Improved Postoperative Speech Recognition and Processor Use with Early Cochlear Implant Activation

Abstract

Objective:

To report speech recognition outcomes and processor use based on timing of cochlear implant (CI) activation.

Study Design:

Retrospective cohort.

Setting:

Tertiary referral center.

Patients:

604 adult CI recipients from October 2011 to March 2022, stratified by timing of CI activation (group 1: ≤10 days, n=47; group 2: >10 days, n=557).

Main Outcome Measures:

Average daily processor use; CNC and AzBio in quiet at 1-, 3-, 6-, and 12-month visits; time to peak performance.

Results:

The groups did not differ in sex (p=0.887), age at CI (p=0.109), preop CNC (p=0.070), or preop AzBio in quiet (p=0.113). Group 1 had higher median daily processor use than group 2 at the 1-month visit (12.3 vs. 10.7 hours/day, p=0.017), with no significant differences at 3, 6, and 12 months. The early activation group had superior median CNC performance at 3 months (56% vs. 46%, p=0.007) and 12 months (60% vs. 52%, p=0.044). Similarly, the early activation group had superior median AzBio in quiet performance at 3 months (72% vs. 59%, p=0.008) and 12 months (75% vs. 68%, p=0.049). Both groups were equivalent in time to peak performance for CNC and AzBio. Earlier CI activation was significantly correlated with higher average daily processor use at all follow-up intervals.

Conclusions:

CI activation within 10 days of surgery is associated with increased early device usage and superior speech recognition at both early and late follow-up visits. Timing of activation and device usage are modifiable factors that can help optimize postoperative outcomes in the CI population.

INTRODUCTION

Cochlear implants (CI) have become the standard treatment option for patients with moderate-to-profound hearing loss who derive limited benefit from their hearing aids. Several factors can influence outcomes after CI surgery: age at implantation, duration of deafness, preoperative hearing, cognitive status, and device utilization (1–10). With innovative healthcare delivery models such as same-day CI consultation and surgery, where sound processor activation is performed within a few days after implantation, understanding how the timing of activation impacts postoperative outcomes is imperative (11, 12).

A healing period of three to six weeks has been traditionally allotted before device activation in order to ensure appropriate wound healing and to prevent infections related to the skin flap (13–15). More recently, the use of a smaller postauricular incision has been suggested to result in faster healing and less soft tissue swelling (16, 17). Several studies have highlighted the safety of device activation within the first ten days after CI surgery, including as early as one day post-implantation (18–22). Moreover, similar electrode impedances have been found among individuals undergoing early activation, in comparison to standard activation, at early- and long-term follow-up (23, 24). Two papers, with small sample sizes of 40 and 127 patients, have reported the impact of early activation (less than 8 days in Gunther et al. and within 14 days in Bruschke et al.) on device usage and speech outcomes (22, 25). Both studies found that the early activation group had lower processor use at initial follow-up and equivalent speech recognition at all visits up to one year after surgery.

The present study aims to evaluate postoperative speech recognition and processor use based on the timing of device activation using the largest cohort to date. We hypothesized that early activation within 10 days after surgery would be associated with superior postoperative speech performance, with the primary driver of this relationship being higher early device usage. This study’s findings can help guide clinical practice and patient counseling and offer few modifiable factors (e.g. activation timing, device usage) for optimization of postoperative speech perception in the CI population.

METHODS

After Institutional Review Board approval (#221883), a retrospective chart review was performed to identify adult patients who underwent CI surgery between October 2011 and March 2022 at our institution. Exclusion criteria included age less than 18 years, revision surgery, lack of postoperative testing, and absence of daily device usage data (usually related to lack of processor feature). To assess the impact of activation timing, patients were stratified based on activation being up to 10 days (early activation group) versus greater than 10 days (standard activation group) after CI surgery. The 10-day cut-off was chosen based on patients typically being activated between 2 to 4 weeks at our institution. Other studies looking at early implant activation have similarly noted standard activation to occur between 10 and 28 days (13, 18, 23, 25, 26). The early and standard activation groups received equivalent counseling before and after surgery regarding expectations and aural rehabilitation, and all three manufacturers’ devices were used for both groups.

