Skip to contents

Confidence intervals for coherent network models

Usage

# S3 method for class 'nma_coherent'
confint(
  object,
  parm,
  level = 0.95,
  type = c("coefficients", "contrasts"),
  adjust = c("none", "bonferroni"),
  inference = NULL,
  test = NULL,
  robust_type = NULL,
  cluster = NULL,
  hksj_adjust = NULL,
  ...
)

Arguments

object

A fitted coherent network model.

parm

Coefficient names or indices for type="coefficients", or row indices for type="contrasts". Selection occurs after defining the full family.

level

Confidence level strictly between zero and one.

type

coefficients for basic effects and moderator coefficients, or contrasts for all pairwise treatment effects at moderator zero.

adjust

none for pointwise intervals or bonferroni for simultaneous intervals across all coefficients or all reported pair/stratum contrasts.

inference

Optional summary-only model or robust inference framework.

test

Optional summary-only coefficient/total-contrast test: z or Satterthwaite; robust inference also accepts naive-t and naive-tp.

robust_type

Optional summary-only sandwich covariance, e.g. CR2.

cluster

Optional fitted contrast-data cluster column, default study_id.

hksj_adjust

Unsupported for networks; supplied values raise an error.

...

Reserved; unused arguments are rejected.

Value

For coefficients, a two-column matrix with coefficient row names. For contrasts, a data frame with treatments, outcome/time strata, estimate, standard error, df, lower and upper limits, and family metadata attributes.

Details

Intervals condition on fitted heterogeneity. Model inference uses normal quantiles; CR1 uses a t reference with number of studies minus one df. Interval adjustment is explicit and independent of the fit's p-value adjustment. Holm-adjusted p-values do not imply Holm-adjusted confidence intervals.

References

Higgins, J. P. T., Jackson, D., Barrett, J. K., Lu, G., Ades, A. E., and White, I. R. (2012). Consistency and inconsistency in network meta-analysis: concepts and models for multi-arm studies. Research Synthesis Methods, 3, 98–110. doi:10.1002/jrsm.1044 .

White, I. R., Barrett, J. K., Jackson, D., and Higgins, J. P. T. (2012). Consistency and inconsistency in network meta-analysis: model estimation using multivariate meta-regression. Research Synthesis Methods, 3, 111–125. doi:10.1002/jrsm.1045 .

Examples

d <- data.frame(s=1:6,a=rep(c("A","A","B"),2),
  b=rep(c("B","C","C"),2),g=c(.2,.5,.3,.3,.6,.2),v=.04)
fit <- network_meta(d,"s","a","b","g","v",adjust="holm")
confint(fit)
#>           2.5%     97.5%
#> d_B 0.04034952 0.4929838
#> d_C 0.30701619 0.7596505
#> attr(,"level")
#> [1] 0.95
#> attr(,"adjust")
#> [1] "none"
#> attr(,"family_size")
#> [1] 2
confint(fit,type="contrasts",level=.90,adjust="bonferroni")
#>   treatment_1 treatment_2  estimate std_error  df      lower     upper
#> 1           A           B 0.2666667 0.1154701 Inf 0.02094117 0.5123922
#> 2           A           C 0.5333333 0.1154701 Inf 0.28760784 0.7790588
#> 3           B           C 0.2666667 0.1154701 Inf 0.02094117 0.5123922