A Prospective, Multicentre Case-Control Trial Examining Factors That Explain Variable Clinical Performance in Post Lingual Adult CI Recipients

Abstract

This study investigated which of a range of factors could explain performance in two distinct groups of experienced, adult cochlear implant recipients differentiated by performance on words in quiet: 72 with poorer word scores versus 77 with better word scores. Tests measured the potential contribution of sound processor mapping, electrode placement, neural health, impedance, cognitive, and patient-related factors in predicting performance. A systematically measured sound processor MAP was compared to the subject's walk-in MAP. Electrode placement included modiolar distance, basal and apical insertion angle, and presence of scalar translocation. Neural health measurements included bipolar thresholds, polarity effect using asymmetrical pulses, and evoked compound action potential (ECAP) measures such as the interphase gap (IPG) effect, total refractory time, and panoramic ECAP. Impedance measurements included trans impedance matrix and four-point impedance. Cognitive tests comprised vocabulary ability, the Stroop test, and the Symbol Digits Modality Test. Performance was measured with words in quiet and sentence in noise tests and basic auditory sensitivity measures including phoneme discrimination in noise and quiet, amplitude modulation detection thresholds and quick spectral modulation detection. A range of predictor variables accounted for between 33% and 60% of the variability in performance outcomes. Multivariable regression analyses showed four key factors that were consistently predictive of poorer performance across several outcomes: substantially underfitted sound processor MAP thresholds, higher average bipolar thresholds, greater total refractory time, and greater IPG offset. Scalar translocation, cognitive variables, and other patient related factors were also significant predictors across more than one performance outcome.

Keywords: cochlear implant; cognition; electrode placement; neural health; predictive factors; sound processor mapping; speech perception.

Figure 1.

Scatter plots for the prediction of LIT performance for predictor variables with a significant association (uncorrected p < .05). Good performer (GP) and poor performer (PP) group data is shown. The regression lines, Pearson correlation coefficients, and associated p-values are shown.

Figure 2.

CI performance across different outcome measures as a function of modiolar distance for perimodiolar arrays. For each outcome measure the uncorrected correlation (r), p-value, and number of good performers (GP) and poor performers (PP) are shown. Lower scores are better for AMDT and sentences in noise. AMDT=Amplitude Modulation Detection Test.