Optimising medication data collection in a large-scale clinical trial

Jessica E. Lockery, Jason Rigby, Taya A. Collyer, Ashley C. Stewart, Robyn L. Woods, John J. McNeil, Christopher M. Reid, Michael E. Ernst, on behalf of the ASPREE Investigator Group

Research output: Contribution to journalArticleResearchpeer-review

3 Citations (Scopus)

Abstract

Objective Pharmaceuticals play an important role in clinical care. However, in community-based research, medication data are commonly collected as unstructured free-text, which is prohibitively expensive to code for large-scale studies. The ASPirin in Reducing Events in the Elderly (ASPREE) study developed a two-pronged framework to collect structured medication data for 19,114 individuals. ASPREE provides an opportunity to determine whether medication data can be cost-effectively collected and coded, en masse from the community using this framework. Methods The ASPREE framework of type-to-search box with automated coding and linked free text entry was compared to traditional method of free-text only collection and post hoc coding. Reported medications were classified according to their method of collection and analysed by Anatomical Therapeutic Chemical (ATC) group. Relative cost of collecting medications was determined by calculating the time required for database set up and medication coding. Results Overall, 122,910 participant structured medication reports were entered using the type-tosearch box and 5,983 were entered as free-text. Free-text data contributed 211 unique medications not present in the type-to-search box. Spelling errors and unnecessary provision of additional information were among the top reasons why medications were reported as freetext. The cost per medication using the ASPREE method was approximately USD $0.03 compared with USD $0.20 per medication for the traditional method. Conclusion Implementation of this two-pronged framework is a cost-effective alternative to free-text only data collection in community-based research. Higher initial set-up costs of this combined method are justified by long term cost effectiveness and the scientific potential for analysis and discovery gained through collection of detailed, structured medication data.

Original languageEnglish
Article numbere0226868
Number of pages12
JournalPLoS ONE
Volume14
Issue number12
DOIs
Publication statusPublished - 27 Dec 2019

Cite this