The primary endpoints were speech recognition measures at each postoperative visit (approximately at 1, 3, 6 and 12 months): i) aided Consonant-Nucleus-Consonant (CNC) words using a 50-item list and ii) AzBio in quiet using a 20-sentence list (27, 28). Testing was performed with recorded stimuli that originated from a single loudspeaker at zero-degree azimuth. Stimuli were presented in the sound field at 60 dB sound pressure level (SPL) in a quiet testing condition, and the contralateral ear was plugged or masked for testing in the CI-alone condition. In addition, average daily device usage was extracted from data logs in the manufacturer software. Demographic variables; dates of CI surgery, device activation, and each postoperative visit; and preoperative CNC and AzBio scores were also recorded from each patient’s medical record. Finally, time to peak performance in months was calculated by subtracting dates of the highest CNC and AzBio scores from date of surgery for each patient.

Medians and interquartile ranges (IQR) were used to summarize continuous variables that were not normally distributed. Categorical variables were described with frequency counts and percentages. The Mann-Whitney U test and chi-square tests were used to determine correlations between continuous and categorical variables, respectively. Spearman correlations were conducted to evaluate relationships between device usage, timing of activation, age at implantation, and preoperative and postoperative speech recognition measures. Furthermore, mediation analysis was performed using PROCESS (v4.2) to evaluate the mediator effect of early device usage on the association between timing of activation and 12-month speech recognition scores (29). All analyses were completed using SPSS software, version 26.0 (IBM Corp., Armonk, NY), and p values less than 0.05 were considered statistically significant.

RESULTS

Table 1 shows baseline characteristics for 604 ears (47 with early activation, 557 with standard activation) from 579 patients who met inclusion criteria. Median timing of activation was 4 days (IQR 0—7) and 19 days (IQR 16—23) for the early and standard activation groups, respectively (p<0.001). Age at implantation, sex, preoperative CNC and AzBio scores, laterality of CI, follow-up at 12 months postop, and bilateral CI status did not significantly differ between the groups (p>0.05).

Table 1.

Baseline characteristics.

	Early Activation (n=47)	Standard Activation (n=557)	p
Age at CI surgery (years)	66.0 (46.2—75.0)	68.7 (57.0—77.0)	0.109
Male sex	27 (57%)	314 (56%)	0.887
Preoperative CNC words (%)	20.0 (0.0—36.0)	15.0 (0.0—32.0)	0.070
Preoperative AzBio in quiet (%)	23.0 (0.8—60.3)	14.0 (1.0—35.3)	0.113
Laterality			0.964
Left	21 (45%)	247 (44%)
Right	26 (55%)	310 (56%)
Timing of CI activation (days)	4.0 (0.0—7.0)	19.0 (16.0—23.0)	<0.001
Available 12-month testing	30 (64%)	426 (76%)	0.075

Data are presented as median (interquartile range) or incidence (frequency).

AzBio indicates Arizona Biomedical; CI, cochlear implant; CNC, Consonant-Nucleus-Consonant.

Significant p values are in bold.

Regarding postoperative speech recognition performance (Table 2), the early activation group had significantly higher median CNC scores in the CI alone condition at 3 (56% vs. 46%, p=0.007) and 12 (60% vs. 52%, p=0.044) months. Similarly, patients activated within 10 days after surgery had superior median AzBio in quiet performance at 3 (72% vs. 59%, p=0.008) and 12 (75% vs. 68%, p=0.049) months. The early activation group had a higher median daily processor use than the standard activation group at the 1-month visit (12.3 vs. 10.7 hours/day, p=0.017), with no significant differences in average daily device use at 3, 6, and 12 months. Both groups were equivalent in terms of time to peak performance for CNC and AzBio in quiet within the first year of surgery (p>0.05).

Table 2.

Postoperative speech recognition outcomes and device usage, stratified by timing of activation.

