Hearing impairment is prevalent among the elderly population but commonly underdiagnosed.
To review the accuracy and precision of bedside clinical maneuvers for diagnosing hearing impairment.
MEDLINE and EMBASE databases (1966 to April 2005) were searched for English-language articles related to screening for hearing impairment.
Original studies on the accuracy or precision of screening questions and tests were included. Articles that used unaccepted reference standards or contained insufficient data were excluded. Medical Subject Headings or keywords used in the search included hearing loss, hearing handicap, hearing tests, tuning fork, deafness, physical examination, sensitivity, specificity, audiometry, tuning fork tests, Rinne, Weber, audioscope, Hearing Handicap Inventory for the Elderly–Screening version, whispered voice test, sensorineural, and conductive.
One author screened all potential articles and 2 authors independently abstracted data. Differences were resolved by consensus. Each included study (n = 24) was assigned a methodological grade.
A yes response when asking individuals whether they have hearing impairment has a summary likelihood ratio (LR) of 2.5 (95% confidence interval [CI], 1.7-3.6); a no response has an LR of 0.13 (95% CI, 0.09-0.19). A score of 8 or greater on the screening version of the Hearing Handicap Inventory for the Elderly (HHIE-S) has an LR of 3.8 (95% CI, 3.0-4.8); a score less than 8 has an LR of 0.38 (95% CI, 0.29-0.51). An abnormal Weber tuning fork test response has an LR of 1.6 (95% CI, 1.0-2.3); a normal response has an LR of 0.70 (95% CI, 0.48-1.0). An abnormal Rinne tuning fork test response has LRs ranging from 2.7 to 62; a normal response has LRs from 0.01 to 0.85. Inability to perceive a whispered voice has an LR of 6.1 (95% CI, 4.5-8.4); normal perception has an LR of 0.03 (95% CI, 0-0.24). Not passing the audioscope test has an LR of 2.4 (95% CI, 1.4-4.1); passing has an LR of 0.07 (95% CI, 0.03-0.17).
Elderly individuals who acknowledge they have hearing impairment require audiometry, while those who reply no should be screened with the whispered-voice test. Individuals who perceive the whispered voice require no further testing, while those unable to perceive the voice require audiometry. The Weber and Rinne tests should not be used for general screening.