Single-sided deafness changes how a cochlear implant is used, and how it should be followed

Design

Retrospective longitudinal cohort at a tertiary academic centre, 2018-2024

Population

360 adult cochlear implant recipients grouped by contralateral pure-tone average (81 single-sided deafness, 41 asymmetric, 238 bilateral non-serviceable)

Primary outcome

Daily device use by datalogging and consonant-nucleus-consonant word recognition to 12 months

Effect

Full-time use in 50 per cent of single-sided deafness vs 82 per cent bilateral (P < 0.001); single-sided deafness OR 4.38 (95 per cent CI 2.03 to 9.48) for limited use

Three hundred and sixty adults implanted at one academic centre between 2018 and 2024 were grouped by the hearing in the other ear: single-sided deafness with a contralateral pure-tone average of 30 dB HL or better (n = 81), asymmetric hearing loss between 30 and 50 dB (n = 41), and bilateral non-serviceable hearing above 50 dB (n = 238). Processor datalogging gave daily use, and consonant-nucleus-consonant word scores were taken at 1, 3, 6 and 12 months.

The single-sided deafness group behaved differently on every measure. Only 50 per cent reached full-time use of eight hours a day or more, against 82 per cent of the bilateral group (P < 0.001), and single-sided deafness was the only independent predictor of limited use (OR 4.38, 95 per cent CI 2.03 to 9.48). Word scores stayed lower across the first year and did not cross the 50 per cent responder threshold at 12 months. Most usefully, the early-performance shortcut broke: one-month scores predicted 12-month outcomes in the asymmetric and bilateral groups but not in single-sided deafness.

That last point is the change to practice. The routine of reassuring a slow starter that early scores predict the year, or of taking a good one-month score as evidence that rehabilitation is on track, does not apply to a patient with a normal contralateral ear. These recipients need configuration-specific counselling before implantation about what use will realistically look like, and closer surveillance from activation rather than at the standard intervals - because the usual early signal will not tell you who is drifting.

• Counsel single-sided deafness candidates before surgery that only about half reach full-time use in the first year.
• Do not use one-month word scores to predict the year in this group; the correlation is absent.
• Bring follow-up forward from activation rather than relying on the standard schedule.
• Check datalogging at every visit, since falling use is the earliest measurable signal.
• Keep expectations for word recognition separate from expectations for sound localisation and listening in noise, which are the usual reasons to implant here.

The statistics, in plain English

An odds ratio of 4.38 with an interval from 2.03 to 9.48 is a solid finding in 360 patients, and the 50 against 82 per cent difference in full-time use is large enough to matter clinically as well as statistically. The failure of one-month scores to predict 12-month outcomes in this group is an absence of correlation, which is harder to interpret than a difference - it may reflect genuinely different rehabilitation dynamics or simply a smaller, more variable subgroup. This is retrospective single-centre data, so the exact proportions will not transfer, but the direction of all three findings is consistent.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.