Data-driven approaches to identifying patterns of asthma medication adherence

A study published in Scientific Reports uses data-driven research to identify patterns of behaviour in children and how they follow their asthma medication regime

A new study published this week in Scientific Reports uses data-driven research to identify patterns of behaviour in children and how they follow their asthma medication regime.

The paper identifies groups of behaviour types more clearly than previous research. The groups are categorised into three types of adherence measures which concisely describe the diversity of how patients with asthma take, or don’t take, their medication.

Comparing inhaler use

The research used data from a study in New Zealand of 211 children, with over 35,000 days of data. All children were on a twice a day treatment (morning and evening).

The researchers compared inhaler use in five different ways:

  • The percentages of all the doses that a child should have taken that were taken
  • The percentages of days on which no doses were taken
  • The percentages of days on which both doses were taken
  • Where there were 5 or more days where no doses were taken (a treatment gap)
  • The number of treatment gaps per 100 study days and how long they lasted altogether

Cluster analysis

Using cluster analysis, the researchers were able to identify groups of children and how well they stuck to their inhaler regimes.

These groups were separated into those who adhered to their regime: poorly; moderately well; and very well.

Future research

The researchers hope to use this study to see if the same groupings are useful in studies with adults or if other patterns emerge. Other research could involve working out how these or other patterns of taking medicine can affect symptoms and risk of asthma attacks.

Information like this could help researchers, healthcare professionals, patients and families find the best ways to stick to asthma medication regimes.

The paper was led by Asthma UK Centre for Applied Research affiliate PhD student, Holly Tibble and co-authored by Centre members Rob Horne and Athanasios Tsanas, PhD student Elsie Horne and Director Aziz-Sheikh.

Read the article

Read the article in Scientific Reports.

Cite as

Tibble, H., Chan, A., Mitchell, E.A. et al. A data-driven typology of asthma medication adherence using cluster analysis. Sci Rep 10, 14999 (2020). https://doi.org/10.1038/s41598-020-72060-0

See Holly Tibble's student profile