Predictive Factors for Preoperative Percutaneous Endoscopic Gastrostomy Placement

Autor: Munique Maia, Ashley R. Chandler, Jason Weissler, Raman R. Sharma, Armen K. Kasabian, Katie E. Weichman, Neil Tanna, Douglas Frank, Denis Knobel, Benjamin D. Smith
Rok vydání: 2015
Předmět:
Zdroj: Journal of Craniofacial Surgery. 26:2124-2127
ISSN: 1049-2275
DOI: 10.1097/scs.0000000000002132
Popis: OBJECTIVE The treatment of head and neck cancer has varying impact on postoperative recovery and return of swallowing function. The authors aim to establish screening tools to assist in preoperatively determining the need for gastrostomy tube placement. METHODS The authors prospectively assessed all patients undergoing complex head and neck reconstructive surgery during a 1-year study period. Only patients tolerating an oral diet, without preoperative gastrostomies, were enrolled for study. Eight parameters were assessed including: body mass index (BMI), prealbumin, albumin, smoking history, comorbidities [including coronary artery disease (CAD), chronic obstructive pulmonary disease (COPD), and diabetes mellitus (DM)], age, use of microvascular reconstruction, and type of defect. Two specific screening tools were assessed. In the first, a multivariate logistic regression model was employed to determine factor(s) that predict postoperative gastrostomy tube. In a second screening tool, the 8 parameters were scored between 0 to 1 points. The total score obtained for each patient was correlated with postoperative gastrostomy placement. RESULTS Out of the 60 study patients enrolled in the study, 24 patients (40%) received a postoperative gastrostomy. In the logistic regression model, albumin level was the only factor that was significantly associated with need for postoperative gastrostomy (P < 0.0023). A score of 4 or greater was determined to have a sensitivity of 83% and specificity of 61% for postoperative gastrostomy. CONCLUSIONS Patients with a score of 4 or more with this screening scoring system or those patients with an albumin level
Databáze: OpenAIRE