Association of Smoking Status With Mortality and Hospitalization in Hemodialysis Patients

Rationale & Objective: The relationship between tobacco smoking and comorbid condition outcomes in hemodialysis (HD) patients is not well understood. This study examined the association of tobacco smoking status with hospitalization and mortality in HD patients. Study Design: Retrospective cohort study. Setting & Participants: Adult HD patients at 2,223 US dialysis centers with HD vintage of 30 days or less who completed a tobacco smoking status survey as part of standard care between April 2013 and June 2015. Predictor: Tobacco smoking category: never smoked, currently living with smoker, former smoker, moderate smoker (<1 pack per day), or heavy smoker (>= 1 pack per day). Outcomes: Death and hospital admissions within 2 years of the tobacco smoking survey. Analytical Approach: Kaplan-Meier analysis and Cox proportional hazards regression for time to death; cumulative incidence function and Cox proportional hazards regression for time to first hospitalization; negative-binomial regression for number of hospitalizations. Results: Of 22,230 patients studied, 13% were active smokers. Mortality probabilities increased with greater exposure to smoking (17%, 22%, 23%, and 27% for never, moderate, former, and heavy smokers, respectively; P < 0.001), as did incidence rates for first hospitalization (23%, 27% 27% and 30%, respectively; P < 0.001). Compared to never smoked, heavy smokers had the highest mortality rate (HR for heavy smokers, 1.41 [95% CI, 1.18-1.69]; HR for moderate smokers, 1.39 [95% CI, 1.24-1.55]; HR for former smokers, 1.19 [95% CI, 1.11-1.28]). Living with a smoker was not associated with mortality (HR, 0.93; 95% CI, 0.72-1.22). HRs for first hospitalization followed similar patterns. The incidence rate of mortality for active smokers with diabetes was 173.7/1,000 patient-years and 103.5/1,000 patient-years for those who never smoked (incidence rate ratio, 1.68; P < 0.001). Limitations: Self-reported survey without detailed history of smoking/cessation. Conclusions: Risks for death and hospitalization are elevated among HD patients who smoke, being highest among younger individuals and those with diabetes. Second-hand smoke was not associated with poor clinical outcomes.