COVID-19 can cause severe respiratory failure and manifest as acute respiratory distress syndrome (ARDS) (1). Mechanical ventilation (IMV) is the treatment of choice in most severe cases of ARDS, and those who receive IMV are at increased risk of mortality (2, 3). Notwithstanding, the great magnitude of the first wave of the pandemic encouraged the creation of intermediate respiratory care units (IRCU) boosting the use of non-invasive ventilation (NIV) to attend the high demand of the health system (4, 5).

Lung ultrasound (LUS) application has been clearly intensified during COVID-19 pandemic as a powerful noninvasive bedside tool that allows a quick and safe assessment to diagnose interstitial-alveolar syndrome, lung consolidation or pleural effusion (6, 7). Likewise, application of diaphragm ultrasound gained importance to detect structural deteriorations due to prolonged IMV and also due to pathophysiological mechanisms observed in COVID-19 (8–10). In patients in the critical care unit, diaphragm dysfunction due to COVID-19 was estimated in 10% (9). Despite the use of diaphragm and LUS is widely accepted in the acute phase of COVID-19 (8), its utility still needs to be defined in the long-term follow-up.

In COVID-19 survivors, the persistence of symptoms still places a burden on the society. Even individuals infected with mild COVID-19 depict prolonged symptoms that affect basic and instrumental activities of daily living (11–13). Approximately half of patients recovering from COVID-19 report chronic dyspnea 2–3 months after infection (14, 15). On top of that, presence of chronic symptoms reduces the quality of life and work productivity in survivors (14, 16).

To date, some studies have reported the long-term outcomes on physical fitness of patients severely affected by COVID-19 (17–20). However, the available evidence is conflicting on the fully fitness recovery of these patients after hospital discharge. Furthermore, the relationship between lung morphological alterations and functional findings after discharge of critically ill patients with COVID-19 has not been yet elucidated thoroughly. We hypothesized that lung morphological alterations caused by severe COVID-19 pneumonia could be detected 2–3 months after discharge and related to changes in lung function and physical fitness. The objective of this study was to describe the clinical and functional status of patients with severe COVID-19 pneumonia at 2–3 months post discharge from a Spanish critical care unit.

In this single-center retrospective descriptive study, the medical records of patients admitted to the intensive care unit (ICU) or the intermediate respiratory care unit (IRCU) of Granollers General Hospital (Spain) were analyzed. The study protocol was approved by the Research Ethics Committee of Granollers General Hospital with an exemption for informed consent. The study was conducted in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement (21).

We included patients admitted to the ICU or IRCU from March to May 2020 due to COVID-19 associated severe respiratory failure who required use of IMV or NIV. The exclusion criteria were patients treated with oxygen therapy only, those who presented any contraindication to perform the follow-up functional tests, and those who did not attend the follow-up visit 2–3 months after discharge.

The total sample consisted of 58 patients with a mean age of 62.2 (SD 10.2) years, mostly male (67.2%) with the following signs and symptoms at 2–3 months of follow-up: mucus hypersecretion (31.0%) dyspnea (29.3%), and dry cough (20.7%). The average length of stay in the critical unit was 29.9 days (SD 20.3). Participants with abnormal findings in LUS stayed longer in the UCI or IRCU when compared to those with normal LUS findings (p = 0.026). The most frequent comorbidities were hypertension (46.6%), diabetes (29.3%), and obesity (17.2%), as described in Table 1.

The blood test analyses revealed a slight increase in D-dimer values 2–3 months after the onset of COVID-19. Other proinflammatory and coagulation parameters showed values within the normal range (Table 2).

Lung ultrasound images revealed normal pattern in 28 patients (51.9%) as represented by A-lines or <3 separated B-lines, whereas 21 patients (38.9%) showed large consolidations in the images, and 5 (9.3%) presented mild morphological alterations (Table 3). Four patients (out of 58 in the total sample) were not assessed by lung ultrasound because CT scan was used instead.

Diaphragm exams performed by ultrasound showed excursions ranging from 0.5 cm–3.8 cm in quiet breathing, 1.6 cm–7.6 cm in deep breathing, and 1.3 cm–3.8 cm in voluntary sniff, whereas thickness ranged from 1.1 to 3.0 mm. Patients with normal and abnormal lung ultrasound findings presented statistical difference in the diaphragmatic excursion in deep breathing (p = 0.029). No difference was observed in terms of thickness. The mean values and standard deviations are shown in Table 4.

