GWG (version 1.0)

OW_above: Sensitivity, specificity, positive and negative likelihood ratios for prediction of excessive GWG for overweight women

Description

This dataset include the sensitivity and specificity for prediction of excessive GWG at the end of pregnancy from exceeding the respective cut-off values in each respective month as well as the likelihood ratio to assess how much the information on exceeding the cut-off values improves the a priori probabilities for excessive and adequate total GWG for overweight women (starting with the 2nd month or week 4/1-8/0). The data is based on 204 overweight women from two German cohorts (see Description).

Usage

data("OW_above")

Format

A data frame with 9 observations on the following 13 variables.

month

month of pregnancy

se.est

vector of sensitivity estimators

se.lower

vector of lower bound of the confidence interval of the sensitivity estimators

se.upper

vector of upper bound of the confidence interval of the sensitivity estimators

sp.est

vector of specificity estimators

sp.lower

vector of lower bound of the confidence interval of the specificity estimators

sp.upper

vector of upper bound of the confidence interval of the specificity estimators

lr.pos.est

vector of positive likelihood ratio estimators

lr.pos.lower

vector of lower bound of the confidence interval of the positive likelihood ratio estimators

lr.pos.upper

vector of upper bound of the confidence interval of the positive likelihood ratio estimators

lr.neg.est

vector of negative likelihood ratio estimators

lr.neg.lower

vector of lower bound of the confidence interval of the negative likelihood ratio estimators

lr.neg.upper

vector of upper bound of the confidence interval of the negative likelihood ratio estimators

Source

Knabl J, Riedel C, Gmach J et al. (2013). Prediction of excessive or inadequate gestational weight gain from week-specific IOM/NRC cut-off values. submitted.

Details

Exact binomial 95% confidence intervals (CI) were calculated for sensitivity and specificity and the 95% CI of the likelihood ratios were calculated as suggested by Simel et al. (1991).

References

Simel D, Samsa G, Matchar D (1991). Likelihood ratios with confidence: Sample size estimation for diagnostic test studies. Journal of Clinical Epidemiology 44.p 763 - 770.

Examples

data("OW_above")

#Plot of positive likelihood ratios for excessive GWG for overweight women 
plot(OW_above$month, OW_above$lr.pos.est,
     xlim=c(1,10),
     ylim=c(0,20),
     main="LR+ for excessive GWG",
     ylab="likelihood ratio",
     xlab="month",
     cex.main=2,
     font.main=1,
     cex.lab=1.9,
     cex.axis=1.5)
lines(OW_above$month, OW_above$lr.pos.est, lty=1,lwd=4)
lines(OW_above$month, OW_above$lr.pos.lower, lty=2,lwd=4)
lines(OW_above$month, OW_above$lr.pos.upper, lty=2,lwd=4)
abline(h=2,lwd=1, col="grey") 
abline(h=4,lwd=1, col="grey") 
abline(h=6,lwd=1, col="grey") 
abline(h=8,lwd=1, col="grey") 
abline(h=10,lwd=1, col="grey") 
abline(h=12,lwd=1, col="grey") 
abline(h=14,lwd=1, col="grey") 
abline(h=16,lwd=1, col="grey") 
abline(h=18,lwd=1, col="grey") 
axis(1, c(3,5,7,9), cex.axis=1.5)
legend("topleft", c("lr+ estimate", "lr+ confidence interval"),
       col=c("black", "black"), lty=c(1,2),
       bg="white", cex=1.5, lwd=c(3,3))