Mixed methods evaluation of a clinical decision support system to reduce variation in healthcare

Abstract

Evidence that clinical decision support systems (CDSS) reduce unwarranted variation in care and improve patient outcomes remains limited, particularly for acute, high-risk conditions where deviations from clinical guidelines are costly. We evaluated utilisation of a CDSS-embedded care plan within an electronic medical record designed to support timely, protocol-driven management of a complex metabolic emergency. Among 345 episodes, the CDSS was used in just over half of cases; yet a 1% increase in utilisation was associated with a reduction in hospital length of stay by approximately 1.3%, and any use was linked to a 3.3% reduction in in-hospital mortality. Fourteen clinician interviews supported these findings but highlighted barriers such as limited awareness and lack of training that constrain meaningful use. Quantitative and qualitative evidence together suggest that well-designed CDSS care plans can reduce variation and improve outcomes, while emphasising the need for awareness, education, and workflow alignment to unlock greater value.