Nucleus 24 advanced encoder conversion study: performance versus preference.

OBJECTIVE: The Nucleus 24 Advanced Encoder Conversion Study was designed to determine the safety and effectiveness of the advanced combination encoder (ACE) and continuous interleaved sampling (CIS) speech coding strategies compared with that of the spectral peak (SPEAK) strategy in a large sample of postlinguistically deaf adults. Data from this study were analyzed to test the hypothesis that the group of subjects who prefer a given strategy for use in everyday life obtain significantly higher speech recognition scores (as a group) with the preferred strategy than with nonpreferred ones.

DESIGN: The first 100 adults implanted with the Nucleus 24 Cochlear Implant System who had a minimum of 3 mo experience with the device were invited to participate. Those who accepted were randomly assigned to one of two groups for an initial 6-wk use of either the ACE or the CIS strategy; the other strategy was used during the second 6-wk period. Parameters in subjects' SPrint speech processor programs were adjusted to maximize perceived benefit with each strategy in everyday life. Recognition of medial consonants and vowels, CNC words, CUNY sentences in quiet and at + 10 dB signal to noise ratio, and HINT sentences in quiet was initially evaluated at the beginning of the study with the SPEAK strategy and at the end of the two 6-wk periods with the ACE and CIS strategies. Then subjects' processors were programmed with all three strategies for use in everyday life. After 3-wk use, a final evaluation of speech recognition with the HINT sentences in quiet and CUNY sentences at +10 dB signal to noise ratio was performed with each strategy. Subjects also responded to a questionnaire giving their strategy preference for most listening situations, the percentage of time they used each strategy, and the strategy they found gave them the best hearing and understanding of speech in 19 listening situations.

RESULTS: Of the 62 subjects who participated, 56 subjects reported that they preferred one strategy for most listening situations (ACE strategy: 37 [59.7%]; SPEAK strategy: 14 [22.6%]; CIS strategy: 5 [8.0%]) and six subjects did not prefer a single strategy (9.7%). For the group who preferred one strategy, the preferred strategy resulted in higher scores than for one of the other strategies at the initial evaluation on CUNY sentences in quiet and noise and at the final evaluation on HINT sentences in quiet and CUNY sentences in noise for approximately two-thirds of the subjects. Strategy preference and performance were not significantly related for the remaining dependent measures. There also was strong agreement between the preferred strategy, percentage time this strategy was used, and the number of specific listening situations the preferred strategy was chosen for best hearing and understanding of speech. Although the majority of subjects strongly preferred a single strategy, some preferred to use two or three strategies, and a few were not sure which strategy they preferred for the majority of listening situations. Of the 19 subjects who reported that it was useful to use different strategies for different listening situations, only 5 of the 13 subjects, who responded to a follow-up questionnaire sent 18 mo later, continued to use multiple strategies.

CONCLUSIONS: There was a significant relation between subjects' strategy preference based on experience in everyday life and their performance on the sentence tests, particularly sentences in noise. Important individual differences in strategy preference as well as in rate and number of channels stimulated per cycle within the ACE and CIS strategies emerged during the study. At the end of this process, over half of the subjects preferred the ACE strategy, and over double the number preferred the SPEAK strategy compared with the CIS strategy. To provide newly implanted recipients with as much benefit as possible, it is important that the speech processor program with each strategy be adjusted to maximize perceived benefit sequentially and then the three strategies need to be compared. With the four memories of the SPrint processor and a recipient who adapts quickly to hearing sound with different speech coding strategies, it may be possible to accomplish this comparison clinically through weekly fitting sessions plus listening in everyday life over a period of approximately 6 wk. At the end of this fitting process, most recipients probably will prefer to use one strategy, whereas some may prefer two or all three strategies to maximize their ability to hear in different listening situations.