Polio is a viral disease that invades the nervous system and can cause paralysis in a matter of hours, though most poliovirus infections are asymptomatic [
After IPV immunization, antibodies are produced in the blood in response to the inactivated virus, protecting the individual from disease; however, viral replication in the gut is still possible, with the potential for asymptomatic transmission to the community [
In contrast, OPV induces mucosal immunity in the gastrointestinal tract, which is important for community protection [
Due to these risks, most developed countries have switched from OPV to IPV [
Today, WPV remains endemic in 3 countries—Pakistan, Afghanistan, and Nigeria [
In the United States, children currently receive 3 routine doses of IPV at 2 months, 4 months, and 6–18 months, and a booster at 4–6 years [ [although] militaries primarily engage in traditional major combat operations, they are increasingly involved in humanitarian assistance missions. Such missions permit extensive interaction with the local populace and environment, greatly increasing the chance of acquiring locally endemic infectious diseases and necessitating the management of diseases in the local populace that are not traditionally seen in military personnel [
Thus, while blanket vaccination of all soldiers may be epidemiologically redundant and cause unnecessary expenditure of resources, risk-based vaccination driven by travel to polio-endemic areas can be appropriate. However, evaluating multiple vaccination strategies once deployment is underway may result in higher-than-necessary disease incidence and cost in an effort to control transmission. In lieu of this, predictive modelling of polio transmission allows for the exploration of vaccination strategies through simulation of multiple scenarios and their outcomes prior to putting troops and mission objectives at risk.
In recent years, published mathematical models for polio transmission have focused primarily on the role of OPV in attaining eradication of the disease. Several mathematical models have also explored vaccine-derived polioviruses [
Each of these modelling studies focuses on issues among multigenerational populations with varying levels of immunization coverage and background immunity; none have addressed the unique circumstances affecting polio transmission in highly mobile military populations, which experience conditions that directly affect the spread of disease, both beneficially and detrimentally. US soldiers are vaccinated against polio at recruitment and again prior to deployment to at-risk areas; however, immunization with IPV may leave troops at risk of VDPV or subject to asymptomatic transmission. Quantifying this risk is crucial to evaluating the elimination of mandatory blanket polio vaccination and switching to a solely risk-based vaccination policy.
Mathematical models can be immensely useful in examining the impact of vaccines on disease transmission and are frequently used to inform response policy. For deployed military populations, these models can also evaluate the relative change in transmission risk associated with multiple vaccination scenarios by employing data on specific demographics, epidemiology, and the effects of timing of response. For this analysis, we modified an existing military-specific model system to accommodate polio transmission and vaccination to maximize the achievement of deployed mission objectives, while minimizing the possibility of transmission to troops both abroad and at home.
The military population structure for this study was built upon previous analyses [
The 10-year simulated deployment period began in 2015, with individual soldiers rotating annually. In- and outbound rotation rates (
Change in deployed population size over the 10-year duration of the deployment action [
The local population was modelled without structure and based on the demographic characteristics of Afghanistan as estimated by the United Nations Population Division [
Polio transmission was modelled as a modified compartmental Susceptible-Exposed-Infected-Removed model (
Variable definitions for polio transmission model.
Variable | Definition |
---|---|
|
Number of unprotected susceptible individuals in subpopulation |
|
Number of partially protected susceptible individuals in subpopulation |
|
Number of unprotected exposed individuals in subpopulation |
|
Number of partially protected exposed individuals in subpopulation |
|
Number of unprotected infected individuals in subpopulation |
|
Number of partially protected infected individuals in subpopulation |
|
Number of recovered or removed individuals in subpopulation |
|
Total population size in subpopulation |
Parameter definitions for polio transmission model for military (MIL) and local (LOC) populations.
