Guidelines for Drug Therapy of Hypertension: Multi-Site Implementation
Mary K. Goldstein MD
VA Palo Alto Health Care System HSR&D COE
Palo Alto, CA
Funding Period: October 2000 - September 2005

BACKGROUND/RATIONALE:
Hypertension, the most commonly reported medical problem in veterans, is a major risk factor for heart disease and stroke. Lowering blood pressure decreases the risk of these adverse clinical outcomes. Widely promoted evidence-based clinical practice guidelines set target blood pressures for adequate control, yet most hypertensives, including VA patients, do not meet the targets. Guidelines also call for use of specific drugs depending on the patient's pattern of comorbid characteristics; yet, clinicians often prescribe drugs that are not guideline-concordant.

OBJECTIVE(S):
The long term objective of this work is to contribute to the VA's ability to respond flexibly to rapidly evolving medical knowledge by establishing a system guidelines that can be used throughout the VA nationally for implementing multiple different clinical practice. In collaboration with Stanford Medical Informatics we developed an automated decision support system for hypertension management, known as ATHENA DSS built with EON technology for guideline-based decision support. ATHENA DSS incorporates hundreds of knowledge rules to operationalize guidelines for hypertension.

METHODS:
ATHENA DSS combines patient information from VistA with an automated knowledge base of hypertension to generate patient-specific recommendations for management of hypertension that are displayed to primary care clinicians in pop-up windows in the VA’s Computerized Patient Record System (CPRS) when the record for appropriate patients is opened on the day of scheduled primary care clinic visits. The ATHENA DSS pop-up provides advice on adequacy of control of blood pressure and specific recommendations for drug therapy of hypertension, a visual display of the patient’s medication history and concurrent blood pressures, evidence supporting the main recommendations, and other information. We deployed the system at three VA medical centers--Durham, San Francisco, and Palo Alto—and conducted a clinician-randomized trial. We logged data on use of the system, monitored comments entered by clinicians, and conducted a questionnaire survey of clinicians. We planned analyses of impact on clinician prescribing and patient blood pressures. We planned preparation for dissemination of the system to additional VA medical centers.

FINDINGS/RESULTS:
We have established the feasibility of developing a knowledge base of the hypertension guideline, integrating it with patient data from VistA, and deploying this system that provides recommendations for management of a chronic disease, based on complex reasoning, in actual primary care clinics in geographically diverse VA settings. Baseline data showed that clinicians frequently overestimated their guideline adherence (Steinman et al). Over a 15-month trial period, we displayed advisories to 91 primary care clinicians about 10,806 distinct patients with hypertension. Clinicians interacted with the system far more extensively than is typical for automated decision support for chronic problems, suggesting that they find it both useful and usable (Goldstein et al). We showed how an encoded knowledge base in a clinical domain may be as part of automated system for quality assessment (Advani et al). Results for clinician prescribing and impact on blood pressure will be reported at a professional meeting in late 2005. Monitoring of clincian comments entered to the pop-up window allowed early detection and correction of rare data problems suggesting that on-going monitoring should be part of implementation of new health information technology (Chan et al 2005). We are preparing materials for dissemination including a streamlined installation procedure and a documentation manual. We are working with VA Office of Information to improve integration with CPRS including real-time data extraction and write-back of blood pressures.

IMPACT:
This project has shown that it is feasible to deploy an automated decision support system for clinical practice guideline implementation in a chronic disease domain, and that clinicians use the system extensively, speaking to its usefulness and usability. Such systems can potentially improve guideline concordance of clinical practice. The project has wide-reaching implications because the technology used to develop the ATHENA DSS can also be used to develop automated decision support in other clinical domains, as part of quality improvement strategies for chronic disease.

PUBLICATIONS:

Journal Articles

1. Bosworth HB, Olsen MK, Goldstein MK, Orr M, Dudley T, McCant F, Gentry P, Oddone EZ. The veterans' study to improve the control of hypertension (V-STITCH): design and methodology. Contemporary Clinical Trials. 2005; 26(2): 155-68.
2. Steinman MA, Fischer MA, Shlipak MG, Bosworth HB, Oddone EZ, Hoffman BB, Goldstein MK. Clinician awareness of adherence to hypertension guidelines. American Journal of Medicine. 2004; 117(10): 747-54.
3. Advani A, Jones N, Shahar Y, Goldstein MK, Musen MA. An intelligent case-adjustment algorithm for the automated design of population-based quality auditing protocols. Medinfo. 2004; 11(Pt 2): 1003-7.
4. Chan AS, Coleman RW, Martins SB, Advani A, Musen MA, Bosworth HB, Oddone EZ, Shlipak MG, Hoffman BB, Goldstein MK. Evaluating provider adherence in a trial of a guideline-based decision support system for hypertension. Medinfo. 2004; 11(Pt 1): 125-9.
5. Goldstein MK, Coleman RW, Tu SW, Shankar RD, O'Connor MJ, Musen MA, Martins SB, Lavori PW, Shlipak MG, Oddone E, Advani AA, Gholami P, Hoffman BB. Translating research into practice: organizational issues in implementing automated decision support for hypertension in three medical centers. Journal of The American Medical Informatics Association : JAMIA. 2004; 11(5): 368-76.
6. Tu SW, Musen MA, Shankar R, Campbell J, Hrabak K, McClay J, Huff SM, McClure R, Parker C, Rocha R, Abarbanel R, Beard N, Glasgow J, Mansfield G, Ram P, Ye Q, Mays E, Weida T, Chute CG, McDonald K, Molu D, Nyman MA, Scheitel S, Solbrig H, Zill DA, Goldstein MK. Modeling guidelines for integration into clinical workflow. Medinfo. 2004; 11(Pt 1): 174-8.
7. Siegel D, Lopez J, Meier J, Goldstein MK, Lee S, Brazill BJ, Matalka MS. Academic detailing to improve antihypertensive prescribing patterns. American Journal of Hypertension. 2003; 16(6): 508-11.
8. Szeto HC, Coleman RK, Gholami P, Hoffman BB, Goldstein MK. Accuracy of computerized outpatient diagnoses in a Veterans Affairs general medicine clinic. American Journal of Managed Care. 2002; 8(1): 37-43.
9. Shankar RD, Martins SB, Tu SW, Goldstein MK, Musen MA. Building an explanation function for a hypertension decision-support system. Medinfo. 2001; 10(Pt 1): 538-42.
10. Shankar RD, Tu SW, Martins SB, Fagan LM, Goldstein MK, Musen MA. Integration of textual guideline documents with formal guideline knowledge bases. Proceedings / Amia ... Annual Symposium. Amia Symposium. 2001; : 617-21.

DRA: Chronic Diseases, Health Services and Systems
DRE: Communication and Decision Making, Quality of Care, Technology Development and Assessment
Keywords: Clinical practice guidelines, Decision support, Pharmaceuticals
MeSH Terms: none