Abstract
The study examined the effect of Magnet recognition on hospital-reported rates of smoking cessation advice (SCA) provided to inpatients with community-acquired pneumonia. Logistic and propensity score regression models estimated the Magnet effect on the SCA rate. Predictors were hospital characteristics, socioeconomic factors affecting smoking, and baseline and mid-study SCA rates as variations in hospital-improvement systems. Magnet recognized hospitals reporting higher mid-study SCA rates had higher outcome SCA rates.