Community interventions and the epidemic prevention potential
Introduction
Increasing attention is being given to evaluating community-level effects of vaccination strategies [1], [2], [3], [4], [5] and other interventions [6], [7], [8], [9] in infectious diseases. In light of the current worries about bioterrorism, there is also increased interest in planning interventions in case an infectious agent, such as smallpox, is introduced into a population. Methods to evaluate the effectiveness of such interventions will be required.
Much has been written on group randomized studies where whole social units are assigned to treatment groups [9], [10], [11], [12], [13], [14]. Considerable theoretical research has focused on establishing threshold conditions for transmission in epidemic models [15], [16], [17], [18], [19]. However, little has been written about evaluating interventions when transmission may be a threshold phenomenon.
In this paper, we introduce the concept of the epidemic prevention potential (EPP) as one aspect of overall effectiveness when an intervention prevents an outbreak or limits it to less than a certain defined size. We illustrate estimation of effectiveness measures with simulations of vaccination of children against influenza to reduce attack rates in communities [20], [21], [22], [23], [24]. We differentiate effectiveness measures that take all communities into account whether or not an epidemic has occurred from effectiveness measures that condition on an epidemic reaching a pre-defined size threshold. We constructed a stochastic simulation model of influenza transmission and vaccination in a structured community to illustrate evaluation of the indirect, total, and overall effects, and the EPP. Although the focus here is on vaccination, the concepts are general and could be applied to other interventions, such as antivirals and quarantine.
Section snippets
Scientific questions of interest
Consider a seasonally epidemic infectious agent, such as influenza or cholera. The outcome of interest may be the illness attack rate in the community as a whole or in a stratum within the community, such as in children 18 years old or under. Suppose that several communities receive the intervention of interest and several communities do not receive the intervention. The populations receiving the interventions will be referred to as type A communities, and the control communities as type B
Epidemic prevention potential
A few examples illustrate the EPP measure. Suppose that seasonal transmission occurs reliably in each unvaccinated community type B, but it occurs in only about 20% of the intervention communities A. Then Pr(e)B=1, Pr(e)A=0.20, and EPP=1−0.20/1=0.80. About 80% of epidemics will be prevented by the intervention. Suppose that seasonal outbreaks occur in about 80% of unvaccinated communities, Pr(e)B=0.80 and 20% of the vaccinated communities, Pr(e)A=0.20. Then the EPP=1−0.20/0.80=0.75. The EPP of
Design and analysis issues
In community studies, the unit of inference is the population. Thus, the sample size and power of a study are more closely related to the number of communities in the study than the number of people in each community. The eligible communities need to be transmission-dynamically separate [1], [2]. As in community studies of noninfectious diseases, the sample size required to estimate effectiveness has to do with the size of the effect expected and the intercommunity variability. At low
Simulation of intervention
We simulated studies of community-level effects of vaccinating children using a stochastic simulator of influenza transmission in a structured population. The effectiveness measures of interest were those described in Table 1 as well as the EPP. The intervention of interest was vaccination of children aged 1–18 years old at different levels of coverage. Each simulated study included four intervention and four baseline communities. For each level of coverage and vaccine efficacy, 100 studies
Discussion
We have introduced the concept of the EPP as an effectiveness measure of interest in community interventions in infectious diseases. As a concept to describe the general ability of an intervention to limit or to contain outbreaks to a certain defined threshold size, the term epidemic prevention potential fills a void. In light of the events since 11 September 2001, and worries over bioterrorism, the concept of the EPP has become particularly relevant.
We have delineated three possible goals for
Acknowledgements
This research was partially supported by NIH grant R01-AI32042.
References (39)
- et al.
Design of a group-randomised Streptococcus pneumoniae vaccine trial
Control Clin Trials
(2001) - et al.
Cluster randomization trials in tropical medicine: a case study
Trans R Soc Trop Med Hyg
(1995) - et al.
An optimization model for influenza A epidemics
Math Biosci
(1978) - et al.
The behavior of common measures of association used to assess a vaccination program under complex disease transmission patterns—a computer simulation study of malaria vaccines
Int J Epidemiol
(1990) - et al.
Study designs for dependent happenings
Epidemiology
(1991) - et al.
Study designs for different efficacy and effectiveness aspects of vaccination
Am J Epidemiol
(1997) - et al.
Evaluating new vaccines for developing countries: efficacy or effectiveness?
J Am Med Assoc
(1996) - Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted disease on HIV infection in...
- Hayes R, Mosha F, Nicoll A, Grosskurth H, Newell J, Todd J, et al. A community trial of the impact of improved sexually...
- Wawer MJ, Sewankambo NK, Serwadda, D, et al. Control of sexually transmitted disease for AIDS prevention in Uganda: a...
Design and analysis issues in cluster-randomized trials of interventions against infectious diseases
Stat Methods in Med Res
Some aspects of the design and analysis of cluster randomization trials
Appl. Stat.
Aggregate data studies of disease risk factors
Biometrika
A contribution to the mathematical theory of epidemics
Proc R Soc London Ser A
The outcome of a stochastic epidemic—a note on Bailey’s paper
Biometrika
Cited by (60)
Pediatricians' attitudes about collaborations with other community vaccinators in the delivery of seasonal influenza vaccine
2012, Academic PediatricsCitation Excerpt :Schools provide easy access to children, the potential for rapid delivery and the potential for significant herd immunity benefits to the community.41,42 The idea that influenza vaccination of children should be the focus of overall community strategies for influenza disease control is not universally agreed upon, but has support from both theoretical modeling8,43–46 and observational data.38,41,42,47 However, although schools appear to be ideal for influenza vaccine delivery in many ways, data from the 2010–2011 season suggested only 8.7% of vaccinated children received it at school,48 and important issues need to be solved in order for schools to be a substantial partner in the delivery of influenza vaccine.
Predictors of acceptance of H1N1 influenza vaccination by FDNY firefighters and EMS workers
2011, VaccineCitation Excerpt :Nonetheless, it is far from the nearly 100% achieved by mandatory vaccination programs such as vaccination in the military [13–15]. Moreover, the rate of voluntary acceptance remained far lower than the 70% acceptance rate estimated by some [29,30] as a requirement for achieving “herd” immunity – a rate high enough to prevent contagious disease even in non-vaccinated. In a voluntary program, what else could be done to further improve vaccination rates?
Real-time estimation and prediction for pandemic A/H1N1(2009) in Japan
2011, Journal of Infection and ChemotherapyGlobal asymptotic properties of an SEIRS model with multiple infectious stages
2010, Journal of Mathematical Analysis and ApplicationsSchool-based influenza vaccine delivery, vaccination rates, and healthcare use in the context of a universal influenza immunization program: An ecological study
2010, VaccineCitation Excerpt :Influenza vaccination of all school-age children has recently been recommended in the U.S. [1] based on the recognition that these children often have the highest influenza attack rates and are effective in propagating influenza epidemics [2–9]. Vaccinating children may decrease influenza-related illness rates in both households [10–13] and communities [6,14–23]. School-based influenza immunization has been proposed as an efficient method for vaccinating children [24,25].