This study primarily involved patients who were hospitalized for COVID-19 between February 2020 and February 2021. Among this population, those who had at least one outpatient follow-up visit after recovery from COVID-19 were included in this study. At the outpatient visit, residual symptom investigations, chest CT scans, pulmonary function tests (PFT), and neutralizing antibody tests were performed, and each was compared with the results of testing at the time of COVID-19 (except for PFT). The study period was from January 11 to March 9, 2021. This study is generally a cross-sectional study. However, some data were collected retrospectively, especially from patients who had already had follow-up outpatient visits before the study period. The study subjects were recruited from seven university hospitals in Korea (four in Seoul, one in Cheonan, one in Gwangju, and one in Daegu). Written informed consent was obtained from all study participants, and the Institutional Review Board of Chung-Ang University Hospital approved the study protocol (IRB No. 2010-014-437).

Clinical data during hospitalization were obtained from electronic medical records including demographic information (age and sex), clinical characteristics (comorbidities, symptoms, and symptom onset time), and treatment regimen (corticosteroids, remdesivir, kaletra, antibiotics, anticoagulation, and vasopressor). Based on the World Health Organization’s clinical progression scale [13], COVID-19 severity was defined as follows: Scale 3 (hospitalized patients not requiring supplemental oxygen); Scale 4 (hospitalized patients requiring supplemental oxygen by mask or nasal prongs); Scale 5 (hospitalized patients requiring high-flow nasal cannula or non-invasive ventilation); and Scale 6 (hospitalized patients requiring intubation and mechanical ventilation). For our study, the low- and high-severity groups contained patients with severity scales of 3–4 and 5–6, respectively.

Patients who had follow-up outpatient visits during the study period filled out a questionnaire. The questionnaire was divided into three categories: COVID-19 symptoms, miscellaneous symptoms, and socioeconomic problems. The detailed questions are described in Appendix 1. If a patient who already had a follow-up outpatient visit before the study period, residual symptoms were investigated based on the medical records of the outpatient visits.

Patients who had follow-up outpatient visits during the study period underwent a high-resolution CT (HRCT) of the lungs. Any HRCT images taken before the study period at each hospital were sent to and scored by a radiologist at Chung-Ang University Hospital. For the CT score, each of the five lung lobes was scored from 0 to 5 based on the extent of involvement, as follows: 0 (no involvement), 1 (< 5% involvement), 2 (5%–25% involvement), 3 (26%–49% involvement), 4 (50%–75% involvement), and 5 (> 75% involvement). The total CT score consisted of the sum of each lobar score (ranging from 0 to 25) [14]. We classified patients with CT scores of 0–3 into the low-CT score group and those with CT scores of 4–10 into the high-CT score group.

Patients who had follow-up outpatient visits during the study period underwent PFT. No PFTs were performed before the study period because they could not be safely performed during the acute infectious period due to problems of transmission through the measuring equipment. Spirometry was performed according to the guidelines set out by the American Thoracic Society/European Respiratory Society [15]. We measured forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC), and diffusing capacity for carbon monoxide (DL_CO). Spirometry results were expressed as a percentage of the measured to the predicted values.

Blood samples were obtained to measure the SARS-CoV-2 antibody titers for patients who had follow-up outpatient visits during the study period. Anti-receptor binding domain (RBD) antibodies against the RBD of the SARS-CoV-2 spike protein were detected using an enzyme-linked immunosorbent assay (ELISA). We used a microneutralization assay to measure neutralizing antibodies against SARS-CoV-2, with the neutralizing antibodies titrated on Vero cells and observed a cytopathogenic effect after 72 hours. A detailed description of the test method is presented in the cited study [16]. Because some patients had serum samples stored from another study before the study period, any stored samples were tested with those collected during the study period.

Categorical variables were presented as numbers (percentages) and continuous variables were presented as medians (interquartile ranges [IQRs]). Either the Pearson’s chi-square test or Fisher’s exact test was used for categorical data, while the Mann-Whitney U test was used to compare continuous variables. A one-sample Wilcoxon test was used to check whether the CT score decreased with time after symptom onset. The null hypothesis was defined as a median ratio of the preceding CT score to the subsequent CT score being greater than 1, with a median ratio of less than 1 being defined as the alternative hypothesis. If the p value is < 0.05, the null hypothesis can be rejected, meaning that the CT score decreases with time. The correlation between CT scores with pulmonary functions was evaluated by Spearman’s correlation analysis. All statistical analyses were performed using GraphPad version 9.2.0 (GraphPad Software, La Jolla, CA, USA), with p values of less than 0.05 considered statistically significant.

