Meister et al developed several models for predicting risk of perineal complications for women having a vaginal delivery. These models can help to identify a patient at increased risk for a laceration or anal sphincter injury. The authors are from Washington University in St. Louis.
Patient selection: woman delivering vaginally
Outcomes: any perineal laceration and anal sphincter injury
Parameters:
(1) race
(2) operative assistance to vaginal delivery
(3) fetal body weight in grams
(4) parity
(5) tobacco use
|
Parameter
|
Finding
|
Points 1
|
Points 2
|
|
race
|
African American
|
-0.78
|
-0.99
|
|
|
other
|
0
|
0
|
|
operative delivery
|
yes
|
0.56
|
1.46
|
|
|
no
|
0
|
0
|
|
fetal weight
|
< 3,500 g
|
0
|
0
|
|
|
>= 3,500 g
|
0.61
|
0.2
|
|
parity
|
nulli
|
1.54
|
1.56
|
|
|
multi
|
0
|
0
|
|
tobacco use
|
no
|
0
|
0
|
|
|
yes
|
-0.18
|
-1.06
|
value of X =
= SUM(points for all 5 parameters in points 1 column) + 0.5
probability of any perineal laceration =
= 1 / (1 + EXP((-1) * X))
value of Y =
= SUM(points for all 5 parameters in points 2 column) - 3.68
probability of anal sphincter injury =
= 1 / (1 + EXP((-1) * Y))
Performance:
• The area under the ROC curve for any perineal laceration is 0.70.
• The area under the ROC curve for anal sphincter injury is 0.83.
