Skip to main content

Table 2 Univariable and multivariable regression using the modified PRISMA-NMA

From: Do reporting guidelines have an impact? Empirical assessment of changes in reporting before and after the PRISMA extension statement for network meta-analysis

Covariates

Interpretation of the coefficient

Coefficient (95% CI)

Sample size

Univariable analyses and subgroups

Published after 2015 vs until 2015

Average increase in the score after 2015

1.25 (0.59, 1.91)

Before 2015: 389

After 2015: 755

Year of publication, subgroup: only NMAs published before 2015

Average increase in the score per year

0.96 (0.32, 1.59)

Year of publication, subgroup: only NMAs published after 2015

Average increase in the score per year

0.53 (0.02, 1.04)

Multivariable analyses with year as a continuous variable

Year of publication

Average increase in the score per year

0.34 (0.16, 0.52)

Year 2013: 91 (reference group)

Year 2014: 104

Year 2015: 194

Year 2016: 198

Year 2017: 316

Year 2018: 241

Treatment type

Average increase in the score if network includes pharmacological treatments

 − 0.66 (− 1.34, 0.02)

Pharmacological treatments: 907

Non-pharmacological treatments (reference group): 237

Funding type

Average increase in the score if non-sponsored/publicly sponsored

1.34 (0.56, 2.11)

Non-sponsored/publicly sponsored/not reported: 974

Industry/mixed sponsored (reference group): 170

Review type

Average increase in the score if protocol is not available/reported

 − 5.12 (− 5.74, − 4.49)

With protocol (reference group): 313

Without protocol: 831

Impact factor

Average increase in the score per impact factor increase (1 unit)

0.10 (0.07, 0.13)

 

Multivariable analyses with year as a dichotomous variable

Year of publication

Average increase in the score per year

0.81 (0.23, 1.39)

Before 2015 (reference group):389

After 2015:755

Treatment type

Average increase in the score if network includes pharmacological treatments

 − 0.72 (− 1.40, − 0.04)

Pharmacological treatments: 907

Non-pharmacological treatments (reference group): 237

Funding type

Average increase in the score if non-sponsored/publicly sponsored

1.36 (0.58, 2.14)

Non-sponsored/publicly sponsored/Not reported: 974

Industry/mixed sponsored (reference group): 170

Review type

Average increase in the score if protocol is not available/reported

 − 5.18 (− 5.8, − 4.55)

With protocol (reference group): 313

Without protocol: 831

Impact factor

Average increase in the score per impact factor increase (1 unit)

0.10 (0.06, 0.13)

 
  1. CI confidence interval