Parameter | Definition | Value (range) | Source |
---|---|---|---|
|
Daily inbound rotation rate (MIL)/birth rate (LOC) for subpopulation |
[Array] | |
|
Daily outbound rotation rate (MIL)/background death rate (LOC) for subpopulation |
[Array] | |
|
Daily background casualty rate for subpopulation |
[Array] | |
|
Daily proportion of inbound population with preexisting partial protection for subpopulation |
[Array] | |
|
Daily (OPV) vaccination rate for subpopulation |
[Array] | |
|
Effective polio transmission rate for subpopulation |
[Function] | |
|
Relative susceptibility of |
0.2 (0.2–0.9) | [ |
|
Duration of latent period for unprotected individuals | 3 d (3-4 d) | [ |
|
Duration of latent period for partially protected individuals | 4 d (3-4 d) | [ |
|
Duration of infectious period for unprotected individuals | 27 d (27-28 d) | [ |
|
Duration of infectious period for partially protected individuals | 9 d (9–25 d) | [ |
|
Polio mortality rate | 0.22 (0.02–0.30) | [ |
|
1/duration of OPV or disease-induced immunity |
|
(See text) |
|
Seasonal variation in polio transmission | [Function] | |
|
Proportional change in polio transmission due to seasonality | 0.15 | [ |
|
Polio attack rate | 20/100,000 (0.1/1,000,000–6.8/100,000) | [ |
|
Daily contact rate between subpopulations |
[Array] | [ |
|
Relative infectiousness of |
0.2 (0.2–0.9) | [ |
symp | Proportion of unprotected polio cases that are symptomatic | 0.1 | [ |
importrisk | Probability of polio case importation from outside population | 0.001 | (See text) |
importampl | Amplitude of polio case importation from outside population | 2/100,000 | (See text) |
Schematic diagram of polio transmission with both OPV (
To establish endemicity, polio transmission among locals was simulated for a burn-in period of 35 years prior to the arrival of the deployed military population in 2015. Simulated, combined symptomatic and asymptomatic incidence during this burn-in phase was validated against reported paralytic polio cases as recorded by the WHO [
Qualitative comparison between simulated (gray lines) and historical (blue line) polio cases (symptomatic + asymptomatic) for the local population over a sample of 10 random simulations.
Individuals entered population either via birth (locals) with full susceptibility (
Ten percent (symp) of cases in nonimmune, infected individuals presented with symptoms, with the remainder being subclinical; all infections in partially protected individuals were assumed to be asymptomatic. Polio-related mortality affected only symptomatically infected individuals.
Polio transmission was driven by an attack rate (
Sensitivity and uncertainty analyses were performed on 21 model parameters, with outcomes measured in terms of total symptomatic and asymptomatic polio cases and mean annual incidence among military and local populations.
Total military cases and annual incidence showed direct sensitivity to the relative susceptibility (
Local annual disease incidence was significantly sensitive to the polio death rate (
The duration of the infectious period
Local incidence was inversely affected by changes in the duration of OPV- or disease-induced immunity (1/
The level of residual protection resulting from childhood IPV immunization (childhood) had a moderate impact on polio transmission within military populations in simulations within which no other polio vaccination was administered. In the presence of either blanket or booster vaccination (or both), variation in childhood residual protection had no impact on military polio transmission, since the more recent adult immunization overrode any effects due to vaccination occurring early in life.
Uncertainty in case importation (importrisk and importampl) had a dramatic impact on polio incidence in both military and local populations, overriding even variation in the polio attack rate (
To evaluate the hypothesis that residual immunity from childhood vaccination, combined with risk-based deployment vaccination, is sufficient to protect troops from polio transmission, 3 military IPV scenarios were tested: Blanket immunization at recruitment + booster immunization at deployment + residual childhood protection (Scenario 1, baseline). (Blanket immunization terminated in 2015) + booster immunization at deployment + residual childhood protection (Scenario 2). (Blanket immunization terminated in 2015) + residual childhood protection only (Scenario 3).
As of 2013, the US childhood IPV vaccination coverage rate was approximately 93% [
The childhood IPV coverage rate accounts for both philosophical and medical exemptions to vaccination, including impaired immune status, allergies to vaccine components, or history of vaccine-associated adverse events. Within the 2012-2013 school year, medical exemptions for childhood vaccinations (not specific to polio) ranged from 0.1 to 1.6% (median 0.3%), and nonmedical exemptions ranged from 0.2 to 6.4% (median 1.5%) [
For this analysis, we assumed that military exemption rates sit at the low-end of the range for childhood vaccination exemptions, yielding an overall military IPV exemption rate (exemption) of 0.3% (range 0.3–8.0%, median 1.8%), with the inclusion of a 75% proportional reduction (dfact) in exemption rate (range 0–100%) for deploying personnel.
Military IPV effective coverage rates for blanket and booster immunization were calculated as follows:
As of 2006 [
Scenario-specific parameters are provided in Table
Parameter definitions for scenario calculations.
Parameter | Definition | Value (range) | Source |
---|---|---|---|
efficacy | IPV vaccine efficacy (MIL) | 99% (50–100%) | [ |
childhood | Residual protection level from childhood IPV vaccination | 0.92 (0.0–1.0) | Calculated from [ |
exemption | Overall military vaccination exemption rate (medical + administrative) | 0.003 (0.003–0.08) | Calculated from [ |
dfact | Proportional reduction in exemption rate for deployed personnel (versus nondeployed) | 0.75 (0.0–1.0) | (Estimated) |
blanket | IPV blanket vaccine coverage (when implemented) for all military personnel upon accession | (Function) | |
boost | IPV boost vaccine coverage (when implemented) for deploying personnel | (Function) | |
accession | Military accession rate | 0.19 (0.13–0.19) | [ |
|
Proportion of military population covered by blanket vaccination prior to 2015 | (Function) | |
|
Overall protection of military population from blanket vaccination prior to 2015 and residual childhood immunity | (Function) |
Scenario definitions for deployed and nondeployed military personnel.
