Table 2. Results of Bayesian NMAs of exacerbations, mortality, and adverse events according to the drug class.
| NMA estimate OR (95% CrI) | Placebo | ICS/LAMA/LABA | LAMA/LABA | ICS/LABA | LAMA | LABA | ICS |
|---|---|---|---|---|---|---|---|
| Total exacerbations (208 studies, 226,117 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 0.57 (0.5–0.64) | 1 | |||||
| LAMA/LABA | 0.7 (0.65–0.76) | 1.24 (1.1–1.41) | 1 | ||||
| ICS/LABA | 0.74 (0.68–0.81) | 1.31 (1.16–1.47) | 1.06 (0.96–1.16) | 1 | |||
| LAMA | 0.77 (0.73–0.81) | 1.35 (1.2–1.53) | 1.09 (1.02–1.18) | 1.03 (0.95–1.13) | 1 | ||
| LABA | 0.86 (0.81–0.9) | 1.51 (1.34–1.72) | 1.22 (1.13–1.32) | 1.15 (1.06–1.25) | 1.12 (1.05–1.19) | 1 | |
| ICS | 0.84 (0.76–0.91) | 1.48 (1.28–1.7) | 1.19 (1.07–1.33) | 1.13 (1.02–1.25) | 1.09 (0.99–1.2) | 0.98 (0.89–1.07) | 1 |
| Moderate to severe exacerbations (67 studies, 116,131 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 0.56 (0.47–0.68) | 1 | |||||
| LAMA/LABA | 0.69 (0.6–0.79) | 1.22 (1.05–1.44) | 1 | ||||
| ICS/LABA | 0.71 (0.61–0.82) | 1.27 (1.07–1.48) | 1.04 (0.9–1.18) | 1 | |||
| LAMA | 0.73 (0.66–0.81) | 1.3 (1.1–1.56) | 1.06 (0.95–1.2) | 1.03 (0.9–1.19) | 1 | ||
| LABA | 0.85 (0.76–0.95) | 1.51 (1.27–1.82) | 1.24 (1.09–1.41) | 1.2 (1.05–1.37) | 1.17 (1.05–1.29) | 1 | |
| ICS | 0.87 (0.74–1.02) | 1.54 (1.25–1.91) | 1.26 (1.05–1.51) | 1.22 (1.04–1.44) | 1.19 (1–1.4) | 1.02 (0.87–1.19) | 1 |
| All-cause mortality (190 studies, 221,451 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 0.74 (0.59–0.93) | 1 | |||||
| LAMA/LABA | 0.98 (0.82–1.18) | 1.32 (1.05–1.66) | 1 | ||||
| ICS/LABA | 0.86 (0.76–0.98) | 1.16 (0.95–1.42) | 0.88 (0.73–1.05) | 1 | |||
| LAMA | 0.98 (0.88–1.12) | 1.33 (1.07–1.67) | 1.01 (0.86–1.19) | 1.15 (0.99–1.34) | 1 | ||
| LABA | 0.94 (0.83–1.07) | 1.27 (1.02–1.58) | 0.96 (0.8–1.16) | 1.1 (0.97–1.24) | 0.95 (0.82–1.1) | 1 | |
| ICS | 0.95 (0.82–1.09) | 1.29 (1.01–1.63) | 0.97 (0.79–1.2) | 1.11 (0.94–1.28) | 0.97 (0.8–1.14) | 1.01 (0.87–1.17) | 1 |
| Cardiovascular-disease–related mortality (109 studies, 142,622 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 0.68 (0.36–1.26) | 1 | |||||
| LAMA/LABA | 1.05 (0.66–1.67) | 1.54 (0.84–2.91) | 1 | ||||
| ICS/LABA | 0.81 (0.57–1.07) | 1.19 (0.67–2.09) | 0.77 (0.49–1.17) | 1 | |||
| LAMA | 1.13 (0.83–1.56) | 1.66 (0.87–3.29) | 1.08 (0.67–1.74) | 1.4 (0.98–2.11) | 1 | ||
| LABA | 0.91 (0.69–1.38) | 1.35 (0.73–2.76) | 0.87 (0.54–1.5) | 1.13 (0.84–1.8) | 0.81 (0.56–1.26) | 1 | |
| ICS | 0.85 (0.64–1.19) | 1.25 (0.67–2.49) | 0.81 (0.49–1.38) | 1.05 (0.78–1.58) | 0.75 (0.5–1.16) | 0.93 (0.63–1.29) | 1 |
| MACEs (26 studies, 52,881 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 0.68 (0.44–1.12) | 1 | |||||
| LAMA/LABA | 0.84 (0.59–1.26) | 1.23 (0.86–1.76) | 1 | ||||
| ICS/LABA | 0.68 (0.41–1.14) | 0.99 (0.67–1.43) | 0.81 (0.54–1.19) | 1 | |||
| LAMA | 0.84 (0.65–1.14) | 1.24 (0.82–1.81) | 1.01 (0.74–1.34) | 1.25 (0.79–1.96) | 1 | ||
| LABA | 0.67 (0.44–1.02) | 0.98 (0.58–1.61) | 0.8 (0.51–1.2) | 0.99 (0.58–1.68) | 0.79 (0.52–1.19) | 1 | |
| ICS | 0.33 (0.01–2.15) | 0.48 (0.01–3.01) | 0.4 (0.01–2.47) | 0.49 (0.02–2.99) | 0.39 (0.01–2.5) | 0.5 (0.02–3.16) | 1 |
| Pneumonia (147 studies, 195,628 patients) | |||||||
| Placebo | 1 | ||||||
| ICS/LAMA/LABA | 1.58 (1.26–2) | 1 | |||||
| LAMA/LABA | 1.05 (0.88–1.27) | 0.66 (0.54–0.83) | 1 | ||||
| ICS/LABA | 1.59 (1.36–1.91) | 1.01 (0.84–1.24) | 1.52 (1.29–1.79) | 1 | |||
| LAMA | 0.98 (0.85–1.13) | 0.62 (0.5–0.77) | 0.93 (0.79–1.08) | 0.61 (0.52–0.72) | 1 | ||
| LABA | 1.01 (0.87–1.18) | 0.64 (0.51–0.8) | 0.96 (0.81–1.13) | 0.63 (0.55–0.72) | 1.03 (0.89–1.2) | 1 | |
| ICS | 1.39 (1.15–1.7) | 0.88 (0.68–1.15) | 1.32 (1.05–1.65) | 0.87 (0.72–1.05) | 1.42 (1.15–1.76) | 1.38 (1.15–1.67) | 1 |
Median OR and 95% CrI were calculated as a row to column ratio.
Abbreviations: CrI, credible interval; ICS, inhaled corticosteroid; LABA, long-acting beta-agonist; LAMA, long-acting muscarinic antagonist; MACE, major adverse cardiovascular event; NMA, network meta-analysis; OR, odds ratio