Twenty-seven studies (designs not reported) with 143,825 participants (although one study's numbers of 1,367 counted ears screened not individual participants). Eighteen studies were peer-reviewed publications (52,260 participants); 8 studies were pre-prints (85,419 participants); 2 studies were general reviews for edited volumes (6,103 participants); and 1 study was an unpublished audit report (43 participants).
Neonatal:
All methods of neonatal hearing screensshowed high screen specificity, generally well above 90% after a 'settling-in' period.
Evidence on screen sensitivity for moderate and greater cases of congenital PCHI is less available, but estimates range from 80% to 100%, except for the portable auditory response cradle (PARC) when used with at-risk babies, where sensitivity may fall to unacceptable levels. Generally, programme sensitivity (including cases which were not covered, and/or which are late-onset or progressive) may be estimated to be nearer 80% than 100%.
The two large cohort studies of UNS in the UK, including the controlled trial in Wessex, produce yields of the expected order, that is 1-1.3 per 1000, but decrease the subsequent incremental yields of the HVDT to very low levels.
At-risk neonatal screening has a potential yield of about 60% of all cases. In practice, this is likely to be much lower because of the difficulty of implementing ful coverage for all indicators of at-risk cases; perhaps 45-50% at best.
The median age of identification of those screened neonatally is of the order of 2 months, depending on follow-up procedures and severity of impairment; this is earlier than for cases not screened neonatally.
Coverage of the HVDT screen probably falls in the range 80-95%, although there may be some urban areas where coverage falls to nearer 60%. There is some limited evidence suggesting that coverage and sensitivity is lower in the Asian population.
HVDT:
Sensitivity estimates of the HVDT vary widely, from 18% to 88% (all degrees of loss). Recent studies and more powerful studies are suggestive of poorer levels of sensitivity. Severity of impairment affects screen sensitivity substantially.
Screen-positive, that is, 'fail' rate is of the order of 5-10%. Many of these cases will have fluctuating non-permanent hearing loss, associated with otitis media with effusion (OME). This referral rate has considerable resource implications for services.
HVDT incremental yield may be at very best 40% but it falls off for the best districts to low levels (e.g. 25%) when at-risk neonatal screening is introduced. With UNS, the evidence indicates that the HVDT incremental yield falls to very low levels. (For those Districts who supplied data in the survey of current practice, the average HVDT yield was about 26-28% and at-risk neonatal screening made no difference).
Median age of identification via the HVDT varies from 12 to 20 months. Age of referral is severity-dependent.