Lung function tests results are presented in Table 5. We observed FEV₁ <80% in 15 (25.8%) patients, out of those 11 (73.3%) showed concomitant FEV₁/FVC ≥70%. Overall, restrictive pattern in spirometry (FEV₁ <80% and FEV₁/FVC ≥70%) was observed in 18.9% of patients in our sample. Eighteen patients (58.0%) had abnormal DLCO, out of those 8 (25.8%) had a mild reduction (DLCO 61%–75%), 9 (29.0%) a moderate reduction (DLCO 41%–60%) and 1 (3.2%) a severe reduction (DLCO <40%).

Patients with abnormal LUS findings presented significant statistical differences in FVC% and FEV₁% (p < 0.01) when compared with patients presenting normal LUS findings. No significant difference was observed in the other parameters of lung function (Figure 1).

Variables related to exercise capacity, physical activity levels (IPAQ), and health-related quality of life (K-BILD) are reported in Table 6. The average distance in the 6MWT was 439.9 meters (64.7% predicted), with 32 patients (55.1%) presenting desaturation larger than 3%. The average number of repetitions in the sit-to-stand test was 13.6. Twenty-eight (48.3%) participants self-reported moderate level of physical activity, 16 (27.6%) self-reported intense levels, and 14 (24.1%) low levels of physical activity. The mean total score in the K-BILD questionnaire of quality of life was 79.5 points.

This study followed up patients with severe respiratory failure due to COVID-19 during the first wave of the pandemic at 2–3 months after hospital discharge. Although a significant part of them presented with normal overall functional status, they displayed abnormalities in LUS, pulmonary function, and diaphragmatic excursion.

The acute phase of COVID-19 is characterized by an inflammatory cascade that may induce venous thromboembolic complications (32, 33). Our results show values of prothrombin time, C-reactive protein and tPA within the normal range values, and a trend to normalization of D-dimer after 2–3 months from hospital discharge. Similarly (33), found a slight and persistent alteration of D-dimer, fibrinogen, and IL-6 in a non-negligible percentage of COVID-19 patients (32.2%, 61%, and 32.2%, respectively) 3–6 months after discharge. Although the clinical relevance of these data is not clear, the authors suggest the risk for thromboembolic complications quickly decreases during the convalescence phase.

It is well known that prolonged IMV poses a threat in developing secondary infections and mobility restrictions that may lead to long-term complications (34). In the present study, we found lung ultrasound abnormalities in nearly half (48.1%) of the individuals 2–3 months after hospital discharge. This finding is supported by a similar study carried out in the first Italian epidemic wave showed that 59% of patients had an interstitial pulmonary syndrome observed in LUS 2 months from acute phase (35). Considering all participants in the present study either required invasive or noninvasive mechanical ventilation in their hospital stay, we could attribute the LUS abnormal findings to the degree of COVID-19 severity.

The pulmonary function tests of this study displayed 18.9% of patients with impaired spirometry values, considering FEV1 <80% and FEV1/FVC ≥70% (36). Other studies show a similar proportion (17%) of spirometry alterations among COVID-19 survivors (37, 38). A recent study described that critical patients had greater improvements in pulmonary function tests during the first year after COVID-19, when compared to moderate and severe patients (39).

Diaphragmatic excursion is 1 cm–2 cm during tidal breathing and 7 cm–11 cm during deep inspiration in healthy subjects (40). On the other hand, diaphragm dysfunction is observed in 10% of the patients with COVID-19 in the critical unit and it has been reported that the diaphragm thickness declines during the hospital stay (9, 41, 42). Despite the long-term implications are still unclear, the participants of our study presented significant less excursion in deep breathing consistent with the lung morphological alterations observed in LUS exam. Although no difference was observed in the diaphragm thickness of patients with or without LUS alterations, an apparent normal diaphragm at some point could become exhausted because of less compliant lungs. Therefore, extrapulmonary damage caused by COVID-19 to the diaphragm cannot be ruled out.

