Discussions
The main finding of this study was that the CT findings and symptoms (especially effort dyspnea, fatigue and muscle weakness) may not be totally resolved 6 months after the onset of symptoms in patients who require ICU admission. Also, some of these patients may encounter impaired pulmonary function tests and decreased exercise capacity. Impairment of quality of life was comparable between ICU and non-ICU patients.
Chest CT has been frequently used as a diagnostic tool during the COVID-19 outbreak and CT severity scores were related to disease severity25. In our study patients who were admitted to ICU had higher CT scores than patients admitted to the medical ward and this result was consistent with previous studies25,26. At the follow-up visit patients who were discharged from ICU had higher CT scores than patients who were discharged from the medical ward (6 vs 0). CT findings were totally resolved in most non-ICU patients, but most ICU patients had abnormal CT findings at the follow-up visit. In a previous study most of the CT findings were resolved in non-severe COVID-19 patients within 4 weeks after discharge16,27. However, as the severity of disease increases the recovery time may be longer. In a large cohort, patients with increased disease severity had higher CT scores at follow-up visits28. We found two-thirds of ICU patients had at least one of the CT findings at follow-up visits; irregular lines, subpleural lines and GGO were the most common patterns in chest CT. We also found advanced age and higher CT scores at admission were risk factors for having abnormal CT findings at the follow-up visit. Disease severity was found to be an independent risk factor for percentage change of CT score in the previous study28. Positive pressure ventilation or higher levels of fraction of inspired oxygen (FiO2), which are frequently used in serious patients, may themselves cause lung damage and may inhibit complete recovery29–31.
We found PFT impairment was more frequent in individuals who were discharged from ICU. Patients with severe disease were more prone to PFT impairment at the early convalescence phase and long-term27,28. Patients with PFT impairment had higher CT scores at the follow-up visit despite these patients having had lower CT scores at admission (Table-4). It is difficult to associate changes in pulmonary function tests with COVID-19, since the baseline pulmonary function status of patients is not fully known. Nevertheless, patients with impaired PFT had higher scores at the follow-up CT, suggesting a relationship between partial improvement in CT findings and PFT impairment.
Median distance of 6-min walk was similar in both the patients discharged from ICU and from the medical ward, and these results are consistent with a previous study28. However, in half of the patients discharged from the ICU, when the patients were stratified into age and weight, the 6-minute walk test was found to be less than 80% of the expected value. Immobilization, severity of illness, and use of corticosteroids are risk factors for reduced exercise capacity32. ICU patients had these risk factors, thus reduced exercise capacity could be expected in these patients. Prolonged immobilization after hospital discharge and restrictions to prevent transmission, such as general curfew, may have limited the mobilization of these patients in the recovery phase.
We found 79% of participants had at least one persistent symptom and patients with severe disease had a higher percentage of persistent symptoms. The percentage of residual symptoms in COVID-19 varies from 49% to 79% in previous studies14,28. Effort dyspnea, fatigue and muscle weakness were the most common persistent symptoms in our and a previous study28.
Impairment of quality of life was observed in SF-36 categories, especially social functioning, role limitation due to physical and emotional problems. Impairment of quality of life was similar in both ICU and non-ICU patients. This result is consistent with long term follow up in SARS patients24. The percentage of residual symptoms and impairment of health status were significantly higher in female participants. Female survivors were more prone to depression and anxiety after the previous SARS outbreak33. Severity of disease and female gender were found to be risk factors for persistent psychological symptoms28. Not just disease-related causes, but also social restrictions (such as quarantine, and curfew to prevent spread of the disease), increased stress, anxiety, and depression in females34. Psychological distress, anxiety and depression may aggravate persistent symptoms and influence the impairment of quality of life.
This study has several limitations. This is a single-centre study so these results cannot be generalized to other centres. Although the desired number of patients could not be reached in the ICU patients group, the study has enough power (0.83) to test the difference between the two groups for the primary outcome. We could not measure diffusion of carbon monoxide (DLCO), which is frequently impaired in patients with SARS or COVID-19, due to technical reasons in our pulmonary functional centre. The baseline data of PFT and 6-min walking distance were unknown so we cannot directly associate COVID-19 and PFT or 6-min walking impairment. The fact that the majority of the participants were men may have affected the results especially the assessment of quality of life.