Comparison of Two Sources of Clinical Audit Data to Assess the Delivery of Diabetes Care in Aboriginal Communities

• Main Authors: Timothy Regan, Christine Paul, Paul Ishiguchi, Catherine D’Este, Claudia Koller, Kristy Forshaw, Natasha Noble, Christopher Oldmeadow, Alessandra Bisquera, Sandra Eades
• Format: Article
• Language: English
• Published: MDPI AG 2017-10-01
• Series: International Journal of Environmental Research and Public Health
• Subjects: decision support systems; clinical; community health services; information storage and retrieval; testing and monitoring
• Online Access: https://www.mdpi.com/1660-4601/14/10/1236

The objective of this study was to determine the concordance between data extracted from two Clinical Decision Support Systems regarding diabetes testing and monitoring at Aboriginal Community Controlled Health Services in Australia. De-identified PenCAT and Communicare Systems data were extracted from the services allocated to the intervention arm of a diabetes care trial, and intra-class correlations for each extracted item were derived at a service level. Strong to very strong correlations between the two data sources were found regarding the total number of patients with diabetes per service (Intra-class correlation [ICC] = 0.99), as well as the number (ICC = 0.98–0.99) and proportion (ICC = 0.96) of patients with diabetes by gender. The correlation was moderate for the number and proportion of Type 2 diabetes patients per service in the group aged 18–34 years (ICC = 0.65 and 0.8–0.82 respectively). Strong to very strong correlations were found for numbers and proportions of patients being tested for diabetes, and for appropriate monitoring of patients known to have diabetes (ICC = 0.998–1.00). This indicated a generally high degree of concordance between whole-service data extracted by the two Clinical Decision Support Systems. Therefore, the less expensive or less complex option (depending on the individual circumstances of the service) may be appropriate for monitoring diabetes testing and care. However, the extraction of data about subgroups of patients may not be interchangeable.