bolus: Bolus data

In cold: Count Longitudinal Data

Description

The dataset has the number of requests per interval in 12 successive four-hourly intervals following abdominal surgery for 65 patients in a clinical trial to compare two groups (bolus/lock-out combinations).

Usage

data("bolus")

Format

A data frame with 780 observations on the following 4 variables.

id

identifies de number of the individual profile. This vector contains observations of 65 individual profiles.

group

a factor with levels 1mg and 2mg.

time

a numeric vector that identifies the number of the time points observed.

y

a numeric vector with the number of analgesic doses taken by hospital patients in 12 successive four-hourly intervals.

Details

The group 2mg has 30 patients and the group 1mg has 35 patients.

Source

Weiss, Robert E. (2005). Modeling Longitudinal Data. Springer

https://robweiss.faculty.biostat.ucla.edu/book-data-sets

References

Henderson, R. and Shimakura, S. (2003). A Serially Correlated Gamma Frailty Model for Longitudinal Count Data. Biometrika, vol. 90, No. 2, 355–366

Examples

data(bolus)

## change the reference class
contrasts(bolus$group)
bolus$group<-relevel(factor(bolus$group), ref = "2mg")
contrasts(bolus$group)

## Weiss, Robert E. (2005) pp 353-356, compare with Table 11.2

bol0R <- cold(y ~ time + group, random = ~ 1, data = bolus, 
dependence = "indR")
summary (bol0R)


## reparametrization of time 
bolus$time1 <- bolus$time - 6

 
bol0R1 <- cold(y ~ time1 + group, random = ~ 1,data = bolus, 
dependence = "indR")
summary (bol0R1)

bol1R1 <- cold(y ~ time1 + group, random = ~ 1, data = bolus, 
time = "time1", dependence = "AR1R")
summary (bol1R1)

anova(bol0R1, bol1R1)

plot(bol1R1, which = 1, factor = group, ylab = "Bolus count")

cold documentation built on Aug. 25, 2021, 5:06 p.m.