Abstract
Introduction: Text message-based interventions reduce colonoscopy no-show rates and improve bowel preparation scores. In this non-randomized study, we assessed whether an interactive text messaging system could improve colonoscopy outcomes. Methods: Colonoscopy pre-procedural instructions were programmed into a dedicated software platform created for this study. In the intervention arm, text messages were sent to veterans during a 4-week study period. Validated pre-procedural satisfaction questionnaires were completed by patients during standard protocol and intervention periods. Demographics and colonoscopy outcomes data were compared between the standard protocol and intervention arms, including procedure completion rate on scheduled date, Boston bowel preparation score (BPPS), adenoma detection rate, and satisfaction. Results: Of 241 patients, 128 were in the standard protocol arm and 113 in the intervention arm. Higher proportions of patients receiving text messages underwent colonoscopy on their scheduled date (69.9%) compared to the ones in the standard protocol (50.8%, p = 0.015). Patients with ≥3 interactions with the system had 80.6% likelihood of completing colonoscopy on the scheduled date compared to 56.9% with <3 interactions and 50.8% with standard protocol (p < 0.001). Frequency of interaction with the system was similar between older (>65 years) and younger patients (p = 1.0). Among older patients, colonoscopy was completed successfully in 84.2% when alert-based human interactions occurred compared to 65.6% in those without and 47.9% with standard protocol (p = 0.018). More than 90% indicated they would recommend the system to patients undergoing future colonoscopy. Conclusion: An interactive text messaging system improves successful colonoscopy rates in a VA setting, with greatest impact in older patients.
