Adherence to secondary prophylaxis for rheumatic heart disease is underestimated by register data

Jessica Langloh De Dassel, Marea Therese Fittock, Sagen Cheyenne Wilks, Jane Elizabeth Poole, Jonathan Rhys Carapetis, Anna P. Ralph

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Objective: In high-burden Australian states and territories, registers of patients with acute rheumatic fever and rheumatic heart disease are maintained for patient management, monitoring of system performance and research. Data validation was undertaken for the Australian Northern Territory Rheumatic Heart Disease Register to determine quality and impact of data cleaning on reporting against key performance indicators: overall adherence, and proportion of patients receiving ≥80% of scheduled penicillin doses for secondary prophylaxis. Methods: Register data were compared with data from health centres. Inconsistencies were identified and corrected; adherence was calculated before and after cleaning. Results: 2780 penicillin doses were validated; 426 inconsistencies were identified, including 102 incorrect dose dates. After cleaning, mean adherence increased (63.5% to 67.3%, p<0.001) and proportion of patients receiving ≥80% of doses increased (34.2% to 42.1%, p = 0.06). Conclusions: The Northern Territory Rheumatic Heart Disease Register underestimates adherence, although the key performance indicator of ≥80% adherence was not significantly affected. Program performance is better than hitherto appreciated. However some errors could affect patient management, as well as accuracy of longitudinal or inter-jurisdictional comparisons. Adequate resources are needed for maintenance of data quality in acute rheumatic fever/ rheumatic heart disease registers to ensure provision of evidence-based care and accurate assessment of program impact.

Original languageEnglish
Article numbere0178264
Number of pages9
JournalPLoS One
Volume12
Issue number5
DOIs
Publication statusPublished - 31 May 2017

    Fingerprint

Cite this