Exploring challenges and strategies when using routine data in research: a systematic review

Calanzani, Natalia, Vojt, Gabriele, Weller, David and Campbell, Christine (2017) Exploring challenges and strategies when using routine data in research: a systematic review

Abstract

Routine data are invaluable for improving patient care and building evidence on epidemiology, pharmacology and health services research. However, there are several limitations when using these medical databases for research. It is important that these limitations are known, and optimal strategies to deal with them are adopted. We aimed to review the evidence on these limitations and strategies to inform on a study using UK medical data and analysing Read codes. We adopted the thematic analysis methodology and used narrative synthesis to report the findings in text, diagrams and tables. Identified limitations included the possibility of misdiagnosis, misclassification and missing data. The Read code terminology offers many choices for coding a single diagnosis and this may lead to variations in coding. Methodological concerns such as the possibility of residual confounding and the higher likelihood of obtaining statistically significant findings were also outlined. Identified strategies included ways to address common epidemiological challenges such as bias and confounding, specific strategies to identify cases, validate data, analyse free-text and prescription data, and procedures to aid data pre-processing, management, analysis and reporting. Findings show that there is much to be learned from what has already been published. The available evidence can aid the development of purpose-built, robust data analysis plans and research protocols for studies using different types of medical databases.

This record has no associated files available for download.

Contributors

Author: Gabriele Vojt

Download statistics