gms | German Medical Science

Background: Consistent adherence to evidence-based best practices should significantly improve health. However, systematic reviews and guidelines have covered different clinical questions, omitted some key topics, and used different methodologies, creating some confusion and redundant effort.

Objectives: 1. To compare systematic reviews and guideline recommendations for acute coronary syndrome, congestive heart failure, stroke, sepsis and hospital-acquired infections; 2. To describe efforts and outcomes implementing best practices consistently in a large health care system using available recommendations.

Methods: We searched the Cochrane Library and other major data bases for systematic reviews and guidelines on the above topics. We compared the methodologies used, evidence syntheses and guideline recommendations. We also reviewed relevant process improvement (PI) methods and projects at Kaiser Permanente.

Results: Methodologies varied among SRs and guidelines, yet recommendations converged. There was good evidence for some tests and treatments, but less so for the entire integrated process. Nevertheless, positive results were demonstrated in most areas.

Discussion: Trials generally focus on specific tests or treatments rather than processes of care. Evidence to support health care processes may require chains of evidence, which are indirect by definition. Outcomes of processes of care include positive or adverse events, resource use, and operational efficiency as well as intermediate and final clinical states.

Implications for guideline developers/users: Consistency of methods would allow confident sharing of methods and guidelines, reducing resource and time use. Even when there is good evidence, how to have providers to do the right thing consistently remains a challenge.