Improvement in symptom-related disruptions is associated with fewer days of short-acting beta-agonist use in asthma

Study design

This analysis used retrospective data collected from patients aged ≥12 years with a self-reported history of asthma who utilized a digital self-management platform (Propeller Health, Madison, WI, USA)5 between 2016 and 2019. Patients self-enrolled at health fairs and via social media campaigns or were recruited through clinical programs offered at their healthcare organizations. Patients did not receive monetary compensation for their participation.

Data collection

To set up the digital platform, patients attached an EMM to their SABA inhaler to passively record the date and time of each actuation. The days in which the patient did not use their SABA medication were classified as SFD. Patients also paired their EMM with a companion smartphone app, which helped patients track medication usage and trends and provided evidence-based educational content5.

At enrollment and every month thereafter, patients were prompted to complete an Asthma Control Test (ACT) in the app to assess their own perception of daily functioning, symptom-related disruption, and symptom control in the four preceding weeks12. The first question of the ACT (ACT Q1) focuses on self-reported symptom-related disruptions, asking, “In the past 4 weeks, how much of the time did your asthma keep you from getting as much done at work, school, or at home?”. Responses range from 1 (all of the time) to 5 (none of the time).

Ethics approval

All patients agreed to Propeller’s Terms of Use, which allows for retrospective analysis of de-identified aggregate data. The current retrospective analysis proposal was determined to be exempt by the Copernicus Institutional Review Board (PRH1-18-132).

Retrospective analysis

The relationship between ACT Q1 and SFD was assessed cross-sectionally in patients with ≥30 days of continuous EMM data, which captured SABA use preceding a completed ACT. The Kruskal–Wallis test with Dunn’s post hoc comparison test was used to determine if the percent of SFDs in the 30 days preceding an ACT statistically differed between the five response options of ACT Q1 (all of the time, most of the time, some of the time, a little of the time, none of the time). To account for the multiple statistical tests performed simultaneously, Bonferroni correction was applied for Dunn’s post hoc comparison. We further explored the association between ACT Q1 and SFD by performing an ordinal logistic regression model adjusting for age, gender, baseline ACT, season, enrollment site, and controller inhaler use.

The change in total ACT score, ACT Q1, and SFD over time was evaluated between enrollment and month 3 (days 68–97). Patients needed to have ≥90 days of continuous EMM data, ≥1 EMM-recorded SABA usage, and 2 completed ACTs (at baseline and 3 months) to be included in the analysis. Changes over time in the total ACT and SFD were assessed using Wilcoxon signed-rank test for paired data (e.g., consisting of repeated measurements). For the changes over time in ACT Q1, the Stuart-Maxwell test checked the overall difference of all five response options to ACT Q1 at baseline and at month 3. After that, McNemar’s tests were used to test if there is a statistically significant change between baseline and month 3 in each level.

All statistical tests were two-tailed with an alpha = 0.05 threshold for statistical significance, except for Dunn’s post hoc comparison test (alpha = 0.005). All analyses were conducted in R version 4.1.1 (R Foundation for Statistical Computing).

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

留言 (0)

沒有登入
gif