J Thromb Thrombolysis DOI 10.1007/s11239-014-1094-1
Dual antiplatelet therapy with or without oral anticoagulation in the postdischarge management of acute coronary syndrome patients with an indication for long term anticoagulation: a systematic review Jeffrey B. Washam • Rowena J. Dolor • W. Schuyler Jones • Sharif A. Halim • Victor Hasselblad • Stephanie B. Mayer • Brooke L. Heidenfelder • Chiara Melloni
Ó Springer Science+Business Media New York 2014
Abstract Currently, there is a lack of consensus among guidelines for the postdischarge treatment of patients presenting with acute coronary syndrome (ACS) who have a long-term indication for anticoagulation. We conducted a systematic review comparing the safety and effectiveness of dual antiplatelet therapy (DAPT) and triple therapy (TT; defined as DAPT plus an oral anticoagulant) in patients with ACS and a long-term indication for anticoagulation. We searched for clinical studies in MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews published between January 1995 and September 2013. Each investigator screened and abstracted data, assessed applicability and quality, and graded the strength of evidence. Meta-analysis of direct comparison was performed when outcomes and follow-up periods were comparable. Fourteen observational studies were identified that contained comparative effectiveness data on DAPT versus TT. No difference in the odds of mortality (OR 1.04, 95 % CI 0.59–1.83) or stroke (OR 1.01, 95 % CI 0.38–2.67) at 1–5 years was found between TT and DAPT. Major bleeding at 1–5 years (OR 1.46, 95 % CI 1.07–2.00) and nonfatal MI at 1–5 years (OR 1.85, 95 % CI 1.13–3.02) J. B. Washam (&) Duke Heart Center, Duke University Medical Center, DUMC 3943, Durham, NC 27710, USA e-mail:
[email protected] R. J. Dolor W. S. Jones S. A. Halim V. Hasselblad B. L. Heidenfelder C. Melloni Duke Clinical Research Institute, Duke University Medical Center, Durham, NC, USA S. B. Mayer Division of Endocrinology and Metabolism, Department of Medicine, Virginia Commonwealth University, Richmond, VA, USA
occurred more frequently in patients receiving TT. The results of this systematic review demonstrate that treatment with TT was associated with increased rates of nonfatal MI and major bleeding when compared with treatment with DAPT in the postdischarge management of ACS patients with an indication for oral anticoagulation. Until results of ongoing randomized trials assessing antithrombotic therapies define optimal management strategies, the current analysis suggests using caution when prescribing TT to these patients. Keywords Triple therapy Dual antiplatelet therapy Acute coronary syndrome
Introduction Dual antiplatelet therapy (DAPT) consisting of aspirin plus a P2Y12 receptor antagonist is recommended as the preferred antithrombotic regimen for patients presenting with acute coronary syndrome (ACS) and in those undergoing percutaneous coronary intervention (PCI) [1, 2]. However, ACS patients often have an additional indication for longterm oral anticoagulation (i.e., atrial fibrillation, venous thromboembolism, or mechanical heart valve). Data from the National Cardiovascular Data Registry (NCDR), National Cardiovascular Data Intervention Outcomes Network Registry–Get With the Guidelines (ACTION Registry-GWTG) identified 8.5 % of patients presenting with NSTEMI as being on home warfarin therapy [3]. While the use of triple antithrombotic therapy (DAPT plus an oral anticoagulant) has been associated with an increased risk of bleeding events when compared with DAPT alone, the comparative risk of ischemic or thrombotic events with each strategy is less well defined. In addition, given the
123
J. B. Washam et al.
exclusion of patients with an indication for long-term oral anticoagulation from randomized trials of DAPT in ACS patients, the optimal antithrombotic regimen in this population remains uncertain [4–6]. The American College of Cardiology Foundation/ American Heart Association (ACCF/AHA) guidelines for the management of ST-elevation MI (STEMI) provide a Class I recommendation (level of evidence: C) for the use of anticoagulant therapy with a vitamin K antagonist to patients with STEMI and atrial fibrillation with a CHADS2 score of 2 or more, mechanical heart valves, venous thromboembolism or hypercoagulable state [2]. Alternatively, recent guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction (UA/NSTEMI) provide a Class IIb recommendation for the addition of warfarin in UA/NSTEMI patients who have an indication for anticoagulation [1]. Specific guideline recommendations for the use of warfarin in patients with an indication for chronic anticoagulation undergoing elective PCI are not currently available [7]. Given the current lack of consistency in recommendations across the guidelines, a summative analysis of available data is indicated. The aim of this systematic review was to compare the safety and effectiveness of DAPT (defined as aspirin with oral antiplatelet) and triple therapy (TT; defined as DAPT with an oral anticoagulant) in the postdischarge treatment of patients presenting with ACS who had an indication for chronic anticoagulation.
