Self-Reporting of Smoking Cessation in Cardiac Patients: How Reliable Is It and Is Reliability Associated With Patient Characteristics?

Objectives: The objectives of this study were to determine the accuracy of smoking cessation self-reports by cardiac patients who participated in a smoking cessation program, and to determine which patient characteristics are associated with an inaccurate self-report during a follow-up interview 12 months after the start of the program. Methods: Smoking cessation self-reports (point prevalence abstinence) were validated against salivary cotinine levels. Using chi(2) analyses, patients who reported accurately being a nonsmoker were compared with those who reported inaccurately being a nonsmoker (biochemically verified as smokers) on factors the literature has indicated to be associated with inaccurate self-report in smoking. Potential predictors of inaccurate self-report of smoking (P = 0.20 in univariate analyses) were subsequently tested in a multivariate logistic regression analysis. Results: Of the 95 patients tested, almost 25% inaccurately reported having quit smoking at a cutoff of 10-ng/mL cotinine in saliva. The data show more underreporting of smoking among patients who received a face-to-face counseling intervention and among patients with an intermediate education level. There was significantly less underreporting among patients characterized as having a Type D personality. Conclusions: These findings suggest that underreporting of smoking status in cardiac patients who participate in a smoking cessation program is high, especially in those who receive intensive face-to-face counseling. Having a Type D personality seems to be a protective factor, whereas having an intermediate level of education is a risk factor for inaccurate reporting. Biochemical validation in high-risk populations is highly needed, ideally accompanied by alternative forms of verification. Optimal validation testing in smoking cessation studies of cardiac patients is difficult because of high refusal rates, no-show, and organizational difficulties.