Predicting transfusion in shoulder arthroplasty.

Abstract

This study was conducted to evaluate the incidence of transfusion in shoulder arthroplasty, determine clinical factors associated with increased risk for transfusion, and develop an algorithm to assist the surgeon in preoperative planning with regards to blood management. The study had 2 phases: (1) development of a clinical prediction rule for transfusion using 280 procedures and (2) a validation study of the algorithm applied to 109 new patients. Phase 1 consisted of a retrospective record review of 280 consecutive shoulder arthroplasties to determine risk factors for transfusion. Phase 1 also identified a preoperative hemoglobin level of less than 12.5 g/dL as predictive of the need for blood transfusion. This cutoff was prospectively applied to 109 patients undergoing shoulder arthroplasty in phase 2. The transfusion rate for phase 1 was 19.6%. Preoperative hemoglobin level (P < .001), age (P= .003), and the number of comorbid conditions (P = .005) were statistically significant risk factors. Patients with a preoperative hemoglobin level of less than 12.5 g/dL have a 4-fold increased risk of requiring a blood transfusion. In phase 2, the cutoff of less than 12.5 g/dL yielded a sensitivity of 88%, specificity of 78%, and positive and negative likelihood ratios of 4.0 and 0.15, respectively. Preoperative hemoglobin level, age, and number of comorbid conditions are all predictive of transfusion in shoulder arthroplasty. Tailoring blood ordering based on a preoperative hemoglobin level of 12.5 g/dL is safe and effective. Prognostic study, level 2.