152. A model to characterize the incidence of common adverse events in a vaccine-eligible population
Session: Poster Abstract Session: Adult and Pediatric Vaccines
Thursday, October 3, 2013
Room: The Moscone Center: Poster Hall C
Posters
  • p Bhuyan et al 2013 eControls .pdf (1.9 MB)
  • Background:  The evaluation of safety for randomized controlled vaccine trials is a complex undertaking that requires contextual information  to distinguish between expected and unexpected adverse events.  The incidence of adverse events often comes from large observational databases rather than clinical studies.  A model (eControls) is proposed that extends our previously reported ePlacebo methodology by integrating safety data from the control arms of clinical studies to estimate the background incidence of adverse events for a healthy, vaccine-eligible population. 

    Methods:  Disparate clinical data sources were integrated from finalized studies to include data only from medically stable subjects ≥50 years of age to approximate an older adult, vaccine study population.  Subjects were excluded if they had conditions that would have excluded them from typical vaccine studies. Adverse event incident rates were calculated for up to six months.  A placebo subset was used as an internal indicator of potential bias.

    Results:  We report creation of a database that integrates safety data from 66 randomized clinical trials encompassing ~50,000 medically stable historical controls. The frequent (>2%) adverse events observed in eControls were headache (4.77%), dyspepsia (4.11%), diarrhea (3.57%), upper respiratory infections (2.81%), nausea (2.70%), abdominal pain (2.61%), and peripheral edema  (2.14%).  The event with the greatest difference from the placebo subset was headache.

    Conclusion:  The eControls database  provides large  historical control data to enable better anticipate safety findings in vaccine clinical development programs.  Being derived from clinical studies, this database is also qualitatively different from other large databases typically used to estimate incidence rates.  The eControls approach enables the anticipation of common, expected adverse events for a given population and has the potential to add significant efficiencies to vaccine development.

    Prakash Bhuyan, MD, PhD1, Jigar Desai, PhD2, Martin Carlsson3, Matthew V. St.Louis4, Edward Bowen, MS4 and Michael N. Cantor, MD5, (1)Vaccine Research, Pfizer Inc., Collegeville, PA, (2)Worldwide Research and Development Business Technology, Pfizer Inc, New York, NY, (3)Primary Care Statistics, Pfizer Inc., New York, NY, (4)Worldwide Research and Development Business Technology, Pfizer Inc., Groton, CT, (5)Clinical Informatics and Innovation, Pfizer Inc., New York, NY

    Disclosures:

    P. Bhuyan, Pfizer: Employee, Salary

    J. Desai, Pfizer: Employee, Salary

    M. Carlsson, Pfizer: Employee, Salary

    M. V. St.Louis, Pfizer: Employee, Salary

    E. Bowen, Pfizer: Employee, Salary

    M. N. Cantor, Pfizer: Employee, Salary

    Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.