| Abstract | Objective: To analyze the reliability, observational error, and responsiveness of the Voice Handicap Index (VHI) in measuring vocal handicap in individuals with dysphonia.
Methods: This systematic review followed the recommendations of the Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) and was registered in PROSPERO (CRD42022344633). We included studies that validated the VHI; assessed the measurement properties of reliability, observational error, or responsiveness; and with participants with dysphonia. Searches were conducted in electronic databases: Cochrane Library, EMBASE, LILACS, PubMed, Scopus, and Web of Science. Manual searches were conducted in gray literature through the Biblioteca Digital de Teses e Dissertações and ProQuest Dissertation & Theses, along with citation mapping and consultation with an expert. Two independent and blinded reviewers carried out evidence selection, data extraction, assessing the risk of bias, assessing the certainty of evidence, and psychometric measures. Meta-analysis was performed using the Fisher-transformed correlation coefficient r-to-z and standardized mean difference. Heterogeneity was calculated using Tau² and I² statistical tests by JAMOVI 2.3.2 software.
Results: Eighty researchers were selected. Most studies were classified as inadequate in the reliability (80%) and observational error (97.43%) stages in the bias of risk assessment. In the responsiveness stage, they were classified as doubtful in the sensitivity and specificity sub-boxes (92.85%) and good compared to a gold standard instrument sub-box (44.4%). In the evaluation of good psychometric properties, most studies were classified as indeterminate for reliability (55.84%) and responsiveness (80%). In the meta-analysis of the test-retest reliability (k=77), the estimated average Fisher-transformed correlation coefficient (r-to-z) was 1.76. The result indicates a correlation between the test and retest moments of the VHI (z=28.804, p<0.001). In the responsiveness assessment (k=5), the standardized mean difference was 1.29. This result showed a significant reduction in VHI scores post-intervention compared to the pre-intervention period (z=2.10, p=0.035). In the assessment of evidence certainty, the level of evidence was considered very low, both in the reliability and responsiveness stages.
Conclusion: The VHI proved reliable and responsive in measuring vocal handicap. However, caution is advised, given the heterogeneity and risk of bias.
|
|---|