Level 2 Analyses

Level 2 analyses:

  • Identify cohorts of interest
  • Perform more complex adjustment for confounding
  • Generate effect estimates and confidence intervals 

Below you will find descriptions of the different types of cohort identification strategies and tools for level 2 analysis. 

What this program does:

  • Identifies exposures, follow-up time, outcomes, and covariates
  • Estimates a propensity score (based on predefined covariates and/or via a high-dimensional propensity score approach)
  • Uses matching, stratification, inverse probability of treatment weighting (IPTW), or stratum weighting for confounder adjustment and follows the analytic cohort for outcome assessment in a survival analysis framework

Output metrics include:

  • Tables of patient characteristics (unadjusted and adjusted cohorts)
  • Measures of covariate balance
  • Distribution of propensity score by exposure group
  • Hazard ratios (with 95% confidence intervals)
  • Kaplan-Meier curves

​​​Continue reading about propensity score analysis on Sentinel's Git Repository. 

What this program does: 

  • Conducts covariate stratification within each Sentinel Data Partner site via distributed programming code; returns data to the Sentinel Operations Center (SOC), aggregated, and used to calculate effect estimates
  • Groups patients in exposed and comparator cohorts into strata defined by unique values for any requester-defined combination of sex, age group, and/or year of index date

​​​​​
Continue reading about covariate stratification on Sentinel's Git Repository. 

What this program does:

  • Identifies exposure of a medical product of interest
  • Defines risk and control windows relative to the exposure date
  • Examines the occurrence of health outcomes of interest during the risk and control windows

Output metrics include:

  • Number of exposure episodes
  • Exposed individuals
  • Individuals with a health outcome of interest in the risk and/or control windows
  • Censored individuals overall

Stratified by requester-defined:

  • Age group
  • Sex
  • Year
  • Year-month
  • Time-to-event (in days)
  • Race (available upon request)
  • Ethnicity (available upon request)
  • Geographic region (available upon request)

This program provides estimates of relative risk (RR) and 95% confidence intervals.

Continue reading about self-controlled risk interval design on Sentinel's Git Repository. 

What this program does:

  • Identifies exposures, outcomes, and covariates in a population of pregnant women that had live birth deliveries
  • Estimates a propensity score (based on predefined covariates and/or via a high-dimensional propensity score approach)
  • Uses matching or stratification for confounder adjustment and a binary outcome assignment framework for maternal outcome assessment among the pregnant mothers or birth outcome assessment among the infants

Output metrics include:

  • Tables of patient characteristics (unadjusted and adjusted cohorts)
  • Measures of covariate balance
  • Distribution of propensity score by exposure group
  • Hazard ratios (with 95% confidence intervals)

Continue reading about propensity score analysis on Sentinel's Git Repository. 

What this program does: 

  • Conducts covariate stratification within each Sentinel Data Partner site via distributed programming code; returns data to the Sentinel Operations Center (SOC), aggregated, and used to calculate effect estimates
  • Groups patients in exposed and comparator cohorts into strata defined by unique values for any requester-defined combination of sex, age group, and/or year of index date

Continue reading about covariate stratification on Sentinel's Git Repository. ​​​​​​

Want more details on the functional and technical documentation of each level 2 analysis? Visit Sentinel's Git Repository.   

RSS Feed Scroll to Top