As part of a nationwide project to evaluate the recovery of patients receiving primary allied health care after a COVID-19 infection, a prospective cohort study was set up in collaboration with various patient organizations (ie, the Lung Foundation Netherlands, the Netherlands Patient Federation, and Harteraad) and with input from patients contacted through these organizations [27]. In this prospective cohort study, patients were included at the start of their treatment with 1 or more primary allied health professionals. All treatments offered by primary allied health professionals in daily practice were part of usual care and were preferably based on recommendations and guidelines published by the professional bodies of the respective care providers, as available at the start of the research [23-26]. The inclusion period for the cohort study was between March 29 and June 19, 2021. Primary outcome measures were assessed at baseline (T0) and again after 3 months (T1) and 6 months (T2). The full study protocol with timelines is published elsewhere [27]. In this paper, we report the results of our primary outcome measures at baseline, 3-month follow-up, and 6-month follow-up.

The study protocol was approved by the Medical Ethics Committee of the Radboud University Medical Centre (registration #2020-7278). The study has been registered in the ClinicalTrials registry (NCT04735744). Informed consent was obtained from all patients before enrollment in the study, and all procedures were conducted in accordance with the Declaration of Helsinki.

Adult patients (age≥18 years) were eligible for inclusion in the cohort if they were recovering from COVID-19 and started treatment with 1 or more primary allied health professionals (ie, a dietitian, exercise therapist, occupational therapist, physical therapist, or speech and language therapist). Patients may have received treatment from 1 or more primary allied health professionals during the course of the study. Patients were included regardless of their hospital admission status during the acute phase of COVID-19. Patients who were unable to complete questionnaires in Dutch and patients who were receiving palliative care were excluded from the study.

Patients could enroll in the study by (1) signing up after an invitation from their treating primary allied health professional or (2) signing up on their own initiative, upon which the research team also invited the treating primary allied health professional to participate. The enrollment procedure of this study is described in detail in the published study protocol [27]. Both patients and primary allied health professionals reported data via the specifically designed Your Research app. Patients were asked to download the app on their smartphones or make use of the web version. Questionnaires were sent out through this app at the start of the treatment (baseline) and after 3 and 6 months. Patients unable to participate via digital methods were provided with an opportunity to complete the questionnaires on paper and return them by post. Primary allied health professionals were asked to use the web version of the app.

Data on patient characteristics were collected by the treating primary allied health professionals at the start of the treatment. Patient-reported outcome domains (participation, HRQoL, fatigue, and physical functioning) were assessed at baseline and after 3 and 6 months. Data on psychological well-being were collected at baseline and after 6 months.

Descriptive statistics were used to describe the patient population and to analyze the primary outcome measures at baseline and after 3 and 6 months using numbers and proportions for categorical variables, means (SDs), and medians (IQRs) for continuous variables. Linear mixed model analyses were used to evaluate recovery over time for participation, the HRQoL, fatigue, physical functioning, and psychological well-being. This analysis accounts for correlation between repeated measures on the same subject and uses all available data from this subject. A model with a random intercept and all other variables fixed was also generated. The primary outcomes were used as dependent variables, while time (categorical: baseline, 3 months, and 6 months) was used as a fixed factor.

Uni- and multivariate regression analyses were used to explore which baseline characteristics were associated with changes in the scores of the main outcome measures between baseline (T0) and 6-month follow-up (T2). This analysis used data from complete cases, and missingness at random (MAR) was tested (Table S1 in Multimedia Appendix 1). Univariate analyses were performed to determine which baseline characteristics (ie, age, sex, BMI, hospitalization, comorbidities, baseline score, and smoking status) were associated with the mean change in each outcome measure. Comorbidities were coded into 3 categories: none, 1, and 2 or more comorbidities. Variables with P<.157 in the univariate regression were included in the multivariate model [43]. The backward elimination of variables was then performed in order of statistical significance until only factors that were significantly associated with the outcome remained. A sensitivity analysis was performed by forcing age, sex, and hospital admission into the models as these factors are known to be related with recovery over time. Results of this sensitivity analysis are presented in Table S2 in Multimedia Appendix 1. P<.05 was considered statistically significant for all analyses based on 2-sided testing. All data were analyzed using SPSS Statistics version 25 (IBM Corp).

During the development of this study, we involved patients to provide feedback on the readability and appropriateness of proposed measures. The usability of the smartphone and web-based app versions was also tested by patients. Participating patients received updates on the status of the study via their smartphone or web app. Furthermore, various patient organizations (ie, the Lung Foundation Netherlands, the Netherlands Patient Federation, and Harteraad) participated during routine research meetings.

