ABSTRACT
Objectives
To understand the sociodemographic, audiologic, and performance variables that influence participant attrition in long‐term, repeated‐measures clinical research.
Methods
A retrospective review was completed for adult cochlear implant (CI) recipients who enrolled in an ongoing, repeated‐measures clinical research study investigating outcomes between two mapping procedures. The sample was stratified into those who completed the 12‐month study and those who withdrew prior to the study endpoint. The groups were compared for sociodemographic variables (age at consent, race, ethnicity, biological sex, drive time to the clinic, socioeconomic position, and rurality), cognition, and audiologic variables (device type, ipsilateral and contralateral hearing, device wear‐time, and tonotopic mismatch). Early performance on CNC word recognition was analyzed between groups.
Results
Of the 118 research participants, the attrition rate for the sample was 21%. Sociodemographic variables were at parity between the completed and withdrawal groups. Of the reviewed audiologic variables, the withdrawal group had poorer low‐frequency acoustic hearing thresholds, was less likely to be electric‐acoustic stimulation users than CI alone users (likely due to unaided thresholds), and had greater tonotopic mismatch. Early CNC word recognition was significantly different between the two groups, with the completed group having better early performance than the withdrawal group.
Conclusion
The present findings serve as a reminder of the need for reflection on clinical research study attrition patterns and a call for intentionally designed study inclusion criteria and procedures to support enrollment and retention of a representative sample.
Level of Evidence
3.
Keywords: attrition, clinical research, frequency‐to‐place mismatch, healthcare disparity, hearing healthcare, tonotopic mismatch
In a retrospective review of attrition patterns for participants in a clinical research study, sociodemographic variables were not significantly different between those who elected to withdraw and those who completed the study. Those in the withdrawal group had poorer low‐frequency residual hearing, were less likely to be electric‐acoustic stimulation users, had more frequency‐to‐place mismatch, and had poorer early word recognition scores.
1. Introduction
Equitable representation in clinical research participant samples is essential to ensure the appropriate application and generalization of research findings to the associated patient population. Recently, Nix et al. [1] determined that a study design that focused on participant convenience supported recruitment and enrollment of a sample that was at parity with the patient population at a cochlear implant (CI) center. For clinical research using a repeated‐measures study design, it is also important to understand any barriers to continued research participation, minimize participant attrition, and ensure that equitable representation of participants who enroll in research studies is maintained throughout the study duration.
Research, by design, has an ethical obligation for participant autonomy and voluntary withdrawal [2, 3]. Participants may choose to withdraw for different reasons, including reasons related to the clinical research (such as duration of the study), poor understanding of study purpose and procedures, or concern about adverse events [4, 5]. Individual factors, such as the impact of participation on time, energy, and financial resources, may also contribute to an inability or unwillingness to continue participation in clinical research [6]. Facilitators of participant retention typically involve consistent communication about study procedures and visits, incentives for participation, and transparency of study objectives [7].
Beyond age and biological sex, limited data have been reported on the demographic profiles of participants in CI clinical research [8]. Data on attrition in research studies, particularly in studies investigating specialized medical devices such as CIs, is even more scarce [9]. Improving the understanding of attrition within clinical research may help ensure the internal validity of individual study findings, and more broadly, may contribute toward more equitable outcomes of the studied device [4]. Investigation of withdrawal patterns may also illuminate potential areas for intervention, outreach, and strategic support.
The present study compared sociodemographic, audiologic, and performance‐related variables between two groups of CI clinical research participants: those who completed the 12‐month repeated‐measures study and those who withdrew prior to the study endpoint. The primary aim was to understand what variables may influence participant attrition in long‐term, repeated‐measures clinical research.
