LAUSR

e18542 Background: Oncologists offer essential clinical information to patients face-to-face or by phone. Today, patients commonly prefer digital communication. To test the effectiveness of this in patients, we deployed proprietary secure messaging (SM) to communicate clinical messages to patients regarding prescribed oral chemotherapy in an effort to improve adherence. Methods: Chronic myelogenous leukemia (CML) patients receiving a tyrosine kinase inhibitor (TKI) were evaluated as two groups: a test group of patients starting therapy between Jan - Jun 2018 who received SM within 45 days therapy start who received SM refill reminders, adherence, condition management, laboratory testing reminders every 30 days, and information regarding side effects, versus a control group that included propensity matched patients initiating therapy between Jan - June 2017 and who received only non-clinical text message refill reminders within 45 days of therapy start. Adherence was compared 6 months after the initial prescription. Primary outcomes were medication possession ratio (MPR) (total day supply/ [(last fill date+last fill day supply)-first fill date]), number of prescriptions filled, % of patients optimally adherent (MPR ≥ 85%), and rate of discontinuation after first fill. The secondary outcome was days of therapy. A t-test evaluated MPR and length of therapy; logistic regression compared discontinuation after first fill and % optimally adherent. Results: 478 CML patients opted into SM and 478 where in the control group. MPR improved 1.6% absolute points (93.20 control v 94.80 test, p = 0.04); no change was seen in number of prescriptions filled (5.04 control v 5.17 test, p = 0.19). More patients receiving SM were optimally adherent (83.05 control v 88.31 test, p = 0.03); discontinuation after first fill was unchanged (6.90 control v 6.49 test, p = 0.79), as were days of therapy (156.51 control v 159.22 test), p = 0.35). Conclusions: Patients initiated on TKI therapy with receiving clinical SM had significantly higher adherence than the control group and were more likely to be optimally adherent, despite the high adherence rates of CML patients at baseline. Further analysis is needed to understand the durability of this adherence improvement.