Table 2 Results of Bayesian network meta-analyses for exacerbation, mortality, and adverse events among LAMA and LABA combinations compared to tiotropium monotherapy in the network (A)

Total exacerbation (16 studies, 39,065 patients)

Rank

3

4

1

2

5

6

SUCRA

0.438

0.4346

0.8802

0.5393

0.389

0.3016

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.99 (0.6–1.6)

1

Umeclidinium/Vilanterol

0.42 (0.19–0.97)*

0.43 (0.2–0.94)*

1

Glycopyrrolate/Formoterol

0.92 (0.61–1.41)

0.94 (0.59–1.5)

2.18 (0.95–4.93)

1

Glycopyrrolate/Indacaterol

1.02 (0.74–1.45)

1.04 (0.68–1.66)

2.43 (1.1–5.36)*

1.11 (0.76–1.66)

1

Tiotropium/Salmeterol

1.12 (0.62–2.06)

1.14 (0.56–2.34)

2.66 (1–7.09)*

1.22 (0.62–2.37)

1.09 (0.59–2.02)

1

Moderate-to-severe exacerbation (7 studies, 27,489 patients)

Rank

3

2

4

1

5

–

SUCRA

0.4689

0.5119

0.4639

0.5844

0.3794

–

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.95 (0.18–5.04)

1

Umeclidinium/Vilanterol

1.05 (0–91,490)

1.09 (0–97,430)

1

Glycopyrrolate/Formoterol

0.9 (0.22–4.16)

0.95 (0.26–3.69)

0.86 (0–68,860)

1

Glycopyrrolate/Indacaterol

1.08 (0.26–4.86)

1.13 (0.25–5.38)

1.04 (0–80,330)

1.19 (0.35–4.09)

1

Tiotropium/Salmeterol

–

–

–

–

–

–

All-cause mortality (16 studies, 39,065 patients)

Rank

4

3

1

2

5

6

SUCRA

0.4301

0.4567

0.8358

0.5097

0.4057

0.3083

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.9 (0.05–21.92)

1

Umeclidinium/Vilanterol

0.03 (0–7.15)

0.03 (0–6.88)

1

Glycopyrrolate/Formoterol

0.74 (0.05–14.51)

0.82 (0.03–19.07)

29.98 (0.09–41,250)

1

Glycopyrrolate/Indacaterol

1.05 (0.17–8.7)

1.19 (0.07–18.77)

42.73 (0.18–46,040)

1.43 (0.09–21.95)

1

Tiotropium/Salmeterol

1.68 (0.1–27.64)

1.87 (0.04–59.48)

68.98 (0.16–92,860)

2.3 (0.06–59.11)

1.57 (0.08–22.96)

1

COPD-related mortality (7 studies, 25,900 patients)

Rank

3

1

4

2

5

–

SUCRA

0.4741

0.7123

0.4046

0.6009

0.2731

–

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.01 (0–1439)

1

Umeclidinium/Vilanterol

1.93 (0–237,100)

274 (0–7,690,000,000)

1

Glycopyrrolate/Formoterol

0.05 (0–75,270)

6.77 (0–858,000,000)

0.02 (0–1,910,000)

1

Glycopyrrolate/Indacaterol

41.3 (0–10,600,000)

6241 (0–386,000,000,000)

22.16 (0–451,000,000)

1015 (0–2,200,000,000,000)

1

Tiotropium/Salmeterol

–

–

–

–

–

–

Cardiovascular disease-related mortality (9 studies, 28,131 patients)

Rank

4

1

2

5

3

–

SUCRA

0.3452

0.6024

0.4396

0.3087

0.3966

–

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.04 (0–34)

1

Umeclidinium/Vilanterol

0.34 (0–11,300)

9.08 (0–2,050,000)

1

Glycopyrrolate/Formoterol

1.43 (0.02–244)

33.7 (0.08–125,700)

4.6 (0–158,000)

1

Glycopyrrolate/Indacaterol

1.19 (0–1,010,000)

31 (0–134,000,000)

3.7 (0–41,900,000)

0.77 (0–912,000)

1

Tiotropium/Salmeterol

–

–

–

–

–

–

Major adverse cardiovascular event (9 studies, 20,051 patients)

Rank

3

1

–

2

4

–

SUCRA

0.3162

0.6174

–

0.5209

0.1683

–

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.24 (0.01–8.83)

1

Umeclidinium/Vilanterol

–

–

–

Glycopyrrolate/Formoterol

0.4 (0.01–11.56)

1.64 (0.08–40.72)

–

1

Glycopyrrolate/Indacaterol

1.94 (0.09–88.14)

8.06 (0.35–443)

–

4.88 (0.24–190.6)

1

Tiotropium/Salmeterol

–

–

–

–

–

–

Pneumonia (16 studies, 39,065 patients)

Rank

3

2

1

5

4

6

SUCRA

0.5538

0.5906

0.6822

0.4295

0.4599

0.155

NMA estimate OR (95% CrI)

Tiotropium/Olodaterol

1

Aclidinium/Formoterol

0.85 (0.13–5.22)

1

Umeclidinium/Vilanterol

0.65 (0.04–11.25)

0.75 (0.05–12.24)

1

Glycopyrrolate/Formoterol

1.26 (0.32–5.38)

1.49 (0.25–9.62)

1.94 (0.11–33.85)

1

Glycopyrrolate/Indacaterol

1.17 (0.46–4.38)

1.4 (0.27–9.69)

1.84 (0.12–32.73)

0.93 (0.29–4.04)

1

Tiotropium/Salmeterol

24.85 (0.13–164,500)

30.82 (0.12–224,300)

42.81 (0.09–341,300)

19.79 (0.1–137,800)

20.54 (0.11–132,100)

1