Format

Send to

Choose Destination
J Clin Epidemiol. 2019 May 2;113:114-122. doi: 10.1016/j.jclinepi.2019.04.019. [Epub ahead of print]

Applying sequential surveillance methods that use regression adjustment or weighting to control confounding in a multisite, rare-event, distributed setting: Part 2 in-depth example of a reanalysis of the measles-mumps-rubella-varicella combination vaccine and seizure risk.

Author information

1
Kaiser Permanente Washington Health Research Institute, Seattle, WA, USA; Department of Biostatistics, University of Washington, Seattle, WA, USA. Electronic address: Andrea.J.Cook@kp.org.
2
Kaiser Permanente Washington Health Research Institute, Seattle, WA, USA.
3
Kaiser Permanente Washington Health Research Institute, Seattle, WA, USA; Department of Biostatistics, University of Washington, Seattle, WA, USA.
4
Center for Biologics Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA.
5
Office of Biostatistics, Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA.
6
Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA.
7
Division of Health Care Quality Promotion, Immunization Safety Office, Centers for Disease Control and Prevention, Atlanta, GA, USA.

Abstract

OBJECTIVE:

In-depth example of two new group sequential methods for postmarket safety monitoring of new medical products.

STUDY DESIGN AND SETTING:

Existing trial-based group sequential approaches have been extended to adjust for confounders, accommodate rare events, and address privacy-related constraints on data sharing. Most adaptations have involved design-based confounder strategies, for example, self-controlled or exposure matching, while analysis-based approaches like regression and weighting have received less attention. We describe the methodology of two new group sequential approaches that use analysis-based confounder adjustment (GS GEE) and weighting (GS IPTW). Using data from the Food and Drug Administration's Sentinel network, we apply both methods in the context of a known positive association: the measles-mumps-rubella-varicella vaccine and seizure risk in infants.

RESULTS:

Estimates from both new approaches were similar and comparable to prior studies using design-based methods to address confounding. The time to detection of a safety signal was considerably shorter for GS IPTW, which estimates a risk difference, compared to GS GEE, which provides relative estimates of excess risk.

CONCLUSION:

Future group sequential safety surveillance efforts should consider analysis-based confounder adjustment techniques that evaluate safety signals on the risk difference scale to achieve greater statistical power and more timely results.

KEYWORDS:

Active surveillance; Distributed databases; Electronic health record (EHR); Group sequential analysis; Rare events; Vaccine safety

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center