Self-reported chronic therapy use after 24-weeks of follow-up by participants who completed the simplify randomized, controlled trial

•The CF community desires evidence to inform treatment modification decisions.•We studied treatment use by participants after they completed the SIMPLIFY trial.•Discontinuing a therapy in SIMPLIFY was associated with its non-use after the trial.•SIMPLIFY cohort attributes may limit the generalizability of follow-up observations.•Randomization in a treatment discontinuation trial may affect post-trial behavior. Highly effective CFTR modulator therapy (HEMT) has improved the health of many people with cystic fibrosis (pwCF), offering opportunities to discontinue burdensome therapies. SIMPLIFY included randomized, controlled trials that confirmed non-inferiority of discontinuing versus continuing dornase alfa (DA) or hypertonic saline (HS) for 6 weeks in pwCF on HEMT. In this study of post-trial treatment use by SIMPLIFY participants, we hypothesized that randomization to discontinue DA or HS during the trial would be associated with a higher likelihood of non-use of each medication during follow-up. We electronically surveyed SIMPLIFY participants every 4 weeks for 24 weeks after trial completion but before the main trial results were publicly disclosed. We asked them how often they used medications during the previous week. We estimated covariate-adjusted odds ratios (ORs) of DA or HS non-use by logistic regression with generalized estimating equations. After exclusions mostly due to lack of any surveys, 472 participants were included in the analysis population, 181 from the HS trial and 291 from the DA trial. Approximately half of the analysis population completed all six surveys. At every month of follow-up in both trials, the percentage of individuals reporting non-use of DA or HS during the previous week was greater among those randomized to discontinue therapy. Among participants with responses at 24 weeks, 30/122 (24.6 %) in the HS trial and 79/222 (35.6 %) in the DA trial reported non-use of the respective study medication. After adjusting for covariates, participants randomized to discontinue DA were 8.7-times (95 % CI: 4.3–17.7) more likely to not use DA during follow-up than those randomized to continue DA, and participants randomized to discontinue HS were 5.2-times (95 % CI: 2.1–12.8) more likely to not use HS during follow-up compared to those randomized to continue. In healthy pwCF on ETI, randomization to discontinue DA or HS during SIMPLIFY was associated with greater odds of not using each medication after the trial compared to randomizatio