Laryngectomy and smoking: An analysis of postoperative risk

To investigate the impact of smoking on complication rates following total laryngectomy.

The National Surgical Quality Improvement Program database was queried for all total laryngectomies between 2005 and 2014. Patients were identified as smokers (n = 561) or nonsmokers (n = 513) and compared via univariate and multivariate analyses. A nearest-neighbor propensity score-generating algorithm was used to build a subpopulation (n = 714) of matched cases and evaluated in a similar manner. Additionally, pack-year data was available for select cases and analyzed appropriately.

On multivariate analysis of the unmatched cohort accounting for demographics and confounders, no significant difference in overall medical complications was identified between groups (odds ratio = 0.799, P = 0.495). Propensity matching corrected for all significantly distributed comorbidities, except for alcohol, which remained associated with the smoking group (P = < 0.001). In the matched population, there were no significant differences in complication rates between the two groups. Pack-year data was available for 340 patients. These cases were subdivided into cohorts with < 50 (n = 204) and 51+ (n = 136) pack-years. Postoperative pneumonia, ventilation for more than 48 hours, sepsis, and overall medical complications were associated with > 50 pack-years of smoking. After multivariate regression, sepsis and overall medical complications remained significant for the 51+ pack-year smoking cohort.

After accounting for confounding comorbidities, smoking is found to play an insignificant role in the development of postoperative complications following total laryngectomy. However, those with 51+ pack-years are at an increased risk for postoperative sepsis and overall medical complications following these complex procedures.

4. Laryngoscope, 127:2302-2309, 2017.