Accuracy of Cepstral Peak Prominence in the Identification of Voice Disorders: A Systematic Review With Meta-Analysis

The objective of this study was to evaluate the diagnostic accuracy of Cepstral Peak Prominence (CPP) in the detection of vocal disorders, comparing its discriminatory performance in sustained vowel and connected speech tasks.

A systematic review of diagnostic accuracy, with meta-analysis, was conducted and registered with PROSPERO (CRD42022377837). The systematic search was conducted through November 2025 in the MEDLINE, EMBASE, Web of Science, LILACS, and SCOPUS databases, as well as in the gray literature. We included cross-sectional observational studies that assessed CPP in sustained vowels (CPP_SV) and/or connected speech (CPP_CS), using auditory-perceptual judgment as the reference standard. Risk of bias was assessed using the Joanna Briggs Institute tool. A meta-analysis using random-effects models was performed to estimate sensitivity, specificity, and the Diagnostic Odds Ratio (DOR).

Sixteen primary studies were included, totaling 11,714 participants. Most studies (93.75%) had a low risk of bias. In the meta-analysis, the CPP_SV demonstrated an overall combined sensitivity of 77% (95% confidence interval [CI]: 68-85%) and specificity of 82% (95% CI: 72-89%), resulting in a DOR of 15.26 (95% CI: 5.31-43.87). For connected speech, the CPP_CS showed superior diagnostic performance, with sensitivity and specificity of 82% (95% CI: 70-90% and 71-89%, respectively) and an OR of 20.82 (95% CI: 6.04-71.64). The certainty of the evidence (GRADE) was high for all outcomes.

CPP exhibits excellent diagnostic accuracy, with better performance in connected speech. Although high methodological heterogeneity is a limitation, CPP is a robust acoustic measure and is clinically recommended as a complement to auditory-perceptual judgment.