RESUMO
BACKGROUND: Smoking is a poor prognostic factor for healing after rotator cuff repair and is associated with inferior results. We hypothesized that smokers would have higher recurrent tear rates and more postoperative myotendinous junction (MTJ) retraction in healed repairs than nonsmokers three months postoperatively. METHODS: Rotator cuff repairs (RCRs) were retrospectively reviewed over a 2-year period. Patients underwent magnetic resonance imaging (MRI) within 6 months prior to surgery and again at 3 months postoperatively. Seventy-nine patients were included and stratified by smokers versus nonsmokers. Baseline patient demographics, tear characteristics, and surgical factors were collected. Preoperative and postoperative MRIs were assessed to quantify the MTJ position and to establish the recurrent tear rate. RESULTS: For the total cohort (nonsmokers, n = 56; smokers, n = 23), significant differences in age, race, and traumatic onset of injury existed between groups. There were no significant differences in recurrent tear between smokers (26%) and nonsmokers (27%), but nonsmokers were more satisfied. For patients with healed RCRs (nonsmokers, n = 41; smokers, n = 17), there were significant differences in race. On univariate analysis, nonsmokers had a significantly more lateral MTJ postoperatively (P = 0.05). On multivariable regression analysis, medialized postoperative MTJ position in healed cuffs was driven only by greater preoperative rotator cuff retraction preoperatively. There were no significant differences in MTJ position based on smoking status for patients with healed RCRs. CONCLUSION: Smoking does not appear to be an independent risk factor for postoperative MTJ retraction in healed RCRs, also known as failure in continuity. Preoperative tear size and retraction play the biggest roles in predicting postoperative MTJ position, regardless of smoking status. There are no significant differences in patient-reported outcomes for patients with healed RCRs, but nonsmokers had more satisfaction following RCR in the total cohort. LEVEL OF EVIDENCE: Level III; Retrospective cohort study; Diagnostic study.