Both the 6MWT and the 30-s STS test complement each other as they evaluate different aspects of physical capacity: the 6MWT relates to endurance and cardiovascular fitness, whereas the 30-s STS refers to lower extremity muscle strength and power (26, 28). The participants of our study presented with decreased cardiovascular fitness as they walked an average of 64.7% predicted distance and more than half of them desaturated during the 6MWT. These data are consistent with an Italian observational study which attributed that patients with more severe COVID-19 pneumonia (PaO₂/FiO₂ nadir ≤200) had a lower minimum SpO₂ and a higher desaturation % in the 6MWT at 2- and 6-month follow-up (35). The functional status of COVID-19 survivors at the critical unit and hospital discharge has been reported in a few studies. Worse functional status at hospital discharge was associated with longer IMV duration, older age, higher number of comorbidities, hypertension, diabetes, chronic obstructive pulmonary disease, and immunosuppression (43). A study by (44) found that 80% of COVID-19 recovered cases have diverse degrees of functional restrictions ranging from negligible (63.1%), slight (14.4%), moderate (2%), to severe (0.5%) based on “Post-COVID-19 Functional Status” (PCFS). It has also been shown that a large proportion of patients had worsened quality of life, reduced functional status (assessed by PCFS), and persistent symptoms compared with their pre-COVID-19 status at 6 months after ICU admission (45). We chose to assess our patients with the 30 s-STS test because it allowed optimization of resources in the pandemics context as it requires low time of execution, equipment and space (28). The 30 s-STS power was found to indicate global muscle wasting in COVID-19 survivors one month after hospitalization (28). Conversely, the patients in our study presented average values of 30 s-STS within the normal range for the healthy population, thus suggesting a recovery of lower extremity strength and power at 2–3 months post discharge.

Nearly half (48.3%) of the participants of our study perceived their levels of physical activity as moderate, following by 27.6% who self-reported intense levels, and 24.1% low levels of physical activity, along with normal health-related quality of life. It particularly came to our attention that many participants of our sample were very scared of their prognosis after hospital discharge in the first wave of the pandemic, and likewise grateful for the effort made by the frontline healthcare professionals. Taken together, we believe the lack of significant alterations in their self-reported levels of physical activity and quality of life were due to a positive attitude and self-management towards their own recovery. Contrary to our results, data reported from an Italian cohort of severe COVID-19 patients show deterioration on the physical component of quality of life one year after ICU discharge, attributing the female gender to influence negatively the physical function (20). Considering the high proportion of males (67.2%) enrolled in the present study, the overall normal health-related quality of life self-reported by the patients could have been dampened.

A plethora of evidence sustain that critical COVID-19 patients have prominent long-term deterioration of exercise capacity and physical status (39, 43–45). However, it has been lately noticeable that clinical manifestations of COVID-19 are heterogeneous, and the disease severity does not necessarily impact on exercise capacity of survivors. Some authors attribute this finding to different degrees of adaptation to inflammatory responses affecting cardio-respiratory efficiency (2, 46).

The biggest strength of this study is the comprehensive assessment of patients with severe COVID-19 pneumonia after hospital discharge. Considering the large heterogeneity of information on the subacute clinical and functional manifestations caused by COVID-19, this study reports that impaired pulmonary and diaphragmatic function at 2–3 months post discharge could be related to the persistence of LUS abnormalities in this time window. However, this study presents limitations regarding the limited sample size (single center), therefore multicenter studies are necessary to extrapolate the results to other regions. In this line, low external validity could be considered another limitation because of the convenience sampling employed in this study. Additionally, the data extracted from the medical record was also limited by the retrospective nature of this study. Lastly, no lung function, exercise capacity, or ultrasound results were gathered before COVID-19 pandemic, therefore the lack of baseline data hinders a longitudinal analysis. More extended follow-up is necessary to better understand the possible long-term events caused by COVID-19.

This study provides evidence that patients with severe COVID-19 pneumonia who present abnormalities in LUS imaging concomitantly display significant impaired lung function, and lesser mobility of the diaphragm 2–3 months post discharge. Our findings reaffirm the need of assessing the functional status of the survivors to monitor the evolution of pulmonary and systemic damage, and target specific rehabilitation interventions to tackle COVID-19 functional sequelae.