Study-level covariates can be included in the model to adjust treatment effects following an approach for meta-regression outlined in NICE Technical Support Document 3 (Dias et al. 2011). This can be used to explore and account for potential effect modification.
Following the definition in NICE Technical Support Document 3, network meta-regression can be expressed as an interaction on the treatment effect in arms ≥ 2:
θi, k = μi + (f(x, βai, k) − f(x, βai, 1)) + (ψ1, ai, k − ψ1, ai, 1)
where θi, k is the linear predictor, μi is the baseline effect on arm 1, f(x, βai, k) is the dose-response function at dose x with dose-response parameters βai, k for agent a in arm k of study i. ψ1, ai, k is then the effect modifying interaction between the agent in arm k and the network reference agent (typically Placebo in a dose-response analysis).
To improve estimation:
# Using the SSRI dataset
ssri.reg <- ssri
# For a continuous covariate
ssri.reg <- ssri.reg %>%
dplyr::mutate(x.weeks = weeks - mean(weeks, na.rm = TRUE))
# For a categorical covariate
table(ssri$weeks) # Using 8 weeks as the reference
ssri.reg <- ssri.reg %>%
dplyr::mutate(r.weeks = factor(weeks, levels = c(8, 4, 5, 6, 9, 10)))
# Create network object
ssrinet <- mbnma.network(ssri.reg)
#> Values for `agent` with dose = 0 have been recoded to `Placebo`
#> agent is being recoded to enforce sequential numbering
For performing network meta-regression, different assumptions can be made regarding how the effect modification may be shared across agents:
The least constraining assumption available in MBNMAdose
is to assume that the effect modifier acts on each agent independently,
and separate ψ1, ai, k
are therefore estimated for each agent in the network.
A slightly stronger assumption is to assume that agents within the same class share the same interaction effect, though classes must be specified within the dataset for this.
# Regress for continuous weeks Separate effect modification for each agent vs
# Placebo
ssrimod.a <- mbnma.run(ssrinet, fun = dfpoly(degree = 2), regress = ~x.weeks, regress.effect = "agent")
summary(ssrimod.a)
#> ========================================
#> Dose-response MBNMA
#> ========================================
#>
#> Likelihood: binomial
#> Link function: logit
#> Dose-response function: fpoly
#>
#> Pooling method
#>
#> Method: Common (fixed) effects estimated for relative effects
#>
#>
#> beta.1 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|-------:|
#> |citalopram |beta.1[2] | -0.0924| -0.3600| 0.1629|
#> |escitalopram |beta.1[3] | 0.1228| -0.1396| 0.3956|
#> |fluoxetine |beta.1[4] | 0.2685| 0.0361| 0.5053|
#> |paroxetine |beta.1[5] | -0.1374| -0.4402| 0.1690|
#> |sertraline |beta.1[6] | 1.4649| -7.7676| 10.5839|
#>
#>
#> beta.2 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|------:|
#> |citalopram |beta.2[2] | 0.0678| -0.0037| 0.1401|
#> |escitalopram |beta.2[3] | -0.0057| -0.0930| 0.0784|
#> |fluoxetine |beta.2[4] | -0.0654| -0.1334| 0.0018|
#> |paroxetine |beta.2[5] | 0.0912| -0.0066| 0.1863|
#> |sertraline |beta.2[6] | -0.1973| -1.1940| 0.7963|
#>
#>
#> power.1 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> power.2 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> Meta-regression
#>
#> Covariates interacting with study-level relative effects: x.weeks
#> Common (identical) covariate-by-agent effects
#>
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:------------|-------:|--------:|-------:|
#> |citalopram |B.x.weeks[2] | -0.1602| -0.3322| 0.0142|
#> |escitalopram |B.x.weeks[3] | 0.1632| -0.0216| 0.3602|
#> |fluoxetine |B.x.weeks[4] | -0.0287| -0.1720| 0.1122|
#> |paroxetine |B.x.weeks[5] | 0.0884| -0.0242| 0.1956|
#> |sertraline |B.x.weeks[6] | 1.8568| -17.6337| 21.0663|
#>
#>
#> Model Fit Statistics
#> Deviance = 890.2
#> Residual deviance = 190.5
#> Deviance Information Criterion (DIC) = 967.5
Within the output, a separate parameter (named
B.x.weeks[]
) has been estimated for each agent that
corresponds to the effect of an additional week of study follow-up on
the relative effect of the agent versus Placebo. Note that due to the
inclusion of weeks as a continuous covariate, we are assuming a linear
effect modification due to study follow-up.
