The COVID-19 pandemic has profoundly impacted lives, causing both short-term and long-term health consequences, including a significant socioeconomic burden. Many symptoms related to the aerodigestive system occur in COVID-19 patients, including cough, dyspnea, sore throat, olfactory and gustatory dysfunction, and voice impairment (dysphonia). While dysphonia is a common symptom, it is often underestimated due to its relatively less emergent nature compared to other symptoms. Dysphonia, characterized by vocal functional changes, is mainly attributed to the immune response following infection, leading to poor sound production and negatively impacting quality of life through reduced self-confidence and communication difficulties. Therefore, understanding the prevalence and risk factors of COVID-related dysphonia is crucial for early management and prevention. While several observational studies exist, a synthesis of data on the prevalence of COVID-related dysphonia is lacking, especially regarding long-term sequelae. The persistence of symptoms, defined as "long-COVID" by the World Health Organization (at least three months post-infection, with symptoms lasting more than two months), necessitates investigation into whether dysphonia prevalence changes after recovery from acute infection. This study aimed to investigate the global prevalence of COVID-related dysphonia and associated epidemiological risk factors during and after acute infection to provide insights into this health problem.

The introduction section mentions several studies exploring voice outcomes in COVID-19 patients using various assessment modalities, but highlights the lack of synthesized data on the prevalence of COVID-related dysphonia, particularly concerning long-term sequelae. The need for this meta-analysis is justified by the absence of a comprehensive overview of the prevalence and risk factors of this issue.

This systematic review and meta-analysis followed the PRISMA guideline and was registered with PROSPERO (CRD42022383399). Five electronic databases (PubMed, Embase, ScienceDirect, Cochrane Library, and Web of Science) were searched until December 15, 2022, using keywords such as "COVID-19" or "SARS-CoV-2" and "voice impairment," "dysphonia," or "aphonia." Studies published between December 2019 and December 2022 were included, regardless of language. Inclusion criteria included prospective and retrospective cohort studies, case-control studies, cross-sectional studies, and clinical trials reporting dysphonia with a clear definition and sufficient data on prevalence. Exclusion criteria involved studies with patients under 18, populations mainly composed of occupational voice users, or studies not excluding pre-existing dysphonia. Data extraction included author names, publication year, country, sample size, age, gender, smoking percentage, voice-related comorbidities, assessment modalities, tracheostomy/intubation percentage, and dysphonia prevalence during and after infection. The STROBE checklist was used to assess risk of bias. A random-effects model was used for meta-analysis, with heterogeneity assessed using the I² statistic. Meta-regressions analyzed the impact of clinical factors (age, gender, tracheostomy/intubation, and assessment modality) on dysphonia prevalence. Sensitivity analysis (one-study-removal) and publication bias assessment (funnel plots and Egger's tests) were performed. Statistical significance was set at P < .05.

Twenty-one studies involving 13,948 patients were included. During acute COVID-19 infection, the weighted prevalence of dysphonia was 25.1% (95% CI: 14.9% to 39.0%), significantly lower in males (-0.116, 95% CI: -0.196 to -0.036; P = .004). After recovery, the weighted prevalence decreased to 17.1% (95% CI: 11.0% to 25.8%). In the long-COVID subgroup (at least three months post-recovery), the weighted prevalence of persistent dysphonia was 20.1% (95% CI: 8.6% to 40.2%). Meta-regression analysis showed that only male gender was significantly associated with lower dysphonia prevalence during acute infection. No other significant associations were found with age, tracheostomy/intubation, or assessment modality in either the acute or post-acute phases. High heterogeneity was observed across studies, but sensitivity analyses did not alter the primary findings. Funnel plots suggested potential publication bias, but Egger's tests were not significant for acute infection (P = .889) or long COVID (P = .072).

This meta-analysis is the first to specifically focus on the prevalence and risk factors of COVID-related dysphonia. The findings highlight the substantial impact of COVID-19 on voice, both during acute infection and in the long-term. The higher prevalence in females during acute infection may be explained by differences in immune response, although the mechanisms require further investigation. The lack of significant association between tracheostomy/intubation and dysphonia in the long-term aligns with previous research showing that most post-extubation vocal symptoms are self-limiting. The lack of significant association with assessment methodology suggests that subjective assessments alone might suffice during the acute phase. The persistence of dysphonia in a significant proportion of patients after recovery highlights the need for awareness and management of long-COVID voice sequelae. Possible underlying mechanisms include viral entry into the aerodigestive tract through ACE2 and TMPRSS2 receptors, leading to inflammation in the vocal tract.

This meta-analysis reveals that a significant proportion of COVID-19 patients experience dysphonia during acute infection, with females being more affected. Importantly, a considerable number of these patients continue to suffer from persistent voice problems even after recovery, emphasizing the need for long-term monitoring and treatment. Future studies should investigate the severity of voice impairment, explore the roles of additional factors like smoking and comorbidities, and examine treatment efficacy in the context of long-COVID dysphonia. Further research is also needed to understand the evolving prevalence of dysphonia with changes in COVID-19 variants.

This study's limitations include the lack of assessment of dysphonia severity, limited analysis of certain potentially influential factors (smoking, minor comorbidities) due to insufficient data, absence of treatment analysis (due to challenges with compliance and follow-up), and the potential for the results to reflect the situation during the earlier stages of the pandemic, with the possibility of different prevalence rates in the present context.