2. Materials and Methods
We completed a retrospective review of adult CI recipients who enrolled in an ongoing clinical research study of outcomes for two different mapping procedures (see Dillon et al. [10]). The University of North Carolina Institutional Review Board approved all study procedures (IRB #19‐0136). For that study, criteria for inclusion are: native English speaker (as speech recognition materials are presented in English), between 18 and 80 years of age at cochlear implantation, passed the Mini Mental State Examination [11], received a MED‐EL straight electrode array, and willing to participate in the study procedures. The inclusion criteria were intentionally broad to recruit a sample that is representative of the clinical CI patient population. Potential candidates were excluded if they had a history of vestibular schwannoma, neurofibromatosis type 2, inner ear malformation, meningitis, or revision surgery. Participants include CI and electric‐acoustic stimulation (EAS) users who are randomized to one of the two mapping procedures and are assessed at device activation and 1‐, 3‐, 6‐, and 12‐months post‐activation. These study intervals were selected since they were the same as the recommended clinical follow‐up visits for adult CI recipients at the CI center during the study protocol development. Participants are blinded to their assigned grouping for the duration of the study. For the present study, data were queried for participants who were enrolled between 2019 (study initiation) and December 2023. This time period was selected so that all participants would have reached the study endpoint (12‐months post‐activation) at the time of data review. Data were excluded for participants who were disqualified from the referenced study. Disqualification occurred in instances of mapping adjustments needed to address rare cases, such as facial nerve stimulation. The present study excluded these participants since the decision to discontinue participation was beyond their control.
The queried sample was stratified into two groups: participants who completed the study (completed group) and participants who elected to withdraw prior to the study endpoint (withdrawal group). The groups were compared for the following sociodemographic variables: biological sex, race, ethnicity, age at consent, drive time to the CI center, socioeconomic position, and rurality of the participant's home environment. Drive time to the CI center was approximated using ArcGIS Nearest Facility Tool, using previously described procedures [1]. Socioeconomic position was characterized by the Social Deprivation Index (SDI), which is a grade of socioeconomic deprivation from seven population‐level statistics [Seven population‐level statistics: (1) Percent population less than 100% Federal Poverty Level (FPL), (2) percent population 25 years of age or more with less than 12 years of education, (3) percent non‐employed for population age 16–64 years, (4) percent households living in renter‐occupied housing units, (5) percent households living in crowded housing units, (6) percent single parent families with dependents less than 18 years of age, (7) percent households with no vehicle] based on the individual's Zip Code Tabulation Areas (ZCTA) code [12]. The SDI scores range from 0 to 100, with higher scores indicating higher levels of social deprivation (lower socioeconomic position). The degree of rurality was quantified by the United States Department of Agriculture's Economic Research Services rural–urban commuting area (RUCA) codes, also at ZCTA level [13]. The code classification spans from 1 to 10, with a broader subcategorization of Metropolitan (RUCA code 1–3), Micropolitan (RUCA code 4–6), Small Town (RUCA code 7–9), and Total Rural (RUCA code 10).
The groups were also compared for the following individual, audiologic, and device variables: cognition (as measured with the Hearing‐Impaired version of the Montreal Cognitive Assessment (HI‐MoCA [14])), ipsilateral low‐frequency pure tone average (LFPTA: 125, 250, and 500 Hz), contralateral pure tone average (PTA: 0.5, 1, 2, and 4 kHz), contralateral status (i.e., normal hearing (NH), hearing aid (HA), CI, EAS, or unaided), participant‐reported duration of hearing loss, participant‐reported history of HA use, device type (i.e., EAS or CI), device wear‐time based on datalogging in the Maestro programming software (when available), randomized mapping procedure (i.e., default or place‐based), and frequency‐to‐place (tonotopic) mismatch. Tonotopic mismatch was calculated in semitone deviation between the channel's center frequency of the most apical electrode contact and the cochlear place frequency determined using the spiral ganglion (SG) frequency‐to‐place function [15]. Additionally, CNC word recognition [16] was compared between the groups with the prediction that poorer performers may be more likely to discontinue study participation.
