Skip to contents

Overview summary statistics for the premature-death report

Source: R/pd_OverviewStats.R
pd_OverviewStats.Rd

`r lifecycle::badge("experimental")`

Computes the four headline numbers shown in the report's Overview table from the single classified-cohort source of truth ([pd_Classify()] output), so the table cannot drift from the bucket bars and scatter. "Premature" is the union of the two death categories (the `death30` / `death3190` keys of [pd_CategoryLevels()]); the ineligible share is taken over that same premature set, so numerator and denominator always describe one cohort.

Usage

pd_OverviewStats(dfClassified, dfExclusion = NULL, nWindowDays = 90)

Arguments

dfClassified

`data.frame` Output of [pd_Classify()]: one row per enrolled (randomized) subject, with `subjid`, `invid`, and `Category`.

dfExclusion

`data.frame` (optional) Mapped exclusion data with `subjid` and `Source` (as produced by `EXCLUSION.yaml`). When absent (or lacking a `Source` column), `has_eligibility` is `FALSE` and the ineligible counts are `NA` (the report renders the cell as a dash rather than asserting zero).

nWindowDays

`numeric` Premature-death window in days. Default 90.

Value

A named `list`: `nEnrolled`, `nSites`, `nPremature`, `nPrematureRate`, `nIneligible`, `nIneligibleRate`, `has_eligibility`, `nDeath30` (subjects who died within 30 days), `nDeath3190` (subjects who died within 31-90 days). Invariant: `nDeath30 + nDeath3190 == nPremature`.