An audioscope was more accurate than a questionnaire for screening for hearing loss in older adults
ACP J Club. 1994 July-Aug;121:23. doi:10.7326/ACPJC-1994-121-1-023
McBride WS, Mulrow CD, Aguilar C, Tuley MR. Methods for screening for hearing loss in older adults. Am J Med Sci. 1994 Jan;307:40-2.
To determine the diagnostic accuracy of the Hearing Handicap Inventory for the Elderly-Screening Version (HHIE-S) and the portable Audioscope (Welch-Allyn Inc., Skaneatles Falls, New York) for the detection of hearing loss in older adults.
Comparison of the HHIE-S and the Audioscope with pure tone audiometry, the diagnostic standard.
Primary care clinics at a community health center and a Veterans Affairs medical center.
185 consecutive patients over the age of 60 years (mean age 70 y, 69% men). Exclusion criteria were cerumen impaction or severe comorbid illness.
Description of tests and diagnostic standard
The Audioscope and the HHIE-S were administered in random order. The Audioscope was administered by a research associate, and hearing impairment was defined as a better-ear threshold of ≥ 40 dB at 2 kHz. The HHIE-S, a 10-item questionnaire assessing the perceived handicap that a person experiences as a result of hearing impairment, was self-administered or read aloud if patients were illiterate or had visual problems. Scores ranged from 0 to 40, with < 10 indicating no handicap and > 24 indicating moderate to severe handicap. Pure tone audiometry was done by an audiologist. 3 reference standards for hearing loss were considered: Speech Frequency Pure Tone Average (mean threshold > 25 dB at 0.5, 1.0, and 2.0 kHz in the better ear); High Frequency Pure Tone Average (mean threshold > 25 dB at 1.0, 2.0, and 4.0 kHz in the better ear); and Ventry and Weinstein criteria (40 dB binaural loss at 1.0 or 2.0 kHz or a 40 dB loss at both frequencies in either ear).
Main outcome measures
Sensitivity, specificity, and positive likelihood ratios.
The ranges of sensitivities, specificities, and positive likelihood ratios of the Audioscope for the 3 reference standards were 64% to 96%, 80% to 91%, and 4.9 to 7.5, respectively, compared with 29% to 63%, 75% to 93%, and 2.4 to 4.3, respectively, for the HHIE-S (set at 2 different cutpoints, > 8 and > 24) (Table). Combining the Audioscope and HHIE-S (cut point > 24) resulted in lower sensitivities and slightly improved specificities and positive likelihood ratios compared with the Audioscope alone. 60% of the patients preferred the testing methods of the Audioscope, 13% preferred the HHIE-S, and 27% expressed no preference.
The Audioscope was more sensitive than the Hearing Handicap Inventory for the Elderly-Screening Version and had similar specificities and likelihood ratios for the detection of hearing loss in older adults. The Audioscope was also the testing method preferred by patients.
Source of funding: American College of Physicians' Teaching and Research Scholarship.
For article reprint: Dr. C.D. Mulrow, 7400 Merton Minter Boulevard, Audie Murphy Veterans Affairs Hospital, Ambulatory Care (11C), San Antonio, TX 78284, USA. FAX 210-567-4685.
Table. and Hearing Handicap Inventory in the Elderly Screening Version (HHIE-S) for detecting hearing loss in older adults*
|Test||Diagnostic standard||Sensitivity (95% CI)||Specificity (CI)||+LR||-LR|
|Audioscope||SFPTA||64% (52 to 77)||89% (83 to 94)||5.8||0.4|
|HFPTA||71% (63 to 80)||91% (84 to 97)||7.5||0.3|
|V & W||96% (90 to 100)||80% (74 to 87)||4.9||0.05|
|HHIE-S cutpoint >8||SFPTA||58% (45 to 70)||76% (69 to 84)||2.4||0.5|
|HFPTA||48% (39 to 58)||86% (79 to 94)||3.6||0.6|
|V & W||63% (49 to 76)||75% (68 to 82)||2.5||0.5|
|HHIE-S cutpoint >24||SFPTA||36% (23 to 48)||87% (81 to 93)||2.8||0.7|
|HFPTA||29% (20 to 37)||93% (88 to 99)||4.3||0.8|
|V & W||42% (28 to 56)||88% (82 to 93)||3.4||0.6|
|Audioscope HHIE-S (>24) combined||SFPTA||31% (20 to 44)||96%(91 to 99)||7.8||0.7|
|HFPTA||27% (16 to 39)||98% (92 to 99)||13.||0.7|
|V & W||41% (27 to 51)||96% (86 to 98)||10.3||0.6|
*HFPTA = High frequency pure tone average; SFPTA = Speech frequency pure tone average; V & W = Ventry and Weinstein's criteria of a 40-decibel binaural loss at 1.0 or 2.0 KHZ. Other abbreviations defined in Glossary.
McBride and colleagues have confirmed the validity of the Audioscope and the HHIE-S (1, 2). Strengths of this study are clear and include random ordering of tests and blinding of the audiologist to screening results. Less clear is the selection of ≥ 40 dB at 2 kHz alone to define a positive Audioscope test. Particularly for the Speech Frequency or High Frequency definitions of hearing loss, the 25-dB threshold and a wider range of frequencies may be more suitable and yield higher sensitivities. The rationale of testing the combination of a positive Audioscope test and a score > 24, rather than > 8, on the HHIE-S is also unclear if the objective was to maximize sensitivities for screening.
Studies in other settings have shown low rates of follow-up after screening for hearing impairment (1, 2). McBride and colleagues observed that the HHIE-S identified patients who are more likely to obtain further evaluation, hearing rehabilitation, and benefit from hearing aids. The Canadian Task Force on the Periodic Health Examination (3), and the U.S. Prevention Services Task Force (4) have recommended screening older adults for hearing impairment; the former suggests use of a single question about hearing difficulty or the Audioscope, whereas the latter favors periodic questions about hearing but not hand-held audiometry devices. As shown by McBride and colleagues, the Audioscope equals or exceeds other tests in accuracy and may be preferred by patients, yet self-assessment questionnaire represent the most rapid and least expensive approach.
It remains unknown whether some combination of the Audioscope and the HHIE-S or some other questions may prove to be superior to the Audioscope alone for identifying patients who will require further evaluation (1).
Thomas Jones, MD
The Milton S. Hershey Medical CenterHershey, Pennsylvania, USA