2.1. Data Analysis
Mann–Whitney U tests were used for nonparametric comparisons between the completed group to the withdrawal group for the continuous variables (i.e., age, SDI, drive time, PTA, LFPTA, reported duration of hearing loss, tonotopic mismatch). Chi‐Squared tests were used to compare the categorical data (i.e., race, ethnicity, biological sex, RUCA categorization, randomized mapping procedure, device type, reported HA use). Welch's two‐sample t‐tests were used to compare CNC scores between groups at the preoperative and 1‐month post‐activation intervals. Analyses were conducted with the Stata software (version 17.0). Significance was defined as p < 0.05.
3. Results
At the time of data review, 118 participants met the criteria for inclusion for the present analyses. Of those, 93 (79%) were in the completed group and 25 (21%) were in the withdrawal group. Figure 1 displays the number of participants who withdrew prior to each study interval; the majority (n = 17, 68%) discontinued study participation prior to the 3‐month interval. In most instances (n = 13; 52%), a reason for withdrawal was not specified. However, some individuals (n = 10, 40%) reported competing demands on their time or health as factors limiting their ability to continue participation. Two individuals (8%) elected to withdraw due to relocation during the study.
FIGURE 1.
Time points of study discontinuation for the withdrawal group (n = 25). [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com]
3.1. Sociodemographic Variables
Sociodemographic variables for each group are listed in Table 1. The groups did not differ significantly in age at the time of study consent, z = [−0.603], p = [0.547], biological sex, X 2 (1, N = 118) = 0.424, p = 0.515, SDI, z = [−1.361], p = [0.174], racial distribution, X 2 (4, N = 118) = 1.988, p = 0.738, or ethnic distribution, X 2 (2, N = 118) = 4.01, p = 0.135. Figure 2 plots the geographic distribution of the completed group (blue) and the withdrawal group (yellow). There were no significant differences in drive time between the two groups, z = [−0.800], p = [0.424], or RUCA code category, X 2 (3, N = 118) = 0.915, p = 0.822.
TABLE 1.
Sociodemographic variables for the completed group and withdrawal group.
| Completed group (n = 93) | Withdrawal group (n = 25) | Statistical findings | |
|---|---|---|---|
| Age at consent, years, mean | 61 (SD: 12) | 61 (SD: 15) | z = [−0.603], p = [0.547] |
| Age at consent, years, quintiles | Min: 18 | Min: 26 | |
| 25th: 53 | 25th: 50 | ||
| Median: 63 | Median: 67 | ||
| 75th: 69 | 75th: 75 | ||
| Max: 80 | Max: 79 | ||
| Biological sex, % female | 53% | 60% | X 2 (1, N = 118) = 0.424, p = 0.515 |
| Biological sex, % male | 47% | 40% | |
| Race | AI: 2% | B: 4% | X 2 (4, N = 118) = 1.988, p = 0.738 |
| AS: 2% | W: 96% | ||
| B: 7% | |||
| O: 2% | |||
| W: 87% | |||
| Ethnicity, % NH | 99% | 96% | X 2 (2, N = 118) = 4.01, p = 0.135 |
| SDI, mean | 43 (SD: 28) | 52 (SD: 29) | z = [−1.361], p = [0.174] |
| SDI, quintiles | Min: 5 | Min: 4 | |
| 25th: 20 | 25th: 23 | ||
| Median: 37 | Median: 57 | ||
| 75th: 69 | 75th: 74 | ||
| Max: 93 | Max: 97 | ||
| Drive time to CI center, minutes, mean | 85.39 (SD: 60.41) | 94.33 (SD: 54.17) | z = [−0.800], p = [0.424] |
| Drive time to CI center, minutes, quintiles | Min: 9.58 | Min: 10.16 | |
| 25th: 29.41 | 25th: 64 | ||
| Median: 68.27 | Median: 76.21 | ||
| 75th: 144.87 | 75th: 143.27 | ||
| Max: 242.73 | Max: 185.39 | ||
| RUCA, % from metro | 78.49% | 72% | X 2 (3, N = 118) = 0.915, p = 0.822 |
Abbreviations: AI: American Indian or Alaskan Native, AS: Asian American, B: Black or African American, O: Other, W: White or Caucasian, NH: Non‐Hispanic or Latino, SDI: Social Deprivation Index, RUCA: Rural Urban Commuting Area, Metro: Metropolitan.
