Provider-Level Variation in Smoking Cessation Assistance Provided in the Cardiology Clinics: Insights From the NCDR PINNACLE Registry

Background-Studies show suboptimal provision of smoking cessation assistance (counseling or pharmacotherapy) for current smokers attempting to quit. We aimed to identify smoking cessation assistance patterns in US cardiology practices. Methods and Results-Among 328 749 current smokers seen between January 1, 2013, and March 31, 2016, in 348 NCDR (National Cardiovascular Data Registry) PINNACLE (Practice Innovation and Clinical Excellence)-affiliated cardiology practices, we measured the rates of cessation assistance. We used multivariable hierarchical logistic regression models to determine provider-, practice-, and patient-level predictors of cessation assistance. We measured provider variation in cessation assistance using median rate ratio (the likelihood that the same patient would receive the same assistance at by any given provider; > 1.2 suggests significant variation). Smoking cessation assistance was documented in only 34% of encounters. Despite adjustment of provider, practice, and patient characteristics, there was large provider-level variation in cessation assistance (median rate ratio, 6 [95% CI, 5.76-6.32]). Practice location in the South region (odds ratio [OR], 0.48 [0.37-0.63] versus West region) and rural or suburban location (OR, 0.92 [0.88-0.95] for rural; OR, 0.94 [0.91-0.97] for suburban versus urban) were associated with lower rates of cessation assistance. Similarly, older age (OR, 0.88 [0.88-0.89] per 10-year increase), diabetes mellitus (OR, 0.84 [0.82-0.87]), and atrial fibrillation (OR, 0.93 [0.91-0.96]) were associated with lower odds of receiving cessation assistance. Conclusions-In a large contemporary US registry, only 1 in 3 smokers presenting for a cardiology visit received smoking cessation assistance. Our findings suggest the presence of a large deficit and largely idiosyncratic provider-level variation in the provision of smoking cessation assistance.