Thromboprophylaxis with apixaban in patients undergoing major orthopedic surgery : meta-analysis and trial-sequential analysis

h ps://doi.o g/10.1177/1179545X17704660 C ea i e Commons Non Comme cial CC BY-NC: This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion-NonComme cial 4.0 License (h p://www.c ea i ecommons.o g/licenses/by-nc/4.0/) which pe mi s non-comme cial use, ep oduc ion and dis ibu ion o he wo k wi hou u he pe mission p o ided he o iginal wo k is a ibu ed as speci ied on he SAGE and Open Access pages (h ps://us.sagepub.com/en-us/nam/open-access-a -sage). Clinical Medicine Insigh s: Blood Diso de s Volume 10: 1–8 © The Au ho (s) 2017 Rep in s and pe missions: sagepub.co.uk/jou nalsPe missions.na DOI: 10.1177/1179545X17704660 In oduc ion Venous h omboembolic e en s (VTEs), such as deep enous h ombolysis and pulmona y embolism (PE), a e po en ially a al complica ions o o hopedic su ge y, whose incidence may each 30% o he pa ien s unde going knee o hip eplace- men .1 Fu he mo e, non a al e en s a e equen and a e asso- cia ed wi h signi ican mo bidi y. The de elopmen and clinical Th ombop ophylaxis Wi h Apixaban in Pa ien s Unde going Majo O hopedic Su ge y: Me a-Analysis and T ial-Sequen ial Analysis Daniel Caldei a1,2,3, Filipe B Rod igues1,3,4, Faus o J Pin o5, Joaquim J Fe ei a1,3 and João Cos a1,3,6,7 1Labo a ó io de Fa macologia Clínica e Te apêu ica, Faculdade de Medicina, Uni e sidade de Lisboa, Lisboa, Po ugal. 2Ca diology Depa men , Hospi al Ga cia de O a, Almada, Po ugal. 3Clinical Pha macology Uni , Ins i u o de Medicina Molecula , Lisboa, Po ugal. 4Hun ing on’s Disease Cen e, Ins i u e o Neu ology, Uni e si y College London, London, UK. 5Ca diology Depa men , CCUL, CAML, Faculdade de Medicina, Uni e sidade de Lisboa, Lisboa, Po ugal. 6Cen e o E idence-Based Medicine (CEMBE), Faculdade de Medicina, Uni e sidade de Lisboa, Lisboa, Po ugal. 7Coch ane Po ugal, Faculdade de Medicina, Uni e sidade de Lisboa, Lisboa, Po ugal. ABSTRACT BACkgROUnd: Venous h omboembolism (VTE) is a po en ially a al complica ion o o hopedic su ge y, and un il ecen ly, ew an i h om- bo ic compounds we e a ailable o pos ope a i e h ombop ophylaxis. The in oduc ion o he non– i amin K an agonis s o al an icoagu- lan s (NOAC), including apixaban, has ex ended he he apeu ic a mamen a ium in his ield. The e o e, es ima ion o NOAC ne clinical bene i in compa ison wi h he es ablished ea men is needed o in o m clinical decision making. OBjeCTi eS: Sys ema ic e iew o assess he e icacy and sa e y o apixaban 2.5 mg wice a day e sus low-molecula -weigh hepa ins (LMWH) o h ombop ophylaxis in pa ien s unde going knee o hip eplacemen . dATA SOURCeS: MEDLINE, Embase, and CENTRAL we e sea ched om incep ion o Sep embe 2016, o he sys ema ic e iews, e e - ence lis s, and expe s we e consul ed. STUdy eligiBiliTy CRiTeRiA, PARTiCiPAnTS, And inTeR enTiOn: All majo o hopedic su ge y andomized con olled ials com- pa ing apixaban 2.5 mg wice daily wi h LMWH, epo ing h ombo ic and bleeding e en s. dATA exTRACTiOn: Two independen e iewe s, using a p ede e mined o m. STUdy APPRAiSAl And SynTheSiS MeThOdS: The Coch ane ool o assess isk bias was used by wo independen au ho s. Re Man so wa e was used o es ima e pooled isk a io (RR) and 95% con idence in e als (95% CI) using andom-e ec s me a-analysis. T ial sequen ial analysis (TSA) was pe o med in s a is ical signi ican esul s o e alua e whe he cumula i e sample size was powe ed o he ob ained e ec . O e all con idence in cumula i e e idence was assessed using he G ading o Recommenda ions Assessmen , De elop- men , and E alua ion (GRADE) Wo king G oup me hodology. ReSUlTS: Fou s udies compa ing apixaban 2.5 mg wice daily wi h LMWH we e included, wi h a o al o 11.828 pa ien s (55% unde going knee and 45% hip eplacemen ). The o e all isk o bias ac oss s udies was low. In compa ison wi h LMWH (all egimens), apixaban showed a signi ican ly lowe isk o VTE e en s and o e all mo ali y combined (RR: 0.63, 95% CI: 0.42-0.95, I2 = 84%, n = 8346), bu no o majo VTE e en s (RR: 0.62, 95% CI: 0.32-1.19, I2 = 63%, n = 9493), o o symp oma ic VTE e en s and VTE- ela ed mo ali y combined (RR: 1.14, 95% CI: 0.68-1.90, I2 = 0%, n = 11 879). T ial sequen ial analysis showed ha he isk educ ion ob ained o VTE and mo ali y was based on unde powe ed cumula i e sample size and e ec dimension. Subg oup analysis acco ding o LMWH egimens showed ha apixaban educed he isk o VTE e en s and o e all mo ali y, and majo VTE e en s, when compa ed wi h LMWH once daily, wi hou di e ences be ween apixaban and LMWH wice daily. COnClUSiOnS: The e is low o mode a e e idence ha in pa ien s unde going knee o hip eplacemen , apixaban seems equally e ec i e and sa e o LMWH wice a day. When compa ed wi h LMWH once a day, apixaban seems a supe io h ombop ophylaxis op ion. Howe e , he esul s a e unde powe ed which p ecludes de ini e answe s ega ding he ue ne clinical bene i o apixaban e sus LMWH in his clini- cal con ex . keyWORdS: Venous h omboembolism, non– i amin K o al an icoagulan s, apixaban, hepa in, low-molecula -weigh , sys ema ic e iew ReCei ed: No embe 07, 2016. ACCePTed: Ma ch 13, 2017. PeeR Re ieW: Fou pee e iewe s con ibu ed o he pee e iew epo . Re iewe s’ epo s o aled 351 wo ds, excluding any con iden ial commen s o he academic edi o . TyPe: Re iew FUnding: The au ho (s) ecei ed no inancial suppo o he esea ch, au ho ship, and/o publica ion o his a icle. deClARATiOn OF COnFliCTing inTeReSTS: The au ho (s) decla ed no po en ial con lic s o in e es wi h espec o he esea ch, au ho ship, and/o publica ion o his a icle. CORReSPOnding AUThOR: Daniel Caldei a, Labo a ó io de Fa macologia Clínica e Te apêu ica, Faculdade de Medicina, Uni e sidade de Lisboa, A . P o . Egas Moniz, Lisboa 1649-028, Po ugal. Email: dgcaldei a@ho mail.com 704660BDX0010.1177/1179545X17704660Clinical Medicine Insigh s: Blood Diso de sCaldei a e al esea ch-a icle2017 2 Clinical Medicine Insigh s: Blood Diso de s launch o new o al an icoagulan s also called non– i amin K an agonis o al an icoagulan s (NOACs) has inc eased he lim- i ed a senal o an i h ombo ic he apeu ic op ions a ailable o pos su gical h ombop ophylaxis o pa ien s unde going o hopedic su ge ies.2 Non– i amin K an agonis o al an icoagulan s selec i ely inhibi h ombin o ac o Xa, and hei o al egimen and absence o egula hemos a ic pa ame e s e alua ion ep esen a signi ican ad ancemen o an i h ombo ic ea men com- pa ed wi h pa en e al an icoagulan s such as low-molecula - weigh hepa in (LMWH) and i amin K an agonis s. As occu s wi h any pa ien ea ed wi h an icoagulan s, he main objec i e is o o e he ea men ha p o ides he bes ne clinical bene i ou come, ha is, which concedes he bes bal- ance be ween h omboembolism p e en ion and wi h minimal bleeding isk.3 The aim o his wo k was o assess he e icacy and sa e y o apixaban (a Xa inhibi o ) a 2.5 mg wice a day ( he Eu opean Medicines Agency and Food and D ug Adminis a ion– app o ed dosage in VTE p e en ion in o hopedic pa ien s) in compa ison wi h LMWH o h ombop ophylaxis in pa ien s unde going knee o hip eplacemen o help in o ming he p ocess o clinical decision making. Me hods P o ocol and egis a ion This sys ema ic e iew was epo ed in line wi h P e e ed Repo ing I ems o Sys ema ic Re iews and Me a-Analyses (PRISMA) guidelines.4 Repo ing o s a is ical da a ollowed S a is ical Analyses and Me hods in he Published Li e a u e (SAMPL) guidelines.5 Eligibili y c i e ia We adop ed a me hodology simila o p e iously published a icles.6–8 All phase 3 andomized con olled ials (RCTs) compa ing apixaban agains LMWH in pa ien s submi ed o majo o hopedic su ge y we e included. All published RCTs we e conside ed o inclusion i espec i e o backg ound he - apy, ea men du a