FIGURE 2.
Geographic distribution of the completed group (blue) and the withdrawal group (yellow). Shading of individual points represents participant denisty from that region, with lighter shades indicating higher density. Drive time to the cochlear implant (CI) center is indicated by purple shading, with darker shades indicating higher drive times. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com]
3.2. Individual, Audiologic, and Device Variables
Individual, audiologic, and device variables for each group are listed in Table 2. There was no significant difference in HI‐MoCA scores between the completed and withdrawal groups, z = [−0.965], p = [0.334]. Also, there was no significant difference in contralateral PTA, z = [−0.851], p = [0.395], contralateral status, X 2 (5, N = 118) = 4.956, p = 0.423, device wear‐time, z = [0.842], p = [0.405], reported duration of hearing loss, z = [−0.041], p = [0.962], or reported HA use, X 2 (1, N = 118) = 0.875, p = 0.349, between the groups.
TABLE 2.
Individual, audiologic, and device variables for the completed group and withdrawal group. Variables that were significantly different between groups are in bold.
| Completed group (n = 93) | Withdrawal group (n = 25) | Statistical findings | |
|---|---|---|---|
| % investigative group | 47% | 28% | X 2 (2, N = 118) = 6.246, p = 0.044 |
| % control group | 53% | 68% | |
| 4% never assigned | |||
| Absolute tonotopic mismatch for default listeners, semitones, mean | 8.26 (SD: 5.9) | 14.18 (SD: 9.3) | z = [−2.513], p = [0.011] |
| Reported duration of hearing loss, years, mean | 19.98 (SD: 14.45) | 21.71 (SD: 18.60) | z = [−0.041], p = [0.962] |
| Reported history of HA use | Yes: 74% | Yes: 80% | X 2 (1, N = 118) = 0.875, p = 0.349 |
| No: 26% | No: 20% | ||
| Device type | EAS: 44% | EAS: 20% | X 2 (1, N = 118) = 7.973, p = 0.019 |
| CI: 56% | CI: 80% | ||
| Device wear time, hours, mean | 10.15 (SD: 4.09) | 9.64 (SD: 4.32) | z = [0.842], p = [0.405] |
| Ipsilateral LFPTA, dB HL, mean | 60 (SD: 26) | 82 (SD: 29) | z = [−2.813], p = [0.004] |
| Ipsilateral LFPTA, dB HL, quintiles | Min: 15 | Min: 23 | |
| 25th: 40 | 25th: 58 | ||
| Median: 60 | Median: 75 | ||
| 75th: 80 | 75th: 113 | ||
| Max: 120 | Max: 120 | ||
| Contralateral PTA, dB HL, mean | 57 (SD: 33) | 63 (SD: 30) | z = [−0.851], p = [0.395] |
| Contralateral PTA, dB HL, quintiles | Min: 6 | Min: 8 | |
| 25th: 24 | 25th: 39 | ||
| Median: 61 | Median: 65 | ||
| 75th: 83 | 75th: 85 | ||
| Max: 120 | Max: 115 | ||
| Contralateral device type | CI: 12% | CI: 4% | X 2 (5, N = 118) = 4.956, p = 0.423 |
| EAS: 2% | EAS: 4% | ||
| HA: 51% | HA: 56% | ||
| PSAP: 1% | |||
| UHL: 23% | UHL: 12% | ||
| Unaided a : 12% | Unaided a : 24% | ||
| HI‐MoCA score, mean | 25 (SD: 3) | 26 (SD: 2) | z = [−0.965], p = [0.334] |
Abbreviations: SD: standard deviation, dB HL: decibels (hearing level), CI: cochlear implant, EAS: electric‐acoustic stimulation, LFPTA: low‐frequency pure tone average, PTA: pure tone average, HA: hearing aid, PSAP: personal sound amplification product, UHL: unilateral hearing loss, HI‐MoCA: Hearing Impaired version of the Montreal Cognitive Assessment.
