians are at higher risk of functional decline and dependency during hospitalization.⁴ This could be explained in part by the higher prevalence of cognitive impairment, but it would be interesting to study which variables affect this functional decline and to quantify the effect of decreased rehabilitation in this age group, as this is a modifiable factor, since dementia does not appear to be a clear factor in functional deterioration after hip fracture.²⁹ Orthogeriatric interventions have proven to provide the most benefit to the oldest patients and those with functional and cognitive impairment; they have been shown to reduce mortality, complications, length of stay and functional impairment following hip fractures. Centenarians are likely to be among those who would benefit the most from these interventions.²⁷

The in-hospital and one-month mortality were 10.3% and 20.9%, respectively, and was higher as the age of the groups increased. Ogawa T et al showed that age was an independent predictor of in-hospital mortality.¹⁰ Thirty-day mortality has been reported to be approximately 8–34% in older patients.^{6–8,11,12,18,23} Our findings showed that female sex, time to surgery >48 hours, baseline ambulatory capacity and cognitive function, anaesthetic risk, type of anaesthesia and peripheral nerve block, and age group were related to 30-day mortality over the entire sample. However, only the ASA score was associated with 30-day mortality in centenarians. In contrast with previous studies, we did not find an association with type of fracture. Mortality after subtrochanteric fractures has been described to be almost twice as high as after other types of fractures.^30,31 Also, while other authors have concluded that arthroplasty reduced the risk of mortality compared to internal fixation,⁵ we were unable to find differences between different surgical interventions. Regarding anaesthesia, evidence has been conclusive as to the effect of nerve blocks on the reduction of pain and delirium, but their association with functional recovery and mortality was less clear.³²

The results of the present work increase the evidence for the effect of age on in-hospital and short-term mortality in elderly patients with hip fracture. Despite this, age per se should not be considered a limiting factor to surgery in very elderly patients. As occurs in younger age groups, surgical treatment of hip fractures has been shown to reduce the risk of mortality in centenarians, compared to those managed non-operatively.²⁴ In the case of centenarians, we have found that their previous medical condition is a prognostic factor and, therefore, the assessment of surgical risk in this age group becomes even more important. The involvement of geriatricians in this decision-making process is crucial in the very elderly, as is taking into consideration the patient’s rights and wishes.³³ Furthermore, the multidisciplinary team should work to minimize surgery time, optimize the patient prior to surgery, mobilize the patient early and prevent postoperative complications.

We are aware of several limitations of our study. There are a significant number of variables that are not collected in the registry, for example, geriatric syndromes such as nutritional status³⁴ or delirium^35,36 that also contribute to mortality and length of stay and are modifiable factors in older persons. Neither does the registry include perioperative variables such as the need for transfusion or time of surgery, nor postoperative weight-bearing status, likely related with destination and mortality at discharge. The latter has recently been added to the registry’s list of variables. Finally, the number of centenarians in our analysis was lower than the number of patients in other age groups, limiting statistical power and thus increasing the likelihood that associations were not found despite them existing. However, our series is one of the largest published to date, using data collected prospectively from a large number of hospitals in Spain, increasing the external validity of our conclusions. Furthermore, Spain is one of the countries with the highest life expectancy and largest number of centenarians worldwide, with the RNFC providing an ideal framework to study hip fractures in this age group. In contrast to other studies using automated datasets, the RNFC includes follow-up after discharge collected by participating clinicians, providing insight into the functional decline, mortality and institutionalization due to the fracture.

Conclusions

Centenarians were more likely to suffer intertrochanteric fractures and had worse baseline functional status, cognitive status, and short-term outcomes than other age groups. However, nearly 4 out of 5 centenarians were alive in the first month after surgery. One-month survival seemed to be affected more by baseline comorbidity as estimated by anaesthetic risk than by other factors.