1. Introduction
The COVID-19 pandemic in Switzerland started in the beginning of February 2020 as a regional sub-happening of the global outbreak of the respiratory disease COVID-19 and was based on infections with the SARS-CoV-2 virus, which had emerged in late 2019. The COVID-19 pandemic spread from the Chinese metropolis of Wuhan, Hubei Province, starting in December 2019. Beginning on 11 March 2020, the World Health Organization (WHO) classified the outbreak of the novel coronavirus as a pandemic [
1]. The number of people infected with coronavirus in Switzerland has been increasing since the end of February 2020. Up until 12 December 2022, there were approximately 4.35 million confirmed cases of the disease in Switzerland. Due to or with co-occurring lung disease, more than 13,700 people have died in Switzerland so far.
During the pandemic, epidemiologists distinguished different waves of the spread that led to hospitalization of different population groups. The first wave occurred in spring 2020, the second period at the end of summer and in autumn 2020, and the third at the beginning of 2021. In Switzerland, a fourth wave started in October 2021 and lasted until June 2022 [
2].
COVID-19 is caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). The pathophysiologic mechanism of this virus is mainly related to acute respiratory distress syndrome (ARDS) and systemic severe inflammatory reaction [
3]. One of the first observations at the beginning of the pandemic was that patients with comorbidities, including previous cardiovascular disease, were more likely to present with worse clinical outcomes, including a higher risk of hospitalization and death [
4,
5]. Accordingly, the first three waves mainly affected people with severe courses of the disease who were already severely ill in advance. However, after an effective vaccine became available, the composition of the COVID-19 cohort with severe courses changed: those who had not been vaccinated predominated [
6].
Pulmonary rehabilitation (PR) has been prioritized in COVID-19 patient care from the beginning [
7,
8]. Not only has it been used to provide patients with the usual rehabilitative services to promote recovery following a severe course, but in some countries (e.g., Switzerland, Germany, and Italy), it also, in part, has been used to relieve the acute hospitals and maintain their admission facilities and inpatient capacities [
9].
Little has been described about the impact of changes in the composition of patients in the post-COVID-19 stage on rehabilitation outcomes, which is why, in this study, we analyzed the commonalities and differences in PR outcomes of the different patient groups.
2. Materials and Methods
Between February 2020 and June 2022, 676 patients with post-acute COVID-19 were admitted to an inpatient PR program at the Zurich RehaCenter Wald Clinic in Switzerland. For several reasons, data for only 483 patients could be included (Wave 1: 51, Wave 2: 202, Wave 3: 84, Wave 4: 146). Data collection was conducted prospectively. All included patients gave their written informed consent. For data storage, the clinic information system INES® (INES Schweiz GmbH, Bottighoferstrasse 6, 8596 Scherzingen, Switzerland) was used. The study protocol was approved by the local ethics committee (BASEC-No 2020-01061) and was registered at the German Clinical Trials Register (DRKS00024613).
All necessary assessments were performed upon admission to PR and 1–2 days before discharge. The contents of the assessments consisted of the Hospital Anxiety and Depression Scale (HADS), laboratory values, the Chronic Respiratory Disease Questionnaire (CRQ), the Cumulative Illness Rating scale (CIRS), and pulmonary function testing (PFT). The following assessments were performed twice, once upon admission and the second time upon discharge, in order to record changes: Feeling Thermometer (FT), Functional Independence Measure (FIM), and 6-min walk test.
2.1. Pulmonary Rehabilitation Program
The contents of the PR program are described extensively elsewhere and are thus described only cursorily at this point [
7].
All patients participated in an inpatient PR program of 3 weeks on average. The various sessions included individualized training, consisting of gymnastics and stretching, treadmill training or ergometer training, and walking training on terrain or indoors (three different levels of intensity), as well as strength training adapted to the performance capacity of the patients. The intensity level was based on the severity of the cardio-pulmonary effects as well as the functional limitations of the patients.
Many patients showed significant limitations at the beginning of PR, so it was usually necessary to start with a low intensity level and successively increase the load, depending on the patient’s condition and tolerance. The gymnastic exercises addressed not only an improvement in endurance performance but also coordination and strength exercises, as well as exercises that trained flexibility and balance. Respiratory physiotherapy included cough control exercises in addition to breathing control and energy conservation techniques. All patients participated in educational sessions and nutritional interventions. If needed, patients participated in a structured smoking cessation program, received psychosocial support, or received diabetes counseling.
2.2. Six-Minute Walk Test (6-MWT)
Exercise capacity was measured at hospital admission and discharge using the 6-MWT according to the guidelines of the American Thoracic Society (ATS) and carried out by experienced, well-instructed examiners [
10]. According to the ATS, this exercise test provides valuable data for patients with cardiac or pulmonary diseases, representing functional, therapeutic response as well as prognostic data. Due to its reproducibility and simplicity, the 6-MWT is frequently used and delivers a good overview of the cardiopulmonary and musculoskeletal status. This test is safe and well-tolerated by most patients at any stage of disease, with the test being highly reflective of usual daily activity and exercise performance.
