A conditional maximized sequential probability ratio test for Pharmacovigilance

Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, 133 Brookline Ave., 6th floor, Boston, MA 02215, USA.
Statistics in Medicine (Impact Factor: 2.04). 01/2009; 29(2):284-95. DOI: 10.1002/sim.3780
Source: PubMed

ABSTRACT The importance of post-marketing surveillance for drug and vaccine safety is well recognized as rare but serious adverse events may not be detected in pre-approval clinical trials. In such surveillance, a sequential test is preferable, in order to detect potential problems as soon as possible. Various sequential probability ratio tests (SPRT) have been applied in near real-time vaccine and drug safety surveillance, including Wald's classical SPRT with a single alternative and the Poisson-based maximized SPRT (MaxSPRT) with a composite alternative. These methods require that the expected number of events under the null hypothesis is known as a function of time t. In practice, the expected counts are usually estimated from historical data. When a large sample size from the historical data is lacking, the SPRTs are biased due to the variance in the estimate of the expected number of events. We present a conditional maximized sequential probability ratio test (CMaxSPRT), which adjusts for the uncertainty in the expected counts. Our test incorporates the randomness and variability from both the historical data and the surveillance population. Evaluations of the statistical power for CMaxSPRT are presented under different scenarios.

  • [Show abstract] [Hide abstract]
    ABSTRACT: We conducted weekly surveillance for pre-specified adverse events following receipt of the 2012-2013 influenza vaccines in the Vaccine Safety Datalink (VSD). For each outcome, risk intervals (i.e., period after vaccination with a potentially increased risk) were defined on the basis of biologic plausibility and prior literature. Seizures following inactivated influenza vaccine (IIV) were monitored in children in three age groups (6-23 months, 24-59 months, and 5-17 years) using a self-controlled risk interval design. We also monitored for Guillain-Barré syndrome, encephalitis, and anaphylaxis following IIV in patients ≥6 months of age using a cohort design with historical controls. In the risk intervals following live attenuated influenza vaccine (LAIV), we collected weekly counts of Guillain-Barré syndrome, encephalitis, and anaphylaxis in patients ages 2-49. Among LAIV vaccinees, numbers of expected events based on rates in historical controls were calculated, adjusted for age and site. At the end of surveillance, approximately 3.6 million first doses of IIV and 250 000 first doses of LAIV had been administered in the VSD. No elevated risks were identified in risk intervals following 2012-2013 IIV, as compared with a self-matched control interval or to historical controls. For each outcome, fewer than three events occurred in the risk interval following 2012-2013 LAIV, and we thus were unable to estimate measures of relative risks. No increased risk was identified for any of the pre-specified outcomes following 2012-2013 influenza vaccinations in the VSD. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
    Pharmacoepidemiology and Drug Safety 05/2014; 23(5). DOI:10.1002/pds.3575 · 3.17 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In 2008, a diphtheria, tetanus, acellular pertussis, and inactivated poliovirus combined vaccine (DTaP-IPV) was licensed for use in children 4 through 6 years of age. While pre-licensure studies did not demonstrate significant safety concerns, the number vaccinated in these studies was not sufficient to examine the risk of uncommon but serious adverse events. To assess the risk of serious adverse events following DTaP-IPV vaccination. The study was conducted from January 2009 through September 2012 in the Vaccine Safety Datalink (VSD) project. In the VSD, electronic vaccination and encounter data are updated and aggregated weekly as part of ongoing surveillance activities. Based on previous reports and biologic plausibility, eight potential adverse events were monitored: meningitis/encephalitis; seizures; stroke; Guillain-Barré syndrome; Stevens-Johnson syndrome; anaphylaxis; serious allergic reactions other than anaphylaxis; and serious local reactions. Adverse event rates in DTaP-IPV recipients were compared to historical incidence rates in the VSD population prior to 2009. Sequential probability ratio testing was used to analyze the data on a weekly basis. During the study period, 201,116 children received DTaP-IPV vaccine. Ninety-seven percent of DTaP-IPV recipients also received other vaccines on the same day, typically measles-mumps-rubella and varicella vaccines. There was no statistically significant increased risk of any of the eight pre-specified adverse events among DTaP-IPV recipients when compared to historical incidence rates. In this safety surveillance study of more than 200,000 DTaP-IPV vaccine recipients, there was no evidence of increased risk for any of the pre-specified adverse events monitored. Continued surveillance of DTaP-IPV vaccine safety may be warranted to monitor for rare adverse events, such as Guillain-Barré syndrome.
    Vaccine 03/2014; 32(25). DOI:10.1016/j.vaccine.2014.03.063 · 3.49 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The Vaccine Safety Datalink (VSD) is a collaborative project between the Centers for Disease Control and Prevention (CDC) and 9 health care organizations. Established in 1990, VSD is a vital resource informing policy makers and the public about the safety of vaccines used in the United States. Large linked databases are used to identify and evaluate adverse events in over 9 million individuals annually. VSD generates rapid, important safety assessments for both routine vaccinations and emergency vaccination campaigns. VSD monitors safety of seasonal influenza vaccines in near-real time, and provided essential information on the safety of monovalent H1N1 vaccine during the 2009 pandemic. VSD investigators have published important studies demonstrating that childhood vaccines are not associated with autism or other developmental disabilities. VSD prioritizes evaluation of new vaccines; searches for possible unusual health events after vaccination; monitors vaccine safety in pregnant women; and has pioneered development of biostatistical research methods.
    Vaccine 08/2014; 32(42). DOI:10.1016/j.vaccine.2014.07.073 · 3.49 Impact Factor