Dataset Information

Exploratory Data Mining Techniques (Decision Tree Models) for Examining the Impact of Internet-Based Cognitive Behavioral Therapy for Tinnitus: Machine Learning Approach.

ABSTRACT:

Background

There is huge variability in the way that individuals with tinnitus respond to interventions. These experiential variations, together with a range of associated etiologies, contribute to tinnitus being a highly heterogeneous condition. Despite this heterogeneity, a "one size fits all" approach is taken when making management recommendations. Although there are various management approaches, not all are equally effective. Psychological approaches such as cognitive behavioral therapy have the most evidence base. Managing tinnitus is challenging due to the significant variations in tinnitus experiences and treatment successes. Tailored interventions based on individual tinnitus profiles may improve outcomes. Predictive models of treatment success are, however, lacking.

Objective

This study aimed to use exploratory data mining techniques (ie, decision tree models) to identify the variables associated with the treatment success of internet-based cognitive behavioral therapy (ICBT) for tinnitus.

Methods

Individuals (N=228) who underwent ICBT in 3 separate clinical trials were included in this analysis. The primary outcome variable was a reduction of 13 points in tinnitus severity, which was measured by using the Tinnitus Functional Index following the intervention. The predictor variables included demographic characteristics, tinnitus and hearing-related variables, and clinical factors (ie, anxiety, depression, insomnia, hyperacusis, hearing disability, cognitive function, and life satisfaction). Analyses were undertaken by using various exploratory machine learning algorithms to identify the most influencing variables. In total, 6 decision tree models were implemented, namely the classification and regression tree (CART), C5.0, GB, XGBoost, AdaBoost algorithm and random forest models. The Shapley additive explanations framework was applied to the two optimal decision tree models to determine relative predictor importance.

Results

Among the six decision tree models, the CART (accuracy: mean 70.7%, SD 2.4%; sensitivity: mean 74%, SD 5.5%; specificity: mean 64%, SD 3.7%; area under the receiver operating characteristic curve [AUC]: mean 0.69, SD 0.001) and gradient boosting (accuracy: mean 71.8%, SD 1.5%; sensitivity: mean 78.3%, SD 2.8%; specificity: 58.7%, SD 4.2%; AUC: mean 0.68, SD 0.02) models were found to be the best predictive models. Although the other models had acceptable accuracy (range 56.3%-66.7%) and sensitivity (range 68.6%-77.9%), they all had relatively weak specificity (range 31.1%-50%) and AUCs (range 0.52-0.62). A higher education level was the most influencing factor for ICBT outcomes. The CART decision tree model identified 3 participant groups who had at least an 85% success probability following the undertaking of ICBT.

Conclusions

Decision tree models, especially the CART and gradient boosting models, appeared to be promising in predicting ICBT outcomes. Their predictive power may be improved by using larger sample sizes and including a wider range of predictive factors in future studies.

SUBMITTER: Rodrigo H

PROVIDER: S-EPMC8596228 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundTinnitus is a symptom that can be very distressing owing to hearing sounds not related to any external sound source. Managing tinnitus is notoriously difficult, and access to evidence-based care is limited. Cognitive behavioral therapy (CBT) is a tinnitus management strategy with the most evidence of effectiveness but is rarely offered to those distressed by tinnitus. The provision of internet-based CBT for tinnitus overcomes accessibility barriers; however, it is not currently readily available in the United States.ObjectiveThe aim of this study is to investigate the efficacy of internet-based CBT compared with that of weekly monitoring for the management of tinnitus in reducing tinnitus distress; reducing tinnitus-related comorbidities, including tinnitus cognitions, insomnia, anxiety, and depression; and assessing the stability of the intervention effects 2 months after the intervention.MethodsA 2-arm randomized clinical trial comparing audiologist-guided internet-based CBT (n=79) to a weekly monitoring group (n=79) with a 2-month follow-up assessed the efficacy of internet-based CBT. Eligible participants included adults seeking help for tinnitus. Recruitment was conducted on the web using an open-access website. Participants were randomized via 1:1 allocation, but blinding was not possible. The study was undertaken by English or Spanish speakers on the web. The primary outcome was a change in tinnitus distress as measured using the Tinnitus Functional Index. Secondary outcome measures included anxiety, depression, insomnia, tinnitus cognition, hearing-related difficulties, and quality of life.ResultsInternet-based CBT led to a greater reduction in tinnitus distress (mean 36.57, SD 22) compared with that in weekly monitoring (mean 46.31, SD 20.63; effect size: Cohen d=0.46, 95% CI 0.14-0.77) using an intention-to-treat analysis. For the secondary outcomes, there was a greater reduction in negative tinnitus cognition and insomnia. The results remained stable over the 2-month follow-up period. No important adverse events were observed. Further, 16% (10/158) of participants withdrew, with low overall compliance rates for questionnaire completion of 72.3% (107/148) at T1, 61% (91/148) at T2, and 42% (62/148) at T3.ConclusionsThis study is the first to evaluate and indicate the efficacy of audiologist-delivered internet-based CBT in reducing tinnitus distress in a US population. It was also the first study to offer internet-based CBT in Spanish to accommodate the large Hispanic population in the United States. The results have been encouraging, and further work is indicated in view of making such an intervention applicable to a wider population. Further work is required to improve compliance and attract more Spanish speakers.Trial registrationClinicalTrials.gov NCT04004260; https://clinicaltrials.gov/ct2/show/NCT04004260.