Methods This analysis was part of broader systematic review assessing the effectiveness and safety of antiplatelet and anticoagulant medications used for the treatment of patients with UA/NSTEMI, funded by the Agency for Healthcare Research and Quality. The full comparative effectiveness research report includes an extensive description of the methods and results, including evidence tables [8]. Data sources We searched for clinical studies in MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews, from January 1, 1995, through September 2013. Complete search details are in the full AHRQ report [8]. This timeframe was selected to include studies that had assessed contemporary antiplatelet and anticoagulant strategies in patients presenting with UA/NSTEMI. We also examined the grey literature of study registries and conference abstracts for relevant articles from completed trials,
123
including ClinicalTrials.gov; metaRegister of Controlled Trials; ClinicalStudyResults.org; WHO International Clinical Trials Registry Platform Search Portal; and ProQuest COS Conference Papers Index. In addition, we reviewed reference lists of relevant review articles to provide additional citations. Figure 1 depicts the flow of articles through the search and screening process including the broader systematic review. Study selection Two investigators independently reviewed titles, abstracts, and full-text articles for eligibility. For this analysis, criteria for inclusion were comparative assessments of antithrombotic strategies for UA/NSTEMI patients with an indication for long term oral anticoagulant at hospital discharge. We excluded screening a study if (1) the population was composed entirely of STEMI or stable angina populations, (2) the study design did not include an active comparator, (3) the intermediate or final outcomes of interest (all-cause mortality, cardiovascular death, myocardial infarction, stroke, stent thrombosis, repeat revascularization, and safety outcomes [bleeding events]) were not reported, or (4) it was not a clinical study. Data abstraction, quality, and applicability assessment Standardized forms were used to abstract the study data (general study characteristics, patient characteristics, treatment strategy, clinical outcomes, and safety). Predefined criteria were used to assess study quality using methods described by the AHRQ’s ‘‘Methods Guide for Effectiveness and Comparative Effectiveness Reviews’’ [9]. Methodological quality was based on criteria derived from the core elements described in the Methods Guide and theses criteria were used as a guidance to determine the rating of good (low risk of bias), fair (moderate risk of bias), or poor (high risk of bias). To assess study applicability, we evaluated the study eligibility criteria, demographic characteristics of the enrolled population compared with the target population, the intervention used compared with technologies currently in use, and clinical relevance and timing of the outcome measures. Data synthesis and analysis We synthesized the primary literature by continuous data (e.g., age, event rates) and categorical data (e.g., race/ethnicity, presence of coronary disease risk factors). We determined the feasibility of completing a quantitative synthesis (meta-analysis), based on the volume of relevant literature, conceptual homogeneity of the studies, and completeness of the reporting of results. We considered
Dual antiplatelet therapy with or without oral anticoagulation Fig. 1 Literature flow diagram
26,278 citations identified by literature search: MEDLINE: 19,317 Cochrane: 158 Embase: 6,803
3,206 duplicates
Manual searching: 42
23,114 citations identified
21,539 abstracts excluded
1,575 passed abstract screening
1,274 articles excluded: Not English language: 1 Not a clinical study: 102 Not a full publication, not original data, not peer reviewed literature, or not gray literature meeting specified criteria: 56 Study population did not have UA/NSTEMI: 256 Did not include an active comparator: 578 Did not include primary or secondary outcomes of interest: 281
301 articles representing 174 studies passed full-text screening
14 studies assessed DAPT versus TT in the postdischarge management of ACS patients with an indication for long-term oral anticoagulation
meta-analysis for comparisons where at least three studies reported the same outcome at a similar time of follow-up. Meta-analyses were based on the nature of the outcome variable, but random-effects models were used for all outcomes because of the heterogeneity of the studies. Dichotomous outcome measures comparing two treatments were combined using unadjusted odds ratios and a randomeffects model as implemented in Comprehensive MetaAnalysis Version 2 (Biostat; Englewood, NJ). We tested for statistical heterogeneity between studies (Q and I2 statistics) while recognizing that the power to detect such heterogeneity may be limited. Potential heterogeneity between studies was reflected through the confidence intervals (CIs) of the summary statistics obtained from a random-effects approach. We present summary estimates, standard errors, and CIs in our data synthesis. Two investigators independently assessed the strength of evidence (SOE) using the four required domains: risk of bias, consistency, directness, and precision [9]. The SOE was graded for each outcome evaluated; thus, a given study may be of different quality for two individual outcomes
reported within that study. When appropriate, the studies were evaluated for the presence of confounders that would diminish an observed effect, strength of association (magnitude of effect), and publication bias. These domains were considered qualitatively, and a summary rating of high (evidence reflects the true effect), moderate (further research may change the estimate), low (further research is likely to change the estimate) or insufficient (an estimate of effect is not possible with the available data).