“Unaided” indicates individuals with diagnosed hearing loss who are not utilizing any amplification despite clinical recommendation.
The groups differed significantly for device type (i.e., EAS or CI), X 2 (1, N = 118) = 7.973, p = 0.019, with the withdrawal group having more CI users (80%) than EAS users, while the completed group was relatively evenly split between CI users (56%) and EAS users (45%). Further, the groups differed significantly for LFPTA in the implanted ear, z = [−2.813], p = [0.004]. The completed group had better low‐frequency hearing than the withdrawal group, with a median LFPTA of 60 dB HL and 75 dB HL, respectively. Also, there was a significant difference in the randomized mapping procedure between the groups, with a higher proportion of individuals in the withdrawal group listening with a default map, X 2 (2, N = 118) = 6.246, p = 0.044. For participants listening with default maps, the magnitude of absolute tonotopic mismatch was significantly greater in the withdrawal group (median: 12 semitones) as compared to the completed group (median: 8 semitones), z = [−2.513], p = [0.011].
3.3. Performance Variables
Figure 3 plots the individual and group‐level CNC scores across intervals. Symbol‐coding, as specified in the legend, indicates the last completed interval for each participant in the withdrawal group. As observed in Figure 1, most (68%) individuals in the withdrawal group discontinued study participation prior to the 3‐month interval. Considering the magnitude of difference between sample sizes at the later post‐activation intervals, CNC scores were compared between the groups at the pre‐operative and 1‐month intervals only. Pre‐operatively, there was a significant difference in CNC scores between groups, t(44.90) = 2.59, p = 0.013, with better pre‐operative scores for the completed group (M = 9%, SD = 12%) than the withdrawal group (M = 4%, SD = 8%). At 1‐month post‐activation, there was also a significant difference in CNC scores between the groups, t (22.10) = 2.26, p = 0.034, with better scores for the completed group (M = 31%, SD = 21%) than the withdrawal group (M = 21%, SD = 15%).
FIGURE 3.
CNC word recognition across study intervals for the completed group (n = 93; open box plots) and withdrawal group (filled box plots). Withdrawal points for the withdrawal group are coded by symbol shape, as defined in the figure legend. Comparisons were completed for the pre‐operative and 1‐month intervals only due to smaller sample sizes with available data for the withdrawal group at later intervals (i.e., n = 6 at 3 months and n = 4 at 6 months post‐activation).
4. Discussion
Recruitment and maintenance of a diverse sample within any clinical research study is paramount in ensuring appropriate generalization and application of study findings. Thusly, analyzing patterns related to participant attrition may promote intervention to maintain a representative sample in active studies, as well as improve the design of future iterations of studies in development.
For the present sample, sociodemographic factors—including race, ethnicity, drive time to the CI center, socioeconomic position, and degree of urbanization—were not significant determinant factors in early withdrawal from study participation. This finding is promising, as it maintains the notion that intentional study design and recruitment strategies (e.g., study interval alignment with clinical recommendations, research center location, broad inclusion criteria) help obtain [1] and maintain a sample that is representative of the larger population. However, it is critical to note that motivators to either continue or discontinue participation in clinical research are likely highly individual.
Notably, the withdrawal and completed groups were significantly different for device type, ipsilateral LFPTA, and magnitude of tonotopic mismatch on the most apical electrode contact. The withdrawal group had poorer low‐frequency acoustic hearing thresholds, was less likely to be EAS users, and had greater tonotopic mismatch. The LFPTA determined the type of device used; that is, participants in the withdrawal group with functional acoustic low‐frequency hearing were fit with an EAS device. It is unclear how poorer LFPTA and/or being a CI user would influence willingness to continue participation in clinical research. Potentially, EAS users, and thus those with better LFPTA, are more frequently counseled on the need for follow‐up assessment and mapping of the device. For tonotopic mismatch, poorer speech recognition with tonotopic mismatch has been demonstrated in studies using CI or EAS simulations and with actual CI and EAS users [10, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]. An interpretation of these data is that the motivation to continue study participation and clinical follow‐up may be associated with poorer early outcomes due to tonotopic mismatch. This hypothesis is also echoed in the early speech recognition data, with the withdrawal group having significantly poorer CNC scores at the 1‐month post‐activation interval. Our ability to understand the impact of performance on attrition in this sample is limited, since most participants who withdrew from the study did so before an interval when asymptotic performance is expected (e.g., 6 months post‐activation [27, 28]). Taken together, these data demonstrate the importance of counseling on post‐activation procedures and realistic expectations, and the need to better understand the variables that influence early performance.
The attrition rate for this sample at the time of analysis was 21%, which is in line with reported attrition rates for other general areas of clinical research [29]. However, ideal attrition rates for clinical research are reportedly below 5% [30], albeit across a variety of subspecialties. As investigation into withdrawal patterns continues, there is a constant search for interventions to ease any non‐intentional burdens of research participation. Given the magnitude of the distribution of patients across the state (visualized in Figure 2), it is anticipated that there is a wide range of individualized needs to consider in order to ensure continued retention of a representative, heterogeneous sample. It is also likely that the attrition rate may vary based on demand and duration of participation in individual research studies, which should also be considered during study design.
There are limitations of this analysis that are worth consideration. First, some of the data available for this analysis were restricted to generalizations based on group‐level factors (i.e., RUCA, SDI, and drive‐time). While most data were collected at a granular ZCTA level, this information is still not inherently specific to the individual. Also, these data were reviewed retrospectively from a study that did not include a questionnaire of the reason(s) to withdraw. That is, if a participant reported a reason for electing to withdraw then it was documented; participants who did not provide a reason when they elected to withdraw were not pursued further to provide a rationale. It is possible that there are individual variables related to withdrawal that were not captured within the present analyses. Additionally, the sample size of the withdrawal group (n = 25) was much smaller than that of the completed group (n = 93). While non‐parametric statistical tests were used to analyze between‐group differences, studies with larger sample sizes are needed to corroborate the present findings.
5. Conclusions
Intentional study design and recruitment strategies may be effective in obtaining and maintaining a participant sample that is representative of the clinical population. These present data serve as a reminder of the need for reflection on clinical research study attrition patterns throughout a study's duration, and a call for intentionally‐designed study inclusion criteria and procedures to ensure broader application of the study findings. Attrition in clinical research studies assessing CI patient performance may be associated with audiologic variables (e.g., device type, LFPTA, and degree of tonotopic mismatch).
Funding
This project was supported in part by a research grant from MED‐EL Corporation. Amanda D. Sloop, Margaret E. Richter, Samantha P. Scharf, Meredith A. Rooth, and Margaret T. Dillon are/were supported by a research grant from MED‐EL Corporation. Kevin D. Brown serves on advisory boards for Advanced Bionics and MED‐EL Corporation and is a consultant for Cochlear Corporation.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
The authors thank Michelle Bocanegra and Krystal Rivera for their assistance in coordinating participant recruitment, scheduling study visits, and assisting with data management throughout the study.
Data Availability Statement
Research data are not shared.
References
- 1. Nix E. P., Davis A. G., Brown K. D., et al., “Demographic Representation in Clinical Research Relative to a Cochlear Implant Patient Population,” Laryngoscope 134, no. 9 (2024): 4101–4110, 10.1002/lary.31467. [DOI] [PubMed] [Google Scholar]
- 2. “Withdrawal of Subjects From Research Guidance,” accessed August 4, 2025, https://www.hhs.gov/ohrp/regulations‐and‐policy/guidance/guidance‐on‐withdrawal‐of‐subject/index.html.