ion, o ollow-up. Only ials epo ing h ombo ic e en s and/o a al and non a al bleeding e en s we e included. Ou ou comes o in e es we e as ollows: isk o VTE and all-cause mo ali y (p ima y e icacy ou come), majo VTE, symp oma ic VTE and VTE- ela ed mo ali y, majo bleeding (p ima y sa e y ou come), and su ge y si e bleeding. We used he In e na ional Socie y o Th ombosis and Haemos asis de ini ion o majo bleeding.9 In o ma ion sou ces and sea ch me hod Reco ds o po en ially eligible s udies we e iden i ied h ough an elec onic sea ch o bibliog aphic da abases om incep ion o Sep embe 2016 (MEDLINE, Embase, and CENTRAL a Coch ane Lib a y). Sea ch s a egy de ails a e p o ided in Supplemen a y Da a 1. No language es ic ions we e applied. We sc eened, c oss-checked, and iden i ied sys ema ic e iews and me a-analyses e alua ing NOACs, as well as e e ence lis s o epo s o po en ial eligible s udies. S udy selec ion, da a collec ion p ocess, and da a i ems Ti les and abs ac o eco ds ob ained om he sea ch p ocess we e sc eened by 2 in es iga o s (D.C. and F.B.R.). Doub s and disag eemen s we e sol ed by consensus. Whene e needed a hi d elemen was consul ed (J.C.). Selec ed s udies we e assessed in ull ex o de e mine i s app op ia eness o inclu- sion. Da a om included s udies we e independen ly ex ac ed by 2 au ho s (D.C. and F.B.R.) o a p epilo ed elec onic o m. Re ie ed da a i ems we e as ollows: s udy design, yea o pub- lica ion, pa ien s’ cha ac e is ics, in e en ions es ed, s udies’ ou comes, and da a o equi ed ou comes. Da a we e double- checked o so wa e en y be o e analyses by an addi ional au ho (J.C.). Risk o bias in indi idual s udies We used he Coch ane ool o assessing isk o bias o included s udies.10 The 6 p ede ined speci ic domains o anal- ysis we e as ollows: andom sequence gene a ion, alloca ion concealmen , blinding o pa icipan s and pe sonnel, blinding o ou come assessmen , incomple e ou come da a, and selec- i e epo ing. Fo -p o i bias domain was added. Two inde- penden e iew au ho s (D.C. and F.B.R.) pe o med c i ical assessmen s o each domain o he isk o bias ool. Any disa- g eemen was sol ed by discussion be ween he 2 e iewe s and, i necessa y, eached consensus wi h he pa icipa ion o a hi d e iewe (J.C.). The isk o bias was quali a i ely e al- ua ed as high, unclea , o low isk. Risk o bias g aphs we e de i ed om hese ools. Summa y measu es All ou comes da a we e summa ized as dicho omous da a. The e ec measu emen es ima e chosen was isk a io (RR) because ela i e es ima es a e mo e simila ac oss s udies wi h di e en designs, popula ions, and leng hs o ollow-up han absolu e e ec s.11 Syn hesis esul s We used Re Man 5.3.3 so wa e (The No dic Coch ane Cen e, The Coch ane Collabo a ion, 2014) o s a is ical analysis and o de i e o es plo showing he esul s o indi- idual s udies and pooled analysis. We compa ed apixaban 2.5 mg wice a day wi h LMWH, h ough andom-e ec s me a- analysis weigh ed by he Man el-Haenszel me hod o es ima e Caldei a e al 3 pooled RR and 95% con idence in e als (95% CI). He e ogenei y measu ed as he pe cen age o o al a ia ion be ween s udies due o he e ogenei y was assessed h ough he I2 es .12 We used andom-e ec s model independen ly o he exis ence (I2 ⩾ 50%) o no o subs an ial he e ogenei y be ween s udies’ esul s because we pooled esul s o s udies wi h di e en designs and pa ien s’ cha ac e is ics. When sig- ni ican di e ences we e ound, we also de e mined he num- be needed o ea (NNT) and 95% CI aking in o accoun he baseline isk (p opo ion o e en a e in con ol g oup).10 P especi ied sensi i i y (by excluding s udies a a highe isk o bias) and/o subg oup (by conside ing di e en egimens o LMWH) analyses we e pe o med o explain and explo e he ou come es ima es po en ially associa ed wi h s a is ical and/o clinical he e ogenei y. T ial sequen ial analyses (TSAs) we e pe o med o p i- ma y ou