	Early Activation	Standard Activation	p
CI alone CNC words (%)
1-mo	37.0 (20.0—53.0)	30.0 (14.0—48.0)	0.120
3-mo	56.0 (34.0—72.0)	46.0 (26.0—60.0)	0.007
6-mo	51.0 (40.5—70.0)	48.0 (28.0—64.0)	0.166
12-mo	60.0 (47.0—70.0)	52.0 (36.0—68.0)	0.044
CI alone AzBio in Quiet (%)
1-mo	26.0 (24.0—82.0)	47.0 (17.5—68.0)	0.970
3-mo	72.0 (50.0—91.0)	59.0 (35.0—80.0)	0.008
6-mo	74.5 (45.0—83.0)	64.0 (36.0—82.0)	0.156
12-mo	74.5 (59.8—88.8)	68.0 (45.0—85.0)	0.049
Daily device usage (hours/day)
1-mo	12.3 (9.5—13.5)	10.7 (8.0—12.6)	0.017
3-mo	12.8 (10.8—13.8)	11.5 (8.1—13.6)	0.061
6-mo	12.3 (10.3—14.0)	11.5 (8.1—13.6)	0.270
12-mo	12.7 (10.2—14.5)	11.9 (8.0—13.8)	0.229
Time to peak CNC (months)	6.0 (4.0—12.0)	7.0 (4.0—12.0)	0.857
Time to peak AzBio in quiet (months)	7.0 (4.0—12.0)	7.0 (4.0—12.0)	0.426

Data are presented as median (interquartile range).

AzBio indicates Arizona Biomedical; CI, cochlear implant; CNC, Consonant-Nucleus-Consonant.

Significant p values are in bold.

On Spearman correlational analyses, earlier CI activation was significantly correlated with higher average daily processor use at 1, 3, 6, and 12 months (Table 3). In addition, higher 12-month CNC scores were significantly associated with younger age at CI surgery (Spearman ρ value of −0.23, p<0.001) and higher preoperative CNC scores (Spearman ρ value of 0.16, p=0.001). Better 12-month AzBio in quiet performance was also associated with younger age at CI surgery (Spearman ρ value of −0.29, p<0.001) and higher preoperative AzBio scores (Spearman ρ value of 0.20, p<0.001).

Table 3.

Spearman product component correlations and associated p values for timing of activation and average daily CI use.

	Postoperative Measure	Correlation	p
Timing of activation	Average daily CI use at 1 month	−0.11	0.044
	Average daily CI use at 3 months	−0.13	0.020
	Average daily CI use at 6 months	−0.16	0.005
	Average daily CI use at 12 months	−0.13	0.022

AzBio indicates Arizona Biomedical; CI, cochlear implant; CNC, Consonant-Nucleus-Consonant.

Significant p values are in bold.

Mediation analysis was then used to better understand how timing of activation and 1-month daily device usage interacted with one another in their associations with 12-month CNC and AzBio in quiet scores (Figure 1). Age at CI surgery and preoperative speech recognition scores were included as covariates. The total and direct effects of timing of activation on 12-month CNC and AzBio scores were not statistically significant after controlling for age and preoperative speech scores (p>0.05). However, the indirect effect of activation timing mediated by early 1-month device usage after surgery was significant (CNC: β = −0.10, 95% confidence interval: −0.19 to −0.02; AzBio: β = −0.09, 95% confidence interval: −0.17 to −0.04). Therefore, in this cohort, increased early device usage was a significant mediator between early activation and better 12-month speech recognition outcomes.

Figure 1.

Mediation analysis of timing of activation, 1-month daily cochlear implant use, and 12-month CNC (A) and AzBio in quiet (B) scores. Age at surgery and preoperative speech recognition score were included as covariates. Data are presented as beta coefficients with 95% confidence intervals, and bolded values are significant. Grey text and dotted lines depict indirect effect while black text and lines depict direct and total effects. AzBio indicates Arizona Biomedical; CNC, Consonant-Nucleus-Consonant.

DISCUSSION

This study of 604 ears represents the largest cohort to date for evaluation of postoperative speech performance and device usage based on timing of CI activation. Two previous studies, with 127 and 40 patients, have reported lower daily processor use at initial follow-up for the early activation group and equivalent speech outcomes at all visits (22, 25). In contrast, we found that patients undergoing device activation within 10 days after CI surgery had significantly higher daily processor use at the initial 1-month visit and higher CNC and AzBio scores at 3 and 12 months. These contradictory findings may be attributed to a considerably larger number of patients in the present study, which allowed for rigorous analyses that could detect significant results. Moreover, one study evaluated daily device usage at initial follow-up, which was 3 days after activation, and then 3, 6, and 12 months (25). The early group was thought to wear their device less in both studies since their postauricular wound was still fresh and covered in dressing from surgery. By looking at 1-month outcomes, this study provided an additional early time period to assess differences in device usage based on timing of activation.