Project description:Background Patients often report that living with a condition such as tinnitus can be debilitating, worrying, and frustrating. Efficient ways to foster management strategies for individuals with tinnitus and promoting tinnitus self-efficacy are needed. Internet-based cognitive behavioral therapy (ICBT) for tinnitus shows promise as an evidence-based intervention in Europe, but is not available in the United States. The aim of this pilot study was to evaluate the feasibility of an ICBT intervention for tinnitus in the United States. Method This study reports the Phase 1 trial intended to support implementation of a larger randomized clinical trial (RCT) comparing ICBT to a weekly monitoring group. As a pilot study, a single-group pretest-posttest design was used to determine outcome potential, recruitment strategy, retention, and adherence rates of ICBT for tinnitus. The primary outcome was a change in tinnitus distress. Secondary outcome measures included measures of anxiety, depression, insomnia, tinnitus cognitions, hearing-related difficulties, and quality of life. Results Of the 42 screened participants, nine did not meet the inclusion criteria and six withdrew. There were 27 participants who completed the intervention, with a mean age of 55.48 (± 9.9) years. Feasibility was established, as a large pretest-posttest effect size of d = 1.6 was found for tinnitus severity. Large pretest-posttest effect sizes were also found for tinnitus cognitions and hearing-related effects, and a medium effect was found for insomnia and quality of life. Treatment adherence varied with a retention rate of 85% (n = 23) at post-intervention assessment and 67% (n = 18) for the follow-up assessment. Conclusions This pilot study supported the feasibility of ICBT for tinnitus in the United States. Ways of improving intervention retention and recruitment rates need to be explored in future ICBT studies. Protocol refinements that were identified will be implemented prior to further RCTs to investigate the efficacy of ICBT for tinnitus in the United States. Supplemental Material https://doi.org/10.23641/asha.15501135.

Project description:BackgroundInternet-based cognitive behavioral therapy (ICBT) for tinnitus is an evidence-based intervention. The components of ICBT for tinnitus have, however, not been dismantled and thus the effectiveness of the different therapeutic components is unknown. It is, furthermore, not known if heterogeneous tinnitus subgroups respond differently to ICBT.AimsThis dismantling study aimed to explore the contribution of applied relaxation within ICBT for reducing tinnitus distress and comorbidities associated with tinnitus. A secondary aim was to assess whether outcomes varied for three tinnitus subgroups, namely those with significant tinnitus severity, those with low tinnitus severity, and those with significant depression.MethodsA parallel randomized controlled trial design (n = 126) was used to compare audiologist-guided applied relaxation with the full ICBT intervention. Recruitment was online and via the intervention platform. Assessments were completed at four-time points including a 2-month follow-up period. The primary outcome was tinnitus severity as measured by the Tinnitus Functional Index. Secondary outcomes were included for anxiety, depression, insomnia, negative tinnitus cognitions, health-related quality of life, hearing disability, and hyperacusis. Treatment engagement variables including the number of logins, number of modules opened, and the number of messages sent. Both an intention-to-treat analysis and completer's only analysis were undertaken.ResultsEngagement was low which compromised results as the full intervention was undertaken by few participants. Both the ICBT and applied relaxation resulted in large reduction of tinnitus severity (within-group effect sizes d = 0.87 and 0.68, respectively for completers only analysis), which were maintained, or further improved at follow-up. These reductions in tinnitus distress were greater for the ICBT group, with a small effect size differences (between-group d = 0.15 in favor of ICBT for completers only analysis). Tinnitus distress decreased the most at post-intervention for those with significant depression at baseline. Both ICBT and applied relaxation contributed to significant reductions on most secondary outcome measures, with no group differences, except for a greater reduction of hyperacusis in the ICBT group.ConclusionDue to poor compliance partly attributed to the COVID-19 pandemic results were compromised. Further studies employing strategies to improve compliance and engagement are required. The intervention's effectiveness increased with initial level of tinnitus distress; those with the highest scores at intake experienced the most substantial changes on the outcome measures. This may suggest tailoring of interventions according to tinnitus severity. Larger samples are needed to confirm this.

Project description:BackgroundInternet-delivered cognitive behavioral therapy (ICBT) can improve access to mental health care for students, although high attrition rates are concerning and little is known about long-term outcomes. Motivational interviewing (MI) exercises and booster lessons can improve engagement and outcomes in face-to-face cognitive behavioral therapy.ObjectiveThis study aimed to examine the use of pretreatment MI exercises and booster lessons in ICBT for postsecondary students.MethodsIn this factorial trial (factor 1: web-based MI before treatment; factor 2: self-guided booster lesson 1 month after treatment), 308 clients were randomized to 1 of 4 treatment conditions, with 277 (89.9%) clients starting treatment. All clients received a 5-week transdiagnostic ICBT course (the UniWellbeing course). Primary outcomes included changes in depression, anxiety, and perceived academic functioning from before treatment to after treatment and at the 1-month and 3-month follow-ups.ResultsOverall, 54% (150/277) of students completed treatment and reported large improvements in symptoms of depression and anxiety and small improvements in academic functioning after treatment, which were maintained at the 1-month and 3-month follow-ups. Pretreatment MI did not contribute to better treatment completion or engagement, although small between-group effects favored MI for reductions in depression (Cohen d=0.23) and anxiety (Cohen d=0.25) after treatment. Only 30.9% (43/139) of students randomized to one of the booster conditions accessed the booster. Overall, no main effects were found for the booster. Subanalyses revealed that clients who accessed the booster had larger decreases in depressive symptoms (Cohen d=0.31) at the 3-month follow-up. No interactions were found between MI and the booster.ConclusionsRather than offering MI before treatment, clients may experience more benefits from MI exercises later in ICBT when motivation wanes. The low uptake of the self-guided booster limited our conclusions regarding its effectiveness. Future research should examine offering a booster for a longer duration after treatment, with therapist support and a longer follow-up period.Trial registrationClinicalTrials.gov NCT04264585; https://clinicaltrials.gov/ct2/show/NCT04264585.