comes using TSA e sion 0.9 be a (Copenhagen T ial Uni , Cen e o Clinical In e en ion Resea ch, Copenhagen, Denma k, 2011) o explo e whe he cumula i e da a we e ade- qua ely powe ed o e alua e ou comes.13,14 The equi ed in o - ma ion size and he O’B ien-Fleming adjacen ial sequen ial alpha spending moni o ing bounda ies we e calcula ed based on a 2-sided 5% isk o a ype I e o , 20% isk o a ype II e o (powe o 80%), isk educ ion based on pooled analysis, he weigh ed incidence o e en s in he con ol g oup, and he e o- genei y. Powe o he p ima y ou comes indings was in e - p e ed i signi icance was eached wi h ei he a minimum sample size o c ossing ial sequen ial alpha spending moni- o ing bounda y. Assessmen o con idence in cumula i e e idence As ecommended by he G ading o Recommenda ions Assessmen , De elopmen , and E alua ion (GRADE) Wo king G oup me hodology,15,16 2 e iewe s independen ly assessed all he c i ical ou comes in he ollowing domains: isk o bias, inconsis ency, indi ec ness, imp ecision, and publica ion bias. In case o disag eemen , he au ho s eached consensus, consul ing an independen hi d e iew, i necessa y. Fo his pu pose, we used he GRADEp o ile (GRADEp o) so wa e ool, which was hen ex ac ed in o he o m o a summa y o indings able o inclusion in o he e iew manusc ip . We applied he s anda d de ini ions o he quali y o e idence17 and explici c i e ia o ensu e he consis - ency and ep oducibili y o GRADE judgmen s o each domain and o all key compa isons o he c i ical ou comes (Supplemen a y Da a 2). Resul s S udy selec ion O e all, 1161 e e ences we e e ie ed om he elec onic sea ch (425 MEDLINE, 492 Embase, and 244 CENTRAL). A e manual and au oma ic deduplica ion, 743 i les and abs ac s we e sc eened o ull- ex e iew (Figu e 1). Fou s udies wi h o e all 11 828 pa ien s unde going knee (55%; 6496 pa ien s) o hip eplacemen (45%; 5332 pa ien s), ea ed wi h apixaban 2.5 mg wice daily o LMWH (2 s udies wi h enoxapa in 40 mg once daily, and 2 s udies wi h enoxapa in 30 mg wice daily), we e selec ed o be included.18–21 Thei main ea u es a e b ie ly cha ac e ized in Table 1. Risk o bias wi hin s udies The o e all isk o bias ac oss s udies was low (Figu e 2). No s udy had high o low isk o bias o e e y e alua ed domain. One s udy (APROPOS18) did no epo he me hods o an- domiza ion. All s udies used a cen alized me hod o pa ien alloca ion. The blinding o pa icipan s, s udy pe sonnel, and ou come assesso s was o low isk o bias ac oss RCTs. In a single s udy (APROPOS), i was no possible o e alua e he in luence o he imbalances p esen on some o ea men a m in he s udy esul s. All s udies we e o high isk o o -p o i bias because he s udies we e unded and sponso ed by he Figu e 1. P e e ed Repo ing I ems o Sys ema ic Re iews and Me a-Analyses (PRISMA) lowcha o s udies selec ion. 4 Clinical Medicine Insigh s: Blood Diso de s companies who owned apixaban’s pa en , bu one (APROPOS), whe e s udy unding was no decla ed. Syn hesis o esul s Apixaban 2.5 mg showed a 37% signi ican isk educ ion (RR: 0.63, 95% CI: 0.42-0.95) o he composi e ou come o all VTE e en s o mo ali y (Figu e 3) in compa ison wi h LMWH. Acco ding o his es ima e, he NNT wi h apixaban would be 26 pa ien s (95% CI: 16-190) o a weigh ed mean pe iod o 73 days. Howe e , he s a is ical he e ogenei y was e y high (I2 = 84%) which can be pa ially a ibu ed o he di e en egimens o LMWH (enoxapa in) in he con ol a ms. In subg oup analysis, apixaban signi ican ly educed VTE e en s o mo ali y compa ed wi h once-daily LMWH egimens (RR: 0.50, 95% CI: 0.41-0.61; I2 = 68%), bu no signi ican di e ence was ound when compa ing wi h wice-daily LMWH egimens (RR: 0.96, 95% CI: 0.73-1.25, I2 = 33%) (Table 2). Rega ding TSA analysis, RR educ ion (RRR) o 37% was assumed based on he RR o 0.63 ound in he me a- analysis o VTE and all-cause mo ali y. The cumula i e e idence eached 59% o