Predicting Mortality in Acute Cholangitis Using the Scoring System of Gigot et al.

Parameters studied with logistic regression:

(1) acute renal failure

(2) associated liver abscesses

(3) high malignant biliary stricture

(4) associated liver cirrhosis

(5) radiologic cholangitis

(6) gender

(7) age of the patient

mortality =

= (0.401 * (points for acute renal failure)) + (0.312 * (points for liver abscesses)) + (0.242 * (points for high malignant biliary stricture)) + (0.173 * (points for liver cirrhosis)) + (0.1121 * (points for radiologic cholangitis)) + (0.045 * (points for gender)) + (0.0012 * (points for age)) + 0.046

where:

• points for first 5 parameters are: 0 if absent, 1 if present

• points for gender: 0 if male, 1 if female

• points for age: 0 if < 50 years, 1 if >= 50 years

NOTE: If the score is calculated as described, then values > 1.00 can be achieved if multiple risk factors are present; the maximum score is 1.33. I also tried putting the result in the equation 1 / (1 + EXP((-1) * (mortality))), but this did not give meaningful results.

A scoring system can be derived if each regression coefficient is multiplied by 20, then rounded to an integer.

after Table 6, page 437

score =

= SUM(points for all 7 parameters)

Interpretation:

• minimum score: 0

• maximum score: 27

• The higher the score the greater the mortality rate, with scores >= 7 indicating a high risk population.