Several factors were found to correlate significantly with 12-month speech recognition outcomes. Consistent with other studies, younger age at CI surgery was associated with better postoperative speech recognition (2, 3, 30). This relationship has been shown to persist even after matched audiometric hearing, suggesting that declines in speech perception in older individuals is at least partly caused by cognitive and perceptual changes that occur with age (31, 32). Moreover, higher preoperative speech recognition scores significantly correlated with better 12-month CNC and AzBio scores in this cohort. Prior studies parallel this finding, where better preoperative AzBio scores and pure-tone averages have been linked with superior postoperative speech measures (2, 3, 30, 33).

Independent of age at CI surgery and preoperative CNC and AzBio scores, early activation was found to have significant indirect effects on 12-month speech recognition scores, mediated by 1-month daily processor usage (Figure 1). A direct relationship between average daily CI usage and speech outcomes has been well-established (8–10). This study further highlights that early processor use influences long-term speech recognition outcomes. While timing of activation did not independently predict 12-month CNC and AzBio scores on mediation analysis, earlier CI activation was found to significantly correlate with higher average daily processor use at all follow-up intervals (Table 3). Altogether, these results support our initial hypothesis that early activation is associated with higher early device usage which, in turn, has a large impact on postoperative speech outcomes. Unlike age at implantation and preoperative speech recognition thresholds, earlier timing of activation and consistent processor use are modifiable factors that can help optimize postoperative outcomes in the CI population.

In addition to benefits in device usage and speech recognition, financial advantages can exist from earlier CI activation. With a shift towards patient-centered care, healthcare delivery models, such as same-day CI consultation and surgery, have been designed to bundle care and overcome barriers to implantation including travel burden (11, 12). Indirect costs to implantation increase considerably as patients living far from their CI center often stay in the area or travel back-and-forth for their activation appointment. As such, Hajr et al. have suggested that earlier CI activation can alleviate some of this financial burden (34). In this study, the median timing of activation after surgery was notably decreased from 19 days for the standard group to 4 days for the early group.

Psychological benefits may also be an important consideration for patients and families based on the timing of activation. The question of whether a CI will provide significant communication benefit is answered sooner with earlier activation. Previous studies have shown that postoperative switch-on of the device within 24 hours results in a shortened period of uncertainty and worry for patients and their caregivers (24, 35, 36). Furthermore, earlier activation can help patients return to their normal activity and life as soon as possible, allowing them to adapt sooner with their CI (13, 37). Extremely favorable rates of patient satisfaction have also been found, with 84% and 100% of patients reporting “highly satisfied” or “satisfied” with early CI activation at initial follow-up and 3 months, respectively (22).

Several limitations of this study should be noted. In addition to those inherent to a retrospective chart review, this study’s conclusions are limited to one year of follow-up after CI surgery. Other variables, such as cognitive status, neural health, and underlying patient motivation, which can impact postoperative speech outcomes and device usage were not assessed. Medical comorbidities and postoperative complications (e.g. pain, wound healing issues, balance) were not collected and could potentially influence feasibility of early activation for certain patients. However, several studies have already demonstrated the safety of early activation (18–21, 25, 35). Future studies evaluating timing of CI activation can involve randomization into early and standard activation groups; more comprehensive variables including markers for cognitive status and neural health, comorbidities, and postoperative complications; follow-up beyond one year; and patient satisfaction questionnaires. With this paper’s findings serving as a catalyst, future studies can ultimately help derive a recommendation for when patients should ideally be activated after implantation.

CONCLUSION

CI activation within 10 days of surgery is associated with increased early device usage and superior speech recognition at both early and late follow-up visits. Independent of age at implantation and preoperative hearing status, higher early device usage is significantly associated with better long-term postoperative speech recognition performance. Both timing of activation and device usage are modifiable factors that can help optimize postoperative outcomes in the CI population. As such, motivated patients with routine healing may benefit from earlier device activation.