minimum in o ma ion size equi ed (14 138 pa ien s) adjus ed o he ob ained RRR and he e o- genei y (Figu e 4). As s a is ical signi icance was ob ained be o e he in o ma ion size has been eached, i was impo an o e alua e whe he an adjus men o signi icance bounda ies (O’B ien-Fleming bounda ies) o he sample size s ill esul s in s a is ical signi ican es ima es. The TSA g aph shows ha cumula i e es ima es we e no obus enough o de e mine he p ema u e s a is ically signi ican esul s (ie, he blue line did no c oss he do ed o ange line in Figu e 4). Conside ing he isk o majo VTE, he e we e no di e - ences be ween apixaban and LMWH (RR: 0.62, 95% CI: 0.32-1.19) (Figu e 3). Simila o he esul s ound o he p i- ma y e icacy ou come, subs an ial s a is ical he e ogenei y was also no iced (I2 = 63%). Subg oup analysis acco ding o LMWH egimens showed a signi ican isk educ ion in majo VTE o apixaban compa ed wi h once-daily LMWH (RR: 0.45, 95% CI: 0.27-0.74; I2 = 0%), bu no signi ican di e - ences when compa ed wi h wice-daily LMWH (RR: 0.91, 95% CI: 0.30-2.83; I2 = 48%). Table 1. Main cha ac e is ics o included s udies compa ing apixaban 2.5 mg wice daily e sus LMWH. TRIALS PATIENTS ORTHOPEDIC CONDITION MEAN AgE LMWH (ENOXAPARIN) PRIMARy OUTCOME MEAN FOLLOW- UP, D APROPOS18 303aElec i e knee eplacemen 67 30 mg bid Majo bleeding 42 ADVANCE-119 3195 Elec i e knee eplacemen 66 30 mg bid VTE e en s and all-cause mo ali y 72 ADVANCE-220 3057 Elec i e knee eplacemen 67 40 mg od VTE e en s and all-cause mo ali y 72 ADVANCE-321 5407 Elec i e hip eplacemen 61 40 mg od (ex ended p ophylaxis egimen—35 d) VTE e en s and all-cause mo ali y 95 Abb e ia ions: bid, wice daily; LMWH, low-molecula -weigh hepa in; od, once daily; VTE, enous h omboembolism. aFo he p e ended compa ison. Figu e 2. Risk o bias g aph. Caldei a e al 5 The o e all isk o symp oma ic VTE and VTE- ela ed dea h, majo bleeding, and su ge y si e bleeding was no di e - en be ween apixaban and LMWH (conside ing all egimens) (Figu es 3 and 5). The sensi i i y analyses excluded he APROPOS s udy which was he only phase 2 ial and he only s udy wi h an unclea isk o bias in 3 o he 7 i ems analyzed. Such sensi i - i y analyses did no esul in any s a is ically signi ican esul s, and s a is ical he e ogenei y did no change subs an ially o all ou comes’ es ima es (Table 3). Table 4 de ails he GRADE app oach o he quali y o he a ailable e idence which was conside ed o be low o mode a e. Discussion The main indings o his e iew we e as ollows: (1) he e is low o mode a e quali y e idence compa ing apixaban wi h LMWH; (2) apixaban 2.5 wice daily dec eases he isk o VTE o all-cause mo ali y, and majo VTE, mos ly due o he esul s o ials compa ing apixaban wi h once-daily 30 mg enoxapa in; (3) his isk educ ion is, howe e , unde powe ed acco ding o he TSA analysis; (4) he isk o symp oma ic VTE and VTE- ela ed dea h wi h apixaban was simila o LMWH; (5) he e we e no signi ican di e ences be ween bo h in e en ions conce ning majo bleeding and su ge y si e bleeding e en s. The in e p e a ion o cu en da a is no as op imis ic as epo ed in a p e ious sys ema ic e iew.22 Despi e he o e all s a is ical signi ican educ ion in he isk o VTE and all- cause mo ali y, i seems easonable o assume ha his di e - ence was mainly d i en by he isk o majo VTE, pa icula ly in compa ison wi h he “Eu opean” once-daily egimen o enoxapa in, which may be an impo an gain in he ca e o hese pa ien s. I is impo an o s a e ha he signi icance o such ou come and isk educ ion es ima es a e Figu e 3. Fo es plo o e icacy ou comes. Table 2. Subg oup analysis pe di e en egimens o LMWH in he con ol a ms. OUTCOME RR (95% CI) FOR APIXABAN VS LMWH 40 Mg OD; I2 (%) RR (95% CI) FOR APIXABAN VS LMWH 30 Mg BID; I2 (%) P VALUE FOR RR INTERACTION All VTE o all-cause mo ali y 0.50 (0.41–0.61); I2 = 68 0.96 (0.73–1.25); I2 = 33 <.001 Majo VTE 0.45 (0.27–0.74); I2 = 0 0.91 (0.30–2.83); I2 = 48 .26 Symp oma ic VTE o VTE- ela ed dea h 0.92 (0.40–2.08); I2 = 0 1.31 (0.68–2.51); I2 = 0 .50 Majo bleeding 0.94 (0.51, 1.73); I2 = 30 0.50 (0.24–1.02); I2 = N/A .18 Su ge y si e bleeding 0.96 (0.56–1.65); I2 = 3 0.57 (0.24–1.35); I2 = N/A .31 Abb e ia ions: bid, wice daily; CI, con idence in e al; LMWH: low-molecula -weigh hepa in; N/A, no applicable; od, once daily; RR: isk a io; VTE: enous h omboembolism. 