Heterogeneity in clinical research data quality monitoring: A national survey

Lauren Houston; Ping Yu; Allison Martin; Yasmine Probst

doi:10.1016/j.jbi.2020.103491

Heterogeneity in clinical research data quality monitoring: A national survey

J Biomed Inform. 2020 Aug:108:103491. doi: 10.1016/j.jbi.2020.103491. Epub 2020 Jun 20.

Authors

Lauren Houston¹, Ping Yu², Allison Martin³, Yasmine Probst⁴

Affiliations

¹ School of Medicine, Faculty of Science, Medicine, and Health, University of Wollongong, Wollongong, Australia; Illawarra Health and Medical Research Institute, University of Wollongong, Australia. Electronic address: lah993@uowmail.edu.au.
² Illawarra Health and Medical Research Institute, University of Wollongong, Australia; School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Wollongong, Australia.
³ School of Medicine, Faculty of Science, Medicine, and Health, University of Wollongong, Wollongong, Australia.
⁴ School of Medicine, Faculty of Science, Medicine, and Health, University of Wollongong, Wollongong, Australia; Illawarra Health and Medical Research Institute, University of Wollongong, Australia.

PMID: 32574794
DOI: 10.1016/j.jbi.2020.103491

Abstract

Introduction: Clinical research is vital in the discovery of new medical knowledge and reducing disease risk in humans. In clinical research poor data quality is one of the major problems, affecting data integrity and the generalisability of the research findings. To achieve high quality data, guidance needs to be provided to clinical studies on the collection, processing and handling of data. However, clinical trials are implementing ad hoc, pragmatic approaches to ensure data quality. This study aims to explore the procedures for ensuring data quality in Australian clinical research studies.

Material and methods: We conducted a national cross-sectional, mixed-mode multi-contact (postal letter and e-mail) web-based survey of clinical researchers associated with clinical studies listed on the Australian and New Zealand Clinical Trials Registry.

Results: Of the 3689 clinical studies contacted, 589 (16%) responded, 570 (97%) consented and 441 (77%) completed the survey. 67% clinical studies reported following national and/or international guidelines for data monitoring, with the National Statement (86%) and Good Clinical Practice Guidelines (55%) most common. Source data were most likely to be recorded on one instrument (46%), of which paper (77%) being most common. 46.4% studies did not use data management software and 55% monitored data via traditional approaches (e.g. source data verification). Training on data quality was only provided to less than half of the staff responsible for data entry (43.9%) and data monitoring (37.5%). Regression analysis on 179 (33%) respondents found a borderline significant association between intervention trials and a definition for protocol deviation and/or violation (odds 3.065, p = 0.096). This may suggest when clinical trials are provided with additional guidance and resources, they are more likely to implement required procedures. Statistical strength of the full regression model was not significant χ2 (13, 179) = 15.827, p = 0.259.

Conclusion: Small single-site academic clinical studies implemented ad hoc procedures to ensure data quality. Education and training are required to promote standardised practices to ensure data quality in small scale clinical trials.

Keywords: Clinical research; Data management; Data monitoring; Data quality; Education; Training.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Australia
Cross-Sectional Studies
Data Accuracy*
Data Management*
Humans
New Zealand