Results Fourteen studies met the inclusion criteria [10–23]. Each of the 14 studies was observational in design and contained comparative effectiveness data on the use of antithrombotic combinations in the postdischarge treatment of 97,067 total patients. Indications for long-term anticoagulation included atrial fibrillation, presence of a prosthetic valve, chronic deep venous thrombosis, or hypercoagulable states (e.g., protein C or S deficiency).
123
J. B. Washam et al.
Seven studies compared DAPT (aspirin and thienopyridine) with TT (with oral anticoagulant, aspirin, and thienopyridine) [11–14, 19, 20, 23]. One study compared DAPT (aspirin with clopidogrel) with two TT arms—one consisting of oral anticoagulant, aspirin, and clopidogrel, and one consisting of low molecular weight heparin, aspirin, and clopidogrel [17]. Two studies compared aspirin and/or thienopyridine versus oral anticoagulant with or without an antiplatelet agent [21, 22]. One study compared TT (oral anticoagulant, aspirin, and clopidogrel) with warfarin plus aspirin or thienopyridine [18]. One study compared warfarin versus no warfarin among patients with atrial fibrillation–complicating an UA/NSTEMI event [16]. One study with five treatment arms compared aspirin, warfarin, aspirin plus warfarin, aspirin plus a thienopyridine (DAPT), and aspirin plus warfarin plus a thienopyridine (triple therapy) [10]. One study compared monotherapy with aspirin, oral anticoagulant, or clopidogrel; aspirin plus oral anticoagulant; aspirin plus clopidogrel (DAPT); oral anticoagulant plus clopidogrel; and aspirin plus oral anticoagulant plus thienopyridine (triple therapy) [15]. A summary of individual study characteristics is provided in Table 1. Of the 14 studies, 10 (71 %) were rated good quality, 3 (21 %) were rated fair quality, and 1 (7 %) was rated poor quality. Sample sizes for individual studies ranged from 102 to 27,972 patients. Study duration ranged from 30 days to 5 years. The mean age of study participants ranged from 61 to 80 years of age and the proportion of female patients ranged from 28 to 51 %. A total of 75,571 patients (77.9 %) had a diagnosis of ACS on the index hospitalization. Four studies (33 %) reported the racial and/or ethnic demographics of study participants. Two studies (14 %) were conducted within the United States or Canada, eight studies (57 %) were conducted in Europe, two were international (14 %), one was conducted in Asia (7 %), and one was conducted in Israel (7 %). Funding source was reported in seven studies (50 %), with two studies (14 %) funded by an industry source. Effect on all-cause mortality at 1 to 5 years Eight studies including 41,192 patients reported the comparative effectiveness of TT versus DAPT on the outcome of all-cause mortality at 1–5 years [11–15, 17, 19, 20]. Of the patients included in the 8 studies, 31,197 (75.7 %) were managed with DAPT while 3,513 (8.5 %) received TT. A random-effects meta-analysis found that the odds ratio (OR) for TT compared with DAPT was 1.04 (95 % CI 0.59–1.83) (Fig. 2). Two studies [21, 22] reported all-cause mortality but were not included in the analysis because they had different treatment comparison groups: aspirin and/or thienopyridine versus oral anticoagulant with or without an antiplatelet