6 Clinical Medicine Insigh s: Blood Diso de s unde powe ed and u he in es iga ion would be equi ed o de ini ely es ablish he ue e ec size o apixaban bene i . Gómez-Ou es e  al23 in hei sys ema ic e iew sugges ed ha he po en ial bene i s in bleeding isk (clinically ele an Figu e 4. T ial sequen ial analysis o all enous h omboembolism e en s and all-cause mo ali y. RR indica es isk a io. Figu e 5. Fo es plo o sa e y/bleeding ou comes. CI indica es con idence in e al; LMWH, low-molecula -weigh hepa in. Table 3. Resul s o he sensi i i y analyses ha excluded he APROPOS s udy. EXCLUDINg APROPOS OUTCOMES RR (95% CI) FOR APIXABAN VS LMWH I2 (%) All VTE o all-cause mo ali y 0.63 (0.39–1.01) I2 = 89 Majo VTE 0.66 (0.31–1.39) I2 = 74 Symp oma ic VTE o VTE- ela ed dea h 1.20 (0.70–2.04) I2 = 0 Majo bleeding 0.76 (0.43, 1.33) I2 = 45 Su ge y si e bleeding 0.83 (0.53–1.31) I2 = 0 Abb e ia ions: CI, con idence in e al; LMWH, low-molecula weigh hepa in; RR, isk a io; VTE, enous h omboembolism. Caldei a e al 7 bleeding) and po en ial p o h ombo ic e ec s (pa icula ly PE) could be he esul om delaying he adminis a ion o apixaban i s pos ope a i e dose o 18 hou s. Ou da a show ha majo bleeding and su gical si e bleeding isks a e simila be ween apixaban and enoxapa in egimens, and ha majo VTE was no inc eased wi h apixaban, despi e he end owa d a p o ec i e e ec . Fu he mo e, as occu s wi h all NOACs, he e is a signi i- can dec ease in in ac anial hemo hage, and he pha macoki- ne ic and pha macodynamic p o iles o hese d ugs,7 including apixaban, disclose a p edic able an icoagulan e ec , he eby dis- missing egula e alua ions o hemos a ic pa ame e s. O e all, apixaban 2.5 mg wice daily seems o be a aluable op ion o he h ombop ophylaxis o pa ien s unde going elec i e knee o hip eplacemen . The e a e s ill some unanswe ed ques ions, such as he exis ence o no o ne clinical bene i s o apixaban compa ed wi h LMWH wice daily (besides he in ac anial hemo hage isk educ ion7), and he op imal iming o apixaban adminis- a ion a e su ge y, as well as he op imal ea men du a ion o each condi ion (knee o hip eplacemen ). I is s ill unknown whe he using an ul aspeci ic ac o Xa/ h ombin es s o LMWH moni o ing (4 hou s a e he adminis a ion) could imp o e he ou comes in he con ol a m, as sugges ed o some subg oups o pa ien s wi h enous h omboembolism. The esul s o his e iew should be in e p e ed in line wi h he limi a ions inhe en o sys ema ic e iews and me a-analysis. The he e ogenei y in he an icoagula ion du a ion and in con ols somehow impai s he obus ness o he p esen ed da a. The compa a o s we e also di e en as low-dose enoxapa in (30 mg) was gi en wice daily in 1 ial, whe eas he emaining ials engaged o a once-daily 40 mg o enoxapa in o h ombop ophylaxis. The ollow-up was di e en acco ding o he condi ions ( ials ended ea lie in he knee eplacemen s udies) as well as he measu emen o ou comes ha equi ed examina ions such as enog aphy. These limi a ions may a leas pa ially explain o some ex en he high s a is ical he e ogenei y ound in mos e icacy ou comes. E en hough we conside he esul s a e eassu ing o e icacy and sa e y. Despi e he a ailable da a and p e ious me a-analysis, none o hem had analyzed he powe o he signi ican esul s ob ained in hei me a-analysis. The e o e, he me hods and conclusions de i ed om TSA a e impo an o elucida ing he obus ness o he da a ega ding apixaban e icacy in pa ien s unde