The American College of Epidemiology: co-sponsored by
Statistics in Epidemiology Section of the ASA
Session Slot: 10:30-12:20 Monday
Estimated Audience Size: 80 - 150
AudioVisual Request:
Session Title: Statistical Evaluation of Vaccine Effects
Vaccination is the most effective and cost-efficient way to control an
infectious disease. The basic vaccine efficacy statistic has been
around for over 80 years. However, the assessment of the properties of
this statistic and the development of alternative measures of vaccine
effects are relatively new research topics for biostatisticians and
epidemiologists.
This session will focus on issues related to the statistical evaluation
of vaccine effects, such as sample size and power calculations, errors
in measurement of exposure status and the use of stochastic models in
making policies regarding the design of vaccination programs.
Theme Session: no
Applied Session: yes
Session Organizer: Haber, Michael Emory University
Address: Department of Biostatistics Rollins School of Public Health Emory University Atlanta, GA 30322
Phone: 404-727-7698
Fax: 404-727-1370
Email: mhaber@sph.emory.edu
Session Timing: 110 minutes total (Sorry about format):
Opening Remarks by Chair - 0 minutes First Speaker - 25 minutes Second Speaker - 25 minutes Third Speaker - 25 minutes Fourth Speaker - 25 minutes Floor Discusion - 10 minutes
Session Chair: Haber, Michael Emory University
Address: Department of Biostatistics Rollins School of Public Health Emory University Atlanta, GA 30322
Phone: 404-727-7698
Fax: 404-727-1370
Email: mhaber@sph.emory.edu
1. Preventing Epidemics in Heterogeneous Communities
Britton, Tom, Uppsala University
Address: Department of Mathematics Uppsala University Box 480, S-751 06 Uppsala, Sweden
Phone: 46-18-471-3222
Fax: 46-18-471-3201
Email: tom.britton@math.uu.se
Abstract: The main practical motivation for the study of epidemic models lies in the insight they provide about the control of infectious diseases. The present paper concerns estimation procedures for the basic reproduction number, R, in closed heterogeneous communities. R is important when designing vaccination policies in terms of how many and which individuals should be vaccinated. The estimation procedures are based on stochastic models, thus allowing confidence regions, incorporating both individual heterogeneities, e.g. age and sex, and social heterogeneities such as households.
2. Exact Power and Sample Sizes for Vaccine Efficacy Studies
Chan, Ivan S. F., Merck Research Laboratories
Address: Merck Research Laboratories BL 3-2 West Point, PA 19486
Phone: 610-397-3391
Fax: 610-397-2931
Email: ivan_chan@merck.com
Bohidar, Norman R., Merck Research Laboratories
Abstract: In vaccine efficacy studies, the goal is to show that the vaccine reduces the incidence of the disease of interest when compared to placebo. Various methods of calculating sample size and power based on large sample approximation are available for these efficacy trials (Farrington and Manning 1990; Blackwelder 1993). In this paper we describe two procedures for calculating sample size and power based on exact distributions. In small studies where the disease incidence and the anticipated vaccine efficacy are both high, an unconditional exact procedure may be desirable because it guarantees the level of the test and loses little sensitivity. In large studies where the disease is rare, a Poisson approximation to the number of events is reasonable and an exact test is simple to construct conditional on the total number of events. One can use this exact procedure in study planning to guarantee the power, and the Poisson model is readily extended to handle different follow-up times between groups. We compare the power and type I error rate of these two exact methods to the method based on the normal approximation for varying disease incidence and sample size.
3. Semiparametric Models for Mismeasured Exposure Information in Vaccine Trials
Golm, Gregory T., Emory University
Address: Department of Biostatistics Rollins School of Public Health Emory University Atlanta, GA 30322
Phone: 404-727-7697
Fax: 404-727-1370
Email: golm@sph.emory.edu
Halloran, M. Elizabeth, Emory University
Longini, Ira M., Emory University
Abstract: Exposure to infection information is important for estimating vaccine efficacy, but is difficult to collect and inherently prone to missingness and mismeasurement. It is therefore not feasible to collect good exposure information on all participants in a large vaccine trial. We discuss study designs which collect detailed exposure information for only a small subset of trial participants, while collecting crude exposure information on all participants, and treat estimation of vaccine efficacy in the missing data/measurement errors framework. We demonstrate with the example of an HIV vaccine trial the improvements in bias and efficiency when combining different levels of exposure information to estimate vaccine efficacy for reducing both susceptibility and infectiousness. We compare the performance of recently developed semiparametric missing data methods of Pepe and Fleming (1994) and Robins, Hsieh and Newey (1995).
4. Statistical Evaluation of Preventive Vaccine Efficacy: Current Methods
Address: FDA/CBER HFM-215 1401 Rockville Pike Rockville, MD 20852-1448
Phone: 301-827-6068
Fax: 301-827-3529
Email: HORNE@cber.fda.gov
Blackwelder, William C., NIH/NIAID
Abstract: Preventive vaccines are generally given to healthy humans to prevent disease or infection. Consequently, a relatively low risk/benefit ratio is required compared to many therapeutic agents. Thus, statistical evaluation depends primarily on estimation of efficacy, with a suitably high lower bound on the corresponding confidence interval. (A p-value from testing a null hypothesis of no difference from placebo is generally of little interest.)Vaccine efficacy is routinely estimated as 1-R, where R may be a ratio of risks, incidence rates, or hazards in the vaccinated relative to control subjects. There are several methods for estimating R and for obtaining its confidence interval, which will be reviewed here.
Traditionally, preventive vaccine efficacy trials have not used an intention-to-treat (ITT) approach as the primary analysis. The reason for this practice is that many preventive vaccine trials give little cause for concern about bias due to dropouts and other sources of noncompliance, unlike most therapeutic trials. However, ITT is always an appropriate analysis to perform (perhaps as a co-primary analysis) and in some vaccine trials in which there may be considerable concern about bias, ITT may be the only appropriate analysis.
Sometimes it is not possible to evaluate vaccine efficacy directly via observation of who succumbs to the disease of interest. In such trials, which will be reviewed here, immunological surrogates of efficacy are used. Vaccine efficacy trials which use these surrogate endpoints are often problematic when no immunological correlate of protection is known. These trials are often designed and analyzed as equivalence trials.
List of speakers who are nonmembers: None