Large-scale evaluation of the Digit Triplet Test for school-age hearing screening of 5th & 9th grade school children

Given that some types of hearing impairment (HI) are only acquired in the early years of life, and given the well-known negative consequences of (even minimal) HI, accurate hearing screenings at a time when adequate hearing is of educational importance are needed as a follow-up to the newborn hearing assessment. Unfortunately, pure-tone screening (PTS) protocols — which are most commonly used — lack accuracy in a school screening environment. Better screening protocols need to be developed.

In Flanders, school-age hearing screenings (SHS) are mandatorily organized by regional school health services in 1st graders (5-6 yrs), 5th graders (10-11 yrs) and 9th graders (14-15 yrs). In these last two cohorts, PTS has recently been replaced by a speech-in-noise (SPIN) screening protocol using the Flemish Digit Triplet Test (DTT). This automated self-test has a very high sensitivity and specificity (> 90%) to detect mild HI in adults, as well as a high test-retest accuracy. Furthermore, it has been shown that bilateral speech reception thresholds (SRT) can be obtained within 4-6 minutes. In the current study, large-scale applicability for SHS was investigated and age-appropriate pass/fail-criteria were determined.

Eleven school health service centers participated. Children who failed the test, based on preset cut-off SRTs of -7.2 dB (5th graders) and -8.5 dB (9th graders), were referred for a diagnostic evaluation with a full audiogram assessment. We aimed for a referral rate of 5% to receive sufficient audiograms enabling pass/fail-criteria to be fine-tuned.

Out of the 6499 children tested (3304 5th graders, 3195 9th graders), 4.1 and 3.6% failed the test, respectively. We obtained audiograms of half of the referred children (48% in 5th graders, 52% in 9th graders). Elevated audiometric thresholds were confirmed in only 1/5 (5th graders) and 1/3 (9th graders) of the children. Because of this low positive predictive value and relatively high referral rate, pass/fail-criteria of the DTT must be tuned less strict. When a fail is only registered in case of an SRT being 2.5 interquartile ranges above the median group-SRT (-9.7 dB in 5th graders, -10.5 in 9th graders), referral rates drop to 2.6 and 3%, respectively, whilst positive predictive values rise by 10 and 14%.

Effective data registration will enable monitoring of the SRT over the school years. Furthermore, this will enable to track the performance of the screening program. By collecting epidemiological data, we will be able to estimate the prevalence of acquired HI in school children.