Evolution of COVID-19-related olfactory disorders: A systematic review and meta-analysis

Abstract

Recently acquired olfactory dysfunction (OD) has emerged as one of the hallmark manifestations of the novel coronavirus disease (COVID-19), but the evolution of its spontaneous recovery has remained inconclusive, with reports of persistence of OD beyond 6 months of onset. We undertook this systematic review and meta-analysis with a view of generating a pooled recovery rate of COVID-19-associated ODs and attempt to examine the predictors of olfactory recovery. A systematic search of Scopus, Google Scholar, and PubMed databases, comprising all longitudinal studies reporting the trajectory of COVID-19-related OD, was carried out. The pooled recovery rate was estimated with random-effects model, and the potential heterogeneity of the subgroup sources was analyzed using a meta-regression test. After the Preferred Reporting Items for Systematic Reviews and Meta-Analysis selection process, 28 studies from 16 countries were included, with a total of 5,175 OD patients, among 11,948 COVID-19 cases. The estimated global pooled recovery rate of OD was 82.7% (95% confidence interval, 77.46–88.04%), with a pooled median duration of OD of 11.6 days. Only 2 out of 28 studies had recovery data beyond a period of 2 months. But no significant difference was found in the recovery rate regarding the length of follow-up (P = 0.840). Studies that conducted objective olfactory assessments showed a significantly higher recovery rate than those with subjective assessments (P = 0.001). Although 10 studies (36%) reported >90% recovery, 9 studies (32%) documented persistence of OD in >25% of their patients. Five out of six studies showed that hyposnia tended to show complete recovery than anosmia. Age, co-morbidities, and intranasal treatments had no effects. Test of homogeneity between subgroups using Cochran’s Q test was not significant (Q = 0.69, P = 0.40). Our meta-analysis revealed high rates of early- and medium-term recovery of COVID-19-related OD. However, it also showed disturbing rates of persistence of OD. Anosmia tended to be predictive of residual OD than hyposmia. Age, comorbidities, intranasal corticosteroid, and decongestants had no effects on OD recovery.