going majo elec i e o hopedic su ge ies such as knee o hip eplacemen . Conclusions Apixaban 2.5 mg wice daily is a aluable and p ac ical op ion o h ombop ophylaxis wi h bleeding isks simila o LMWH and an e icacy likely o be be e han cu en enoxapa in egimens. Table 4. gRADE summa y o indings able—apixaban compa ed wi h LMWH o enous h omboembolism p ophylaxis a e majo o hopedic su ge y. OUTCOME NO. OF PARTICIPANTS (STUDIES) RELATIVE EFFECT (95% CI) ANTICIPATED ABSOLUTE EFFECTS (95% CI)* QUALITy WITHOUT APIXABAN WITH APIXABAN DIFFERENCE All VTE and all-cause dea h No. o pa icipan s: 8346 (4 RCTs) RR: 0.63 (0.42–0.95) 10.5% 6.6% (4.4–9.9) 3.9% ewe (6.1 ewe -0.5 ewe ) ⊕⊕ Lowa,b Majo VTE No. o pa icipan s: 9493 (4 RCTs) RR: 0.62 (0.32–1.19) 1.6% 1.0% (0.5–1.9) 0.6% ewe (1.1 ewe -0.3 mo e) ⊕⊕ Lowa,b Symp oma ic VTE and dea h om VTE No. o pa icipan s: 11 879 (4 RCTs) RR: 1.14 (0.68–1.90) 0.5% 0.5% (0.3–0.9) 0.1% mo e (0.2 ewe -0.4 mo e) ⊕⊕⊕ Mode a eb Majo bleeding No. o pa icipan s: 11 828 (4 RCTs) RR: 0.76 (0.43–1.33) 0.9% 0.7% (0.4–1.2) 0.2% ewe (0.5 ewe -0.3 mo e) ⊕⊕ Lowa,b Su ge y si e bleeding No. o pa icipan s: 11 828 (4 RCTs) RR: 0.83 (0.53–1.31) 0.7% 0.6% (0.4–0.9) 0.1% ewe (0.3 ewe -0.2 mo e) ⊕⊕⊕ Mode a eb Abb e ia ions: CI, con idence in e al; RCTs, andomized con olled ials; RR, isk a io; VTE, enous h omboembolism. * The isk in he in e en ion g oup (and i s 95% con idence in e al) is based on he assumed isk in he compa ison g oup and he ela i e e ec o he in e en ion (and i s 95% CI). gRADE Wo king g oup g ades o e idence: high quali y, we a e e y con iden ha he ue e ec lies close o ha o he es ima e o he e ec ; mode a e quali y, we a e mode a ely con iden in he e ec es ima e: he ue e ec is likely o be close o he es ima e o he e ec , bu he e is a possibili y ha i is subs an ially di e en ; low quali y, ou con idence in he e ec es ima e is limi ed: he ue e ec may be subs an ially di e en om he es ima e o he e ec ; e y low quali y, we ha e e y li le con idence in he e ec es ima e: he ue e ec is likely o be subs an ially di e en om he es ima e o e ec . aS a is ical he e ogenei y supe io o 40%. bMinimal in o ma ion size no me . 8 Clinical Medicine Insigh s: Blood Diso de s Au ho Con ibu ions DC con ibu ed o he concep and design, da a acquisi ion, da a analysis, and in e p e a ion o he da a; w o e he i s d a o he manusc ip ; c i ically e ised he manusc ip ; and ga e inal app o al o he submi ed manusc ip . FR con- ibu ed o he da a analysis and in e p e a ion; con ibu ed o he i s d a ; c i ically e ised he manusc ip ; and ga e inal app o al o he submi ed manusc ip . JC con ibu ed o he concep and design, and in e p e a ion o he da a; c i i- cally e ised he manusc ip ; and ga e inal app o al o he submi ed manusc ip . FJP and JJF con ibu ed o he in e - p e a ion o da a, c i ically e ised he manusc ip , and ga e inal app o al o he submi ed manusc ip . DC and JC a e he gua an o s. Re eRenCes 1. Gee s WH, Be gq is D, Pineo GF, e al. P e en ion o enous h omboembo- lism: Ame ican College o Ches Physicians E idence-Based Clinical P ac ice Guidelines (8 h Edi ion). Ches . 2008;133:381S–453S. 2. Falck-Y e Y, F ancis CW, Johanson NA, e al. P e en ion o VTE in o hope- dic su ge y pa ien s: an i h ombo ic he apy and p e en ion o h ombosis, 9 h ed: Ame ican College o Ches Physicians E idence-Based Clinical P ac ice Guidelines. Ches . 2012;141:e278S–e325S. 3. Caldei a D, Cos a J, Fe ei a JJ, Pin o FJ. Ne clinical bene i ou come should be s anda dized in ials e alua ing an i h ombo ic d ugs: he example o NOACs in a ial ib illa ion. In J Ca diol. 2014;174:405–406. 4. Libe a i A, Al man DG, Te zla J, e al. The PRISMA s a emen o epo ing sys ema ic e iews and me a-analyses o s udies ha e alua e heal hca e in e - en ions: explana ion and elabo a ion. BMJ. 2009;339:b2700. 