RobergeEtAl2017 {bayesmeta}

Aspirin during pregnancy example data

Description

Numbers of cases (patients) and events (preeclampsia (PE) or fetal growth restriction (FGR)) in experimental and control groups of 45 studies.

Usage

data("RobergeEtAl2017")

Format

The data frame contains the following columns:

study	character	publication identifier (first author and publication year)
year	numeric	publication year
N	numeric	number of patients
onset	factor	treatment onset (up to 16 weeks' gestation or later)
dose	numeric	dose (mg/day)
control	factor	type of control group
asp.PE.events	numeric	number of PE events in aspirin group
asp.PE.total	numeric	number of PE cases in aspirin group
cont.PE.events	numeric	number of PE events in control group
cont.PE.total	numeric	number of PE cases in control group
asp.FGR.events	numeric	number of FGR events in aspirin group
asp.FGR.total	numeric	number of FGR cases in aspirin group
cont.FGR.events	numeric	number of FGR events in control group
cont.FGR.total	numeric	number of FGR cases in control group

Details

A systematic literature review was performed in order to summarize the evidence on effects of aspirin administered during pregnancy. Of particular interest were occurrences of preeclampsia (PE) and fetal growth restriction (FGR). A total of 45 relevant randomized controlled trials (RCTs) were found, out of which 40 reported on PE, and 35 reported on FGR. Besides event rates, the mode of administration (treatment onset (early vs. late) and dose (in mg)) was also recorded for each study.

Source

S. Roberge, K. Nicolaides, S. Demers, J. Hyett, N. Chaillet, E. Bujold. The role of aspirin dose on the prevention of preeclampsia and fetal growth restriction: systematic review and meta-analysis. American Journal of Obstetrics & Gynecology, 216(2):110-120, 2017. doi:10.1016/j.ajog.2016.09.076.

References

C. Roever, T. Friede. Using the bayesmeta R package for Bayesian random-effects meta-regression. Computer Methods and Programs in Biomedicine, 299:107303, 2023. doi:10.1016/j.cmpb.2022.107303.

See Also

bmr, escalc, model.matrix.

Examples

# load data:
data("RobergeEtAl2017")
str(RobergeEtAl2017)
head(RobergeEtAl2017)

# compute effect sizes (log odds ratios) from count data
# (using the "metafor" package's "escalc()" function);
# preeclampsia (PE):
es.pe  <- escalc(measure="OR",
                 ai=asp.PE.events,  n1i=asp.PE.total,
                 ci=cont.PE.events, n2i=cont.PE.total,
                 slab=study, data=RobergeEtAl2017,
                 subset=complete.cases(RobergeEtAl2017[,7:10]))
# show forest plot:
forestplot(es.pe, title="preeclampsia (PE)")
# show "bubble plot" (bubble sizes are
# inversely proportional to standard errors):
plot(es.pe$dose, es.pe$yi, cex=1/sqrt(es.pe$vi),
     col=c("blue","red")[as.numeric(es.pe$onset)],
     xlab="dose (mg)", ylab="log-OR (PE)", main="Roberge et al. (2017)")
legend("topright", col=c("blue","red"), c("early onset", "late onset"), pch=1)

# fetal growth restriction (FGR):
es.fgr <- escalc(measure="OR",
                 ai=asp.FGR.events,  n1i=asp.FGR.total,
                 ci=cont.FGR.events, n2i=cont.FGR.total,
                 slab=study, data=RobergeEtAl2017,
                 subset=complete.cases(RobergeEtAl2017[,11:14]))
# show forest plot:
forestplot(es.fgr, title="fetal growth restriction (FGR)")
# show "bubble plot":
plot(es.fgr$dose, es.fgr$yi, cex=1/sqrt(es.fgr$vi),
     col=c("blue","red")[as.numeric(es.fgr$onset)],
     xlab="dose (mg)", ylab="log-OR (FGR)", main="Roberge et al. (2017)")
legend("topright", col=c("blue","red"), c("early onset", "late onset"), pch=1)

## Not run: 
# set up regressor matrix (common intercept and slope):
X01 <- model.matrix(~ dose, data=es.pe)
colnames(X01) <- c("intercept", "slope")
print(X01)

# perform regression:
bmr01 <- bmr(es.pe, X=X01)
bmr01$summary

# set up alternative regressor matrix
# (individual intercepts and slopes for two subgroups):
X02 <- model.matrix(~ -1 + onset + onset:dose, data=es.pe)
colnames(X02) <- c("intEarly", "intLate", "slopeEarly", "slopeLate")
print(X02)

# perform regression:
bmr02 <- bmr(es.pe, X=X02)
bmr02$summary

# derive predictions from the model;
# specify corresponding "regressor matrices":
newx.early <- cbind(1, 0, seq(50, 150, by=5), 0)
newx.late  <- cbind(0, 1, 0, seq(50, 150, by=5))
# (note: columns correspond to "beta" parameters)

# compute predicted medians and 95 percent intervals: 
pred.early <- cbind("median"=bmr02$qpred(0.5, x=newx.early),
                    bmr02$pred.interval(x=newx.early))
pred.late <- cbind("median"=bmr02$qpred(0.5, x=newx.late),
                    bmr02$pred.interval(x=newx.late))

# draw "bubble plot": 
plot(es.pe$dose, es.pe$yi, cex=1/sqrt(es.pe$vi),
     col=c("blue","red")[as.numeric(es.pe$onset)],
     xlab="dose (mg)", ylab="log-OR (PE)", main="Roberge et al. (2017)")
legend("topright", col=c("blue","red"), c("early onset", "late onset"), pch=1)
# add predictions to bubble plot:
matlines(newx.early[,3], pred.early, col="blue", lty=c(1,2,2))
matlines(newx.late[,4], pred.late, col="red", lty=c(1,2,2))


## End(Not run)