123
agent. Stenestrand et al. found that the mortality rate at 1 year was significantly lower in patients in the oral anticoagulant arm (22.4 vs. 31.4 %, RR 0.73; 95 % CI 0.62–0.86, p B 0.001) [22]. The other study showed that in patients at high bleeding risk (HAS-BLED C 3) those on an oral anticoagulant had a lower rate of death (9.3 vs. 20.1 %, hazard ratio 0.45; 95 % CI 0.26–0.78, p \ 0.01) [21]. The SOE for the outcome all-cause mortality at 1–5 years was rated insufficient (Table 2). Effect on nonfatal myocardial infarction at 1–5 years A random-effects meta-analysis of four observational studies [12, 13, 17, 20] including 1,425 patients reporting nonfatal MI at 1–5 years found that the OR for TT compared with DAPT was 1.85 (95 % CI 1.13–3.02), favoring DAPT (Fig. 3). The SOE for the outcome of nonfatal MI at 1–5 years was rated low (Table 2). Effect on stroke at 1–5 years A random-effects meta-analysis of four observational studies [12, 13, 15, 17] including 6,485 patients reporting stroke at 1–5 years found that the OR for TT compared with DAPT was 1.01 (95 % CI 0.38–2.67) (Fig. 4). The SOE for the outcome of stroke at 1–5 years was rated as insufficient (Table 2). Effect on revascularization up to 5 years Four observational studies [12, 13, 18, 20] reported revascularization rates between 6 months and 5 years of follow-up. One study [18] comparing TT versus warfarin plus single antiplatelet found no difference in the rate of repeat revascularization (unscheduled PCI) at 6 months between the two groups (10.6 vs. 12.5 %, p = 0.50). The second study [13] comparing TT with DAPT found no difference in the rate of target vessel revascularization (TVR) at 1 year between the two treatment groups (11.0 vs. 7.5 %, OR 1.5; 95 % CI 0.8–2.9, p = 0.21). A third study [12] comparing TT with DAPT found no significant difference in the rate of revascularization (TVR) at 3 years between the two treatment groups (4.3 vs. 1.2 %, p = 0.13). The fourth study [20] again comparing DAPT with TT found no difference in the rate of revascularization (TVR) between the two treatment groups (8.4 vs. 7.1 %, p = 0.3). The SOE for the outcome of revascularization was rated insufficient (Table 2). Effect on stent thrombosis at 1–3 years Two studies reported stent thrombosis at C 1 year. The first study compared TT with DAPT and found no
Jang [12]
Fosbol [11]
Major bleeding
(N = 20)
(N = 195)
Timeframe: 01/1996–03/1996
Timing: 3 yr
Stroke Revascularization Major bleeding Minor bleeding
17 % NSTEMI
19 % STEMI
43 % Stable CAD
Race: NR
Female: 32 %
Mean age: 68
Stroke
Minor bleeding
Major bleeding
Stent thrombosis
Revascularization
Nonfatal MI
Total mortality
Individual
Total mortality Nonfatal MI
21 % UA
Total N: 362
Revascularization (secondary)
Population
Total mortality Nonfatal MI
Composite (primary)
Clopidogrel, ASA
Timeframe: 01/2005–12/2005
(N = 278)
(N = 84)
Major bleeding
Individual
Stroke (any kind)
Funding: NR
Placebo
Warfarin
Observational
5 sites in Asia
Race: NR
Female: 48 %
Median age: 80
Total N: 7619
100 % UA/NSTEMI
Population
Timeframe: 1/2003–12/2006
Total mortality Nonfatal MI
(primary)
ASA ? clopidogrel ? warfarin (N = 731)
Funding: Private foundation, Industry
ASA ? warfarin (N = 1271)
ASA ? clopidogrel (N = 2841)
514 sites in U.S.
