The quality of diagnostic studies was not reported. The quality of other types of study was generally poor. Methodological limitations included lack of: sample size planning; blinded outcome assessment; adjustment for confounding factors; and reporting of uninterpretable tests and tests not performed. One treatment study showed minor deficiencies; all other studies had major deficiencies.
Diagnostic accuracy: nine studies
Eight studies (n=1,801 patients, range 64 to 444) compared otoacoustic emissions with automated auditory brainstem response. Studies were heterogeneous. Sensitivity ranged from 0.50 (95% CI 0.35 to 0.65) to 1.00 (95% CI 0.03 to 1.00). Specificity ranged from 0.49 (95% CI 0.36 to 0.63) to 0.97 (95% CI 0.94 to 0.99).
One study (n not reported) compared two-stage screening using otoacoustic emissions plus auditory brainstem response and reported a sensitivity of 0.92 (95% CI 0.74 to 0.98) and specificity of 0.99 (95% CI 0.98 to 0.99). Children who screened negative were not followed up.
Screening versus no screening: two studies (n=101 and 50)
Studies compared hearing-impaired children from hospitals, time periods or regions with and without screening. Test procedures were specified. Screening was associated with improvements in receptive language development (two studies reported significant effects), expressive language development (one study reported a significant effect of screening and the other reported a non-significant trend), total language development (one study reported a significant effect of screening) and communicative abilities and spontaneous speech (one study reported a significant effect of screening).
Early versus later treatment: six studies (n not reported)
Five studies were retrospective analyses and one was a population-based cohort study. Treatments varied and included cochlear implants, hearing aids, early intervention sessions and a parent/infant programme. One study reported a significant benefit in language intelligibility in the early screening group. Four of five studies reported improvements or significant improvements in receptive language development in early screening groups. One study reported no significant difference.