The expanding use of Electronic Medical Records (EMR) makes it possible to 1) examine practice variationin applying standard guidelines, 2) to precisely determine areas of high and low performance, 3) todetermine the relation of practice variation to differences in medical outcomes and utilization, and 4) to feedthis information back to improve the quality of care. This study will use the Prevention Index (PI)methodology and the Cancer Research Network (CRN) Virtual Data Warehouse (VDW) to develop and applya set of Cancer Prevention Indices (CPI) to assess the quality of primary, secondary, and tertiary preventivecare for cancer. These CPI measures determine the variations in application by clinicians and healthsystems of evidence-based guidelines. The indices are based on EMR data and include quality measuresfor individual services, for groups of services (subscales) and a combined score that summarizes the overallquality of care in participating health care systems. We will: 1) identify retrospectively the variation in CPIscores across clinics and clinical practices and determine the association of these variations to selectedevent rates several years later. These event rates will include primary prevention (smoking and obesityrates), secondary prevention events (stage at diagnosis of new cancers) and tertiary prevention events(stage at diagnosis of seco'nd primary cancers and recurrences) if numbers permit. The epidemiologicalcomponent of the study uses a retrospective-prospective design to determine the association of clinicianadherence to guidelines to subsequent events among their patients. The intervention component of thestudy will analyze performance scores for primary care practices in each health care system and providepersonalized feedback on performance variation at the system, clinic, and clinician levels to system andclinic managers. The emphasis will be on identifying system problems that can be remedied easily and atreasonable cost. Using a randomized clinical trial design, we will then determine if such feedback leads tochanges in care delivery and reductions in performance problems. Finally, we will provide assistance andconsultation to systems that would like to use this methodology to monitor future quality of care. Twelvemonths after the end of the intervention trial, we will repeat the analyses to determine the persistence ofintervention effects in the involved systems and assess the degree to which the methods have beenadopted.
Showing the most recent 10 out of 188 publications
