Health related quality of life in COVID-19 survivors discharged from acute hospitals: results of a short-form 36-item survey

Introduction

More than 100 million cases of coronavirus disease 2019 (COVID-19) have been reported worldwide since January 2020.¹ Recent publications have described various persistent physical, cognitive and psychological symptoms in COVID-19 survivors, also named “long-COVID”,² following their discharge from acute hospital units.^3–5 These symptoms include fatigue, difficulty in breathing, difficulty in concentration and memory, pain, anxiety, depression and symptoms of post-traumatic stress disorder. Some authors have found an association of these symptoms to reduced functioning and performance in activities of daily living (ADL)⁶ with an improvement when a period of rehabilitation was offered.^6,7 These findings appear to confirm what was expected based on previous literature on post ITU syndrome^8–10 and pulmonary rehabilitation^11,12 and have an important implication for implementing follow up services and rehabilitation pathways.^13,14

Health-related quality of life (HRQL) is an important domain for evaluating the impact of a disease in the longer term, which can be measured by a large variety of either disease-specific or generic questionnaires. The Short Form 36 Health Survey (SF-36) is one of the most widely used generic HRQL questionnaires. It was developed in the USA for use in the RAND corporation's health insurance experiment¹⁵ and has subsequently been used around the world to gauge the health of populations and to help with service planning. A validated Italian version is available.¹⁶ The SF-36 contains eight scales, assessing the quality of life across different domains with physical and mental components. The SF-36 has been found to be a valid instrument to measure HRQL in patients with other chronic respiratory problems such as COPD¹⁷ and idiopathic pulmonary fibrosis.¹⁷ HRQL has been evaluated in a post COVID-19 population at one month following their discharge from acute hospitals in Wenzhou, China,¹⁸ showing significantly poorer SF 36 scores, particularly in the dimensions of physical and emotional role (RP and RE) as well as social functioning (SF), when compared to the Chinese population norm. Based on their findings, these authors highlighted the importance of addressing physical and psychological long-term suffering in post COVID-19 patients.

Preliminary results have been published to evaluate the response to pulmonary rehabilitation in a cohort of post COVID-19 elderly patients,⁷ showing a positive effect both on respiratory function as well as on HRQL.

The focus during the acute COVID-19 infection is on the survival of the patient followed by physical functioning in basic ADLs in the post-acute phase.¹⁹ In the authors’ opinion other domains beyond basic ADLs, such as role limitations, vitality, and social and emotional functions become at least as important once patients are discharged home expecting to resume their previous roles and functioning.

The aim of our study is to determine the quality of life in a group of COVID-19 survivors following their discharge from acute hospitals using the SF-36 questionnaire.

Methods

Results

Baseline demographic and clinical characteristics of the population are reported in Table 1.²⁷ A total of 45 out of 80 (56%) patients were eligible for the study and agreed to participate, of which 35 patients completed the SF-36 questionnaire. Of these, 12 were female and 23 were male, with a mean age of 60 years (±15) ranging from 30 to 85 years; 63% were below 65 years. The average CIRS score was 9 (±4.3), ranging from 3 to 25.

At the time of the SF-36 administration, the average time in days since the initial hospital admission for COVID-19 related symptoms was 61.8 days (±15.5), ranging from 20 and 95, while the average time from the discharge from acute hospitals was 39 days (±14.8), ranging from 15 and 71 days.

Table 2 shows the scores for the different SF-36 domains. The mean (SD) SF-36 scales scores in ascending order (from the lowest to the highest) were as follows: limitations in role physical RP = 6.4 (17.5); limitations in role emotional problems RE = 30.5 (27.2); social functioning SF = 48.2(19.2); general health perceptions GH = 42.4 (17.7); energy and fatigue EF = 48.4 (15.3); physical functioning PF = 51.6 (24.4); emotional well-being EWB = 60.2 (16.8); bodily pain BP = 60.8 (19.2). All scores were significantly different from the normative data (p < 0.01) except emotional wellbeing (t = −1.79, p = 0.07).

Higher age was significantly associated with higher physical functioning (t = 4.34, p = 0.01; r(32) = 0.57, p = 0.01) and higher energy fatigue (t = 3.14, p = 0.05; r(32) = 0.49, p = 0.03). Whilst age was significantly associated to higher general health (t = 3.09, p = 0.05), the correlation is non-significant (p = 0.09). Higher performance on the 30sSTS test was significantly associated to higher physical functioning (t = 4.01, p = 0.02; r(23) = 0.63, p = 0.01). Finally higher CIRS was significantly associated with lower physical functioning (t = -3.63, p = 0.03; r(33) = -0.58, p < 0.01), as well as being significantly correlated with poorer general health (r(33) = -0.57, p = 0.01). Time between acute admission and SF-36, as well as time from discharge to SF-36, sex, MRC-D, HADS A and HADS D showed non-significant correlation to SF-36 scores. Correlations for the components of the SF-36 with demographic and clinical factors are shown in Table 3. After adjusting for age, only age remained significantly associated with the SF-36 scores. The full results from univariate and multivariate analysis are provided as supplementary material.

