By your bootstraps.
The Study: This simulated model using cross sectional data aims to tease out the reasons underlying this trend. Besides just joint pains and hot flashes, that is. 216 women were questioned at a single time point at any follow-up clinic visit within the window recommended for upfront maintenance endocrine therapy for localized breast cancer. The questionnaire included when they started +/- when they stopped endocrine therapy as well as scales of how much they were affected by 13 potential treatment toxicities. Of note, a true early discontinuation rate was not evaluable as the vast majority of patients were sampled prior to year 5. The simulated models are complicated but help tease out which adverse events impacted discontinuation at different time points, if at all. Turned out they all had a significant impacts at some point, with the three heaviest hitters being cost, joint pain, and emotional distress. It’s perhaps no shocker that “dislike” and “forgetting” had significant impacts on never even starting endocrine therapy. Cost lead to a significant drop in usage at year 3, and cognitive effects and weight gain both impacted discontinuation in year 4. While these specific findings are interesting, the authors are more proud of their novel use of “iterative bootstrapping” to produce informative results from an easily administered snapshot questionnaire—and we’re guessing this informatics journal was smitten, too.
Bottom Line: Simulated models may offer a cheap, easy, and effective gauge of patient compliance influencers. | Shinn, JCO Clin Cancer Inform 2019