Lung Function In Discharged Patients After Coronavirus Infection (SARS, MERS, COVID-19): A Systemic Review And Meta-Analysis

Abstract

\n Background: To date, coronaviruses have caused three pandemics. Fewer studies concentrated on the prognosis of lung function.Objective: To summarize the lung function of the discharged after coronavirus infection.Methods: We systematically searched PubMed, Cochrane Library, Web of Science, and EMBASE. Two authors independently screened articles and extracted data. On average, predicted values and damage rates of seven lung function indices were pooled by single-arm meta-analysis. And, in severe/critical vs. non-severe/critical and one-year follow-up, they were pooled by two-arm meta-analysis. The source of high heterogeneity was explored by meta-regression or subgroup analysis. Results: Of the 7798 articles identified, 34 studies were included. On average, the pooled predicted values of the seven indices were within normal except for DLCO (79.2, 95% CI (76.2－82.2 )). Damage of lung function indices accounted for 6.2－35.2% of the discharged with DLCO most, and 83－100% of the damage was mild. Meta-regression showed that different viruses, countries, disease settings, and measurement times were not the source of high heterogeneities. In severe/critical illness vs. non-severe/critical, predicted values of seven indices were significantly lower (largest gap in DLCO (WMD -11.60, 95% CI -14.23－-8.98)). However, damage rates got rises only in DLCO (RR 1.74, 95% CI 1.46－2.07) and TLC, having no differences in the other indices. In one-year follow-up, predicted values were significantly improved in the severe/critical subgroup, while having no change in the non-severe/critical subgroup. Damage rates got no improvement in all indices.Interpretation: A single predicted value or damage rate can t give a clear description of lung function after coronavirus infection, and the trends of the two are sometimes inconsistent. We suggest more prospective cohort or follow-up studies in the future to lessen the influence of differences in lung function measurements across studies.Registration: PROSPERO (CRD42020192843)