HIV: HIV viral load study

In qrNLMM: Quantile Regression for Nonlinear Mixed-Effects Models

HIV viral load study

Description

The data set belongs to a clinical trial (ACTG 315) studied in previous researches by Wu (2002) and Lachos et al. (2013). In this study, we analyze the HIV viral load of 46 HIV-1 infected patients under antiretroviral treatment (protease inhibitor and reverse transcriptase inhibitor drugs). The viral load and some other covariates were mesured several times days after the start of treatment been 4 and 10 the minimum and maximum number of measures per patient respectively.

Usage

data(HIV)

Format

This data frame contains the following columns:

patid

a numeric vector indicating the patient register number.

ind

a numeric vector indicating the number patient on which the measurement was made. It represents the subject number in the study.

day

time in days.

cd4

cd4 count in cells/mm^{3}.

lgviral

viral load in log10 scale.

cd8

cd8 count in cells/mm^{3}.

Details

In order to fit the nonlinear data we sugest to use the Nonlinear model proposed by Wu (2002) and also used by Lachos et al. (2013).

Source

Wu, L. (2002). A joint model for nonlinear mixed-effects models with censoring and covariates measured with error, with application to aids studies. Journal of the American Statistical association, 97(460), 955-964.

Lachos, V. H., Castro, L. M. & Dey, D. K. (2013). Bayesian inference in nonlinear mixed-effects models using normal independent distributions. Computational Statistics & Data Analysis, 64, 237-252.

Examples

## Not run: 
data(HIV)
attach(HIV)

y     = lgviral            #response
x     = day/100            #time
covar = cd4/100            #covariate

#Nonlinear model used in Lachos(2013)

#Full Nonlinear expression
exprNL = expression(log(exp(fixed[1]+random[1])*exp(-(fixed[2]+random[2])*x)+
                    exp(fixed[3]+random[3])*exp(-(fixed[4]+random[4]+fixed[5]
                    *covar[1])*x))/log(10))

#Initial values for fixed effects
initial = c(12,31,6,-2,0.6)

#A median regression (by default)
median_reg = QRNLMM(y,x,ind,initial,exprNL,covar)

## End(Not run)

qrNLMM documentation built on Aug. 18, 2022, 5:05 p.m.