2.3. Chronic Respiratory Disease Questionnaire (CRQ)
To assess health-related quality of life (HRQoL) the Chronic Respiratory Disease Questionnaire (CRQ) was used. The questionnaire measures eight dimensions of HRQoL and allows for the calculation of two summary scales of physical and mental experience [
11]. The 20 items were completed by the patients individually. They represent the four areas of dysfunction (fatigue, emotional functioning, mastery, and dyspnea) in patients with chronic pulmonary diseases. The patients graded their symptoms according to a 7-point Likert scale for the domains, which included 4–7 items.
2.4. Functional Independence Measure (FIM)
The Functional Independence Measure (FIM) is an 18-item measurement tool that explores an individual’s physical, psychological, and social functioning. In this study, the FIM was used to assess changes in patient functioning, specifically with the aim of evaluating response to rehabilitation or medical intervention. The FIM uses level of assistance and individual needs to grade functional status from total independence to total support [
12].
2.5. Hospital Anxiety and Depression Scale (HADS)
The HADS is questionnaire of 14 items (7 questions focussing on depression and 7 questions on anxiety). Scores range from 0 to 21, with higher scores representing greater disruption. Additionally, the questionnaire can be used as a diagnostic tool to assess depression in patient with severe physical restrictions [
13]. Completion took about 2–5 min. In this study, the HADS was used to evaluate the response of depression and anxiety to PR.
2.6. Cumulative Illness Rating Scale (CIRS)
The CIRS was used to assess a patient’s level of disability and as an indicator of health status, including predicted 18-month mortality and social function [
14]. It is a comprehensive method for recording diseases in 14 organ systems on the basis of an evaluation of 0 to 4 points, which is used to calculate a cumulative score. The range of the total score is 0–56 points. When evaluating the CIRS, each individual illness in the corresponding organ system must be classified. If there were different diseases within the same organ system, only the disease that was most pronounced was evaluated. The calculated CIRS at admission is useful for predicting important hospital outcomes such as high risk of death or long stays and to better anticipate end-of-life issues.
2.7. Feeling Thermometer (FT)
The FT analyzes and evaluates the emotional status of the patients regarding their current well-being. Therefore, the patients rate their feelings according to a rating scale in terms of degrees from 0–100, representing their mood corresponding to temperatures [
15]. A higher temperature indicates a better mood.
2.8. Pulmonary Function Tests (PFT) and Blood Gas Analysis
Upon each patient’s discharge from PR, we conducted comprehensive pulmonary functioning testing using Body-Plethysmography and Spirometry (Master Screen Body; Jaeger GmbH, Hoechberg, Germany). The tests were carried out considering the current ATS-ERS Guidelines. The following blood tests were analyzed once upon admission: arterial blood gas analysis (inhouse analysis: Radiometer ABL800, Willich, Germany) [
16,
17], blood cell counts, leukocytes, and C-reactive protein (CRP) (external laboratory analysis: Medica, Medizinische Laboratorien Dr. F. Kaeppeli AG, Zurich).
2.9. Statistical Analysis
Binary variables were presented as relative and absolute frequencies, and Fisher’s exact test was used for group comparisons. Continuous variables were visually inspected using a Q-Q plot to verify normal distribution. Differences between groups in normally distributed variables were analyzed using one-way ANOVA with Tukey’s HSD as a post hoc test. To calculate the difference between groups of non-normally distributed data, the Kruskal-Wallis test with Dunn’s test and the Bonferroni correction as a post hoc test was used.
Using combined data across all waves, multivariable regression was used to determine variables independently associated with rehabilitation outcomes ∆6-MWT and ∆FIM. To ensure no multicollinearity between independent variables, the variance inflation factor (VIF) was calculated, and variables with values > 5 were further investigated for multicollinearity and not used in the regression model. Continuous variables describing patient’s characteristics were entered into the multivariable regression model, and non-significant variables were eliminated individually using backward subtraction.
4. Discussion
To the best of our knowledge, this is the first study with detailed characterization of patients from waves 1–4 of the coronavirus pandemic describing the impact on outcomes of an inpatient PR program. Patients differed significantly according to their anthropometric data, incidence of comorbidities, and impact of the infection. All patient groups achieved clinically relevant and significant functional improvements during PR, with significantly higher improvements in Wave 3 + 4.
Patients of Wave 3 + 4 were significantly younger than patients of Wave 1 + 2. A reason for this might be the fact that the percentage of vaccinated persons in Switzerland as of 1 March 2021 increased with age, leading to more protection of elderly patients in Wave 3 + 4. In our study, only 29.1% of Wave 4 patients were vaccinated, while during the same period, the vaccination status of the population of Switzerland had already risen up to 71.7% [
18]. As a result, we saw younger unvaccinated and severely ill patients in Wave 3 + 4. In an observational study regarding a similar time period (October 2018 to September 2021), Sleffel et al. found in a comparable age structure in 1520 PR participants [
19]. The vaccination status might have led to this shift in age, but the awareness of elderly people at risk, the protection measures available, and, in some cases, initial infection during Wave 1 + 2 can potentially be additional factors.