5. Lang TA, Al man DG. Basic s a is ical epo ing o a icles published in Biomedical Jou nals: he “S a is ical Analyses and Me hods in he Published Li e a u e” o The SAMPL Guidelines.” In: Science Edi o s’ Handbook. Eu opean Associa ion o Science Edi o s; 2013. h p://www.equa o -ne wo k.o g/wp- con en /uploads/2013/07/SAMPL-Guidelines-6-27-13.pd . 6. Caldei a D, Ba a M, San os AT, e al. Risk o d ug-induced li e inju y wi h he new o al an icoagulan s: sys ema ic e iew and me a-analysis. Hea . 2014;100: 550–556. 7. Caldei a D, Ba a M, Pin o FJ, Fe ei a JJ, Cos a J. In ac anial hemo hage isk wi h he new o al an icoagulan s: a sys ema ic e iew and me a-analysis. J Neu ol. 2015;262:516–522. 8. Caldei a D, Ba a M, Da id C, Cos a J, Fe ei a JJ, Pin o FJ. The p e alence o o al an icoagula ion in pa ien s wi h a ial ib illa ion in Po ugal: sys ema ic e iew and me a-analysis o obse a ional s udies. Re Po Ca diol. 2014;33:555–560. 9. Schulman S, Kea on C. De ini ion o majo bleeding in clinical in es iga ions o an ihemos a ic medicinal p oduc s in non-su gical pa ien s. J Th omb Haemos . 2005;3:692–694. 10. Higgins JPT, G een S. Coch ane Handbook o Sys ema ic Re iews o In e en ions. 5.1.0 ed. Hoboken, NJ: John Wiley & Sons; 2011. 11. Deeks JJ. Issues in he selec ion o a summa y s a is ic o me a-analysis o clini- cal ials wi h bina y ou comes. S a Med. 2002;21:1575–1600. 12. Higgins JP, Thompson SG. Quan i ying he e ogenei y in a me a-analysis. S a Med. 2002;21:1539–1558. 13. B ok J, Tho lund K, Gluud C, We e sle J. T ial sequen ial analysis e eals insu - icien in o ma ion size and po en ially alse posi i e esul s in many me a-analyses. J Clin Epidemiol. 2008;61:763–769. 14. Caldei a D, Rod igues FB, Ba a M, e al. Non- i amin K an agonis o al an i- coagulan s and majo bleeding- ela ed a ali y in pa ien s wi h a ial ib illa ion and enous h omboembolism: a sys ema ic e iew and me a-analysis. Hea . 2015;101:1204–1211. 15. Alonso-Coello P, Oxman AD, Mobe g J, e al. GRADE E idence o Decision (E D) amewo ks: a sys ema ic and anspa en app oach o making well in- o med heal hca e choices. 2: clinical p ac ice guidelines. BMJ. 2016;353:i2089. 16. Alonso-Coello P, Schunemann HJ, Mobe g J, e al. GRADE E idence o Decision (E D) amewo ks: a sys ema ic and anspa en app oach o making well in o med heal hca e choices. 1: in oduc ion. BMJ. 2016;353:i2016. 17. Balshem H, Hel and M, Schunemann HJ, e al. GRADE guidelines: 3. Ra ing he quali y o e idence. J Clin Epidemiol. 2011;64:401–406. 18. Lassen MR, Da idson BL, Gallus A, Pineo G, Ansell J, Dei chman D. The e i- cacy and sa e y o apixaban, an o al, di ec ac o Xa inhibi o , as h ombop ophylaxis in pa ien s ollowing o al knee eplacemen . J Th omb Haemos . 2007;5:2368–2375. h p://onlinelib a y.wiley.com/o/coch ane/clcen- al/a icles/373/CN-00619373/ ame.h ml. 19. Lassen MR, Raskob GE, Gallus A, Pineo G, Chen D, Po man RJ. Apixaban o enoxapa in o h ombop ophylaxis a e knee eplacemen . N Engl J Med. 2009;361:594–604. h p://onlinelib a y.wiley.com/o/coch ane/clcen al/a i- cles/569/CN-00700569/ ame.h ml. 20. Lassen MR, Raskob GE, Gallus A, Pineo G, Chen D, Ho nick P; ADVANCE-2 In es iga o s. Apixaban e sus enoxapa in o h ombop ophylaxis a e knee eplacemen (ADVANCE-2): a andomised double-blind ial. Lance . 2010; 375:807–815. 21. Lassen MR, Gallus A, Raskob GE, Pineo G, Chen D Rami ez LM; ADVANCE-3 In es iga o s. Apixaban e sus enoxapa in o h ombop ophy- laxis a e hip eplacemen . N Engl J Med. 2010;363:2487–2498. 22. Li XM, Sun SG, Zhang WD. Apixaban e sus enoxapa in o h ombop ophy- laxis a e o al hip o knee a h oplas y: a me a-analysis o andomized con olled ials. Chin Med J (Engl). 2012;125:2339–2345. 23. Gómez-Ou es A, Te lei a-Fe nandez AI, Sua ez-Gea ML, Va gas-Cas illon E. Dabiga an, i a oxaban, o apixaban e sus enoxapa in o h ombop ophy- laxis a e o al hip o knee eplacemen : sys ema ic e iew, me a-analysis, and indi ec ea men compa isons. BMJ. 2012;344:e.3675.