Baseline Predictors of Discontinuation of Prescription Drug Therapy for IBS-C: Cohort Analysis at an Integrated Healthcare System.

Abstract

Effective prescription drug treatment of constipation-predominant irritable bowel syndrome (IBS-C) requires patients to remain on daily therapy, yet predictive factors to optimize treatment selection are unknown. We assessed whether common comorbidities including chronic overlapping pain conditions (COPCs), mood disorders, or concurrent medications influence the risk of discontinuing IBS-C prescription drug therapy. We included all IBS-C patients who initiated treatment with the secretagogues linaclotide or lubiprostone across the Michigan Medicine healthcare system between 2012 and 2016. A Cox proportional hazards model was constructed to model time-to-treatment discontinuation as a valid, quantifiable measure of IBS medication persistence using hazards ratios (HR) with 95% confidence intervals (CI). Our cohort included 225 patients on linaclotide and 492 on lubiprostone (mean age 48.3years, 86.9% women, 46.6% with at least one COPC, 60.3% with at least one mood disorder) with an average follow-up of 2.1years. Patients with at least one COPC (HR = 0.566; 95%CI = 0.371-0.863) and also women (HR = 0.535; 95%CI = 0.307-0.934) had a lower risk of discontinuing linaclotide, while COPCs predicted a trend toward increased discontinuation of lubiprostone (HR = 1.254; 95%CI = 0.997-1.576). Age, comorbid mood disorders, and baseline use of narcotics or benzodiazepines did not significantly mediate the risk of treatment discontinuation; our findings remained stable in univariate and multivariable analyses. COPCs and sex appear to influence the likelihood of discontinuation of two commonly prescribed secretagogues, while mood disorders, narcotics, and benzodiazepines may not. Routine assessment for comorbid COPCs prior to initiating therapy may optimize IBS-C treatment selection and outcomes in practice.