Calculate Person-Years by Age Strata

This function calculates person-years of follow-up by age strata for cohort studies. It handles age transitions during follow-up by splitting person-time contributions across age groups as individuals age during the observation period.

Usage

calculate_person_years_by_age_strata(
  .df,
  age_cut_points = c(0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80,
    85, 90, 95, 100, 150)
)

Arguments

.df

A data frame containing individual-level follow-up data

age_cut_points

Numeric vector of age breakpoints defining the age strata. Default is 5-year age groups from 0 to 100+: c(0, 5, 10, ..., 95, 100, 150). This matches the WHO 2000-2025 Standard Population age groups. The final age group is displayed as "[100+)" regardless of the upper cut point value.

Value

A tibble containing:

• age_group: Age group labels (factor)

• person_years: Total person-years of follow-up in each age group

• events: Number of events in each age group

Details

The function requires the following columns in .df:

• patient_id: Unique identifier for each individual

• dob: Date of birth (Date class)

• study_entry_date: Date of study entry (Date class)

• study_exit_date: Date of study exit/end of follow-up (Date class)

• event_status: Event indicator (integer: 1 = event occurred, 0 = censored)

The function automatically handles individuals who transition between age groups during follow-up by splitting their person-time contributions appropriately.

Examples

# Example data
cohort_data <- data.frame(
  patient_id = 1:3,
  dob = as.Date(c("1960-01-15", "1955-06-20", "1972-09-01")),
  study_entry_date = as.Date(c("2020-03-01", "2021-01-01", "2019-01-01")),
  study_exit_date = as.Date(c("2024-06-15", "2025-12-31", "2022-04-10")),
  event_status = c(1L, 0L, 1L)
)

# Calculate person-years by 5-year age groups
result <- calculate_person_years_by_age_strata(cohort_data)

# Custom age groups
custom_ages <- c(0, 20, 40, 60, 80, 100)
result_custom <- calculate_person_years_by_age_strata(cohort_data, custom_ages)