loading  Checking for direct PDF access through Ovid


To the Editor: We read the paper by Jayakody et al. (1) from the current issue of Otology & Neurotology with great interest.
The paper is dealing with the question, whether the cognitive performance in older adults improves after cochlear implantation, which is a topic of eminent importance. However, we have some concerns and questions regarding this paper.
To begin with a major concern, there are many dropouts. In this study, cochlear implant (CI) recipients were compared with CI candidates. There were baseline assessments in the case of CI recipients before CI and follow up assessments 6 and 12 months after CI. According to the data provided in Table 2 the dropout rate was at least 52% in the group of CI candidates (sample size at baseline: n = ?, after 6 months: n = 23, after 12 months: n = 11) and at least 31% in the group of CI recipients (sample size at baseline: n = ?, after 6 months: n = 16, 12 months: n = 11). The authors did not inform comprehensively about the reasons for these dropouts: “The decrease in numbers was partially due to CI candidates receiving an implant after 6 months and therefore had to be excluded from further analysis” (p. e290, left column, last line). How many CI candidates were excluded because of cochlear implantation. How many CI candidates were excluded because of other reasons? Why did five CI recipients dropout? As a consequence of these dropouts the demographic data of both groups changed considerably. For example, at the time of the follow up after 6 months the mean age of the CI recipients was 61.75 ± 15.62 years, the mean age of the CI candidates was 69.04 ± 12.35 years. At the time of the follow up after 12 months the mean age of the CI recipients was 67.27 ± 15.94 years, the mean age of the CI candidates was 67.27 ± 15.94 years. Additionally, we did not find any information about the age of the study participants at baseline.
These unexpected and distinctive changes of group-sizes and characteristics may bias the results. Furthermore, the question arises if a pre- and post-assessment is still justified.
Other important concerns relate to the statistics of the study. The authors reported that t tests were calculated to analyze group differences, which can only be used in normally distributed data, which was either not tested or not reported. Given the small sample size, nonparametric statistical testing is recommended. Furthermore, there was no controlling for social background and education which is crucial in respect of cognition (2). Another important issue is that multiple hypothesis testing was not considered, which could change the levels of significance dramatically.
The authors report a significant decline in stress scores, but taking into account the obviously high standard deviations and the range of the confidence intervals these results should be interpreted with caution. The results could alternatively be explained by changes of group-size or effect of sex, education, or two cases which “have gone through life changing events” (see p. e292, left column).
Furthermore, we think that the title of the paper is not justified (“Impact of Cochlear Implantation on Cognitive Functions.”). The authors used t tests to compare the differences between baseline and 6 months, and between baseline and 12 months. T tests do not allow statements regarding cause effect relations.
Minor concerns relate to the sample characteristics. Inclusion and exclusion criteria were not defined in this study. We also miss information regarding comorbidity and pharmacological treatment, which has known effects on cognition (3). Furthermore, cognitive tests were performed three times within a time span of 12 months. Under these circumstances a certain learning effect cannot be ruled out, which could also bias the results.

Related Topics

    loading  Loading Related Articles