Effects of Preoperative Factors on the Learning Curves of Postlingual Cochlear Implant Recipients.

This was a single-center retrospective record analysis, so the people and care pathway may not match other clinics. The learning curves excluded 30 people who did not meet fitting criteria, and 75 rehabilitation-time values were discarded because they exceeded the 96-month cutoff or were non-learners, so the analysed group was smaller than the 533 included. Duration of deafness was self-reported using inability to use a phone, which may be unreliable. Etiology was often unknown and not modelled; cochlear abnormalities were not included because of low incidence and heterogeneity. Preoperative predictors were highly collinear, and the authors removed some correlated variables, which weakens interpretation of individual factors. Residual distributions were skewed because scores and times were capped, which may affect regression estimates. The study analysed adults with unilateral implants under Dutch reimbursement rules, so results may not apply to bilateral implant users or other healthcare systems. The authors did not stratify cases by outcome level, which could have changed the regression results.

Advanced Bionics, whose implants were used in 83% of cases, cofounded the work and gave a nonrestrictive grant to H.C.S., J.H.M.F., and J.J.B.; J.H.M.F. is a member of the European medical advisory board of Advanced Bionics.

Do not use age, duration of deafness, best-aided speech scores, or education to predict an individual's post-implant speech outcome. The models explained only 2%, 10%, or 19% of differences, and the authors state these factors are not clinically useful.