A Dynamic Clinical Decision Support Tool to Improve Primary Care Outcomes in a High-Volume, Low-Resource Setting

Citations Over TimeTop 10% of 2024 papers

Abstract

The Family Health Centers at New York University Langone (FHC), a federally qualified health center network in New York City, created a novel clinical decision support (CDS) tool that alerts primary health care providers to patients' gaps in care and triggers a dynamic, individualized order set on the basis of unique patient factors, enabling providers to readily act on each patient's specific gaps in care. FHC implemented this tool in 2017, starting with 15 protocols for quality measures; as of February 2024, there are 30 such protocols. During a patient visit with a provider, when there is a gap in care, a best-practice alert (BPA) fires, which includes an order set unique to the patient and visit. The provider can bypass the alert (not open it) or acknowledge the alert (open it). The provider may review the content of the order set and accept it as is or with modifications, or they can decline its recommendations if they believe it is not appropriate or plan to address the gap in care another way during the visit. To accept the dynamic order set is the intended workflow. The authors present data from September 2019 to January 2023 totaling 171,319 patient visits with at least one open gap in care among providers in pediatrics, family medicine, and internal medicine. The rate at which providers acknowledged the BPA in the first 6 months was 45% and steadily increased. In the last 6 months of the period, providers acknowledged the BPA 78% (19,281 of 24,575) of the time. Similarly, in the first 6 months, in all encounters in which a BPA was fired, 28.8% (8,585 of 29,829) had an order placed via the dynamic order set (accepted); that rate increased to 49.7% (12,210 of 24,575) during the last 6 months. This order set completion rate is notable given that most CDS use rates are low. Gap closure was higher when providers acknowledged the alert. In an analysis of all encounters with at least one open gap, spanning 2019–2023, 46% (48,431 of 105,371) of the time, at least one gap was closed when the alert was acknowledged compared with 33% (21,993 of 65,948) when the alert was bypassed (and the recommendations of the dynamic order set were never followed). The authors show that CDS tools can be successfully implemented in a high-volume, low-resource setting if designed with efficiency in mind, ensuring provider utilization and clinical impact through closing care gaps. CDS tools that are dynamically patient specific can help improve quality of care if they are part of a broader culture of quality improvement.

Related Papers

• → A four-phase model of the evolution of clinical decision support architectures(2008)148 cited
• Development, deployment and usability of a point-of-care decision support system for chronic disease management using the recently-approved HL7 decision support service standard.(2007)
• → Evaluating Clinical Decision Support Rules as an Intervention in Clinician Workflows With Technology(2010)12 cited
• → Methodologies of Legacy Clinical Decision Support System(2017)1 cited