Woo et al developed a discriminant score based on biochemical measures of malnutrition and diabetes to predict the risk of short term mortality for an elderly patient in a chronic care facility. This can help identify patients who may benefit from more aggressive management and closer monitoring. The authors are from the Chinese University of Hong Kong.

Patient selection: elderly male and female patients in Hong Kong, so probably mostly Chinese



(1) fructosamine in mmol/L (affected by both protein malnutrition and hyperglycemia)

(2) transferrin in g/L (marker of protein calorie malnutrition)

(3) prealbumin in mg/L (marker of protein calorie malnutrition)

(4) glycosylated hemoglobin in percent from 0 to 100

(5) hemoglobin in g/dL (marker of protein calorie malnutrition)


discriminant score =

= (9.2 * (fructosamine)) + (2.38 * (transferrin)) + (0.08 * (prealbumin)) - (1.309 * (glycosylated hemoglobin)) - (0.2 * (hemoglobin)) - 6.029



• A score of 0 indicates a 50:50 chance of survival.

• A negative score indicates an increased risk of short term mortality.

• A positive score indicates an increased chance of short term survival.


Discriminant Score

Outcome at 3 Months

- 5.0

all died


most died, some survived


half lived, half died


most survived, a few died


all survived

after Table 4, page 1244


Analytical methods:

• Most of the biochemical tests were performed on a Parallel Analyzer from American Monitor Corporation.

• The reference ranges for the different tests were not reported.

• The methods for glycosylated hemoglobin and fructosamine are described in Woo et al. Diabetologia. 1987; 30: 863-868.

To read more or access our algorithms and calculators, please log in or register.