daoh

R package for calculating Days Alive and Out of Hospital (DAOH)

Supports all three hospital-time algorithms (nights, days, exact) and all three death-handling approaches (midday, midnight, zero), with tools for comparison, and visualisation.

Installation

# From source (once on CRAN or GitHub)
# install.packages("daoh")

# From local source
devtools::install("path/to/daoh_package/daoh")

Quick start

library(daoh)

# Load the built-in day-stay example
ex <- load_example("daystay")

# Nights algorithm: same-day admission = 0 nights → DAOH = 30
calc_daoh(ex$events, ex$index_dates, period = 30, method = "nights")

# Days algorithm: same-day admission = 1 day → DAOH = 29
calc_daoh(ex$events, ex$index_dates, period = 30, method = "days")

Key functions

Function	Purpose
`calc_daoh()`	Main DAOH calculation
`load_example()`	Load synthetic example datasets
`bland_altman_daoh()`	Bland–Altman agreement statistics
`daoh_icc()`	Intraclass correlation across methods
`daoh_reclassify()`	Quartile reclassification analysis
`plot_daoh_dist()`	Distribution plot (Figure 2 style)
`plot_daoh_ba()`	Bland–Altman plot
`plot_daoh_reclassify()`	Reclassification heatmap

Examples

Three synthetic datasets are included:

"daystay" — single day-stay illustrating the nights vs days difference
"death" — four admissions + in-period death (Figure 1 worked example)
"population" — 500-patient cohort for method comparison

See vignette("getting_started", package = "daoh") for a full walkthrough.

Citation

DOI: 10.5281/zenodo.20671491