Level 1 Modular Program Queries

Level 1 modular program queries identify cohorts of interest and, for some cohorts, can perform unadjusted and minimally adjusted (i.e., by Data Partner, age group, sex, and year) analyses. The table below provides a description of the different types of cohort identification strategies and tools that can be used to perform a level 1 analysis.

For more details on the functional and technical documentation of each type, please visit Sentinel's Git Repository located externally. The Git Repository serves as Sentinel's version control tracking system for analytic packages and technical documentation.

Cohort
Identification Type

Cohort
Identification 

Strategy

Description

Type 1

Calculate Background Rates

The program identifies an event (exposure, outcome, condition) and calculates the rate of that event in the Sentinel Distributed Database (SDD). Output includes the number of individuals with the event, eligible members, and eligible member-days. Rates are calculated and reported overall and stratified by requester-defined age group, sex, year, and year-month. An attrition table is provided upon request.

Continue reading about Background Rate Calculation on Sentinel's Git Repository.

Type 2

Identify
Exposures and
Follow-Up Time

The program identifies an exposure of interest, determines exposed time (either requester-defined number of days after treatment initiation or based on drug dispensings’ days supply), and looks for the occurrence of a health outcome of interest (HOI) during exposed time. Output metrics include number of exposure episodes and number of individuals, number of events, and days at-risk. Events per days at-risk are calculated and reported overall and stratified by requester-defined age group, sex, year, and year-month. Incidence rate ratios (IRRs) can be calculated using two identified cohorts (e.g., exposed vs. active-comparator cohort). Unadjusted IRRs and adjusted IRRs (adjusted by age group, sex, year, and Data Partner using Poisson regression) are reported upon request. An attrition table is also provided upon request.

Continue reading about Exposures and Follow-Up Time on Sentinel's Git Repository.

Type 2: Concomitant Use Tool Characterize Concomitant Exposures and Follow-Up Time

The Concomitant Use Tool identifies concomitant exposure to two medical products, creates concomitant treatment episodes, and looks for the occurrence of a health outcome of interest (HOI) during exposed time. Concomitant exposure is defined as overlapping exposure to two medical products, with options to add exposure extension periods and to restrict to episodes of a minimum duration. Output metrics include number of treatment episodes and number of patients, number of events, and days at risk. Events per days at risk are calculated.

Continue reading about Concomitant Use Tool on Sentinel's Git Repository.

Type 2: Multiple Events Tool Characterize Secondary Events
Relative to the Exposure
 

The Multiple Events Tool allows the requester to specify a primary treatment episode, define an observation window relative to that primary episode, and evaluate the occurrence of multiple secondary events. Events can be defined as an interval (i.e., an episode) or as a single point in time. The tool gives users the flexibility to specify the observation window to be before, during or after the primary treatment episode. Secondary cohort events are only considered if they fall in a requester-defined observation window. 

Continue reading about Multiple Events Tool on Sentinel's Git Repository.

Type 2: Treatment Overlap Tool Characterize the Overlap between
Primary and Secondary Exposures 

The Overlap Tool can be used to characterize the overlap between primary and secondary treatment episodes during the observation window. The observation window is user-defined relative to the first primary treatment episode, during which the occurrence of secondary episodes are evaluated. The tool gives users the flexibility to specify the observation window to be before, during or after the primary treatment episode. Secondary episodes are only considered if they fall in a requester-defined observation window. Along with overlap between primary and secondary treatment episodes during the observation window, the user is also able to optionally assess for “adherence” to user-defined thresholds for the % or days overlap between two treatment episodes.

Continue reading about Treatment Overlap Tool on Sentinel's Git Repository.

Type 4 Identify
Pregnancy Episodes and Medical
Product Use

The program identifies live births, computes pregnancy episodes based on those live birth events, and assesses the use of specific medical products both during pregnancy episodes and in a comparator group of women likely to not have delivered a live birth during the same time frame. Output includes the number of pregnancy episodes stratified by year, maternal age, and existence of a pre-term or postterm pregnancy code. Medical product use is reported for both pregnancy episodes and comparator episodes according to trimester of use, gestational week, maternal age, and calendar year of delivery.

Continue reading about Pregnancy Episodes on Sentinel's Git Repository.

Type 5 Identify
Medical
Product Utilization

The program identifies the “first valid” exposure episode (i.e., the first episode during the query period that meets cohort entry criteria) as the index date, and then includes all subsequent exposure episodes. Output metrics include the number of patients, episodes, dispensings, and days supply by sex, age group and month of study start (for the first patient episode or all observed episodes during the query period); number of episodes by episode number, episode length, sex and age group, reason(s) for censoring; number of episode gaps by gap number, gap length, sex and age group.

Continue reading about Medical Product Utilization on Sentinel's Git Repository.

Type 6 Identify
Manufacturer-Level Product
Utilization and Switching Patterns

The program identifies product groups by user-defined lists of product codes (e.g., NDCs) grouped together to represent distinct manufacturer-level products and then characterizes patterns of drug use. Output metrics include counts of users and dispensings, days supplied per dispensing, episode duration, as well as time to uptake. The Cohort Identification and Descriptive Analysis (CIDA) tool also performs a product switching analysis that evaluates patient-level switching behavior between manufacturer-level product groups.

Continue reading about Manufacturer-Level Product Utilization and Switching Patterns on Sentinel's Git Repository.