Abstract
Introduction
The COVID-19 pandemic has not only been associated with significant morbidity and mortality internationally, studies have also reported reduced physical activity, poorer mental health and sleep in the general population during social lockdowns. 1 Even after the mandate on lockdowns was lifted, social distancing measures, in an attempt to limit the spread of the pandemic, were employed when individuals were not within the confines of their homes. This social distancing had thus attributed to poorer quality of life (QoL) and functional health.1,2
These effects are amplified in vulnerable populations with various chronic diseases as they were found to have poorer mental health, lower QoL, disabilities in activities of daily living, and poorer access to healthcare.3–8 The geriatric population, another at-risk group, were found to have the poorest QoL and highest anxiety levels during quarantine. 9 With the closure of community facilities, the elderly were found to be involved in less physical activities and also experienced increased anxiety symptoms with poorer emotional well-being.10,11
Osteoporotic hip fracture patients, a subset of this geriatric population, have also been implicated in the effects of the COVID-19 pandemic. Existing literature has shown that during the pandemic, these patients had poorer post-operative functional scores such as the Modified Barthel index (MBI) and Parker mobility score (PMS).12,13 Le Brun et al. found that hip fracture patients not only had poorer Charlson Comorbidity Index (CCI) and American Society of Anaesthesiologists (ASA) scores during the pandemic, but also poorer ambulatory status. 14
However, there has not been any study which objectively measured how pre-morbid functional status and QoL of hip fracture patients are affected by a pandemic. The aim of this study is to compare the pre-morbid functional status and QoL of patients with osteoporotic hip fractures during the COVID-19 pandemic with those presenting before the pandemic using objective measures. We hypothesise that the COVID-19 pandemic results in a more deconditioned population with poorer scores in functional status and QoL.
Methods
This is a retrospective cohort study performed in accordance with the Declaration of Helsinki and approved by our institutional review board, with the need for informed consent exempted. We recruited consecutive patients admitted to a tertiary institution, for a surgically managed hip fracture, between January 2018 and December 2021, with consent taken prior to obtaining data. We included patients aged 60-years and above, with an isolated intra- or extra-capsular hip fracture sustained from low-energy trauma. Low-energy trauma is defined as a fall from standing height or less. 15 Exclusion criteria included atypical femoral fractures, pathological fractures and polytrauma cases.
The study cohort was grouped into three phases based on Singapore’s response to the COVID-19 pandemic: “Pre-COVID” group included patients admitted between 2018 and 2019, before the advent of the COVID-19 pandemic; “Early-COVID” group included patients admitted in year 2020, when global healthcare systems were adjusting to the uncertainties of the pandemic and when lockdown was mandated; “Late-COVID” group included patients admitted in year 2021, after a year of enforced lockdowns and social distancing.16,17
Demographic data including age, gender and ethnicity were collected. ASA Classification was used to determine each patient’s physiological status and hence operative risk. Types of hip fracture included intra-capsular (neck of femur) and extra-capsular (basi-cervical, inter-trochanteric, subtrochanteric) proximal femoral fractures, and types of surgery included arthroplasty (bipolar hemiarthroplasty or total hip arthroplasty) and surgical fixation (intramedullary or extramedullary devices). On admission, all patients were surveyed by a trained physiotherapist in the ward via immediate recall. Objective scoring systems were used to determine each patient’s pre-morbid functional status and QoL. Patients’ main caregivers were contacted in cases of cognitive impairment. Functional status was assessed using the PMS and MBI, and QOL was assessed using the Short Form-36 (SF-36) and EuroQol- 5 Dimension (EQ5D)).
