Nutritional screening with Subjective Global Assessment predicts hospital stay in patients with digestive diseases

Nutrition. 2007 Sep;23(9):634-9. doi: 10.1016/j.nut.2007.06.005. Epub 2007 Jul 13.

Abstract

Objective: Nutritional status is an important factor that determines hospital stay, and the Subjective Global Assessment (SGA) is a candidate tool for nutritional screening on admission. However, the significance of the SGA has not been evaluated well in the ward for digestive diseases. We conducted the present study to test whether the SGA predicts hospital stay of these patients.

Methods: Two hundred sixty-two patients with digestive diseases were consecutively enrolled between July 2004 and April 2005. They consisted of 145 males and 117 females and included 110 patients with cancer. Disease category was gastrointestinal in 94, hepatic in 111, and biliary/pancreatic in 57. The SGA was performed by a certified dietician. Effects of SGA and other nutritional parameters on hospital stay were examined by simple and multiple regression analysis.

Results: Among tested variables, simple regression analysis identified the SGA, disease category, presence of malignancy, serum albumin level, percent triceps skinfold thickness, and percent arm muscle circumference as significant predictive parameters for hospital stay. Multiple regression analysis revealed that the SGA had the best predictive power, followed by the presence of malignancy and disease category.

Conclusion: The SGA is a simple and reliable predictor for hospital stay in patients with digestive diseases.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Anthropometry
  • Blood Chemical Analysis
  • Digestive System Diseases / complications*
  • Female
  • Gastrointestinal Neoplasms / complications
  • Humans
  • Length of Stay / statistics & numerical data*
  • Male
  • Malnutrition / diagnosis*
  • Malnutrition / etiology
  • Mass Screening / methods*
  • Middle Aged
  • Nutrition Assessment*
  • Nutritional Status*
  • Predictive Value of Tests
  • Regression Analysis
  • Sensitivity and Specificity