Transfusion rates for patients undergoing PCI vary widely in the U.S., according to a study published Feb. 26 in JAMA. The variation may be driven more by local practice patterns than clinical need.
The decision to use or not use transfusions in PCI procedures poses a potential quandary for interventional cardiologists. Some observational studies have associated transfusions with worse clinical outcomes, and clinical trial results showing benefits or risks are limited. That paucity of high-level evidence is reflected in guidelines, which “make cautious recommendations for restricted transfusion strategies in hospitalized patients with a history of coronary artery disease and make no recommendation on transfusion in the setting of ACS [acute coronary syndrome],” wrote the authors, led by Matthew W. Sherwood, MD, of the Duke Clinical Research Institute in Durham, N.C.
Sherwood and colleagues used the CathPCI Registry to identify 2,258,711 PCI patient visits at 1,431 hospitals between July 2009 and March 2013. Their primary outcome was transfusion in the patient population by looking at the rate of transfusions overall and by hospital. They also examined outcomes such as in-hospital MI, congestive heart failure and death.
They calculated rates by quarter, by hospital site, by patients with or without bleeding complications and divided hospitals into low-, medium- or high-transfusing groups.
Sherwood et al found a total of 48,430 patient visits in which there was a transfusion, for an overall rate of 2.14 percent. The transfusion rate declined over time, from 2.11 percent to 2.04 percent by the last quarter. Patients who were older, female or had comorbidities were more likely to receive a transfusion.
Transfusion use varied greatly, with 25.5 percent of hospitals giving transfusions to less than 1 percent of patients and 3.7 percent of hospitals providing transfusions to more than 5 percent of patients. The risk-standardized rates of transfusion ranged from 0.3 percent to 9.3 percent, with a median of 2.5 percent.
Hospitals in the high-transfusing group provided transfusions to all hemoglobin values more frequently than the medium and low groups. They had a higher transfusion threshold (between 9 g/dL and 10 g/dL) than the low group (between 8 g/dL and 9 g/dL). The authors argued that the median odds ratio for the likelihood of a transfusion (1.85) and between hospital variation (0.42) showed that the hospital was the driver behind “a significant amount of the variation seen in transfusion rates.”
Patients given transfusions were more likely to have an in-hospital MI, stroke, congestive heart failure, cardiogenic shock or die. Transfusions were associated with an increased risk of in-hospital MI, stroke and death after adjustments, even in analyses stratifying patients by bleeding events and preprocedure hemoglobin values. The one exception involved patients with bleeding and hemoglobin values of less than 10 g/dL, who saw a decreased risk with transfusion use.
Their findings indicate that blood transfusion practices vary widely across the nation, Sherwood and colleagues proposed, even when adjusting for patient differences.
“Within this variation there appeared to be patients who underwent transfusion in the absence of clinical bleeding events and patients who underwent transfusion with nearly normal postprocedure hemoglobin values. These patient-level data, as well as our finding that transfusions were more common across all hemoglobin values at some hospitals, suggest that thresholds for transfusion may have been driven more by local practice patterns than by clinical necessity,” they wrote.
They suggested that the variation also may be due to physician beliefs in the benefit of transfusions, and that the declining transfusion rate over time may reflect recent studies challenging that belief. Given the lack of definitive evidence, they recommended that an adequately powered randomized clinical trial of transfusion strategies be conducted to better inform guidelines and physician practice.