RESUMEN
BACKGROUND: Dysphagia is common after stroke, requiring exclusive enteral feeding in 20% of patients. Recovery of oral feeding is associated with increased quality of life, better functional outcomes, and decreased mortality rates. However, evidence is heterogeneous and not conclusive on which factors are predictive of oral feeding recovery for stroke patients in rehabilitation units. AIM: To investigate predictors of complete oral feeding recovery. DESIGN: Retrospective study. SETTING: Intensive inpatient rehabilitation hospital. POPULATION: Poststroke dysphagic individuals with enteral feeding. METHODS: Retrospective chart review of demographic, clinical, rehabilitation, and swallowing factors. Univariate analysis and multivariate regression analysis were used to compare variables between the oral feeding recovery group and the enteral feeding group at discharge. RESULTS: One hundred thirty-nine patients were included in the analysis. A total of 61.9% of the sample population resumed complete oral intake at discharge. There were statistically significant differences between the 2 groups in Functional Independence Measure cognitive score, clinical swallow evaluation, and instrumental swallow evaluation at admittance, and dysphagia rehabilitation. Multiple logistic regression analysis identified the absence of aspiration signs with liquids associated with a higher probability of the resumption of complete oral feeding (odds ratio [OR] 3.57; 95% confidence interval [CI] 1.07-11.89). Age between 73 and 79 years (OR .96; 95% CI .01-.58), the presence of aspiration and/or penetration (OR .22; 95% CI .07-.72), and the presence of residue (OR .14; 95%CI .04-.43) during fiberoptic endoscopic evaluation of swallowing presented lower probability of returning to complete oral feeding. CONCLUSION: Several demographic and swallowing characteristics predicted oral feeding recovery. Absence of dysphagia signs documented on fiberoptic endoscopic evaluation of swallowing was the strongest predictor of complete oral feeding resumption.