Venkat et al developed 2 models for predicting mortality following pancreaticoduodenectomy. The authors are from the Johns Hopkins University.
Patient selection: pancreaticoduodenectomy
Outcome: 30-day mortality
Parameters:
(1) age in years
(2) gender
(3) tumor diameter in cm
(4) serum albumin in g/dL
(5) type of surgery
|
Parameter |
Finding |
Points |
|
age in years |
<= 60 years |
0 |
|
|
61 to 70 years |
4 |
|
|
71 to 80 years |
6 |
|
|
>= 81 years |
7 |
|
gender |
female |
0 |
|
|
male |
1 |
|
tumor diameter |
< 3 cm |
0 |
|
|
>= 3 cm |
2 |
|
serum albumin |
>=3.5 g/dL |
0 |
|
|
2.5 to 3.4 g/dL |
2 |
|
|
< 2.5 g/dL |
3 |
|
type of surgery |
pancreaticoduodenectomy |
0 |
|
|
total pancreatectomy |
2 |
where:
• Removal of the tail only is not scored.
total score =
= SUM(points for all 5 parameters)
Interpretation:
• minimum score: 0
• maximum score: 15
• The higher the score the greater the 30-day mortality.
|
Total Score |
Percent 30-day Mortality |
|
0 to 3 |
< 0.1% |
|
4 to 9 |
0.11 to 0.48% |
|
10 |
0.7% |
|
11 |
0.9% |
|
12 |
1.2% |
|
13 |
1.6% |
|
14 |
2.2% |
|
15 |
2.9% |
Performance:
• The area under the ROC curve is 0.74.
Specialty: Gastroenterology, Surgery, general
