Access to POEMs and Essential Evidence Plus will no longer be included in CMA membership as of Dec. 1, 2023.
Clinical Question
Do patients who are undergoing orthopedic surgery have better outcomes when managed with liberal transfusion policies compared with restrictive policies?
Bottom line
In patients who are undergoing orthopedic surgery, restrictive transfusion policies are associated with more frequent cardiac events than are liberal policies, regardless of whether the patients have pre-existing heart disease. The overlapping thresholds used in the various studies suggest a need for standardization and reassessment of transfusion policies. 1a
Reference
Study design: Meta-analysis (randomized controlled trials)
Funding: Unknown/not stated
Setting: Inpatient (any location)
Synopsis
We have reviewed many studies (see http://www.essentialevidenceplus.com/content/poem/160351 for examples) that report that restrictive transfusion policies (triggered by hemoglobin concentrations < 7 g/dL) result in lower mortality rates (number needed to treat [NNT] = 33). The authors of this meta-analysis, however, raise concerns that restrictive policies increase cardiovascular events, especially in patients undergoing orthopedic surgery. So, they searched PubMed, Embase, and the Cochrane Central Register of Controlled Trials to identify randomized trials that compared restrictive and liberal transfusion policies in adults who are undergoing orthopedic surgery. Two authors independently assessed which studies to include, extracted data, assessed the risk of bias of the included studies, and used discussion to resolve discrepancies. Ultimately they included 10 studies with 3968 patients. Unfortunately, the range of hemoglobin thresholds for the restrictive and liberal policies overlap (7.3 to 9.7 g/dL and 8.9 to 11.3 g/dL, respectively). Half of the studies were at low risk of bias and the remainder were unclear. Eight of the trials with 3618 patients reported a lower rate of cardiac events in liberally transfused patients than in those transfused with restrictive policies (4.5% vs 7.0%; NNT = 41; 95% CI 25 - 103). They authors report no heterogeneity among the data. Furthermore, the rates are comparable for patients with known pre-existing heart disease (4.8% vs 7.0%; NNT = 47) and those without known pre-existing heart disease (4.0% vs 7.0%; NNT = 33). Although they found no difference in the patients who were undergoing elective arthroplasty, the patients undergoing hip fracture surgery had fewer cardiac events when transfused liberally (4.6% vs 7.0%; NNT = 41; 23-165). The authors found no differences in mortality, infections, thromboembolic events, or strokes.
Reviewer
Henry C. Barry, MD, MS
Professor
Michigan State University
East Lansing, MI
Comments
These results are somewhat conflicting given overlapping “liberal” and “restrictive” transfusion targets. Nevertheless it follows that the hip fracture population would more poorly tolerate anemia given greater frailty.
This is very interesting and of specific relevance to a subset of patients, rarely if ever seen in general practice. However, there may be questions on counselling what to do and it is useful information to present alternatives. More work in this area is needed.
Well is not relevant since I am retired but I find study useless and confusing there is so many factors make the anesthesiologist call for transfusion level of hgb bp heart rate o2 saturation , these decisions are individually based.
While not relevant toy practice,it's interesting
Interesting. Opposite of GI bleed data.
good poem
One would expect a certain symmetry between the end points of mortality and cardiac events. Perhaps a univariate analysis is limited in its applicability here, given the other patient demographic and medical characteristics which could influence both end points.
The biggest hurdle will be getting the Anaesthetists to comply. They tend to be very conservative with transfusion these days.