PMS was first derived in 1993 based on the degree of independence of an individual’s mobility status and has been shown to be both a useful outcome measure in terms of measuring preinjury mobility levels and reliable predictor of mortality after a hip fracture.18–20
MBI is a scale used to measure the ability to perform basic activities of daily living (ADL) and has been shown in various studies to be a useful indicator of independence, good predictor of mobility and mortality of hip fracture patients.21–23
SF-36 is a quality of health measure consisting of eight different scales which are tabulated to a Physical Component Summary (PCS) and a Mental Component Summary (MCS), showing that poorer pre-morbid SF-36 scores (PCS 38.0–39.4 and MCS 43.8–48.2) are predictive of longer length of hospitalisation after hip fracture surgery.24,25
EQ5D is a measure commonly used to assess health-related quality of life (a value of −0.594 indicates the worst state of health, and a value of one indicates no problems in every dimension), and has been shown to have good association with outcomes of hip fracture patients.26,27
The data were retrospectively processed using Pearson’s chi-squared test to compare categorical data between groups whilst the one-way ANOVA test was used to compare continuous variables between groups. Post-hoc Bonferroni analysis was performed if significant differences were found via one-way ANOVA tests. Statistical significance was defined as
Results
1539 hip fractures were admitted to our institution between January 2018 and December 2021. Of the 1539 patients, 1466 met the study criteria, with an additional 47 patients excluded due to incomplete data not available for analysis, giving rise to a final outcome measure completion of 96.8%. The remaining 1419 patients were divided into three groups, with 721 (50.8%) patients in the Pre-COVID group, 366 (25.8%) patients in the Early-COVID group and 332 (23.4%) patients in the Late-COVID group. Figure 1 shows the summary flow-diagram of the patients involved. Flowdiagram of the patients included in our study.
Demographic data, co-morbidities and pre-morbid scores.
ASA – American Society of Anaesthesiologists; CI – Confidence Interval; EQ-5D – EuroQol- 5 Dimension; MCS – Mental Component Score; MBI – Modified Barthel Index; PCS – Physical Component Score; PMS – Parker Mobility Score; SF-36 – Short Form 36.
aand/or bold indicates statistical significance.
In terms of the SF-36 scores, there was a downward trend in both PCS and MCS as the pandemic progressed, but statistical significance was only achieved when comparing the Late-COVID group to the Pre-COVID group, with poorer PCS (44.4 ± 11.2 vs 46.2 ± 10.4, 95% CI 0.05–3.47;
Discussion
To our knowledge, our centre is the first to use its patient registry to evaluate the impact of the COVID-19 pandemic on pre-morbid functional status and QoL in older adults with hip fractures. This data registry enables a robust retrospective cohort analysis with access to a large depository of patients’ data with objective clinical outcome measures of PMS, MBI and SF-36 scores.
Our study has shown that low-energy hip fracture patients presenting in the late phase of the COVID-19 pandemic have significantly poorer SF-36 scores (both PCS and MCS) than those presenting pre-pandemic (Table 1). This difference was not observed in the early-pandemic cohort, suggesting that the impact of a prolonged pandemic manifests after a time delay. This is likely due to multiple factors, such as the disruption of healthcare and social services for people with comorbidities and disabilities, with closure of rehabilitation services accelerating functional decline. 6 In addition, elderly individuals are found to have increased feelings of health anxiety 28 due to impaired stress responses 29 during the pandemic, which negatively impacts disease perception and treatment compliance, while also decreasing adaptability to lifestyle changes. 30 This reduced mental resilience, coupled with cognitive impairment in the elderly can lead to frailty syndrome, 31 that is associated with increased disability, falls and hospital admissions. 32 This may have contributed to poorer SF-36 scores from a downstream accumulation of the negative effects of increased disease severity due to treatment non-compliance, that may have been more prominent in the late-COVID period.
Movement restrictions attributed to mandated lockdowns and self-imposed social isolation have been directly linked to a decrease in overall physical activity, especially in the elderly population. 33 Reduced physical activity from movement restrictions and a lack of access to physical conditioning like day care or public spaces could also have resulted in increased frailty and worsening perception of their health. 34 Increased sedentary behaviour in the elderly has shown to impair physical function, with Meneguci et al. showing the indirect association between sedentary behaviour and BADL and IADL disability through diminished aerobic endurance, agility, dynamic balance, nutritional status and lower limb flexibility. 35 In addition, poor nutritional status from either a poorer appetite due to sedentary behaviour or lack of access to food sources in the elderly can lead to deconditioning and eventual osteosarcopenia. 36 Coupled with the reduced environmental interactions from lockdowns or social isolation, this accentuates loss of coordination, disequilibrium, proprioception and spatial disorientation 37 which worsens as the time spent in isolation increases. The perceived “balance loss” thus prequels the fear of falling. This results in a vicious cycle of reduced physical activity from the fear of falling and vice versa. These are similar to our findings, with patients presenting post-lockdown having significantly worse scores in physical domains of the SF-36 (Table 1).
Social isolation due to social restrictions and lockdowns could have attributed to the detrimental effects on the mental wellbeing of the patients in our study. Literature has outlined the deleterious effects of social isolation and loneliness on mental health, with stress, anxiety and depression being the most affected.33,38–40 Though Bu et al. has shown that loneliness from social isolation does not affect the elderly as much as other demographics, 41 the PIONEER-COVID population study on the effects of the lockdown on adults ≥60 years and above showed a significant increase in both loneliness and depressive symptoms, 42 suggesting the role which social isolation plays in worsening the mental quality of life of patients in our study. Colucci et al. 43 observed that the lockdown also resulted in a decrease in quality of life, perceived health and well-being in the elderly. These were mostly attributed to lower levels of physical activity, poorer mood, increased medical burden, memory problems, social isolation and lower energy levels after the pandemic. These same factors could have contributed to the trend observed in our patient population.
Of the different indices, only the SF-36 with the PCS and MCS showed a significant association with the progression of the pandemic. While the SF-36 and EQ-5D are overall similarly robust and reliable in making predictions for QoL, 44 some domains like Vitality, Role-Physical and Role-Emotional are not significantly correlated with the EQ-5D as shown by Rowen et al. Research by Liao et al. 45 has also shown how the EQ-5D is less sensitive to health status variations due to potential ceiling effect, which may account for why a significant association is not seen in this cohort, where pre-morbid functioning tends to yield higher scores. PMS can accurately measure mobility, 46 but lacks the comprehensiveness of non-physical domains of the SF-36, which also applies to the MBI.
This is the first study to objectively document poorer pre-morbid QoL in patients admitted with hip fractures in the late COVID-19 pandemic period (Year 2021). Though we have emerged out of the pandemic, its lingering impact on the frail and elderly presenting for medical and surgical care cannot be underestimated. As this could translate into downstream poorer outcomes, and increased healthcare costs and burden, it would be prudent to explore appropriate interventions for the elderly. Furthermore, this study also identifies specific domains in the health related QoL of the patients affected, namely the physical functioning and mental health components, and seeks to suggest fine-tuning in the allocation of appropriate resources and interventions to better fill in the gap with public health implementation. Access to mental health care can also be protocolised in this digital age. Some recommendations include –providing instructional videos through an institutional application that are downloaded during inpatient stays, and expanding to either online or physical sessions to ensure continuity of care and accountability on the patient’s part with regard to physical engagement.
The limitations of this study include data being obtained from a single institution, which may limit its application to other institutions. Our study cohort also lacked data on history of COVID-19 infections, which could shed light on its immediate impact on the pre-fracture physical function and quality of life of such patients. A systematic review and meta-analysis by Marjenberg et al. 47 showed significantly worse health-related quality of life in hospitalised COVID-19 patients, and patients suffering from long COVID, with poorer scores in mobility, pain and psychological health. However, prior COVID-19 infections were not factored into our study as doing so required an extensive consideration of factors not within the scope of this study – duration of infection, need for hospitalisation, need for ICU admission, presence of long COVID, vaccination history etc. In addition, there is a lack of data on outcomes post-surgery that may be affected by their respective pre-morbid status, that may have potentially skewed the outcomes – more in-depth reviews of the post-operative outcomes should thus be looked into to ensure better reproducibility of predictions. We also acknowledge that there is a lack of determination of clinical significances in our patient-reported outcomes. Till date, there has been no study on the minimal clinically important difference (MCID) of hip fracture patients, as existing literature involves only that after hip arthroplasty. As MCID is unique to the condition, we are unable to superimpose the derived data in current literature on the hip fractures in our study. Further prospective studies on MCID of hip fractures can be studied to determine the effect of hip fractures on the QoL of patients.
Conclusion
In conclusion, the COVID-19 pandemic and its associated disruptions to rehabilitation services and social interactions have been shown to be associated with a poorer premorbid quality of life of elderly, low-energy, hip fracture patients. Adequate and regular access to both virtual and in-person quality rehabilitative modalities is paramount to ensuring that the elderly do not suffer from preventable decrease in their quality of life.
