OBJECTIVES: To assess the efficacy and safety of venous thromboembolism prophylaxis in people undergoing elective total hip replacement.
METHODS: Systematic review and Bayesian network meta-analyses of randomized controlled trials were conducted for 3 outcomes: deep vein thrombosis (DVT), pulmonary embolism (PE), and major bleeding (MB). MEDLINE, EMBASE, and Cochrane Library (CENTRAL) databases were searched. Study quality was assessed using the Cochrane risk-of-bias checklist. Fixed- and random-effects models were fitted and compared. The median relative risk (RR) and odds ratio (OR) compared with no prophylaxis, with their 95% credible intervals (CrIs), rank, and probability of being the best, were calculated.
RESULTS: Forty-two (n = 24 374, 26 interventions), 30 (n = 28 842, 23 interventions), and 24 (n = 31 792, 15 interventions) randomized controlled trials were included in the DVT, PE, and MB networks, respectively. Rivaroxaban had the highest probability of being the most effective intervention for DVT (RR 0.06 [95% CrI 0.01-0.29]). Strategy of low-molecular-weight heparin followed by aspirin had the highest probability of reducing the risk of PE and MB (RR 0.0011 [95% CrI 0.00-0.096] and OR 0.37 [95% CrI 0.00-26.96], respectively). The ranking of efficacy estimates across the 3 networks, particularly PE and MB, had very wide CrIs, indicating high degree of uncertainty.
CONCLUSIONS: A strategy of low-molecular-weight heparin given for 10 days followed by aspirin for 28 days had the best benefit-risk balance, with the highest probability of being the best on the basis of the results of the PE and MB network meta-analyses. Nevertheless, there is considerable uncertainty around the median ranks of the interventions.
ESSENTIALS: Despite trial data, guidelines have not endorsed direct oral Xa inhibitors above other options. We provide profiles of venous thromboembolism and hemorrhage risk for 12 options. Direct oral Xa inhibitors had a favorable profile compared with low-molecular-weight heparin. Other options did not have favorable profiles compared with low-molecular-weight heparin.
SUMMARY:
BACKGROUND: There are numerous trials and several meta-analyses comparing venous thromboembolism (VTE) prophylaxis options after total hip and knee replacement (THR and TKR). None have included simultaneous comparison of new with older options. Objective To measure simultaneously the relative risk of VTE and hemorrhage for 12 prophylaxis options. METHODS: We abstracted VTE and hemorrhage information from randomized controlled trials published between January 1990 and June 2016 comparing 12 prophylaxis options. We then constructed networks to compute the relative risk for each option, relative to once-daily dosing with low-molecular-weight heparin (LMWH) Low. RESULTS MAIN: Relative to LMWH Low, direct oral Xa inhibitors had the lowest risk of total deep vein thrombosis (DVT)-asymptomatic and symptomatic- (odds ratio [OR], 0.45; 95% confidence interval [CI], 0.35-0.57), translating to 53-139 fewer DVTs per 1000 patients. Vitamin K antagonists (VKAs) titrated to International Normalized Ratio [INR] 2-3 predicted 56% more DVT events (OR, 1.56; 95% CI, 1.14-2.14). Aspirin performed similarly (OR, 0.80; 95% CI, 0.34-1.86), although small numbers prohibit firm conclusions. Direct oral Xa inhibitors did not lead to significantly more bleeding (OR, 1.21; 95% CI, 0.79-1.90). Secondary: Relative to LMWH Low, direct oral Xa inhibitors prevented 4-fold more symptomatic DVTs (OR, 0.25; 95% CI, 0.13-0.47). CONCLUSIONS: Relative to LMWH Low, direct oral Xa inhibitors had a more favorable profile of VTE and hemorrhage risk, whereas VKAs had a less favorable profile. The profile of other agents was not more or less favorable. Clinicians should consider these profiles when selecting prophylaxis options.
