Characteristics of some prediction rules that stratify patients who are undergoing cardiopulmonary bypass surgery for risk of major bleeding9
| Characteristics . | Biancari et al7 . | Klein et al47 . | Ranucci et al48 . | Alghamdi et al49 . | Goudie et al50 . | Karkouti et al51 . |
|---|---|---|---|---|---|---|
| Outcome predicted | >4 units of RBC | ≥1 unit RBC | ≥1 unit RBC | ≥1 unit RBC | ≥1 unit RBC;≥4 units RBC | ≥5 units RBC |
| Sample size | N = 3744 | N = 20 036 | N = 8989 | N = 11 113 | N = 33 960 for any transfusion; N = 27 353 for ≥4 units RBC | N = 10 667 |
| Origin of dataset | 16 cardiac surgery centers in 6 European countries | 10 cardiac surgery centers in the United Kingdom | 1 Italian institution | 1 Canadian institution | 3 United Kingdom and 1 Italian institutions | 1 Canadian institution |
| c Statistic (95% CI where provided) | 0.72 (0.69‐0.76) | 0.76 (0.75‐0.77) | 0.73 (0.72‐0.74) | 0.79 (0.78‐0.80) | 0.78 (0.77‐0.78); 0.81 (0.80‐0.82) | 0.88 |
| Hosmer and Lemeshow P value | .033 | .28 | .372 | .70 | .001; .32 | .20 |
| Variables scored | Sex, preoperative anemia, eGFR, antiplatelet drugs discontinued <5 d, critical preoperative state, acute coronary syndrome, use of low molecular weight heparin/fondaparinus/unfractionated heparin | Age, body surface area, sex, hemoglobin, type of surgery, creatinine, log Euroscore | Age, weight, sex, hemoglobin, type of surgery | Age, weight, sex, hemoglobin, type of surgery, creatinine, elective status, previous cardiac surgery | Age, weight, sex, creatinine, preoperative intra-aortic balloon pump, various comorbidities, previous cardiac surgery, various surgical procedure characteristics, elective status, presence of cardiogenic shock | Age, body surface area, hemoglobin, platelet count, preoperative shock, complex procedure, high blood loss, surgeon, redo surgery, elective status, duration of circulatory arrest and CPB, nadir CPB hemoglobin |
| Characteristics . | Biancari et al7 . | Klein et al47 . | Ranucci et al48 . | Alghamdi et al49 . | Goudie et al50 . | Karkouti et al51 . |
|---|---|---|---|---|---|---|
| Outcome predicted | >4 units of RBC | ≥1 unit RBC | ≥1 unit RBC | ≥1 unit RBC | ≥1 unit RBC;≥4 units RBC | ≥5 units RBC |
| Sample size | N = 3744 | N = 20 036 | N = 8989 | N = 11 113 | N = 33 960 for any transfusion; N = 27 353 for ≥4 units RBC | N = 10 667 |
| Origin of dataset | 16 cardiac surgery centers in 6 European countries | 10 cardiac surgery centers in the United Kingdom | 1 Italian institution | 1 Canadian institution | 3 United Kingdom and 1 Italian institutions | 1 Canadian institution |
| c Statistic (95% CI where provided) | 0.72 (0.69‐0.76) | 0.76 (0.75‐0.77) | 0.73 (0.72‐0.74) | 0.79 (0.78‐0.80) | 0.78 (0.77‐0.78); 0.81 (0.80‐0.82) | 0.88 |
| Hosmer and Lemeshow P value | .033 | .28 | .372 | .70 | .001; .32 | .20 |
| Variables scored | Sex, preoperative anemia, eGFR, antiplatelet drugs discontinued <5 d, critical preoperative state, acute coronary syndrome, use of low molecular weight heparin/fondaparinus/unfractionated heparin | Age, body surface area, sex, hemoglobin, type of surgery, creatinine, log Euroscore | Age, weight, sex, hemoglobin, type of surgery | Age, weight, sex, hemoglobin, type of surgery, creatinine, elective status, previous cardiac surgery | Age, weight, sex, creatinine, preoperative intra-aortic balloon pump, various comorbidities, previous cardiac surgery, various surgical procedure characteristics, elective status, presence of cardiogenic shock | Age, body surface area, hemoglobin, platelet count, preoperative shock, complex procedure, high blood loss, surgeon, redo surgery, elective status, duration of circulatory arrest and CPB, nadir CPB hemoglobin |
The c statistic is a measure of model discrimination, and it is also referred to as the area under the receiver operating characteristic curve. A nonsignificant P value for the Hosmer and Lemeshow test indicates adequate model calibration. RBC, red blood cell; eGFR, estimated glomerular filtration rate; CPB, cardiopulmonary bypass.