Timing: 30 days, 1 yr Composite
Warfarin (N = 563)
Aspirin (N = 2213)
Observational
Race: NR
Female: 43 %
Median age 74
Total N: 21,443
NR
(primary)
ASA
Warfarin
Funding: Government Minor bleeding
Composite
Thienopyridine
(N = 34)
(N = 656)
Population: NR
Timing: 2 yr
Warfarin
ASA
ASA
Observational
Single site in Canada
Buresly, [10]
Timing outcomes reported
Co-interventions
Comparator (N)
Intervention (N)
Study details
Study
Table 1 Study characteristics
Good
Fair
Good
Quality
Dual antiplatelet therapy with or without oral anticoagulation
123
123
Konstantino [14]
Timing: Discharge, 1 yr
Nonfatal MI Revascularization Stent thrombosis (secondary) Stroke Major bleeding Individual Stroke
18 % UA
24 % NSTEMI
12 % STEMI
Total N: 478
Mean age: 70
Female: 26 %
Race: NR
Nonfatal MI
(N = 76)
(N = 2,661)
Race: NR
Female: 21 %
Mean age: 61 to 64
Total N: 2737
56 % STEMI
42 % NSTEMI
100 % ACS
Population
Timeframe: 2000–2004
Total mortality
Rehospitalization
Major bleeding
Stroke
Individual
ASA, ticlopidine/clopidogrel ? warfarin
ASA ? ticlopidine/ clopidogrel
NR sites in Israel
Funding: NR
Triple therapy
Dual therapy
Observational
Timing: In-hospital, 30 days, 6 mo
Stent thrombosis
Revascularization
Nonfatal MI
Total mortality
Major bleeding
Total mortality
Composite
Population
Clopidogrel, ASA
Clopidogrel, ASA (primary)
(N = 239)
(N = 239)
3 sites in Europe
Timing outcomes reported
Timeframe: 2003–2004
Placebo
Warfarin
Observational
Karjalainen [13]
Co-interventions
Funding: Private foundation
Comparator (N)
Intervention (N)
Study details
Study
Table 1 continued
Good
Good
Quality
J. B. Washam et al.
Maegdefessel [17]
Lopes [16]
Observational
Lamberts [15]
Major bleeding
Nonfatal MI Stroke CV mortality
Population
40 % UA
Race: White 100 %
Female: 28 %
Mean age: 70.3
Total N: 159
14 % Stable CAD
14 % STEMI
32 % NSTEMI
Major bleeding
Individual
Timeframe: 1999–2004
(N = 42)
(N = 103)
Single site in Europe
Funding: NR
Clopidogrel
Clopidogrel
Observational
Race: Black 4 %, White 91 %, Other 5 %
Female: 32 %
Timing: 1.4 yr
Stroke
ASA, Enoxaparin, Warfarin
Major bleeding
Total N: 23,208
Median age: 69
Individual
Total mortality Nonfatal MI
ASA 89.0 %, clopidogrel 26.4 %
Timeframe: 1995–2003
N = 917 NSTEMI
(primary)
Clopidogrel, ASA
Funding: NR
Population
Composite
(N = 793)
(N = 124)
Setting: NR
ASA 62.9 %, clopidogrel 10.5 %
Placebo
Warfarin
Observational
Race: NR Timing: In-hospital, 6 mo
Stroke
Total N: 12,165
Mean age: 75.6 ± 10.3 Female: 29 %
Total mortality
PCI 10 %
Nonfatal MI Individual
Clopidogrel, ASA
Timing: 1 yr Composite
MI 90 %
(N = 1,896)
(N = 3,590)
NR
Timing outcomes reported
Total mortality
TT (ASA ? clopidogrel ? oral anticoagulant)
DAPT (ASA ? clopidogrel)
Co-interventions
Population:
Comparator (N)
Intervention (N)
Timeframe: 1/2001–12/2009
Funding: private foundation
Denmark
Study details
Study
Table 1 continued
Good
Fair
Good
Quality
Dual antiplatelet therapy with or without oral anticoagulation
123
123
Observational
113 sites in U.S., Europe, S. America, Australia/NZ
Nguyen [18]
GRACE
Rossini [23]
(N = 26,789)
(N = 1,183)
(secondary) Total mortality Nonfatal MI Stroke Individual Major bleeding Minor bleeding
45 % UA/NSTEMI
34 % STEMI
21 % Stable CAD
Total N: 204
Mean age: 68
Female: 20 %
Race: NR
Major bleeding Minor bleeding
Population
Composite
Timing: 30 days,18 mo (primary)
Clopidogrel 300 mg loading dose/75 mg/day, ASA 100 mg/day, warfarin
Any bleeding
Major bleeding
Stroke
Total mortality
Individual
Nonfatal MI
Total mortality
(primary)
Composite
Timing: 1 yr
Revascularization
Total mortality
Major bleeding
CHF
Stroke
Nonfatal MI
Individual
Timing: In-hospital, 6 mo
Timing outcomes reported
Timeframe: 10/2005–08/2006
(N = 102)
(N = 102)
Clopidogrel, ASA, unfractionated heparin, low molecular weight heparins
Warfarin
Co-interventions
Funding: NR
Dual therapy
Triple therapy
Observational
3 sites in Europe
Race: NR
Female: 28 %
Median age: 56 to 59
Total N: 27,972
8 % Stable CAD
12 % STEMI
79 % UA/NSTEMI
Population
Timeframe: 1997–2005
Funding: Government, Private foundation
Placebo
Warfarin
(N = 220)
(N = 580)
20 sites in Europe
ASA or thienopyridine
ASA ? thienopyridine
Observational
Dual therapy
Triple therapy
RIKS-HIA and SCAAR
Comparator (N)
Intervention (N)
Persson [19]
