An evaluation of the validity of nutrition screening and assessment tools in patients admitted to a vascular surgery unit.

Vascular surgery patients are nutritionally vulnerable. Various malnutrition screening and assessment tools are available; however, none has been developed or validated in vascular patients. The present study aimed to: (1) investigate the validity of four commonly administered malnutrition screening tools (Malnutrition Screening Tool (MST), Malnutrition Universal Screening Tool (MUST), Nutrition Risk Screen-2002 (NRS-2002) and the Mini-Nutritional Assessment - Short Form (MNA-SF) and an assessment tool (the Patient-Generated Subjective Global Assessment (PG-SGA)) compared against a comprehensive dietitian's assessment and (2) evaluate the ability of the instruments to predict outcomes. Vascular inpatients were screened using the four malnutrition screening tools and assessed using the PG-SGA. Each was assessed by a dietitian incorporating nutritional biochemistry, anthropometry and changes in dietary intake. Diagnostic accuracy, consistency and predictive ability were determined. A total of 322 (69·3 % male) patients participated, with 75 % having at least one parameter indicating nutritional deficits. No instrument achieved the a priori levels for sensitivity (14·9-52·5 %). Neither tool predicted EuroQoL 5-dimension 5-level score. All tools except the MNA-SF were associated with length of stay (LOS); however, the direction varied with increased risk of malnutrition on the MUST and NRS-2002 being associated with shorter LOS (P=0·029 and 0·045) and the reverse with the MST and PG-SGA (P=0·005 and <0·001). The NRS-2002 was associated with increased risk of complications (P=0·039). The MST, NRS-2002 and PG-SGA were predictive of discharge to an institution (P=0·004, 0·005 and 0·003). The tools studied were unable to identify the high prevalence of undernutrition; hence, vascular disease-specific screening and/or assessment tools are warranted.