Coronavirus disease 2019 (COVID-19) is produced by the infection of a virus of the coronavirus family [severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)] identified in December 2019 in China (1). Patients with COVID-19 could develop severe pneumonia in 20% of the cases (2, 3), which is defined by radiographic pneumonia plus respiratory rate >30 breaths⋅min^–1 or severe respiratory distress or oxygen saturation measured by pulse oximetry (S_pO₂) ≤93% on room air at rest (1, 4). Furthermore, 5% of patients with COVID-19 pneumonia may develop multiple organ dysfunction requiring intensive care unit (ICU) admission (2, 3, 5, 6) and life support, including mechanical ventilation, for prolonged periods (2). The most frequent ICU complications related to COVID-19 are acute respiratory distress syndrome (ARDS), in up to 29% of patients (2), and intensive care unit (ICU)-acquired weakness (ICUAW) (7).

Intensive care unit-acquired weakness is a critical illness characterized by polyneuropathy, myopathy, and neuromyopathy that may develop in approximately 40% of patients requiring mechanical ventilation (8–10). In previous coronavirus epidemics, severe acute respiratory syndrome, and middle east respiratory syndrome, around 50% of subjects presented a significant decrease of diffusion capacity, limitation of exercise capacity, and decreased quality of life after several months of hospital discharge (11–14). The first available data related to post-COVID-19 consequences show abnormal pulmonary function and muscle weakness, related to subject functional limitation in their everyday lives (15, 16), especially in elderly patients (17). In this regard, rehabilitation has shown lung function improvement and quality of life recovering in ICU survivors from other diseases (18) and ARDS fatigue (19).

Early rehabilitation has been associated with a reduction of the likelihood of developing ICUAW (20). Initially, rehabilitation in patients with COVID-19 (1, 4) was not performed due to subject and medical staff safety concerns, as there is a possibility of generating airborne aerosols of SARS-CoV-2 particles (21, 22). Undoubtedly, the healthcare work overload and the lack of personal protection equipment at the beginning of the pandemic also played a role (5). However, following the recommendations of main medical societies (23–26) and experts (27–30), early rehabilitation was initiated in stable subjects. Previous data on subacute rehabilitation in COVID-19 have reported promising clinical benefits with no adverse effects reported on either subjects or medical staff (7, 31–34). Nevertheless, studies exploring the early initiation of physical therapy are lacking. Therefore, this study evaluated the impact of rehabilitation on severe COVID-19 pneumonia survivors and compared the potential benefits depending on the time of therapy initiation.

All the 208 severe COVID-19 subjects who survived the acute phase and were admitted to the ICU or IMCU from March 13th until May 15th of 2020 were included. Of them, 49 who did not require HFNC or MV support were excluded. Of the 159 subjects analyzed, 72% were men and the median age was 61 years old. Hypertension (50%) was the most frequent comorbidity, followed by dyslipidemia (44%), obesity (29%), and diabetes mellitus (27%). Physical therapy was performed in 108 subjects (68%). These subjects were significantly older than those in the non-physiotherapy group (62 vs. 55 years old; p-value 0.01), and developed a more severe disease with longer ICU admission time, mechanical ventilation, and tracheostomy requirement, as well as worse laboratory profile (low minimum albumin, higher maximum CRP, LDH, and D-dimer blood levels). No physiotherapist was infected by COVID-19 and no adverse effect or worsening in subjects was detected in our study. The non-physical therapy group required significantly less time than the physical therapy group from admission to achieving the sitting position than the physical therapy group (10 vs. 19 days; p < 0.001) and to hospital discharge (14 vs. 31 days, p < 0.001). Subject characteristics are shown in Table 1.

Among 108 subjects who underwent physical therapy, 32 (30%) started on the first 7 days of admission (early) and 76 (70%) started after the first week of admission (non-early). Their characteristics are depicted in Table 2. Initial admission to the ICU was higher in the early physiotherapy group (56% early vs. 18% non-early; p-value <0.001), who had shorter time from admission until physical therapy started (4 vs. 13 days, respectively; p < 0.001) and sitting position (9.5 vs. 22 days; p < 0.001). The length of hospital stay was 19 and 34 days (IQR 36.25 vs. 27.25; p = 0.001) in early and non-early physiotherapy groups, respectively.