Alternatively, the effect modification for different agents versus the network reference agent can be assumed to be exchangeable/shared across the network about a common mean, ψ̂, with a between-agent standard deviation of τψ:
ψ1, ai, k ∼ N(ψ̂, τψ2)
# Regress for continuous weeks Random effect modification across all agents vs
# Placebo
ssrimod.r <- mbnma.run(ssrinet, fun = dfpoly(degree = 2), regress = ~x.weeks, regress.effect = "random")
summary(ssrimod.r)
#> ========================================
#> Dose-response MBNMA
#> ========================================
#>
#> Likelihood: binomial
#> Link function: logit
#> Dose-response function: fpoly
#>
#> Pooling method
#>
#> Method: Common (fixed) effects estimated for relative effects
#>
#>
#> beta.1 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|------:|
#> |citalopram |beta.1[2] | -0.1103| -0.4700| 0.1732|
#> |escitalopram |beta.1[3] | 0.2579| -0.0151| 0.7551|
#> |fluoxetine |beta.1[4] | 0.2391| -0.0372| 0.4998|
#> |paroxetine |beta.1[5] | -0.0872| -0.4046| 0.2959|
#> |sertraline |beta.1[6] | 0.6106| 0.0431| 1.1175|
#>
#>
#> beta.2 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|------:|
#> |citalopram |beta.2[2] | 0.0736| -0.0027| 0.1730|
#> |escitalopram |beta.2[3] | -0.0408| -0.2114| 0.0504|
#> |fluoxetine |beta.2[4] | -0.0560| -0.1340| 0.0250|
#> |paroxetine |beta.2[5] | 0.0787| -0.0417| 0.1820|
#> |sertraline |beta.2[6] | -0.1032| -0.2082| 0.0125|
#>
#>
#> power.1 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> power.2 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> Meta-regression
#>
#> Covariates interacting with study-level relative effects: x.weeks
#> Random (exchangeable) covariate-by-treatment effects
#>
#>
#> |Regression effect |Parameter | Median| 2.5%| 97.5%|
#> |:-----------------|:---------|------:|-------:|------:|
#> |Random effect |B.x.weeks | 0.0267| -0.0676| 0.1219|
#>
#>
#> Standard deviation for random covariate-by-treatment effects
#>
#> |Parameter | Median| 2.5%| 97.5%|
#> |:------------|------:|------:|------:|
#> |sd.B.x.weeks | 0.0743| 0.0067| 0.2624|
#>
#>
#> Model Fit Statistics
#> Deviance = 890.2
#> Residual deviance = 190.5
#> Deviance Information Criterion (DIC) = 977.5
In this case only a single regression paramter is estimated
(B.x.weeks
), which corresponds to the mean effect of an
additional week of study follow-up on the relative effect of an active
agent versus Placebo. A parameter is also estimated for the
between-agent standard deviation, sd.B.x.weeks
.
