1046 — Approaches to Design and Implementation of Clinical Decision Support in the VA: Relationship to Staff Sufficiency
Saleem JJ (VA HSR&D Center on Implementing Evidence-Based Practice (CIEBP)), Doebbeling BN
(VA HSR&D CIEBP), Flanagan ME
(VA HSR&D CIEBP), Haggstrom DA
(VA HSR&D CIEBP), Woodward-Hagg H
(VA HSR&D CIEBP), Yano EM
(VA HSR&D CIEBP)
For clinical decision support (CDS), recommended approaches to development include pilot testing, human factors assessment and integration into workflow. The VA is widely believed to be a leader in the deployment of CDS. The VA Clinical Practice Organizational Survey (CPOS), conducted in 2006, measured the implementation of information technology considered important to quality improvement.
In this national organizational survey of VHA Chiefs of Staff, sponsored by the Office of Quality & Performance and HSR&D, respondents were queried regarding the different mechanisms being used at their facilities to develop clinical decision support, specifically computerized clinical reminders and disease-specific templates.
111 Chiefs of Staff participated (86%). Implementation of CDS ranged from 46% of facilities for community-acquired pneumonia (inpatient) to 92% for management of diabetes and hypertension (outpatient). In the development of CDS, most facilities used informal discussions between providers and clinical application coordinators (90%) and analysis of reminder impact on performance improvement (85%). Prior to full-scale implementation, 67% reported asking provider experts for clinical opinion or pilot testing the clinical reminders and/or templates. Only 31% of facilities reported using formal usability testing prior to deployment. After implementation, 53% assessed user satisfaction and 77% assessed impact on performance. The sufficiency of IRM or CPRS technical support staff was rated as completely sufficient at 24% of facilities, mostly sufficient at 40%, somewhat at 22%, and barely or not at all sufficient at 14%. The sufficiency of IRM staff was significantly related to the likelihood of pilot testing (p=0.05) prior to implementation.
CDS tools are widely implemented in the VHA. However, the quality of the CDS tools is uncertain because many facilities do not formally evaluate these tools. We found marked variation in approaches to implementing CDS in VA facilities, including significant deviations from recommended approaches. The availability of technical staff resources predicts the likelihood of piloting CDS prior to implementation.
Significant opportunities exist to improve the CDS implementation process by greater use of formal usability testing, pilot testing, and talking with provider experts to improve the integration of CDS into workflow. These implementation efforts are crucial to increasing the likelihood of successful CDS deployment and sustainability in the VHA.