Pharmacological interventions for the prevention of bleeding in people undergoing elective hip or knee surgery: a systematic review and network meta-analysis
Victoria N Gibbs,Rita Champaneria,Josie Sandercock,Nicky J Welton,Louise J Geneen,Susan J Brunskill,Carolyn Dorée,Catherine Kimber,Antony JR Palmer,Lise J Estcourt
DOI: https://doi.org/10.1002/14651858.cd013295.pub2
IF: 8.4
2024-01-18
Cochrane Database of Systematic Reviews
Abstract:Hip and knee replacement surgery is a well‐established means of improving quality of life, but is associated with a significant risk of bleeding. One‐third of people are estimated to be anaemic before hip or knee replacement surgery; coupled with the blood lost during surgery, up to 90% of individuals are anaemic postoperatively. As a result, people undergoing orthopaedic surgery receive 3.9% of all packed red blood cell transfusions in the UK. Bleeding and the need for allogeneic blood transfusions has been shown to increase the risk of surgical site infection and mortality, and is associated with an increased duration of hospital stay and costs associated with surgery. Reducing blood loss during surgery may reduce the risk of allogeneic blood transfusion, reduce costs and improve outcomes following surgery. Several pharmacological interventions are available and currently employed as part of routine clinical care. To determine the relative efficacy of pharmacological interventions for preventing blood loss in elective primary or revision hip or knee replacement, and to identify optimal administration of interventions regarding timing, dose and route, using network meta‐analysis (NMA) methodology. We searched the following databases for randomised controlled trials (RCTs) and systematic reviews, from inception to 18 October 2022: CENTRAL (the Cochrane Library), MEDLINE (Ovid), Embase (Ovid), CINAHL (EBSCO host ), Transfusion Evidence Library (Evidentia), ClinicalTrials.gov and WHO International Clinical Trials Registry Platform (ICTRP). We included RCTs of people undergoing elective hip or knee surgery only. We excluded non‐elective or emergency procedures, and studies published since 2010 that had not been prospectively registered (Cochrane Injuries policy). There were no restrictions on gender, ethnicity or age (adults only). We excluded studies that used standard of care as the comparator. Eligible interventions included: antifibrinolytics (tranexamic acid (TXA), aprotinin, epsilon‐aminocaproic acid (EACA)), desmopressin, factor VIIa and XIII, fibrinogen, fibrin sealants and non‐fibrin sealants. We performed the review according to standard Cochrane methodology. Two authors independently assessed trial eligibility and risk of bias, and extracted data. We assessed the certainty of the evidence using CINeMA. We presented direct (pairwise) results using RevMan Web and performed the NMA using BUGSnet. We were interested in the following primary outcomes: need for allogenic blood transfusion (up to 30 days) and all‐cause mortality (deaths occurring up to 30 days after the operation), and the following secondary outcomes: mean number of transfusion episodes per person (up to 30 days), re‐operation due to bleeding (within seven days), length of hospital stay and adverse events related to the intervention received. We included a total of 102 studies. Twelve studies did not report the number of included participants; the other 90 studies included 8418 participants. Trials included more women (64%) than men (36%). In the NMA for allogeneic blood transfusion, we included 47 studies (4398 participants). Most studies examined TXA (58 arms, 56%). We found that TXA, given intra‐articularly and orally at a total dose of greater than 3 g pre‐incision, intraoperatively and postoperatively, ranked the highest, with an anticipated absolute effect of 147 fewer blood transfusions per 1000 people (150 fewer to 104 fewer) (53% chance of ranking 1st) within the NMA (risk ratio (RR) 0.02, 95% credible interval (CrI) 0 to 0.31; moderate‐certainty evidence). This was followed by TXA given orally at a total dose of 3 g pre‐incision and postoperatively (RR 0.06, 95% CrI 0.00 to 1.34; low‐certainty evidence) and TXA given intravenously and orally at a total dose of greater than 3 g intraoperatively and postoperatively (RR 0.10, 95% CrI 0.02 to 0.55; low‐certainty evidence). Aprotinin (RR 0.59, 95% CrI 0.36 to 0.96; low‐certainty evidence), topical fibrin (RR 0.86, CrI 0.25 to 2.93; very low‐certainty evidence) and EACA (RR 0.60, 95% CrI 0.29 to 1.27; very low‐certainty evidence) were not shown to be as effective compared with TXA at reducing the risk of blood transfusion. We were unable to perform an NMA for our primary outcome all‐cause mortality within 30 days of surgery due to the large number of studies with zero events, or because the outcome was not reported. In the NMA for deep vein thrombosis (DVT), we included 19 studies (2395 participants). Most studies examined TXA (27 arms, 64%). No studies assessed desmopressin, EACA or topical fibrin. We found that TXA given intravenously and orally at a total dose of greater than 3 g intraoperatively and postoperatively ranked the highest, with an anticipated absolute effect of 67 fewer DVTs per 1000 people (67 fewer to 34 more) (26% chance of ranking first) within the NMA (RR 0.16, 95% CrI 0.02 to 1.43; low‐certainty e -Abstract Truncated-
medicine, general & internal