- 3. Gordon E. J. and Prohaska T. R., “The Ethics of Withdrawal From Study Participation,” Accountability in Research 13, no. 4 (2006): 285–309, 10.1080/08989620600848645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chaudhari N., Ravi R., Gogtay N. J., and Thatte U. M., “Recruitment and Retention of the Participants in Clinical Trials: Challenges and Solutions,” Perspectives in Clinical Research 11, no. 2 (2020): 64–69, 10.4103/picr.PICR_206_19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kim J., Kim M. G., and Lim K.‐M., “Participation in and Withdrawal From Cancer Clinical Trials: A Survey of Clinical Research Coordinators,” Asia‐Pacific Journal of Oncology Nursing 9, no. 4 (2022): 197–201, 10.1016/j.apjon.2021.12.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Ulrich C. M., Ratcliffe S. J., Zhou Q., et al., “Association of Perceived Benefit or Burden of Research Participation With Participants' Withdrawal From Cancer Clinical Trials,” JAMA Network Open 5, no. 11 (2022): e2244412, 10.1001/jamanetworkopen.2022.44412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Abshire M., Dinglas V. D., Cajita M. I. A., Eakin M. N., Needham D. M., and Himmelfarb C. D., “Participant Retention Practices in Longitudinal Clinical Research Studies With High Retention Rates,” BMC Medical Research Methodology 17, no. 1 (2017): 30, 10.1186/s12874-017-0310-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Meinhardt G., Sharrer C., Perez N., et al., “Reporting of Sociodemographic Data in Cochlear Implant Clinical Trials: A Systematic Review,” Otology & Neurotology 44, no. 2 (2023): 99–106, 10.1097/MAO.0000000000003766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Babic A., Tokalic R., Cunha A. S., et al., “Assessments of Attrition Bias in Cochrane Systematic Reviews Are Highly Inconsistent and Thus Hindering Trial Comparability,” BMC Medical Research Methodology 19, no. 1 (2019): 76, 10.1186/s12874-019-0717-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Dillon M. T., Canfarotta M. W., Buss E., et al., “Influence of Electric Frequency‐To‐Place Mismatches on the Early Speech Recognition Outcomes for Electric‐Acoustic Stimulation Users,” American Journal of Audiology 32, no. 1 (2023): 251–260, 10.1044/2022_AJA-21-00254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Folstein M. F., Folstein S. E., and McHugh P. R., ““Mini‐Mental State”. A Practical Method for Grading the Cognitive State of Patients for the Clinician,” Journal of Psychiatric Research 12, no. 3 (1975): 189–198, 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
- 12. Robert Graham Center , “Social Deprivation Index (SDI),” 2018, https://www.graham‐center.org/maps‐data‐tools/social‐deprivation‐index.html.
- 13. “USDA ERS—Rural‐Urban Commuting Area Codes,” accessed October 19, 2023, https://www.ers.usda.gov/data‐products/rural‐urban‐commuting‐area‐codes/.
- 14. Lin V. Y. W., Chung J., Callahan B. L., et al., “Development of Cognitive Screening Test for the Severely Hearing Impaired: Hearing‐Impaired MoCA,” Laryngoscope 127, no. 1 (2017): S4–S11, 10.1002/lary.26590. [DOI] [PubMed] [Google Scholar]
- 15. Stakhovskaya O., Sridhar D., Bonham B. H., and Leake P. A., “Frequency Map for the Human Cochlear Spiral Ganglion: Implications for Cochlear Implants,” Journal of the Association for Research in Otolaryngology 8, no. 2 (2007): 220–233, 10.1007/s10162-007-0076-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Peterson G. E. and Lehiste I., “Revised CNC Lists for Auditory Tests,” Journal of Speech and Hearing Disorders 27 (1962): 62–70, 10.1044/jshd.2701.62. [DOI] [PubMed] [Google Scholar]
