The Swedish version of the patient-generated subjective global assessment of nutritional status: gastrointestinal vs urological cancers

Purpose: To translate and evaluate the Patient - Generated Subjective Globa l Assessment (PG-SGA) a method for the assessment of nutritional status. Method: Eighty-seven patients with gastrointestinal and urological tumours completed four sections and the remaining sections were independently compl eted by a doctor and a dietician. Patients were classified as SGA A (well n ourished), SGA B (moderately/suspected of being malnourished) or SGA C (sev erely malnourished). Results: Interobserver agreement was complete in 90% of the cases. More pat ients with gastrointestinal cancers than with urological cancers were class ified as SGA B and C. Mean levels of S-albumin and P-prealbumin differed be tween the SGA-classes. Multivariate logistic regression analyses showed ind ependent contributions to the overall classification by weight loss in the last 6 months, level of food intake, problems with eating, physical activit y and muscle wastage. Survival was significantly higher in SGA A than in SG A B+C, P < 0.001. Conclusion: The PG-SGA is useful for the assessment of nutritional status. Patients had no problems in answering the questions. The PG-SGA also carrie d prognostic information.