Probably due to the age differences, clinically relevant comorbidities such as COPD, arterial hypertension, or atrial fibrillation were significantly more frequent in Wave 1 + 2 than in Wave 3 + 4. The patient characteristics of Wave 1 + 2 are comparable to the findings in other cohorts of the same time period, which also had numerous comorbidities and were even older on average (72 years) [
20].
It is well known that patients with pulmonary diseases show improvements in performance by participating in PR, e.g., regarding the walking distance according to the 6-MWT [
21]. Functional benefits of PR participation in post-COVID-19 patients have been published previously by our group [
7]. In our present study, patients in all groups showed significant improvements according to the assessments, and the enhancements are comparable to results in other studies (6-MWT 132.8 ± 92.8 m; FIM 18.0 ± 11.4 points) [
22]. In the intergroup comparison of Wave 1 + 2 and Wave 3 + 4, however, there was a significantly and clinically relevant higher improvement found for the 6-MWT and FIM in patients from Wave 3 + 4. As the multivariable model shows, duration of PR, age, and the functional level (6-MWT and FIM) at admission are the major predicting factors for the PR outcome. BMI and HADS-D also play a role, but only for the 6-MWT improvement. The model shows that younger and highly impaired COVID patients with a long PR duration are more likely to benefit from PR. However, ICU days and hospitalization duration showed no further influence on the PR outcomes, while ventilation days positively affected the ∆FT only. These findings are in line with previous studies [
23].
Gabunia S et al. compared 138 patients from the first two waves of the COVID-19 pandemic according to the results achieved while participating in inpatient PR with a length of stay of 11–12 days. The patients in both waves experienced the same functional improvements regarding the GG Self-care and Mobility Activities items with a median GG score change of 3.6 per day and similar discharge GG scores [
24]. However, the patients of the respective waves did not differ significantly, and enhancements during PR were similar. These findings are in line with our results, showing no difference in patient characteristics followed by almost the same PR outcomes. However, our data show that as soon as the vaccination status changed, which was found in Wave 3 + 4, patient characteristics also changed, leading to a positive impact on PR outcome parameters.
It has been explained that COVID-19 infection has numerous manifestations beyond the respiratory system, including cardiac injury (cardiomyopathy, myocarditis, ventricular arrhythmias, and hemodynamic instability in the absence of obstructive coronary artery disease) [
25]; thrombotic complications (including stroke, myocardial infarction, and venous thromboses) [
26]; and renal, gastrointestinal, and neurologic symptoms, among others [
27]. These manifestations are frequently followed by persistence of symptoms and a reduced quality of life [
28]. This is in contrast to our study, wherein the more symptomatic and impaired patient group of Wave 3 + 4 showed better PR outcomes.
However, regarding the well-being of the patients measured by the FT, no significant differences between the groups were observed. This discrepancy is in contrast to the results described by Jacobs LG et al., who found a large amount of prevalent and persistent symptoms according to their observational study and found that the persistence of symptoms has an important impact on general, physical, and mental health status, social functioning, and quality of life within 35 days of discharge after COVID-19 infections [
29]. We assume that the perception of the enhancements while participating in the PR program was mainly reduced due to mood disorders following the COVID-19 infection, though the HADS-D scores were significantly higher in Wave 1 + 2. However, the experience in treating the patients of Wave 3 + 4 was that they noted dissatisfaction with their situation in general and that the expectations projected in the PR were not fulfilled.
As already described, severe COVID-19 infections may lead to persisting lung function impairment, especially regarding the diffusion capacity (DLCO) [
30]. The findings of our analysis confirm the results published by Lenoir A et al. showing a reduction of the DLCO (53.8%pred.) at admission to PR in all waves. Additionally, persistent oxygen dependency was observed in 54.1% of our cohort at discharge from PR. This indicated the severity of the lung damage caused by the COVID-19 infection, which seemed to be higher than that found in other cohorts. The chest x-rays performed at discharge from PR showed significant more pulmonary infiltrates for the Wave 1 + 2 patients. According to recent studies, the reported quota of oxygen dependency after COVID 19-infection differs to a huge extent depending on the observed cohort. For example, Jacobs LG et al. reported 20.3% oxygen dependence 35 days after COVID-19 infection, while Sundh J et al. described 67% following 150 days of the infection [
29,
31]. It seems to be obvious that the quota depends on the initial severity of the infection and the number of pulmonary comorbidities. However, most of those with abnormal pulmonary function tests at 3 months improved subsequently, but only another 29% (6 out of 21) reached normal values at 6 months [
32].
This analysis has some limitations, which have to be discussed. First, this study represents data from a single center. Therefore, results should be applied to other cohorts with caution. However, our rehab center is one of the largest centers in Eastern Switzerland, and the area includes many different acute hospitals that referred patients to PR. This is why we believe that the cohort is representative of patients with severe post-acute COVID-19 infections. Second, since the approach to the COVID-19 patients in this study is observational, it provides no control group. Implementation of a control group was not feasible in this context for legal and ethical reasons. On the one hand, patients have a right to rehabilitative services, and on the other hand, withholding rehabilitation would not be ethically defensible or enforceable in Switzerland. Third, a potential selection bias may have occurred due to the amount of patient data that could not be considered, as described in the results section.
