To investigate real-world botulinum toxin A (BoNT-A) treatment persistence.

Adherence to BoNT-A therapy is critical to achieving treatment goals; however, evidence on BoNT-A treatment persistence and discontinuation drivers is limited.

Mayo Clinic electronic health records were used to characterize BoNT-A persistence in adult patients with ≥1 treatment cycle, overall and across mutually-exclusive International Classification of Diseases 9/10-defined cohorts: stroke, migraine, bladder dysfunction (BD), cervical dystonia (CD), and adult cerebral palsy (CP). Index date was first BoNT-A treatment; patients were followed for two years post-index. Persistence/non-persistence were defined by evaluating if all treatment intervals post-index were ≤80^th percentile of all patients in the cohort (persistent) or if any were >80^th percentile (non-persistent). Reasons for discontinuation were identified by analyzing physicians’ free-text notes.

Of 24,212 patients (mean age 52 years, 73.6% female) persistency was evaluable for 23,746; 11.1% were persistent. Cross-cohorts persistency was 5.7% (stroke, 23/407), 7.4% (migraine, 408/5543), 10.5% (BD, 62/593), 10.3% (CD, 102/989) and 14.8% (CP, 8/54). Persistent (vs non-persistent) patients were generally younger (stroke, BD) and more often female (stroke, BD, CD cohorts); neither trend was seen in the migraine cohort. Mean (95% confidence interval) BoNT-A treatments was 3.35 (3.32–3.39) cycles overall, lowest in the BD cohort (1.99 [1.87–2.11]); highest in the migraine cohort (3.90 [3.83–3.98]). Cardiovascular comorbidities were highest for the stroke cohort; depression and anxiety were common across all cohorts (cross-cohort range: 17.6–35.4% and 28.5–57.4%, respectively). Overall, antidepressant use ranged from 16.7–39.5% across cohorts; higher among persistent (41.0%) vs non-persistent (28.4%) patients. The most common reasons for discontinuation in all cohorts were patient- or physician-described lack of treatment effectiveness (cross-cohort range: 17.4–40.4% and 12.3–32.6%, respectively).

BoNT-A persistence was low across disease cohorts over a 2-year follow-up. Understanding drivers of discontinuation may inform strategies to manage BoNT-A persistence.