Skip to contents

Decompose network residual Q into within-design and between-design variation

Usage

# S3 method for class 'nma_q_decomposition'
print(x, ..., digits = getOption("digits"), row.names = FALSE)

# S3 method for class 'nma_q_decomposition'
summary(object, ..., digits = getOption("digits"))

network_q_decomposition(object)

Arguments

x

A result from network_q_decomposition().

...

Additional arguments forwarded to print.data.frame().

digits

Decimal places to display, an integer from 1 to 22; defaults to getOption("digits"). Stored values retain full precision. P-values below 0.0001 print as <0.0001; other p-values use at least four decimal places.

row.names

Logical; print row names (default FALSE).

object

A fitted coherent network model.

Value

A three-row data frame with component, statistic, df, p_value, and covariance, of class nma_q_decomposition. Printing omits the repeated covariance column and states the sampling-only assumption once above the results; the returned column is retained for programmatic use. Zero residual df has a missing p-value.

Details

The printed question asks whether residual variation occurs within designs, between designs, or both. Each row has a distinct null hypothesis: total: study effects follow the coherent mean model with sampling variation only; within_designs: study effects follow the design-specific mean model with sampling variation only; between_designs: design-specific means are compatible with a coherent network. The printout states these hypotheses before the corresponding component statistics.

Uses the supplied sampling covariance alone, without adding fitted heterogeneity. The coherent mean model is nested inside a design-by-treatment mean model, retaining the original moderator slopes and mean strata. GLS residual sums of squares give total and within-design Q; their difference is between-design incoherence Q. Degrees of freedom use the corresponding mean ranks. For models with shared moderator slopes this is a conditional residual decomposition, rather than a sum of separately fitted design analyses. Existing local design-deviation tests are overlapping and are not additive.

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 .

White, I. R. (2015). Network meta-analysis. The Stata Journal, 15(4), 951–985. Section 3.6, Table 1. doi:10.1177/1536867X1501500403 .

Qin, L., Zhao, S., Guo, W., Tong, T., and Yang, K. (2024). A comparison of two models for detecting inconsistency in network meta-analysis. Research Synthesis Methods, 15(6), 851–871. doi:10.1002/jrsm.1734 .

Examples

dat <- data.frame(study = 1:6, a = rep(c("A", "A", "B"), 2),
                  b = rep(c("B", "C", "C"), 2),
                  yi = c(.2, .5, .3, .3, .6, .4), vi = .04)
fit <- network_meta(dat, "study", "a", "b", "yi", "vi")
network_q_decomposition(fit)
#> Network residual Q decomposition
#> Question: Does the network show residual variation within designs, between designs, or both?
#> H0 (total): Study effects follow the coherent mean model, with sampling variation only.
#> H0 (within_designs): Study effects follow the design-specific mean model, with sampling variation only.
#> H0 (between_designs): Design-specific means are compatible with a coherent network.
#> Sampling covariance only.
#> 
#>        component statistic df   p_value
#>            total 0.4166667  4 0.9810898
#>   within_designs 0.3750000  3 0.9453571
#>  between_designs 0.0416667  1 0.8382565