Relationships Between Objective Laryngostroboscopic Measures and Patient Ratings With Treatment Recommendations for...

To use principal component analysis (PCA), including objective laryngostroboscopic measures and patient-reported quality of life and symptom severity ratings from initial evaluations of patients with presbylarynges as an initial step toward developing a framework for streamlining specific treatment recommendations, especially for those who might benefit initially from procedural intervention (PI), as well as patient compliance with those recommendations.

The design was retrospective, analytical, observational, and cross-sectional.

One hundred eighteen patients with presbylarynges evaluated by one laryngologist participated. Variables included normalized glottal gap area (NGGA), normalized true vocal fold width (NTVFW), total bowing index (Total BI) from single laryngostroboscopic images, multi-item ratings [Glottal Function Index (GFI), Voice-Related Quality of Life (V-RQOL)], single-item ratings (effort, weakness, hoarseness, swallowing, and breathing), and age. Associations between variables and treatment recommendation were assessed using bivariate tests followed by multinomial logistic regression, and between variables and compliance using bivariate tests followed by binomial logistic regression.

V-RQOL, GFI, effort, weakness, and hoarseness significantly differed between treatment groups. More severe ratings of GFI, V-RQOL, effort, weakness, and hoarseness were associated with greater odds of either intervention being recommended over observation. More severe hoarseness ratings were associated with greater odds of recommending PIs over voice therapy (VT). Two principal components (PC1, PC2) were revealed. PC1 represented patient satisfaction with voice quality and less severe vocal symptoms. PC2 represented laryngeal impairment. Using a multivariable logistic model (PC1, PC2, and sex), lower values of PC1 and being male were associated with greater odds of being recommended for either intervention compared with observation. None of the predictors considered were significantly associated with compliance.

This initial attempt using data from a single clinic found significant differences in patient ratings between treatment options and significant associations between variables associated with treatment recommendation. The associations were ultimately not predictive for either treatment recommendations or compliance for patients with presbylarynges.