Discussion

We describe the impairment of HRQL in both physical and psychological functioning in COVID-19 survivors about a month (average 39 days) following their discharge from an acute hospital unit. All eight scales of the SF-36 showed significantly lower scores in comparison to the normative data for the Italian population¹⁶ apart from emotional wellbeing. Participants’ role limitation due to physical problems was particularly impaired, followed by role limitation due to emotional difficulties. The SF-36 role limitation domain evaluates the perception of restriction in people’s own job and daily activities as result of their physical health (RP) or emotional problems (RE), such as anxiety or depression. The restriction is assessed by the SF-36 in term of perceived effort, amount of time spent, quality and efficiency in carrying out work or previous activities.

We found no correlation between the role limitation scores and the baseline characteristics considered in this study. A prolonged time to recovery following COVID-19 infection as well as the difficulty to access pulmonary and vocational rehabilitation might have contributed to this delay in returning to previous roles, independently from the baseline variables being considered.

Cognitive difficulties, particularly memory and concentration deficits, have been described in the spectrum of the post COVID-19 persistent symptoms.⁵ Although our study did not include any cognitive evaluation, the presence of cognitive difficulties and the presence of mental fatigue could have certainly affected the efficiency in carrying out tasks.²⁵ The restrictions imposed by the policies to contain the COVID-19 pandemic might have been an additional barrier for functioning in the social context. It is relevant in this regard that two thirds of the patients were below 65 years-old and hence in the working age range so that their delayed return to work could have caused a significant financial loss for the patients themselves, their families and employers.

A lower physical functioning scale significantly correlated to a higher CIRS as well as a poorer performance in the sit to stand test, suggesting that people with a higher burden of co-morbilities and who experienced a more pronounced physical weakness during their hospital admission, found it more difficult to resume their performance in daily activities.

Surprisingly, we found age positively correlated to physical functioning and vitality. We argue that younger patients might have experienced a bigger change in comparison to their premorbid level of functioning, hence with a more pronounced effect on their perceived quality of life.

A lower level of functioning and impaired performance in activities of daily life in COVID-19 survivors has been recently described as well as the benefit of an early inpatient rehabilitation intervention.⁵

Similarly to our study, Chen et al¹⁷ describe HRQL at 1-month post COVID-19 reporting SF-36 sub-scores (RP, SF and RE) significantly lower than in the Chinese population norm. Interestingly the scores reported by these authors are higher in comparison to the scores of our population across all the sub-scores, however the domains more affected are analogous to our findings. We wonder if the younger age of their population in comparison to ours (mean age 47.2 versus 60 years old) might reflect a less severe presentation of the COVID-19 related illness, hence a more limited impact on HRQL.

Other authors⁶ have reported the results of the SF-36 questionnaire administered to elderly (over 65 year) COVID-19 survivors who underwent 6-week pulmonary rehabilitation, showing an improvement of the HRQL along with pulmonary function. In this study, it is unclear the time when the baseline SF-36 was administered making the findings difficult to compare.

About a third of the patients in our study scored in the range for anxiety and depression, similarly to the patients described by Mazza et al.²⁶ Anxiety and depression scores did not correlate to any of the SF-36 domains. Although this lack of correlation could be likely due to the limited sample size, it is also possible that anxiety and depression might have improved following people returning home and that other kinds of emotional difficulties such as feelings of frustration or isolation, rather than anxiety or depression, could have impacted on their return to activities and roles and on the perception of emotional wellbeing.

Our study’s main limitation is the small sample size, which does not allow strong conclusions or generalizations. Furthermore, the Nave Ospedale unit admits a specific group of patients discharged from the acute setting, including those with mild-moderate symptoms requiring further social isolation in absence of alternative accommodation. This suggests a possible social background bias as patients who are more socially disadvantaged may have greater difficulty in finding an alternative suitable environment and therefore may be more likely to be admitted.

More in-depth questions regarding patients’ jobs and activities on top of the SF-36 could have given a broader insight as well as repeated evaluations of the SF-36 to monitor changes of the HRQL over time.

Conclusions

We describe the severe impact across the physical and psychological domains of HRQL in COVID-19 survivors about one month following their discharge from acute hospitals receiving no support from rehabilitation community services. Muscle weakness and the number of co-morbidities at the time of discharge appeared to be associated with lower physical functioning. Interestingly younger patients were more affected in their perceived physical functioning, as well as in vitality. Despite the small sample size, our study clearly highlights the substantial difficulties that COVID-19 survivors experience after their hospital discharge and stimulates a reflection on the individual and social cost of their delayed return to previous roles and jobs. We suggest that dedicated community-based pulmonary and vocational rehabilitation might not only improve these patients’ outcome but could also be socially and financially strategic.

Data availability