Can Automated Text Messaging Successfully Monitor Antibiotic Adherence for Urban Adolescents and Young Women Managed for Pelvic Inflammatory Disease in the Outpatient Setting

Abstract

Objective This study evaluates patient responsiveness to an automated text messaging system for pelvic inflammatory disease (PID) self-care support, and measures the reliability of text-reported adherence. Design Patients aged 13-25-years with mild to moderate PID were recruited from urban, academic outpatient settings. Enrolled patients received antibiotics and were randomized into a standard of care or intervention group. During a 14-day treatment period, participants in the intervention arm received a community-based nursing visit and daily text message medication reminders with evening prompts to report the number of doses taken. Result Of the 97 participants randomized into the intervention arm, 91 (94%) were eligible for analysis. Most were African American and low income, with a mean age of 18.3 (SD=2.2) years. Participants responded to ~53% (SD=34%) of all dosage inquiry messages. Responsiveness attenuated at approximately 2.2% per day over the treatment period. Ninety-three percent (n=85) of the analyzed intervention arm returned for the two-week follow-up. Despite overall adherence and general responsiveness, text-reported and self-reported medication adherence were not well correlated (r=0.37, p<0.001). Conclusion Our findings show that text messaging is a feasible strategy for reaching urban adolescents being managed for complicated sexually transmitted infections in outpatient settings. However, patient responsiveness to self-care text messages do vary, limiting the adherence monitoring capacity of this technology. Given the number of unanswered text messages and incomplete text-reported adherence data, additional measures to assess adolescents’ adherence to PID medications are needed in clinical trials.