The reliability and validity of patient self-rating of their own voice quality

To provide preliminary data on the reliability and validity of dysphonic patients rating their own voice quality.

Prospective reliability/validity assessment of voice ratings in dysphonic patients.

The Royal Free Hampstead NHS Primary Care Trust.

Thirty-five adult dysphonia patients recruited from ENT referrals to a speech and language therapy department. Exclusion criteria were (i) a hearing impairment which may affect auditory discrimination and (ii) a diagnosis of cognitive impairment which may affect task comprehension.

Patient intra-rater reliability was assessed by test-retest ratings, using G (Grade), R (Rough), B (Breathy), A (Asthenic), S (Strained) (GRBAS). Validity was assessed by comparing (i) patient-clinician inter-rater reliability, (ii) patients' GRBAS ratings with their Vocal Performance Questionnaire (VPQ) responses.

(i) Patients had lower intrarater reliability than clinicians (for G of GRBAS, kappa = 0.51 versus 0.74); (ii) patients consistently rated their voices more severely than clinicians (for G of GRBAS, mean rating = 1.4 versus 1.0); (iii) clinician-patient inter-rater agreement was no better than chance (paired t-test, all P < 0.05); (iv) patient ratings correlated significantly with vocal performance scores (r > 0.4, P < 0.05).

Patients appear to have good validity and consistency using GRBAS as a self-perception tool. However, validity measured in terms of agreement with clinician ratings is poor. Voice patients may rate what they perceive rather than what they hear. Disagreement between patient and clinician ratings has implications for therapy aims, prognosis, patient expectations and outcomes. Where disagreement persists, the clinician may have to determine whether therapy priorities need redesigning to reflect patients' perceived needs, or to evaluate whether patient perceptions and expectations are unrealistic.