In total, 1451 patients were included in this study (Figure 1), receiving 1708 different allied health care treatments. Their mean age was 49 (SD 13) years, and 63.8% (848/1330) of the patient population was female (Table 1). The majority (1015/1315, 77.2%) had not been hospitalized for COVID-19, and 1002/1311 (76.4%) patients had experienced mild-to-moderate severity of symptoms during the infection period. The mean BMI was 28 (SD 6) kg/m², and 68.9% (738/1071) of the patient population was classified as being overweight or obese (BMI>25 kg/m²). In addition, 1 comorbidity was reported by 410/1331 (30.8%) patients, and 2 or more comorbidities were reported by 155/1331 (11.6%) patients. Cardiovascular disease (193/1331, 14.5%) and chronic lung disease (183/1331, 13.7%) were the most prevalent comorbidities. Most patients (1086/1331, 81.6%) had been referred for primary allied health care by their GP.

Table 2 presents data on the outcome measures at baseline and at 3- and 6-month follow-up. Additionally, clinically relevant improvements at 6-month follow-up are presented in Table S3 in Multimedia Appendix 1. After 6 months, the majority of patients showed a clinically relevant improvement on the USER-P restrictions and satisfaction subscales (576/890, 64.7%, patients and 543/891, 60.9%, patients, respectively), while 540/908 (59.5%) patients showed a clinically relevant improvement on the EQ-VAS (mean 67.4, SD 19.1 points) compared to baseline (mean 55.5, SD 17.8 points). Severe fatigue was reported by 1205/1281 (94.1%) patients at baseline, persisting after 6 months in 742/929 (79.9%) patients. A clinically relevant improvement on the FSS mean score was found in 490/904 (54.2%) patients. Based on PROMIS-PF-10b scores, over two-thirds of the patients reported being more than 60% impaired, limited, or restricted in physical functioning at baseline, which decreased to 37.8% (351/929) after 6 months; 517/902 (57.3%) patients experienced a clinically relevant improvement in physical functioning. The majority of patients scored less than 7 points on the HADS anxiety and depression scores both at baseline and at 6 months, which indicates no anxiety disorder or depression. At baseline, the HADS anxiety score indicated a probable clinical diagnosis of anxiety disorder in 292/1271 (23%) patients, which decreased slightly to 170/926 (18.4%) patients after 6 months. A probable clinical diagnosis of depression was indicated by the HADS depression score in 285/1271 (22.4%) patients at baseline, decreasing to 137/926 (14.8%) patients at 6-month follow-up.

Table 3 shows the effect of time on the outcome measures. For all dependent variables, a random intercept model was the best-fitting model. No variables were significantly related to missing values in the outcome measures at any point in time. A significant effect of time was observed for all outcome measures at 3- and 6-month follow-up (P<.001). For participation, estimated mean differences of at least 2.9 points (P<.001) were observed for all 3 subscales at all time points. For the HRQoL, the mean increase was 9.0 points (95% CI 7.8-10.2) at 3 months and 12.3 points (95% CI 11.1-13.6) after 6 months. Furthermore, significant improvements were found for fatigue and physical functioning at all time points. The greatest improvements were seen after just 3 months for all outcome measures measured at both 3 and 6 months. Mean differences of –0.8 (95% CI –1.0 to –0.5) on the HADS anxiety score and –1.6 (95% CI –1.8 to –1.3) on the HADS depression score were observed.

Multivariable regression models were estimated to identify factors associated with changes in scores between baseline and 6-month follow-up for each outcome measure. Tables 4-10 provide an overview of the final regression models. All univariable and multivariable regression models are shown in Tables S4 and S5 in Multimedia Appendix 1. Having a worse baseline score was related to greater improvements for all outcome measures. For all 3 subscales of the USER-P and physical functioning, patients admitted to the hospital during the infection period of COVID-19 showed greater improvements in scores than nonhospitalized patients, even when correcting for baseline scores. In terms of the HRQoL, patients admitted to a hospital ward showed greater improvements than patients who had not been hospitalized, although no associations were found with ICU admissions. Male participants showed greater improvements than female participants in all outcome measures, except for psychological well-being, for which no association was found for sex. The baseline age, BMI, comorbidities, and smoking status were not significantly associated with the mean change in any of the outcome measures in our patient population. In a sensitivity analysis (Table S2 in Multimedia Appendix 1) where age, sex, and hospital admission were forced into the model, additional associations were found between the male sex and satisfaction in participation and between age and frequencies of participation, physical functioning, and symptoms of anxiety.

This study presents the first results of our evaluation of the recovery of our unique cohort of patients with COVID-19 receiving primary allied health care until their 6-month follow-up. We explored which baseline characteristics were associated with changes in the scores of the main outcome measures over this 6-month period. Most patients showed a clinically relevant improvement in all outcome measures; however, despite improvement, many patients still experienced persistent problems in their daily lives, with limitations in physical and mental functioning. A worse baseline score, hospital admission, and, for some outcome measures, the male sex were associated with greater improvement between baseline and 6-month follow-up; however, age, the BMI, comorbidities, and smoking status were not associated with the mean change in any of the outcome measures.

