Background:

Smoking is a serious and common health risk among veterans. Given the press of national initiatives and local incentives to improve smoking cessation care in response to VA performance measures, this study tests a widely applicable approach to clinical practice guidelines implementation, namely evidence-based quality improvement, which is directly relevant to the translation of efficacious treatments into enhancements in VA health care policy and practice. Evidence-Based Quality Improvement (EBQI) focuses on improved provider adherence to smoking cessation guidelines and a decrease in patient smoking rates in a manner designed to produce short- and long-term health improvements and cost benefits at the organizational level.

Objectives:

Adherence to smoking cessation guidelines requires practice changes at the patient, provider, and system levels to achieve optimal quit rates. The objective of this study was to evaluate the effectiveness of evidence-based quality improvement (EBQI)�an expert-designed and locally implemented clinical reorganization of smoking cessation care�on changes in smoking cessation (SC) practice among primary care providers and health outcomes among veteran smokers.

Methods:

An evidence-based quality improvement intervention comprising provision of physician and patient educational materials, local priority setting with leadership and providers, and local adaptation of expert-designed protocols was implemented in experimental VA primary care practices (n=9). VA control sites (n=9), matched on size and academic affiliation, received smoking cessation guideline copies. We randomly sampled, consented, screened and surveyed primary care patients at all 18 sites (n=1,941 smokers) and used computer-assisted telephone interviewing to assess sociodemographics, health status, function, and smoking behavior, attitudes and treatment experience. Post-intervention 12-month follow-up interviews were completed using the same measures (n=1,080). We used multiple imputation using hotdeck techniques and applied both enrollment and attrition weights to the patient-level data. We used weighted logistic regression to evaluate intervention effects, controlling for patient-level predictors of quit attempts and quit status (e.g., level of addiction, readiness to change, age, health).

Status:

The project is completed. Analysis is ongoing for manuscripts.