Screening for hearing and visual loss among elderly with questionnaires and tests: which method is the most convincing for action?

Objective - To examine the extent to which self-report and performance-based measures identify the same people, and also whether one of these methods any more than the other induces the general practitioner (GP) to offer a patient help? Design - Comparison of tests and questionnaire in a screening programme for hearing and visual loss. Setting - Twelve general practices in the western part of The Netherlands. Patients - 1121 people 75 years of age and older. Main outcome measures - Hearing loss according to the whispered voice test and two questions (Q1, Q2) and visual loss according to the Snellen and low vision charts and two questions (Q3, Q4). The extent to which test and questions influence the GP's judgment regarding intervention is calculated using logistic regression. Results ? For hearing loss, the concordance of the whispered voice test with Q1 was 67%, and with Q2, 71%. For visual loss, the concordance of the Snellen chart with Q3 was 79% and of the Low Vision Chart with Q4, 69%. For hearing loss, the results of the test were noticeably more significant for the GP in offering help than the answers to the questions, and for visual loss the results of the Snellen chart slightly more convincing than the answers to the questions. Conclusion ? When in a screening programme with simple instruments for visual and hearing loss the choice has to be made between questionnaire and tests, the tests will be the best choice.