The majority of our patient population showed a clinically relevant improvement 6 months after starting treatment provided by 1 or more primary allied health professionals; nevertheless, a large group of patients experienced persistent problems in their daily lives. The mean EQ-VAS score of our patient population (67 points) remained well below the population norm in the Netherlands, which is 82 points [44]. These results are consistent with previous findings that the HRQoL is impaired in the majority of patients post–COVID-19 [12,15,17,45-48]. Persistent fatigue was highly prevalent among the patients included in our study, with 79.9% still reporting severe fatigue (measured with the FSS) after 6 months. These results are consistent with previous studies on patients recovering from COVID-19, showing that fatigue is the most common complaint [5,10,14,47,49-51], even after 6 months [13,15,52-54]. The mean PROMIS-PF-10b score of our population (mean 43.5, SD 8.5) remained well below the population norm in the Netherlands (mean 50, SD 10). These results are also consistent with previous studies [15,47] and indicate that persistent symptoms due to COVID-19 may lead to patients experiencing limitations in physical functioning.

Relative to other outcome measures, a smaller percentage of patients showed a clinically relevant improvement in psychological well-being. This was due to an observed ceiling effect, as 58.7% and 54.2% of patients showed no indication of an anxiety disorder or depression at baseline, respectively. Data of these patients are still informative, however, as they could also have deteriorated throughout the follow-up period. With scores indicating a probable clinical diagnosis of anxiety disorder or depression in 18.4% and 14.8% of patients, respectively, after 6 months, our findings are similar to those reported in previous studies, which showed prevalence rates ranging from 11% to 40% [8,17,46,48,51,54-57]. Furthermore, we performed an additional subgroup analysis to explore the differences in changes in the scores of the outcome measures between patients who showed indications of depression or anxiety disorder at baseline and patients who did not (data not shown). Based on this analysis, we conclude that whether a patient shows indications of depression or anxiety disorder at baseline has little effect on the change in their scores over time.

We found that male participants showed greater improvements than female participants in participation, the HRQoL, fatigue, and physical functioning. These results are consistent with previous studies showing that female participants experience more persistent symptoms after a COVID-19 infection [10,14,21,45,49,51]. Furthermore, patients admitted to the hospital for COVID-19 showed greater improvements than nonhospitalized patients in terms of participation, the HRQoL, and physical functioning, which is in line with previous studies [12,21,45,51]. We observed no associations between fatigue and hospital admission, age, the BMI, comorbidities, or smoking status, which is also consistent with other studies [10,14,51,53,54], indicating that fatigue is highly prevalent in patients recovering from COVID-19, irrespective of the severity of initial infection and patient characteristics. We found that having a worse baseline score is related to greater improvement in anxiety and depressive symptoms; however, no associations with any patient characteristics were found. Similar to our results, previous studies have found no associations between the frequency of anxiety or depressive symptoms and disease severity or hospital admission [10,46,48,49,57]. In contrast, other studies have found the female sex [54,58,59] and older age [54,56] to be predictors of anxiety or depressive symptoms in patients with COVID-19. Although the association between age and symptoms of anxiety was not evident in our primary analysis, it did show up in our sensitivity analysis (Table S2 in Multimedia Appendix 1).

This study did not include a control group to determine the potential effects of primary allied health care by comparing outcome measures with patients who did not receive this type of care. In addition, with a lack of available pre–COVID-19 data for our population, it was difficult to draw conclusions about the impact of pre-existing conditions versus problems in the daily activities and participation of these patients due to their COVID-19 infection.

For the interpretation of results, it is important to consider that the baseline measurement in this study was taken at the start of the treatment by 1 or more primary allied health professionals. It is possible that a patient had already experienced symptoms for some time and only consulted a primary allied health professional at a later stage. Additionally, it should be considered that not all patients received treatment from 1 or more primary allied health professionals during the entire 6-month follow-up period of this study. Some patients received short-term treatment, while others were still receiving treatment at 6 months.

The characteristics of included patients are comparable to COVID-19 populations described by others [3], and therefore, we assume that the study population is representative of the population in 2021, our inclusion period. Based on the inclusion period, which was between March and July 2021, our population most likely had the Wuhan or Alpha variant of SARS-CoV-2 [60]. Different variants may cause different symptoms, and the recovery of patients infected with other variants (eg, Delta or Omicron) may differ from that of our population. A total of 25 patients dropped out during this study (Figure 1). Although a proportion of the patients did not complete all questionnaires, the response rates were still sufficient: 93% at baseline, 68% after 3 months, and 67% after 6 months [61]. There was no selective missingness of data based on patient characteristics (including disease severity) and scores of the outcome measures (Table S1 in Multimedia Appendix 1).

Future research and in-depth analyses of our data are needed to gain more insight into the outcome measures and recovery of patients after COVID-19 who visit 1 or more primary allied health professionals. Future papers will include the results after a 12-month follow-up, determining the related health care costs and profession-specific outcomes per allied health discipline.

The results of this study show that patients recovering from COVID-19 and receiving primary allied health care make progress in recovery, but many still experience limitations in their daily activities and participation after 6 months. The findings of our study provide reference values for health care providers and health care policy makers about what to expect from the recovery of patients who receive or have received health care from 1 or more primary allied health professionals.