Welcome! In this project I am running secondary analyses of a University of Auckland Electronic Monitoring Device (EMD) intervention study which investigated whether audio-visual reminders can improve children's adherence to twice-daily inhaled corticosteroids. The paper for the original study is entitled "The effect of an electronic monitoring device with audiovisual reminder function on adherence to inhaled corticosteroids and school attendance in children with asthma: a randomised controlled trial" (Chan et al., 2015, 3(3) Lancet Respiratory Medicine).
A conference paper entitled "Heterogeneity in Asthma Medication Adherence Measurement" from the IEEE Bioinformatics and Bioengineering Conference 2019 is available online, which explores adherence measurements and their concordance in this data.
The primary paper relating to this body of work was published in Scientific Reports on September 14th, 2020, and is available here.
Thank you for visiting,
Holly Tibble
Asthma preventer medication non-adherence is strongly associated with poor asthma control. One-dimensional measures of adherence, such as the percentage of days where all prescribed doses are taken, may ignore clinically important nuances in patterns of medication-taking behavior. We sought to construct a data-driven multi-dimensional typology of medication non-adherence in children with asthma.
We analyzed data from an intervention study of electronic inhaler monitoring devices, comprising 211 patients yielding 35,161 person-days of data. Five adherence measures were extracted: the percentage of doses taken, the percentage of days on which zero doses were taken, the percentage of days on which both doses were taken, the number of treatment intermissions per 100 study days, and the duration of treatment intermissions per 100 study days. We applied principal component analysis on the medication adherence measures and subsequently applied k-means to determine cluster membership. We built decision trees from the original variables to identify the measure that could predict cluster assignment with the highest accuracy, increasing interpretability and increasing clinical utility.
We identified three subgroups, which we labeled as poor, moderate, or good adherence. The percentage of prescribed doses taken during the study contributed to the prediction of cluster assignment most accurately (84% in out-of-sample data), and had median values of 19.8%, 69.6%, and 9.40% for each respective subgroup.
We demonstrate the use of adherence measures towards a three-group categorization of medication non-adherence, which succinctly describes the diversity of patient medication taking patterns in asthma.