This is the strongest assumption for network meta-regression, and it implies that effect modification is common (equal) for all agents versus the network reference agent:
ψ1, ai, k = ψ̂
# Regress for categorical weeks Common effect modification across all agents vs
# Placebo
ssrimod.c <- mbnma.run(ssrinet, fun = dfpoly(degree = 2), regress = ~r.weeks, regress.effect = "common")
summary(ssrimod.c)
#> ========================================
#> Dose-response MBNMA
#> ========================================
#>
#> Likelihood: binomial
#> Link function: logit
#> Dose-response function: fpoly
#>
#> Pooling method
#>
#> Method: Common (fixed) effects estimated for relative effects
#>
#>
#> beta.1 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|------:|
#> |citalopram |beta.1[2] | -0.0691| -0.3335| 0.2097|
#> |escitalopram |beta.1[3] | 0.2496| 0.0301| 0.4732|
#> |fluoxetine |beta.1[4] | 0.2223| -0.0183| 0.4646|
#> |paroxetine |beta.1[5] | -0.1144| -0.4138| 0.1934|
#> |sertraline |beta.1[6] | 0.5965| 0.1323| 1.0675|
#>
#>
#> beta.2 dose-response parameter results
#>
#> Pooling: relative effects for each agent
#>
#> |Agent |Parameter | Median| 2.5%| 97.5%|
#> |:------------|:---------|-------:|-------:|-------:|
#> |citalopram |beta.2[2] | 0.0594| -0.0155| 0.1295|
#> |escitalopram |beta.2[3] | -0.0297| -0.1072| 0.0476|
#> |fluoxetine |beta.2[4] | -0.0569| -0.1250| 0.0119|
#> |paroxetine |beta.2[5] | 0.0880| -0.0105| 0.1812|
#> |sertraline |beta.2[6] | -0.1032| -0.2003| -0.0063|
#>
#>
#> power.1 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> power.2 dose-response parameter results
#>
#> Assigned a numeric value: 0
#>
#> Meta-regression
#>
#> Covariates interacting with study-level relative effects: r.weeks4, r.weeks5, r.weeks6, r.weeks9, r.weeks10
#> Common (identical) covariate-by-treatment effects
#>
#>
#> |Regression effect |Parameter | Median| 2.5%| 97.5%|
#> |:-----------------|:-----------|-------:|-------:|-------:|
#> |Common effect |B.r.weeks10 | 0.1480| -0.2519| 0.5494|
#> |Common effect |B.r.weeks4 | -0.7300| -1.4624| -0.0377|
#> |Common effect |B.r.weeks5 | 0.8164| -0.2470| 1.8572|
#> |Common effect |B.r.weeks6 | 0.0533| -0.1328| 0.2522|
#> |Common effect |B.r.weeks9 | 0.4120| -0.4235| 1.2815|
#>
#>
#> Model Fit Statistics
#> Deviance = 890.9
#> Residual deviance = 191.2
#> Deviance Information Criterion (DIC) = 967.2
In this case we have performed the network meta-regression on study follow-up (weeks) as a categorical covariate. Therefore, although only a single parameter is estimated for each effect modifying term, there is a separate term for each category of week and a linear relationship for effect modification is no longer assumed.
Although this is beyond the capability of MBNMAdose
, one
could envision a more complex model in which the interaction effect also
varied by a dose-response relationship, rather than assuming an effect
by agent/class or across the whole network. This would in principle
contain fewer parameters than a fully independent interaction model (in
which a separate regression covariate is estimated for each treatment in
the dataset).
Note that adjusting for aggregated patient-level covariates
(e.g. mean age, % males, etc.) whilst using a non-identity link function
can introduce aggregation bias. This is a form of ecological bias that
biases treatment effects towards the null and is typically more severe
where treatment effects are strong and where the link function is highly
non-linear (Dias et al. 2011). This can be
resolved by performing a patient-level regression, but Individual
Participant Data are required for this and such an analysis is outside
the scope of MBNMAdose
.
Models fitted with meta-regression can also be used to make
predictions for a specified set of covariate values. This includes when
estimating relative effects using get.relative()
. An
additional argument regress.vals
can be used to provide a
named vector of covariate values at which to make predictions.
# For a continuous covariate, make predictions at 5 weeks follow-up
pred <- predict(ssrimod.a, regress.vals = c(x.weeks = 5))
plot(pred)
Predictions are very uncertain for Sertraline, as studies only investigated this agent at 6 weeks follow-up and therefore the agent-specific effect modification is very poorly estimated.
# For a categorical covariate, make predictions at 10 weeks follow-up
regress.p <- c(r.weeks10 = 1, r.weeks4 = 0, r.weeks5 = 0, r.weeks6 = 0, r.weeks9 = 0)
pred <- predict(ssrimod.c, regress.vals = regress.p)
plot(pred)
Note that categorical covariates are modelled as multiple binary dummy covariates, and so a value for each of these must be included.