Meeting Brochure and registration form      SMDM Homepage

Tuesday, October 23, 2007 - 8:45 AM
E-2

PRIORITIZING CLINICAL PRACTICE GUIDELINES IN THE PRIMARY CARE SETTING

Mehul Dalal, MD, MSc, Yale University School of Medicine, New Haven, CT, Elizabeth Bradley, PhD, Yale University School of Public Health, New Haven, CT, and R. Scott Braithwaite, MD, MSc, Yale University / VA Connecticut Healthcare System, West Haven, CT.

Purpose: Time restrictions often force primary care providers to choose among multiple clinical practice guidelines (CPGs) during a particular visit, yet there is little guidance on how to prioritize CPGs. The purpose of our study is to estimate the magnitude of benefit that could be achieved from prioritizing CPGs. Methods: Our base case scenario is a 40 year-old obese female smoker with hypercholesterolemia and hypertension. To identify applicable CPGs we searched the United States Preventive Services Task Force recommendations and the National Guidelines Clearinghouse including recommendations with Class A, B or equivalent recommendations. For each identified CPG we searched a previously published compendium for intervention-specific LE gains. For CPGs in which LE gains were available we retrieved the source publications to stratify estimates and the associated changes in mortality by age, sex and risk factors. We then incorporated these mortality estimates in a Markov model to compare the LE gained from usual versus prioritized application of CPGs over a ten-year time horizon (until CPGs would be expected to change for our hypothetical patient). We incorporated published estimates of CPG-specific adherence to avoid overestimating benefit from CPGs. In the base case, we assumed that 1 visit would occur per year and that 2 CPGs could be addressed per visit. Results: We found 12 CPGs applicable to our base case scenario, of which 6 had available LE gain estimates. We found a 28-fold variability in the magnitude of benefit conferred [smoking cessation, 2.8 life-years (LYs); blood pressure reduction, 1.7 LYs; cholesterol reduction, 1.5 LYs; weight loss 1.1 LYs, cervical cancer screening, 0.22 LYs; breast cancer screening 0.10 LYs]. Without CPGs, our hypothetical patient lived 9.50 LYs and her expected 10-yr mortality was 11.3%. When CPGs were applied but not prioritized, our hypothetical patient lived an additional 0.13 LYs (9.63 LYs total) and her 10-year mortality declined to 7.6%. When the CPGs were prioritized, our patient accrued an additional 0.08 life-years (9.71 LYs total), and her 10-year mortality decreased to 6.6%, an absolute risk reduction of 1%. Conclusions: There is a large variability in benefits conferred by CPGs. Systematic prioritization of CPGs may confer benefits on par with commonly applied interventions.