RESULTS
Figure 1 shows the flow diagram from enrollment to randomization, intervention, follow-up, and analysis.
All the patients were recruited from the waiting list of our service and those who met the inclusion criteria (elderly with mild-to-moderate sensorineural and naïve to sound amplification) were invited to our outpatient clinic to receive their hearing aids. All the appointments (recruitment and follow-up) happened from a July 2015 to July 2016. The follow-up appointment was set in accordance with the protocol of our institution. Those who did not meet the 12 weeks of hearing aids use (after 3 months from the first appointment) were rescheduled to a next appointment as suggested in the current practice of hearing aid fitting of our institution. We stopped the trial because of the time we had to finish this study, which worked as a master’s degree of the main author.
The final sample comprised 22 individuals who had a mean age of 80.4 (± 6.1) years, were predominantly female (63.63%), and had a mean education of 3.8 (± 1.6) years. The participants randomized into the 3 groups (A=8, B=6, and C=8) were analyzed by original assigned groups and did not differ for age, gender, education, cognitive screening scores on the MoCA (except language), hours of hearing aid use (informed and actual recorded), pure-tone audiometry (sensorineural, symmetric, and moderate hearing loss bilaterally), initial results on the NEUPSILIN tasks and auditory evoked potentials as shown in Table 1 and Supplementary Figure 1, which present the baseline characteristics of the three groups.
After fitting the hearing aids, in situ checks (probe measurements) revealed differences between individuals using sound amplification products (Groups A and B) and those using placebo (Group C) (Supplementary Table 1) as expected.
The analysis of IOI-HA responses showed differences in benefit (p<0.001), satisfaction (p=0.007), participation restriction (p=0.012) and quality of life (p=0.037). In addition, mean total scores on that inventory also differed across groups (p=0.032) (Table 1).
As we conducted a pragmatic study, no subject was excluded, and we adopted an intention-to-treat approach in analysis.
The primary outcomes are given in Table 2, showing statistically significant differences in reverse counting (p=0.03), auditory word span (p=0.03) and recognition (p=0.03) on the NEUPSILIN tasks and reduction in N1 wave latency for the /t/ phoneme at a 55 dB intensity (p=0.01).
No potential harm was observed along with the follow-up of the patients. In the end of data collection, those who received the placebo intervention (group C) had their devices fitted according to the exact parameters we fitted the devices of both groups A and B. All the patients kept on ordinary follow-up at the outpatient clinic.
The numerical results are summarized in Supplementary Tables 1 and 2.
Comparison of the numerical variables before and after 12 weeks of hearing aid use shows differences in scores on the NEUPSILIN for reverse counting (p = 0.006, 95%CI 2.2;11.63), digit sequence repetition (p = 0.039, 95%CI -1.9;-0.05), delayed recall memory (p = 0.036, 95%CI -1.2;-0.05), recognition (p = 0.008, 95%CI -2.6;-0.45) and visual memory (p = 0.008, 95%CI -0.9;-0.15) and in latency of potentials for the N1 wave for the /g/ phoneme at a 55 dB intensity (p = 0.016, 95%CI 2.1;18.59) (Supplementary Table 3).
Variables which differed before and after the intervention were submitted to a post hoc analysis, revealing differences between the sound amplification groups and the placebo group on NEUPSILIN scores for reverse counting (p = 0.002, 95%CI 5.9;20.55) and recognition (p = 0.013, 95%CI -6.1;-0.88). This analysis found no differences among the groups for a reduction in N1 wave latency of the /g/ phoneme at 55 dB. Likewise, no difference between groups A and B was found (Supplementary Table 3).