Current literature still reports varying speech perception outcomes across adult cochlear implant (CI) users. The largest proportion of variability in post-implantation outcomes remains unexplained, making it challenging for clinicians to provide adult CI candidates with an accurate expected outcome with their CI. Outcome prediction in this field has been attempted using linear models and machine learning algorithms, which have yet to yield reliable and accurate predictions of CI performance. The predictive value for such models and algorithms in clinical practice also remains to be explored. Available literature does not explore whether clinicians appear to hold additional knowledge on outcome prediction when presented with clinical case studies, compared to the factors currently implemented in outcome prediction models. Knowledge in this area may shed light on current prediction strategies in the clinical context. Our study investigated how accurately clinicians can predict adult CI users’ speech perception in quiet, one-year post-implantation. A survey of 10 clinical case studies of adult CI users was completed by 41 clinicians from around the world. Clinicians were provided with a comprehensive overview of pre-implantation, one-month post-implantation, and 6 months post-implantation information for each case. The case studies ranged in one-year post-implantation performance in quiet, from low performers (5% monosyllabic word score) to high performers (82% monosyllabic word score). Our results revealed that low performance wassignificantly over-predicted. Instead, clinicians appear to predict average performance (49-54% monosyllabic word score) across all cases and not performance variability between cases. Overall, clinicians were confident in their predictions, regardless of their prediction accuracy. Our findings can serve the clinical field by bringing prediction bias towards average performance to light, even when risk of lower performance is suggested. We hope these findings inspire clinicians to continue to be critical thinkers in the pre-operative counselling and expectation management journey, to timeously identify candidates at risk of low performance, and to initiate early, individualized intervention when indicated. Further research into methods of accurate outcome prediction is important for expectation management and counselling (especially given expanded CI criteria for individuals with more residual hearing), increased CI uptake and streamlining clinical care for high and low performers.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860718.

Variability in clinicians’ prediction accuracy for outcomes of adult cochlear implant users

It is with great pleasure and enthusiasm, that on behalf of the Cochlear Implant centers of the Radboud UMC in Nijmegen and Erasmus MC Sophia Children’s Hospital in Rotterdam to invite you to attend the 16th European Symposium on Pediatric Cochlear Implantation to Rotterdam, the Netherlands in 2023.

Yours sincerely, on behalf of the CI Centers at Radboud UMC and Erasmus MC,