Conference Coverage

PBM saves blood, costs in HSCT unit



BOSTON – Implementing a patient blood management (PBM) program for patients undergoing hematopoietic stem cell transplantation (HSCT) resulted in significant reductions in blood product use with an attendant reduction in costs, but without negative effects on patient-centered outcomes, investigators reported.

Dr. Nilesh Jambhekar of the Mayo Clinic, Rochester, Minn. Neil Osterweil/MDedge News

Dr. Nilesh Jambhekar

Since the PBM program began, the number of transfusions and the units transfused declined without affecting mortality, ICU admission rates, or other transfusion-related complications. The program saved the hospital more than $600,000 over 1 year, reported Nilesh Jambhekar, MD, an anesthesiology resident at the Mayo Clinic in Rochester, Minn.

“In general, PBM implementation is probably helpful in reducing both platelet and [packed red blood cell] utilization, but it’s not an easy thing to do. It requires institutional buy-in and key players to make it happen,” he said at AABB 2018, the annual meeting of the group formerly known as the American Association of Blood Banks.

“Ongoing PBM-related activities [such as] surveillance, education, and clinical decision feedback are critical to maintaining success that we’ve had,” he added.

The investigators looked at blood-product use both before and after the Mayo Clinic started a PBM program that included emphasis on AABB best practice guidelines and electronic clinical decision support for transfusion orders.

They analyzed the frequency and proportion of red blood cell (RBC) and platelet transfusions, total transfusion quantities, transfusions that occurred outside of the clinical guidelines, and the activity-based costs of transfusions.

Dr. Jambhekar acknowledged that the study relied on rigid hemoglobin and platelet thresholds when considering transfusions conducted outside of the guidelines, which they defined as RBCs administered for hemoglobin values greater than 7 g/dL and platelet transfusions for platelets counts greater than 10 x 109/L. He noted, however, that they conducted sensitivity analyses to account for exceptions, such as patients with coronary disease or neutropenic fever.

The patient-centered outcomes they evaluated included mortality, hospital and ICU admission rates, transfusion reactions, cerebrovascular and coronary ischemic events, and infections.

The study included data on 360 adults who underwent HSCT in 2013, before the PBM program was implemented, and 368 transplanted in 2015, after implementation. In each cohort, patients were followed out to 90 days after transplant.

The investigators found that the total number of units transfused dropped from 1,660 units of platelets and 1,158 U of RBCs before implementation, to 1,417 U and 826 U, respectively, after PBM implementation.

Significantly, in addition to an overall reduction in units transfused, the investigators saw substantial changes in the proportions of inappropriate (outside guidelines) transfusions of red blood cells between the two time periods, with 94.2% of RBC transfusions occurring outside the guidelines in 2013, compared with 35.4% in 2015 (P less than .0001). Similarly, the proportion of inappropriate platelet transfusions declined from 73.4% to 48.7% over the same time period (P less than .0001).

Also of note was the fact that all-cause mortality at 3 months was significantly lower after the PBM program was introduced. The 3-month mortality rate for the 2013 cohort was 30.7%, compared with 20.2% for the 2015 cohort (P = .001). Neither hospital or ICU admission with 30 days or hospital or ICU lengths of stay differed significantly between the groups.

Dr. Jambhekar noted that in a multivariable analysis accounting for baseline differences between the groups as a possible explanation for the higher mortality in the 2013 cohort, mortality for patients treated before the PBM program remained significantly higher, with an odds ratio of 1.85 (P = .0008).

There were also no significant differences in either MIs, cerebrovascular events, sepsis, and febrile or allergic transfusion reactions.

He noted that in addition to the rigid thresholds used, the study was retrospective in design – and therefore could not fully account for potential confounders – and that it is unclear whether the results could be generalized for adoption by other institutions.

The study was internally funded. Dr. Jambhekar reported having no financial disclosures.

SOURCE: Jambhekar N et al. AABB 2018, Abstract PBM3-ST4-22.

Next Article: