Table 1.
Characteristics of included reviews
| Review | Disease | Treatment | Comparator | Endpoint | Number of RCTs | Number of Observational Studies | RCT pooled effect estimate (95% CI) | Observational pooled effect estimate (95% CI) |
|---|---|---|---|---|---|---|---|---|
| Abuzaid (2018) [17] | Severe aortic stenosis | Dual anti-platelet therapy (DAPT) | Single anti-platelet therapy (SAPT) | 30-day all-cause mortality | 3 | 7 | RR 1.2 (0.5–2.89) | RR 1.19 (0.74–1.91) |
| Abuzaid (2018) [17] | Severe aortic stenosis | DAPT | SAPT | Longest reported all-cause mortality | 3 | 7 | RR 1.14 (0.54–2.42) | RR 1.06 (0.52–2.18) |
| Abuzaid (2018) [17] | Severe aortic stenosis | DAPT | SAPT | Major bleeding | 3 | 7 | RR 1.74 (0.52–5.82) | RR 2.23 (1.36–3.65) |
| Agarwal (2019) [18] | Acute coronary syndrome (ACS), coronary artery disease (CAD) | Dual therapy | Triple therapy | Major bleeding | 3 | 3 | RR 0.53 (0.38–0.76) | RR 0.88 (0.46–1.67) |
| Agarwal (2018) [19] | CAD | DAPT | Aspirin | Primary outcome: mid- to long-term (> 30 days) composite of myocardial infarction (MI), stroke, or death | 8 | 4 | RR 0.43 (0.17–1.11) | RR 0.85 (0.72–1.01) |
| An (2019) [20] | Severe aortic stenosis | Antiplatelet | Anticoagulation | Mortality | 2 | 5 | RR 0.82 (0.33–2.03) | RR 0.47 (0.18–1.22) |
| An (2019) [20] | Severe aortic stenosis | Antiplatelet | Anticoagulation | Stroke/transient ischemic attack (TIA) | 2 | 5 | RR 0.9 (0.35–2.33) | RR 0.57 (0.31–1.03) |
| An (2019) [20] | Severe aortic stenosis | Antiplatelet | Anticoagulation | Thromboembolic events | 2 | 5 | RR 1.13 (0.51–2.49) | RR 0.71 (0.38–1.32) |
| An (2019) [20] | Severe aortic stenosis | Antiplatelet | Anticoagulation | Bleeding | 2 | 5 | RR 0.34 (0.11–1.04) | RR 0.34 (0.2–0.58) |
| Chien (2020)* [21] | Multi-drug resistant gram-negative bacteria (MDR-GNB) infections | Colistin | Other antibiotics | Colistin-associated acute kidney injury (CA-AKI) | 1 | 19 | OR 2.75 (0.43–17.49) | Not reported |
| Chien (2020) [21] | MDR-GNB infections | Colistin monotherapy | Colistin combination therapy | Acute kidney injury (AKI) | 3 | 6 | OR 1.77 (1.17–2.66) | OR 1.15 (0.76–1.76) |
| Chopra (2012)* [22] | Pneumonia | Statin therapy | No statin therapy | Unadjusted all-cause mortality following an episode of pneumonia | 1 | 9 | OR 0.84 (0.32–2.18) | Not reported |
| Chopra (2012) [22] | Pneumonia | Statin therapy | No statin therapy | Adjusted all-cause mortality following an episode of pneumonia | 1 | 11 | OR 0.84 (0.32–2.18) | OR 0.66 (0.55–0.79) |
| Desai (2016) [23] | Ankylosing spondylitis, inflammatory bowel diseases, juvenile idiopathic arthritis, plaque psoriasis, psoriatic arthritis, and rheumatoid arthritis | Adalimumab | Etanercept | Discontinuation due to adverse events | 1 | 3 | RR 0.83 (0.39–1.78) | Adjusted HR 1.67 (1.26–2.22) |
| Desai (2016) [23] | Ankylosing spondylitis, inflammatory bowel diseases, juvenile idiopathic arthritis, plaque psoriasis, psoriatic arthritis, rheumatoid arthritis | Adalimumab | Infliximab | Discontinuation due to adverse events | 1 | 5 | RR 6.17 (0.78–48.71) | Adjusted HR 0.57 (0.46–0.7) |
| Gandhi (2015) [24] | Aortic stenosis | DAPT | Mono-antiplatelet therapy (MAPT). | Combined end point of 30-day major stroke, spontaneous MI, all-cause mortality, and combined lethal and major bleeding | 2 | 2 | OR 0.98 (0.46–2.11) | OR 3.02 (1.91–4.76) |
| Ge (2018) [25] | Atrial fibrillation | Novel oral anticoagulants (NOACs) | Vitamin K antagonists (VKAs) | Major bleeding events (Fixed effects model) | 4 | 25 | OR 0.3 (0.14–0.62) | OR 0.68 (0.48–0.95) |
| Ge (2018) [25] | Atrial fibrillation | NOACs | VKAs | Thromboembolic events (Fixed effects model) | 4 | 25 | OR 0.14 (0.01–1.3) | OR 0.91 (0.49–1.67) |
| Heffernan (2020) [26] | Serious infections | β-lactam/aminoglycoside combination therapy | β-lactam monotherapy | All-cause mortality | 2 | 4 | OR 3.18 (0.79–12.73) | OR 0.79 (0.64–0.99) |
| Ho (2013) [27] | Kidney disease (kidney transplant) | Once daily tacrolimus | Twice daily tacrolimus | Biopsy-proven acute rejection (RCT: at 6 months; Observational: mean follow-up ranges from 3 months to 672 months) | 4 | 5 | RR 1.18 (0.82–1.68) | RR 0.83 (0.39–1.78) |
| Ho (2013) [27] | Kidney disease (kidney transplant) | Once daily tacrolimus | Twice daily tacrolimus |
Biopsy-proven acute rejection (RCT: at 12 months Observational: mean follow-up ranges from 3 months to 672 months) |
2 | 5 | RR 1.24 (0.93–1.65) | RR 0.83 (0.39–1.78) |
| Ho (2013) [27] | Kidney disease (kidney transplant) |
RCT: twice daily tacrolimus Observational: once daily tacrolimus |
RCT: once daily tacrolimus Observational: twice daily tacrolimus |
Patient survival (RCT: at 6 months Observational: mean follow-up ranges from 3.5 months to 12 months) |
2 | 2 | RR 1.03 (1–1.06) | RR 1.02 (0.94–1.1) |
| Ho (2013) [27] | Kidney disease (kidney transplant) |
RCT: twice daily tacrolimus Observational: once daily tacrolimus |
RCT: once daily tacrolimus Observational: twice daily tacrolimus |
Patient survival (RCT: at 12 months Observational: mean follow-up ranges from 3.5 months to 12 months) |
3 | 2 | RR 0.99 (0.97–1.02) | RR 1.02 (0.94–1.1) |
| Khan (2019) [28] | CAD | Proton pump inhibitor (PPI) | No PPI | All-cause mortality | 3 | 24 | RR 1.35 (0.56–3.23) | RR 1.25 (1.11–1.41) |
| Kirson (2013)* [29] | Schizophrenia | Depot antipsychotics | Oral antipsychotics | Varies across studies: hospitalization, relapse, discontinuation | 5 | 8 | RR 0.89 (0.64–1.22) | Not reported |
| Land (2017) [30] | Psychiatric illnesses | Clozapine | Control drugs (other antipsychotics) | Hospitalization | 3 | 19 | RR 0.62 (0.41–0.94) | RR 0.75 (0.69–0.81) |
| Li (2016) [31] | Diabetes | Dipeptidyl peptidase-4 (DPP-4) inhibitors |
RCT: control Observational: sulfonylurea (SU) |
Heart failure | 38 | 1 | OR 0.97 (0.61–1.56) | Unadjusted OR 0.88 (0.22–3.48) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors |
RCT: control Observational: SU |
Heart failure | 38 | 1 | OR 0.97 (0.61–1.56) | Adjusted HR 1.10 (1.04–1.17) |
| Li (2016) [31] | Diabetes |
RCT: DPP-4 inhibitors Observational: sitagliptin |
RCT: control Observational: SU |
Heart failure | 38 | 1 | OR 0.97 (0.61–1.56) | Unadjusted OR 0.39 (0.02–6.26) |
| Li (2016) [31] | Diabetes |
RCT: DPP-4 inhibitors Observational: sitagliptin |
RCT: control Observational: no sitagliptin use |
Heart failure | 38 | 1 | OR 0.97 (0.61–1.56) | Adjusted OR 0.75 (0.38–1.46) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors |
RCT: control Observational: active control |
Hospital admission for heart failure | 5 | 6 | OR 1.13 (1.00–1.26) | Adjusted OR 0.85 (0.74–0.97) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors |
RCT: control Observational: SU |
Hospital admission for heart failure | 5 | 3 | OR 1.13 (1.00–1.26) | Adjusted HR 0.84 (0.74–0.96) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors |
RCT: control Observational: pioglitazone |
Hospital admission for heart failure | 5 | 2 | OR 1.13 (1.00–1.26) | Adjusted HR 0.67 (0.57–0.78) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors |
RCT: control Observational: other oral antidiabetics |
Hospital admission for heart failure | 5 | 1 | OR 1.13 (1.00–1.26) | Adjusted OR 0.88 (0.63–1.22) |
| Li (2016) [31] | Diabetes | DPP-4 inhibitors | Control | Hospital admission for heart failure | 5 | 1 | OR 1.13 (1.00–1.26) | Adjusted HR 0.58 (0.38–0.88) |
| Li (2016) [31] | Diabetes |
RCT: DPP-4 inhibitors Observational: sitagliptin |
RCT: control Observational: no sitagliptin use |
Hospital admission for heart failure | 5 | 2 | OR 1.13 (1.00–1.26) | Adjusted OR 1.41 (0.95–2.09) |
| Li (2016) [31] | Diabetes |
RCT: DPP-4 inhibitors Observational: sitagliptin |
RCT: control Observational: no sitagliptin use |
Hospital admission for heart failure | 5 | 1 | OR 1.13 (1.00–1.26) | Adjusted HR 1.21 (1.04–1.42) |
| Li (2016) [31] | Diabetes |
RCT: DPP-4 inhibitors Observational: sitagliptin |
RCT: control Observational: no sitagliptin use |
Hospital admission for heart failure | 5 | 1 | OR 1.13 (1.00–1.26) | Adjusted OR 1.84 (1.16–2.92) |
| Melloni (2015) [32] | Unstable angina/non–ST-segment–elevation myocardial infarction (UA/NSTEMI) |
RCT: omeprazole Observational: any PPI |
RCT: placebo Observational: no PPI |
Composite ischemic endpoint at ≈ 1 year | 1 | 20 | HR 0.99 (0.68–1.44) | Adjusted HR 1.35 (1.18–1.54) |
| Melloni (2015) [32] | UA/NSTEMI |
RCT: omeprazole Observational: any PPI |
RCT: placebo Observational: no PPI |
Nonfatal MI at ≈ 1 year | 1 | 10 | HR 0.92 (0.44–1.9) | HR 1.331 (1.146–1.547) |
| Miles (2019) [33] | Heart failure | Furosemide | Torsemide | All-cause mortality | 5 | 3 | OR 1.12 (0.7–1.8) | OR 0.97 (0.44–2.13) |
| Miles (2019) [33] | Heart failure | Furosemide | Torsemide | Heart failure readmissions | 4 | 1 | OR 2.04 (1.16–3.60) | OR 2.91 (0.78–10.91) |
| Miles (2019) [33] | Heart failure | Furosemide | Torsemide | New York Heart Association class improvement | 7 | 2 | OR 0.91 (0.61–1.35) | OR 0.65 (0.50–0.85) |
| Mongkhon (2019) [34] | Atrial fibrillation | OAC | Non-OAC | Risk of dementia | 1 | 4 | RR 1.31 (0.79–2.18) | RR 0.75 (0.67–0.83) |
| Mongkhon (2019) [34] | Atrial fibrillation | VKA | Non-VKA | Risk of dementia | 1 | 4 | RR 1.31 (0.79–2.18) | RR 0.71 (0.68–0.74) |
| Raheja (2018) [35] | Aortic stenosis | DAPT | SAPT | All-cause mortality | 3 | 2 | RR 1.07 (0.48–2.41) | RR 1.34 (0.51–3.48) |
| Raheja (2018) [35] | Aortic stenosis | DAPT | SAPT | Stroke or TIA | 3 | 2 | RR 0.93 (0.28–3.06) | RR 1.25 (0.32–4.92) |
| Raheja (2018) [35] | Aortic stenosis | DAPT | SAPT | MI | 3 | 2 | RR 3.62 (0.60–21.76) | RR 1.18 (0.14–9.98) |
| Raheja (2018) [35] | Aortic stenosis | DAPT | SAPT | Major/life-threatening bleeding | 3 | 3 | RR 1.75 (0.88–3.50) | RR 3.24 (1.82–5.75) |
| Ramjan (2014) [36] | HIV | Fixed-dose combination (FDC) antiretroviral therapy (ART) | Separate tablet regimens | Virological suppression | 4 | 2 | RR 1.04 (0.98–1.10) | RR 1.07 (0.97–1.18) |
| Ramjan (2014) [36] | HIV | FDC ART | Separate tablet regimens | Adherence to ART | 5 | 2 | RR 1.1 (0.98–1.22) | RR 1.17 (1.07–1.28) |
| Shi (2014) [37] | Liver cancer | Statins | Placebo/non-use | Liver cancer | 1 | 11 | RR 1.06 (0.66–1.71) | RR 0.57 (0.50–0.64) |
| Teo (2014) [38] | Acute infections | Prolonged infusion, which was defined as administration of either extended infusion or continuous infusion of beta-lactam antibiotics | Identical beta-lactams that were administered as intermittent boluses (20–60 min infusion) according to the manufacturer’s package insert | All-cause in-hospital mortality | 10 | 9 | RR 0.83 (0.57–1.21) | RR 0.57 (0.43–0.76) |
| Teo (2014) [38] | Acute infections | Prolonged infusion, which was defined as administration of either extended infusion or continuous infusion of beta-lactam antibiotics. | Identical beta-lactams that were administered as intermittent boluses (20–60 min infusion) according to the manufacturer’s package insert | Clinical success (cure or improvement) | 14 | 5 | RR 1.05 (0.99–1.12) | RR 1.34 (1.02–1.76) |
| Vinceti (2018) [39] | Cancer | Highest selenium exposure | Lowest selenium exposure | Total (any) cancer incidence | 3 | 7 | RR 1.01 (0.93–1.10) | OR 0.72 (0.55–0.93) |
| Vinceti (2018) [39] | Cancer | Highest selenium exposure | Lowest selenium exposure | Cancer mortality | 1 | 7 | RR 1.02 (0.80–1.30) | OR 0.76 (0.59–0.97) |
| Vinceti (2018) [39] | Colorectal cancer | Highest selenium exposure | Lowest selenium exposure | Colorectal cancer risk | 2 | 1 | RR 0.99 (0.69–1.43) | OR 0.80 (0.68–0.94) |
| Vinceti (2018) [39] | Lung cancer | Highest selenium exposure | Lowest selenium exposure | Lung cancer risk | 2 | 5 | RR 1.16 (0.89–1.50) | OR 0.74 (0.43–1.28) |
| Vinceti (2018) [39] | Breast cancer | Highest selenium exposure | Lowest selenium exposure | Breast cancer risk | 1 | 8 | RR 2.04 (0.44–9.55) | OR 1.09 (0.87–1.37) |
| Vinceti (2018) [39] | Bladder cancer | Highest selenium exposure | Lowest selenium exposure | Bladder cancer risk | 2 | 2 | RR 1.07 (0.76–1.52) | OR 0.65 (0.46–0.92) |
| Vinceti (2018) [39] | Prostate cancer | Highest selenium exposure | Lowest selenium exposure | Prostate cancer risk | 4 | 21 | RR 1.01 (0.90–1.14) | OR 0.84 (0.75–0.95) |
| Wang (2019) [40] | Pneumonia | PPI | No PPI | Pneumonia | 10 | 48 | OR 1.13 (0.71–1.78) | OR 1.45 (1.32–1.59) |
| Wat (2019) [41] | Traumatic brain injury (TBI) | Antiepileptic drugs | Placebo/no treatment | Early seizures after TBI | 3 | 6 | RR 0.58 (0.20–1.72) | RR 0.42 (0.29–0.62) |
| Wong (2017)* [42] | Coronary heart disease/CAD | Macrolides | Placebo/no treatment | Short-term primary outcome (defined as cardiac mortality, cardiovascular mortality, sudden death, cardiac arrest, all-cause mortality, or composite outcomes including death and/or other cardiovascular events or procedures) | 5 | 15 | RR 0.99 (0.74–1.34) | Not reported |
| Wong (2017) [42] | Coronary heart disease/CAD | Macrolides | Placebo/no treatment | Long term primary outcome (defined as cardiac mortality, cardiovascular mortality, sudden death, cardiac arrest, all-cause mortality, or composite outcomes including death and/or other cardiovascular events or procedures) | 14 | 8 | RR 1.03 (0.96–1.10) | RR 1.05 (0.91–1.22) |
| Yang (2019)* [43] | Cancer | Epoetin alfa biosimilar drugs | Epoetin alfa drugs | Mean of hemoglobin increase | 1 | 4 | WMD -0.02 (− 0.38–0.34) | WMD 0.07 (− 0.12–0.25) |
| Yang (2019) [43] | Cancer | Epoetin alfa biosimilar drugs | Epoetin alfa drugs | Hemoglobin response | 1 | 1 | RR 1.09 (0.86–1.38) | RR 1.18 (0.87–1.60) |
| Yang (2019) [43] | Breast cancer | Granulocyte colony-stimulating factor (G-CSF) biosimilar drugs | G-CSF drugs | Febrile neutropenia in cycle 1 | 5 | 3 | RR 1.14 (0.80–1.63) | RR 1.36 (0.84–2.23) |
| Yang (2019) [43] | NHL | G-CSF biosimilar drugs | G-CSF drugs | Febrile neutropenia in cycle 1 | 1 | 1 | RR 0.54 (0.20–1.46) | RR 0.87 (0.20–3.85) |
| Yang (2019) [43] | Cancer | G-CSF biosimilar drugs (filgrastim biosimilars) | G-CSF drugs | Bone pain | 4 | 4 | RR 0.90 (0.78–1.05) | RR 0.86 (0.59–1.24) |
| Yu (2018) [44] | Non-cardiac vascular disease | Statins | Placebo/no statin treatment | All-cause mortality | 3 | 6 | OR 0.62 (0.41–0.92) | OR 0.65 (0.48–0.88) |
| Yu (2018) [44] | Non-cardiac vascular disease | Statins | Placebo/no statin treatment | Primary patency | 1 | 10 | OR 0.39 (0.09–1.65) | OR 0.77 (0.59–0.99) |
| Yu (2018) [44] | Non-cardiac vascular disease | Statins | Placebo/no statin treatment | Amputation | 1 | 10 | OR 0.47 (0.07–2.94) | OR 0.64 (0.50–0.83) |
| Yu (2018) [44] | Non-cardiac vascular disease | Statins | Placebo/no statin treatment | Cardiovascular events | 3 | 2 | OR 0.55 (0.36–0.83) | OR 0.87 (0.16–4.60) |
| Zhang (2019) [45] | Atrial fibrillation | NOAC | Non-NOAC therapy | Renal impairment | 11 | 3 | HR 0.82 (0.71–0.93) | HR 0.64 (0.58–0.69) |
| Zhao (2018) [46] | CAD | DAPT | SAPT | Any bleeding events | 5 | 8 | RR 1.25 (0.98–1.59) | RR 0.87 (0.76–1.01) |
| Zhao (2018) [46] | CAD | DAPT | SAPT | Minor bleeding events | 4 | 3 | RR 1.15 (0.73–1.81) | RR 0.84 (0.37–1.93) |
| Zhao (2018) [46] | CAD | DAPT | SAPT | Major bleeding events | 5 | 6 | RR 1.28 (0.95–1.71) | RR 0.99 (0.66–1.51) |
| Zhao (2018) [46] | CAD | DAPT | SAPT | Major bleeding events during hospitalization (random effects model) | 3 | 3 | RR 1.27 (0.91–1.78) | RR 0.50 (0.12–2.09) |
*Not included in the analysis