Race: NR
Female: 30 %
Median age: 64 to 66
Total N: 800
61 % STEMI
23 % NSTEMI
16 % UA
Population
Timeframe: 04/1999–09/2006
Funding: Industry
Study details
Study
Table 1 continued
Good
Good
Good
Quality
J. B. Washam et al.
Composite (secondary) Total mortality Nonfatal MI
warfarin 56.3 % of total population warfarin ? ASA ? clopidogrel 44.6 %
Funding: NR
Timeframe: 1/2001-3/2008
Major bleeding
Total N: 590
(N = 1,848)
(N = 3,768)
38 sites in Europe
Race: NR
Female: 38 %
Mean age: 75 to 79
Total N: 6275
29 % STEMI
Population
Timeframe: 1995–2002
Funding: Government, Private foundation
OAC
ASA
Observational
Race: NR
Female: 28.8 %
ASA and/or thienopyridine
Total mortality
Individual
Timing: 30 days, 1 yr
Total mortality
23 % STEMI
Thienopyridine
Individual
63 % NSTEMI
Mean age: 72
Target vessel failure
Population
Fair
Poor
ASA aspirin, CAD coronary artery disease, DAPT dual antiplatelet therapy, MACE major adverse cardiac event, MI myocardial infarction, NR not reported, NSTEMI non-ST elevation myocardial infarction, OAC oral anticoagulation, TT triple therapy, UA unstable angina, yr year/years
RIKS-HIA
Stenestrand [22]
Timing: 1 yr
Minor bleeding ASA 89.6 % of total population
Observational
NR sites in Europe
Non-OAC
Major bleeding
Revascularization
Nonfatal MI
Total mortality
Individual
Major bleeding MACE
Stroke
(secondary)
Total mortality Nonfatal MI Revascularization
(primary)
Coumarin ?ASA (N = 8), Coumarin ? clopidogrel (N = 16), coumarin N = 5
Good
Timing: 5 yr Composite
Quality
Timing outcomes reported
Warfarin ? ASA ? clopidogrel (N = 213),
Clopidogrel, ASA
Co-interventions
Clopidogrel 94 % of the total population
Race: NR
Female: 30 %
Mean age: 71
Total N: 426
16 % Stable CAD
20 % STEMI
64 % NSTEMI
Population
Timeframe: 01/2001–12/2006
Warfarin
(N = 184)
(N = 242)
2 sites in Europe
Ruiz-Nodar [21]
ASA
Warfarin
Observational
Ruiz-Nodar [20]
Funding: NR
Comparator (N)
Intervention (N)
Study details
Study
Table 1 continued
Dual antiplatelet therapy with or without oral anticoagulation
123
J. B. Washam et al. Study name
Statistics for each study Odds Lower Upper ratio limit limit p-Value
Konstantino, 2006 Karjalainen, 2007 Maegdefessel, 2008 Ruiz-Nodar, 2008 Jang, 2011 Persson, 2011 Fosbol, 2012 Lamberts, 2013
2.76 5.31 4.00 0.56 0.41 1.33 0.71 0.31 1.04
1.17 1.77 0.96 0.34 0.12 0.98 0.58 0.24 0.59
6.46 15.87 16.68 0.92 1.40 1.82 0.87 0.39 1.83
0.02 0.00 0.06 0.02 0.16 0.07 0.00 0.00 0.89
Events / Total Triple Therapy
Dual Therapy
6 / 76 19 / 219 6 / 56 32 / 178 3 / 84 44 / 1183 133 / 731 76 / 1896
82 / 2661 4 / 227 3 / 103 54 / 195 23 / 278 750 / 26789 679 / 2841 430 / 3590
Odds ratio and 95% CI
0.1 0.2
0.5
Favors Triple
1
2
5
10
Favors Dual
Evidence of extreme heterogeneity, with a Q-value of 87.83 for 7 degrees of freedom, p