Heterogeneity analysis
Subgroup analysis according to severity of disease and sensitivity analysis were carried out evaluating the possible source of heterogeneity. In two studies (13, 14) patients with asymptomatic COVID-19 were studied and in four studies all types of disease (asymptomatic and symptomatic) were studied (15, 17-19). According to meta-analysis the pooled mean of FEV1 in asymptomatic subgroup was 98.59%, 95%CI= (96.96, 100.23), I2=zero. The pooled mean of FEV1 in symptomatic subgroup was 103.91%, 95%CI= (101.08, 106.74), I2=53.27. The pooled mean of FVC in asymptomatic subgroup was 95.17%, 95%CI= (92.80, 97.54), I2=zero. The pooled mean of FVC in symptomatic subgroup was 104.62%, 95%CI= (98.00, 111.24), I2=91.17. The pooled mean of FEV1/FVC in asymptomatic subgroup was 98.28, 95%CI= (86.25, 110.31), I2=zero. The pooled mean of FEV1/FVC in symptomatic subgroup was 94.99%, CI= (88.19, 101.79), I2=95.65.
Meta regression showed neither disease severity nor asthma comorbidity had a significant effect on pooled mean of FEV1 (p value= 0.35 and 0.21 respectively) and FVC (p value= 0.80 and 0.51 respectively). In sensitivity analysis, the effect of each study on the pooled mean was assessed. There was no major deviation from pooled mean by omitting studies in FEV1, FVC FEV1/FVC, FEF25-75 and DLCO outcomes, indicating the stability and robustness of the results (Data not shown).