Accuracy of tympanometry cutoff points in children from Central-Western Brazil.
Cláudia Cruvinel Câmara, Melissa Ameloti Gomes Avelino, Paulo Sergio Sucasas da Costa and 2 others
PMID 42561344WHAT IT FOUND
For 6-to-42-month-old children, a tympanometry cutoff of -180 daPa (not the -100 daPa most Brazilian programs use) best balances catching real middle-ear problems against sending healthy kids for unnecessary follow-up, but it still fails to detect about half of true cases.
Key findings
01The -180 daPa cutoff (suggested by the ROC curve) had the best overall balance: accuracy 85.71%, specificity 93.94%, sensitivity 55.56%, with a Youden index of 0.49.
02Static compliance below 0.2 mL had 98.42% specificity but only 50% sensitivity, meaning it correctly identifies most healthy ears but misses half of the ears with a problem.
03Using the -100 daPa cutoff most commonly applied in Brazilian screening programs generated 38% Refers (32 of 84 ears), while -200 daPa reduced this to 13% (11 of 84).
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
TEOAEs were used as the reference standard, not conventional audiometry (the gold standard), so the 'true' diagnosis is itself uncertain, especially for conductive losses that TEOAEs cannot detect. Single city, single country, single equipment model; 84 ears is a modest sample for estimating sensitivity and specificity with narrow confidence intervals. Sensitivity at the recommended -180 daPa cutoff was only 55.56%, meaning nearly half of children with a genuine middle-ear problem would be missed at screening. Cross-sectional design: one screening visit, no follow-up to confirm whether 'Refer' children actually had a persistent problem. The randomisation described was for the parent micronutrient trial (school-level allocation), not for the hearing-screening analysis; the diagnostic accuracy analysis itself is a single-time-point comparison. Environmental noise in the screening rooms was moderate (mean maximum 73.96 dB), which could affect TEOAE reliability in young children.
Declared interests
Funded by the Brazilian Ministry of Health (DECIT), CNPq, FAPEG, the Goiás State Health Secretariat, and Pontifícia Universidade Católica de Goiás. No industry or device-manufacturer funding is declared. No conflicts of interest beyond the public funding sources are listed.
The easy way to misread this
Do not read the 85.71% accuracy at -180 daPa as 'this test is reliable enough to rule out hearing problems.' Sensitivity was only 55.56%, meaning roughly one in two children with a true middle-ear problem will pass screening and not be referred. Additionally, TEOAEs were the reference standard, not audiometry, so the 'true' diagnosis itself is uncertain. A pass at this cutoff should prompt continued surveillance, not reassurance.