- 17. Lee S. E., Brown K. D., Overton A. B., et al., “Association of Tonotopic Mismatch With the Speech Recognition of Cochlear Implant Users With Unilateral Hearing Loss,” Otology & Neurotology 46, no. 8 (2025): e302–e306, 10.1097/MAO.0000000000004543. [DOI] [PubMed] [Google Scholar]
- 18. Başkent D. and Shannon R. V., “Interactions Between Cochlear Implant Electrode Insertion Depth and Frequency‐Place Mapping,” Journal of the Acoustical Society of America 117, no. 3 (2005): 1405–1416, 10.1121/1.1856273. [DOI] [PubMed] [Google Scholar]
- 19. Başkent D. and Shannon R. V., “Combined Effects of Frequency Compression‐Expansion and Shift on Speech Recognition,” Ear and Hearing 28, no. 3 (2007): 277–289, 10.1097/AUD.0b013e318050d398. [DOI] [PubMed] [Google Scholar]
- 20. Fu Q.‐J., Shannon R. V., and Galvin J. J., “Perceptual Learning Following Changes in the Frequency‐To‐Electrode Assignment With the Nucleus‐22 Cochlear Implant,” Journal of the Acoustical Society of America 112, no. 4 (2002): 1664–1674, 10.1121/1.1502901. [DOI] [PubMed] [Google Scholar]
- 21. Shannon R. V., Zeng F. G., and Wygonski J., “Speech Recognition With Altered Spectral Distribution of Envelope Cues,” Journal of the Acoustical Society of America 104, no. 4 (1998): 2467–2476. [DOI] [PubMed] [Google Scholar]
- 22. Fu Q.‐J., Galvin J. J., and Wang X., “Integration of Acoustic and Electric Hearing Is Better in the Same Ear Than Across Ears,” Scientific Reports 7, no. 1 (2017): 12500, 10.1038/s41598-017-12298-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Willis S., Moore B. C. J., Galvin J. J., and Fu Q.‐J., “Effects of Noise on Integration of Acoustic and Electric Hearing Within and Across Ears,” PLoS One 15, no. 10 (2020): e0240752, 10.1371/journal.pone.0240752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Canfarotta M. W., Dillon M. T., Buss E., Pillsbury H. C., Brown K. D., and O'Connell B. P., “Frequency‐To‐Place Mismatch: Characterizing Variability and the Influence on Speech Perception Outcomes in Cochlear Implant Recipients,” Ear and Hearing 41, no. 5 (2020): 1349–1361, 10.1097/AUD.0000000000000864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Li T. and Fu Q.‐J., “Effects of Spectral Shifting on Speech Perception in Noise,” Hearing Research 270, no. 1–2 (2010): 81–88, 10.1016/j.heares.2010.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Dillon M. T., O'Connell B. P., Canfarotta M. W., Buss E., and Hopfinger J., “Effect of Place‐Based Versus Default Mapping Procedures on Masked Speech Recognition: Simulations of Cochlear Implant Alone and Electric‐Acoustic Stimulation,” American Journal of Audiology 31, no. 2 (2022): 322–337, 10.1044/2022_AJA-21-00123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Lenarz M., Sönmez H., Joseph G., Büchner A., and Lenarz T., “Long‐Term Performance of Cochlear Implants in Postlingually Deafened Adults,” Otolaryngology and Head and Neck Surgery 147, no. 1 (2012): 112–118, 10.1177/0194599812438041. [DOI] [PubMed] [Google Scholar]
- 28. Holden L. K., Finley C. C., Firszt J. B., et al., “Factors Affecting Open‐Set Word Recognition in Adults With Cochlear Implants,” Ear and Hearing 34, no. 3 (2013): 342–360, 10.1097/AUD.0b013e3182741aa7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Hewitt C. E., Kumaravel B., Dumville J. C., and Torgerson D. J., “Assessing the Impact of Attrition in Randomized Controlled Trials,” Journal of Clinical Epidemiology 63, no. 11 (2010): 1264–1270, 10.1016/j.jclinepi.2010.01.010. [DOI] [PubMed] [Google Scholar]
- 30. Schulz K. F. and Grimes D. A., “Sample Size Slippages in Randomised Trials: Exclusions and the Lost and Wayward,” Lancet 359, no. 9308 (2002): 781–785, 10.1016/S0140-6736(02)07882-0. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Research data are not shared.