CONTEXTE: la thromboprophylaxie optimale pour les patients à risque de saignement reste incertain. Cette méta-analyse a évalué si la compression pneumatique intermittente (CPI) des membres inférieurs était efficace dans la réduction de la maladie thromboembolique veineuse et si la combinaison thromboprophylaxie pharmacologique avec l'IPC devrait améliorer son efficacité.
MÉTHODES ET RÉSULTATS: Deux examinateurs ont fouillé Medline, Embase et le registre des essais contrôlés Cochrane (1966-Février 2013) pour des essais contrôlés randomisés et évalué les résultats et la qualité des essais indépendamment. Des essais comparant l'IPC avec une thromboprophylaxie pharmacologique, bas de dissuasion thromboemboliques, aucune prophylaxie, et une combinaison de l'IPC et pharmacologiques thromboprophylaxie ont été pris en compte. Les essais qui ont utilisé CIB <24 heures ou contre différents types d'IPC ont été exclus. Un total de 16 164 patients hospitalisés provenant de 70 essais répondaient aux critères d'inclusion et ont été soumis à une méta-analyse. IPC était plus efficace que l'absence IPC prophylaxie dans la réduction de la thrombose veineuse profonde (7,3% versus 16,7%; réduction du risque absolu, 9,4%, intervalle de confiance à 95% [IC], 07.09 à 10.09; risque relatif, 0,43, IC 95%, 0,36 0,52, p <0,01; I (2) = 34%) et l'embolie pulmonaire (1,2% versus 2,8%; réduction du risque absolu de 1,6%, IC à 95% 0,9-2,3; risque relatif, 0,48, IC 95%, 0,33 0,69, p <0,01; I (2) = 0%). IPC était également plus efficace que les bas de dissuasion thromboemboliques dans la réduction de la thrombose veineuse profonde et semble être aussi efficace que la thromboprophylaxie pharmacologique, mais avec une réduction du risque de saignement (risque relatif: 0,41, IC 95%, 0,25 à ,65; P <0,01; I ( 2) = 0%). Ajout d'une thromboprophylaxie pharmacologique IPC réduit davantage le risque de thrombose veineuse profonde (risque relatif: 0,54, IC95% 0,32 à 0,91, P = 0,02, I (2) = 0%) par rapport à l'IPC seul.
CONCLUSIONS: CIB a été efficace dans la réduction de la maladie thromboembolique veineuse, et la combinaison de la thromboprophylaxie pharmacologique avec l'IPC était plus efficace que d'utiliser IPC seul.
To assess the efficacy and safety of venous thromboembolism prophylaxis in people undergoing elective total hip replacement.
METHODS:
Systematic review and Bayesian network meta-analyses of randomized controlled trials were conducted for 3 outcomes: deep vein thrombosis (DVT), pulmonary embolism (PE), and major bleeding (MB). MEDLINE, EMBASE, and Cochrane Library (CENTRAL) databases were searched. Study quality was assessed using the Cochrane risk-of-bias checklist. Fixed- and random-effects models were fitted and compared. The median relative risk (RR) and odds ratio (OR) compared with no prophylaxis, with their 95% credible intervals (CrIs), rank, and probability of being the best, were calculated.
RESULTS:
Forty-two (n = 24 374, 26 interventions), 30 (n = 28 842, 23 interventions), and 24 (n = 31 792, 15 interventions) randomized controlled trials were included in the DVT, PE, and MB networks, respectively. Rivaroxaban had the highest probability of being the most effective intervention for DVT (RR 0.06 [95% CrI 0.01-0.29]). Strategy of low-molecular-weight heparin followed by aspirin had the highest probability of reducing the risk of PE and MB (RR 0.0011 [95% CrI 0.00-0.096] and OR 0.37 [95% CrI 0.00-26.96], respectively). The ranking of efficacy estimates across the 3 networks, particularly PE and MB, had very wide CrIs, indicating high degree of uncertainty.
CONCLUSIONS:
A strategy of low-molecular-weight heparin given for 10 days followed by aspirin for 28 days had the best benefit-risk balance, with the highest probability of being the best on the basis of the results of the PE and MB network meta-analyses. Nevertheless, there is considerable uncertainty around the median ranks of the interventions.
