Uptake of new drugs in the early post-approval period in the Mini-Sentinel distributed database

Purpose Several factors limit the statistical power of drug safety surveillance during the early post‐approval period, including uptake of the drug and lag in data availability. This study characterized new drug uptake in the Mini‐Sentinel Distributed Database and determined statistical power to det...

Full description

Saved in:
Bibliographic Details
Published in:Pharmacoepidemiology and drug safety 2016-09, Vol.25 (9), p.1023-1032
Main Authors: Mott, Katrina, Graham, David J., Toh, Sengwee, Gagne, Joshua J., Levenson, Mark, Ma, Yong, Reichman, Marsha E.
Format: Article
Language:English
Subjects:
active surveillance
Computer Communication Networks
Databases, Factual
distributed databases
Drug Approval
drug safety
Drug-Related Side Effects and Adverse Reactions - epidemiology
FDA approval
Humans
new molecular entities
Pharmaceutical Preparations - administration & dosage
pharmacoepidemiology
Pharmacoepidemiology - methods
Poisson Distribution
Prescription drugs
Product safety
Product Surveillance, Postmarketing - methods
Product Surveillance, Postmarketing - statistics & numerical data
Risk Assessment - methods
Sentinel
Sentinel Surveillance
Surveillance
Time Factors
United States
United States Food and Drug Administration
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Purpose Several factors limit the statistical power of drug safety surveillance during the early post‐approval period, including uptake of the drug and lag in data availability. This study characterized new drug uptake in the Mini‐Sentinel Distributed Database and determined statistical power to detect levels of risk in post‐launch safety assessments. Methods The cumulative exposure among initiators of 46 new molecular entities approved from 2008 to 2011 was assessed. Using a Poisson estimation method, minimum incidence rate ratios (IRRs) detectable, with 80% power, were calculated under varying background incidence rates. Results Twelve products (26.1%) had more than 15 000 new users after 2 years. With comparator group incidence rate of 1/1000 person‐years, 16 (33.3%) products had enough exposure to detect an IRR of 5 with 24 months of data collected that would be available for assessment at 33 months post‐launch. With an incidence rate of 5/1000 person‐years, 23 (50%) products had enough exposure to detect an IRR of ≥3 with 2 years of data collected. At 33 months post‐launch, only two (4.3%) of the drugs examined had enough data availability to detect IRR of
ISSN:1053-8569
1099-1557
DOI:10.1002/pds.4013