Multivariate analysis showed that factors independently associated with a higher risk of prolonged hospitalization were obesity [odds ratio (OR) 3.21; 95% CI 1.18–9.66], invasive mechanical ventilation (OR 6.25; 95% CI 2.37–18.33), and starting physiotherapy after the first week of admission (OR 3.54; 95% CI 1.29–10.44) (Table 3).

No significant differences were observed at 60 days follow-up between early and non-early physical therapy groups (Table 4). Oxygen therapy and oral corticosteroid treatment could be reduced at follow-up. Both groups reported dyspnea mMRC 1, neuropathy symptoms, and cough as the most frequent complaints at follow-up.

This retrospective study evaluates the features of 108 subjects that started physiotherapy after surviving severe COVID-19 pneumonia and compared the potential benefits depending on the delay in starting this therapy (early versus non-early). The initiation of physical therapy was limited by available resources and safety concerns for subjects and healthcare workers during the first weeks of the pandemic. First and foremost, physical therapy was safe without physiotherapist infection and no adverse effect or worsening in subjects detected. This is consistent with previous rehabilitation studies in critical care units (42), as well as in reported acute (7) and postacute (32, 34, 43–46) COVID-19 rehabilitation studies. Therefore, physical therapy appears to be safe for subjects and healthcare workers.

On the other hand, the benefit of early systemic mobilization in critically ill subjects remains unproven, awaiting longer randomized trials (47), although it is being recommended (17, 28). In this regard, available data on the time of rehabilitation initiation on patients with acute severe COVID-19 pneumonia are still limited (48), with an average of 14 ± 7 days to first mobilization (7). In our study, physiotherapy was initiated within a median of 4 (IQR 2.25) and 13 days (IQR 8.25) from admission in the early and non-early groups of our severe COVID-19 cohort, respectively.

Interestingly, the time from admission until sitting position and the length of hospitalization were significantly shorter (12.5 and 15 days, respectively) in the early physical therapy group. Mc William et al. (7) identified that obesity and older patients were associated with a delay in time to the first mobilization in patients with acute severe COVID-19. Furthermore, another survivor cohort study of severe COVID-19 pneumonia showed that ventilation requirement was associated with a longer hospital stay (43). Similarly, the analysis of prolonged hospitalization performed in our subjects that underwent physiotherapy during admission identified that obese subjects, those who required invasive mechanical ventilation, and those who did not initiate early physiotherapy had higher probabilities of prolonged hospital stay.

Recent pulmonary rehabilitation studies in the post-acute phase of COVID-19 pneumonia reported significant benefits in exercise performance, lung function, and reported quality of life (32, 34, 43–46). In our study, follow-up of subject outcomes did not show differences between early and non-early physical therapy groups in terms of dyspnea, walking capacity, reported quality of life, and persistent chest X-ray infiltrates. However, this study lacks a control group without physical therapy to better assess its benefits. Both early and non-early physical therapy groups showed a decrease in oxygen therapy and corticosteroids requirements at 2 months of follow-up. Neuropathy and cough were the most frequently reported symptoms in both groups. Other COVID-19 postdischarge cohorts (49, 50) reported less fatigue, breathlessness, and better-reported quality of life in those subjects without ICU admission.

There is no evidence that early or non-early physical therapy could be unsafe for healthcare workers or produce adverse effects in patients with severe COVID-19 pneumonia. Furthermore, physical therapy could reduce the time from admission until sitting position and length of hospital stay.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Hospital Ethics Committee (reference code PR168/20). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

JB-M contributed to literature search, data collection, study design, data analysis, manuscript preparation, and manuscript review. AL-G contributed to data collection, manuscript preparation, and manuscript review. AA, FT, and LM contributed to data collection. VV-Z and SS contributed to study design and manuscript review. GB contributed to data collection and manuscript review. MM-M and GS-C contributed to study design, data analysis, manuscript preparation, and manuscript review. RP contributed to study design, manuscript preparation, and manuscript review. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.