Advanced time interval calculator designed for survival analysis, epidemiological studies, and person-time analysis. Features intelligent date parsing, comprehensive data quality assessment, landmark analysis, and robust statistical summaries. Time intervals form the foundation of person-time follow-up calculations, capturing both participant counts and observation duration for accurate incidence rate calculations.
Usage
timeinterval(
data,
dx_date,
fu_date,
time_format = "auto",
output_unit = "months",
use_landmark = FALSE,
landmark_time = 6,
remove_negative = FALSE,
remove_extreme = FALSE,
add_times = FALSE,
include_quality_metrics = FALSE,
confidence_level = 95
)Arguments
- data
The data as a data frame containing date columns for interval calculation.
- dx_date
Column containing start dates (e.g., diagnosis date, study entry, treatment start). Supports various date formats including text and numeric representations.
- fu_date
Column containing end dates (e.g., follow-up date, event date, study exit). Must be in the same format as the start date variable.
- time_format
Date format specification. 'Auto-detect' attempts to identify the format automatically. Manual selection ensures accurate parsing for specific date formats.
- output_unit
Unit for calculated time intervals. Affects person-time calculations and statistical summaries. Choose based on study duration and event frequency.
- use_landmark
Enables conditional analysis from a specific time point. Useful for studying outcomes after a landmark time (e.g., 6-month survivors only).
- landmark_time
Time point for landmark analysis in the specified output units. Only participants surviving past this time are included in analysis.
- remove_negative
Automatically exclude negative time intervals (end date before start date). Recommended for data quality assurance.
- remove_extreme
Identify and flag potentially extreme time intervals for quality review. Uses statistical outlier detection methods.
- add_times
Appends calculated time intervals as a new variable for downstream analysis. Useful for subsequent survival analysis or person-time calculations.
- include_quality_metrics
Provides comprehensive data quality assessment including missing values, negative intervals, and distribution statistics.
- confidence_level
Confidence level for statistical intervals (mean confidence intervals). Standard epidemiological practice uses 95\ confidence intervals.
Value
A results object containing:
results$todo | a html | ||||
results$personTimeInfo | a html | ||||
results$qualityAssessment | a html | ||||
results$summary | a html | ||||
results$calculated_time | an output |
Examples
# Basic time interval calculation:
timeinterval(
data = study_data,
dx_date = "diagnosis_date",
fu_date = "followup_date",
time_format = "ymd",
output_unit = "months"
)
#> Error: object 'study_data' not found
# With landmark analysis:
timeinterval(
data = study_data,
dx_date = "start_date",
fu_date = "end_date",
use_landmark = TRUE,
landmark_time = 6,
output_unit = "months"
)
#> Error: object 'study_data' not found