poutcome {pandemics}R Documentation

Probability of Outcome by Cause Specific.

Description

From the two-dimensional estimated hazards of deaths and recoveries, the probability that a person, who has been in hospital for a number of days, leaves the hospital alive or death, depending on the date of admission.

Usage

poutcome(hi1.zt,hi2.zt,z1)

Arguments

hi1.zt

a matrix with the estimated hazard of deaths (M times M).

hi2.zt

a matrix with the estimated hazard of recoveries (M times M).

z1

(optional) a vector of indexes between 1 and M indicating the admission days to evaluate the probabilities. If missing then z1<-c(seq(1,M-1,by=2),M-1).

Value

alive.zt

a matrix (M times M) with the computed probabilities of leaving the hospital alive (each column corresponds to a day in z1).

death.zt

a matrix (M times M) with the computed probabilities of dying in hospital (each column corresponds to a day in z1).

Note

Evaluate the function hazard2Dmiss to create the estimated hazards for the arguments hi1.zt and hi2.zt.

Author(s)

M.L. Gámiz, E. Mammen, M.D. Martínez-Miranda and J.P. Nielsen.

References

Gámiz, M.L., Mammen, E., Martínez-Miranda, M.D. and Nielsen, J.P. (2024b). Monitoring a developing pandemic with available data. arXiv:2308.09919.

See Also

hazard2Dmiss

Examples


data('covid')
Ei.z<-covid$Hospi   # exposure for survival analysis
Oi1.z<-covid$Death  # deaths
Oi2.z<-covid$Recov  # recoveries
# compute incremental values
Oi1.z<-diff(Oi1.z)
Oi2.z<-diff(Oi2.z)
Ei.z<-Ei.z[-1]     # exposure is left as cumulative
M<-length(Ei.z)
t.grid<-z.grid<-1:M
# notification date (marker)
ddates<-covid$Date

## First compute the estimated hazard
bs<-t(c(150,150))
res.h<-hazard2Dmiss(t.grid,z.grid,Oi1.z,Oi2.z,Ei.z,bs.grid=bs,cv=FALSE)
hi1.zt<-res.h$hi1.zt # 2D-hazard of deaths
hi2.zt<-res.h$hi2.zt # 2D-hazard of recoveries

## Now the probabilities at few values of the marker (admission dates)
z1<-c(13,44,105,197)
zdates<-ddates[z1]
nz<-length(z1)
t.min<-35  # maximum duration is 35
ti<-1:t.min;n0<-length(ti)
res<-poutcome(hi1.zt,hi2.zt,z1)
alive.zt<-res$alive.zt
death.zt<-res$death.zt

# Cause= recovery
plot(ti,alive.zt[1:n0,1],ylim=c(0,1),lwd=2,type='l',
     main='Probability to get out alive',
     ylab='',xlab='Time from admission (days)')
for(i in 2:nz) lines(ti,alive.zt[1:n0,i-1],lwd=3,col=i,lty=i)
legend('bottom',legend=zdates,lty=c(2:nz,1),
  lwd=c(rep(3,nz-1),2),col=c(2:nz,1),bty='n')

# Cause= death
plot(ti,death.zt[1:n0,1],ylim=c(0,1),lwd=2,type='l',
     main='Probability of dying in the hospital',
     ylab='',xlab='Time from admission (days)')
for(i in 2:nz) lines(ti,death.zt[1:n0,i-1],lwd=3,col=i,lty=i)
legend('top',legend=zdates,lty=c(2:nz,1),lwd=c(rep(3,nz-1),2),
       col=c(2:nz,1),bty='n')


[Package pandemics version 0.1.0 Index]