Between February 2020 and February 2021, 1,665 patients were hospitalized at the study hospitals, of whom 1,586 were discharged, and 79 died (in-hospital mortality rate of 4.7%). Of the 1,586 patients, investigations of residual symptoms, chest CT scans, PFTs, and neutralizing antibody tests were performed in 84 (5.3%), 35 (2.2%), 31 (2.0%), and 27 (1.7%) cases, respectively. Supplementary Tables 1 and 2 and Supplementary Figs. 1 and 2 show the study process, test date, and the number of tests. The baseline characteristics of the 84 participants were as follows: median age 64 years (IQR, 55 to 70) with 59.5% (n = 50) male; 48.8% (n = 41) used corticosteroids; 32.1% (n = 27) did not requiring supplemental oxygen, 35.7 (n = 30) required supplemental oxygen, 20.2% (n = 17) required high-flow nasal cannula, and 11.9 (n = 10) required mechanical ventilation.

An investigation of residual symptoms after COVID-19 is presented in Table 1. Of the 84 patients who participated in the investigation, 72 were investigated once and 12 were studied twice. In the periods of 20–90, 91–180, and 181–360 days after the onset of symptoms, 45, 36, and 15 patients were investigated, respectively. Acute symptoms such as cough, sputum, myalgia, fatigue, nausea or vomiting, and diarrhea are rarely observed after 6 months. However, lung-related symptoms such as chest discomfort and dyspnea were observed in 20% to 30% of cases, even after 6 months. Dermatologic symptoms such as rash, urticaria, and hair loss were observed in 10% to 30% of subjects after 6 months. The proportion of participants who complained of socioeconomic problems did not decrease over time. Even after 6 months, 40.0% (6/15) and 26.7% (4/15) of patients experienced financial loss and emotional distress, respectively.

The baseline characteristics of the 35 subjects who underwent HRCT are presented in Table 2. In total, 15 had severity scales of 5–6 (high-severity group). There were no significant differences in age between the high- and low-severity groups (p = 0.064). However, hypertension was significantly more common in the high-severity group compared with the low-severity group (64.3% vs. 15.0%, p = 0.005) (Table 1). The overall median CT score was 6 (IQR, 2.5 to 10), with the high-severity group having a higher median CT score than the low-severity group (8 [IQR, 5.75 to 10.25] vs. 5 [IQR, 0 to 8], p = 0.010). The distribution of CT scores over time after the onset of symptoms is shown in Fig. 1A. To identify the change in CT score over time, the CT score was rearranged into 1–14, 15–60, and 61–365 days after the onset of symptoms (Fig. 1B). We calculated the ratio of changes in CT values between intervals for each patient and obtained the following median values of the ratio of change: 0.65 (the ratio of CT scores at 15–60 to 1–14 days; p = 0.003), 0.39 (the ratio of CT scores at 61–365 to 15– 60 days; p = 0.0002), and 0.20 (the ratio of CT scores at 61–365 to those in 1–14 days; p = 0.0003). These values indicate that the CT score decreases with time. When the ratio of change is divided by severity group, the value of the low-severity group is lower than the high-severity group (Fig. 2). The high-severity group showed a smaller decrease in CT score over time compared with the low-severity group (0.64 vs. 0.68 in Fig. 2A, p = 0.3115; 0.10 vs. 0.48 in Fig. 2B, p = 0.1784; 0.00 vs. 0.38, p = 0.0171), suggesting that the recovery was slower in the high-severity group.

The results of the Spearman’s correlation analysis for the CT scores and PFTs are presented in Fig. 3. Both FVC (R² = 0.2261, p = 0.012), FEV1 (R² = 0.2580, p = 0.007), and DL_CO (R² = 0.4231, p = 0.005) were significantly negatively correlated with CT scores. Compared with the low-CT score group, the high-CT score group showed significantly lower FVC (77% vs. 86%, p = 0.031), FEV1 (81% vs. 91%, p = 0.024), and DL_CO (69.5% vs. 113%, p < 0.001) values (Fig. 4).

SARS-CoV-2 antibody tests, including neutralizing antibodies and anti-RBD antibodies, were performed, producing 106 samples from 31 patients. When days elapsed after the onset of symptoms were divided into seven intervals (Fig. 5), the antibody titer was highest in the 22 to 60 days, and began decreasing after 120 days. From 22 to 60 days, anti-RBD antibody (RBD ELISA) titers tended to be higher in the high-severity group than in the low-severity group (1.53 vs.1.23 in 22–60 days, p = 0.0907; 1.21 vs. 0.50 in 61–120 days, p = 0.0688; 0.73 vs. 0.27 in 121–180 days, p = 0.0571), although there was no statistical significance when comparing neutralizing antibody titers.