Ahmadzia et al reported 2 models for predicting peripartum blood transfusion in a woman undergoing Cesarean delivery. Model 2 relies on antepartum and intrapartum factors. The authors are from George Washington University and Duke University.
Patient selection: woman prior to Cesarean delivery
Parameters:
(1) age in years
(2) body mass index at delivery in kg per square mter
(3) number of previous term deliveries
(4) health insurance
(5) platelet count per microliter
(6) hematocrit in percent
(7) number of previous Cesarean deliveries
(8) history of heart disease
(9) gestational hypertension or preeclampsia
(10) abruptionat time of delivery
(11) nonelective repeat cesarean delivery
(12) general anesthesia
(13) multiple births (twins, triplets, etc)
(14) failure to progress
(15) eclampsia/HELLP
(16) placenta accreta, increta, or percreta
(17) intra-partum antibiotic use
|
Parameter |
Finding |
Points |
|
age in years |
< 21 |
0.1583 |
|
|
21 to 36 |
0 |
|
|
> 36 years |
0.1583 |
|
BMI |
|
-0.0153 * (BMI) |
|
number of previous term deliveries |
< 3 |
0 |
|
|
>= 3 |
0.6304 |
|
health insurance |
private |
0 |
|
|
government |
0.2733 |
|
|
none |
0.2258 |
|
platelet count |
< 150,000 per µL |
0.3706 |
|
|
>= 150,000 per µL |
0 |
|
hematocrit |
< 32 percent |
1.3111 |
|
|
>= 32 percent |
0 |
|
previous Cesareans |
0 |
0.5894 (first CD) |
|
|
1 |
0 |
|
|
2 |
0.4069 |
|
|
>= 3 |
0.18 |
|
history of heart disease |
no |
0 |
|
|
yes |
0.7524 |
|
gestational hypertension or preeclampsia |
no |
0 |
|
|
yes |
0.3479 |
|
abruption at time of delivery |
no |
0 |
|
|
yes |
0.3313 |
|
nonelective repeat cesarean |
no |
0 |
|
|
yes |
0.5607 |
|
general anesthesia |
no |
0 |
|
|
yes |
1.6439 |
|
multiple births |
no |
0 |
|
|
yes |
0.5518 |
|
failure to progress |
no |
0 |
|
|
yes |
0.5205 |
|
eclampsia or HELLP |
no |
0 |
|
|
yes |
1.0587 |
|
placentra accreta, increta, percreta |
no |
0 |
|
|
yes |
4.5154 |
|
intra-partum antibiotic use |
no |
0 |
|
|
yes |
0.2802 |
value of X =
= SUM(points for all of the parameters) - 5.2402
probability of transfusion =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under ROC curve was 0.83.
Specialty: Clinical Laboratory
