Sima et al developed several methods of predicting the risk of perioperative transfusion of a patient undergoing elective hepatectomy. One of these involved a simple score. The authors are from Memorial Sloan-Kettering Cancer Center in New York City.
Patient selection: elective hepatectomy
Outcome perioperative red blood cell transfusion
Parameters:
(1) number of segments resected
(2) diagnosis
(3) resection of extrahepatic organ at the surgery
(4) preoperative hemoglobin (in g/dL) and gender
(5) preoperative platelet count per µL
Parameter |
Finding |
Points |
number of segments resected |
0 or 1 |
0 |
|
2 or 3 |
1 |
|
4 to 6 |
4 |
diagnosis |
benign |
0 |
|
primary liver malignancy |
1 |
|
metastasis |
0 |
extrahepatic organ resected |
no |
0 |
|
yes |
1 |
hemoglobin and gender |
female and hemoglobin >= 12 |
0 |
|
female and hemoglobin < 12 |
2 |
|
male and hemoglobin >= 14 |
0 |
|
male and hemoglobin < 14 |
2 |
platelet count |
>= 125,000 per µL |
0 |
|
< 125,000 per µL |
1 |
total score =
= SUM(points for all 5 parameters)
Interpretation:
• minimum score: 0
• maximum score: 9
• The higher the score the more likely that the patient will require blood transfusion.
Score |
Percent Transfused |
0 or 1 |
16% |
2 or 3 |
29-34% |
4 to 6 |
50% |
7 or 8 |
64% |
9 |
> 64% |
Performance:
• The tabular data in Figure 5 may be misleading. The data recorded for specificity corresponds to (1 – specificity) on the ROC curve.
• The sensitivity, specificity and ROC curve indicate mediocre performance. Likelihood can be condensed to 4 risk categories.
Specialty: Clinical Laboratory, Surgery, general, Anesthesiology, Emergency Medicine, Critical Care