Protection levels (protect) | ||
---|---|---|
Deployed personnel | Nondeployed personnel | |
Scenario 1 (baseline) | blanket | blanket |
Scenario 2 (booster) | boost |
|
Scenario 3 (childhood) |
|
|
Each scenario was run for 1500 simulations to account for stochasticity in polio case importation. Model outputs were measured as total deployed symptomatic and asymptomatic polio cases and average annual incidence (included both symptomatic and asymptomatic cases) for deployed military and local populations.
For all scenarios, local disease dynamics remained fairly consistent across all simulations, tracking with historical cases prior to 2015, then sustaining low endemicity driven by case importation to the end of the simulation period (Figures
Polio disease dynamics among local populations under military immunization Scenarios (a) 1 (blanket + booster + residual childhood), (b) 2 (booster + residual childhood), and (c) 3 (residual childhood).
Similarly, there was insignificant change in polio dynamics among deployed military populations between Scenarios 1 and 2, with a slight increase in infections under Scenario 3 (Figures
Polio disease dynamics among deployed military populations under military immunization Scenarios (a) 1 (blanket + booster + residual childhood), (b) 2 (booster + residual childhood), and (c) 3 (residual childhood).
Stochasticity associated with case importation caused variation between simulation results for total cases and average annual polio incidence for all 3 scenarios (Figures
Distribution over 1500 simulations per scenario of (a) total symptomatic polio cases in deployed military populations; (b) total asymptomatic polio cases in deployed military populations; (c) average annual polio incidence in deployed military populations; and (d) average annual polio incidence in local populations under 3 military immunization scenarios.
Dropping blanket immunization but maintaining predeployment booster immunization had negligible effect on simulated deployed military cases and incidence. Dropping both blanket and predeployment immunizations yielded a 5% increase in polio cases and annual incidence among deployed populations over the baseline scenario of blanket immunization (Table
Median total polio cases and average annual incidence with percentage of change over baseline for deployed military and local populations under 3 military immunization scenarios.
Scenario | Total cases (military) | Average annual incidence | ||
---|---|---|---|---|
Symptomatic | Asymptomatic | Military | Local | |
1 | 0.076 | 0.687 | 0.012/1000 | 0.022/1000 |
2 | 0.076 (+0%) | 0.687 (+0%) | 0.012/1000 (+0.2%) | 0.022/1000 (+0.8%) |
3 | 0.080 (+5%) | 0.721 (+5%) | 0.013/1000 (+5.4%) | 0.022/1000 (+0.1%) |
The total number of symptomatic and asymptomatic polio cases in the deployed military population remained less than one for all 3 immunization scenarios, though fractional cases were still utilized in the calculation of incidence rates. Though frequently undetectable in the field, asymptomatic cases were included in the case-count and incidence calculations to provide a measure of the potential for silent transmission.
For nondeployed personnel, dropping blanket immunization resulted in a decrease in polio disease protection from 99% to 93%. Since IPV vaccination confers full protection from disease but only partial (model assumption: 20%) protection from transmission, this yielded an increase in overall susceptibility to transmission among the nondeployed population from 21% to 26%.
Combining nondeployed susceptibility levels with the average annual polio incidence among deployed populations for each scenario allowed for estimation of the risk of new polio infections within nondeployed personnel due to mixing with infected soldiers returning from deployment. For the blanket immunization scenario (Scenario 1), the risk of new polio infections resulting from reintegrating infected soldiers was predicted to be 0.000504/1,000,000. For Scenario 2—where blanket immunization was terminated but predeployment booster was still employed—the simulated risk of new polio infections increased from 0.000504/1,000,000 to 0.000624/1,000,000 after 10 years without blanket vaccination. For Scenario 3—where both blanket and booster immunizations were terminated—simulated risk of new polio infections among nondeployed personnel increased from 0.000546/1,000,000 to 0.000676/1,000,000 (Table
Modelled risk of new polio infections among nondeployed service members as a result of infected soldiers reintegrating upon return from deployment.
Scenario | % protected from disease |
Average annual incidence |
% protected from disease |
% susceptible to transmission |
Risk of new infections |
---|---|---|---|---|---|
1 | 99% | 1.2/100,000 | 99% | 21% | 0.000504/1,000,000 |
2 | 99% | 1.2/100,000 | 99% |
21% |
0.000504/1,000,000 |
3 | 99% |
1.3/100,000 | 99% |
21% |
0.000546/1,000,000 |
Mathematical models can help guide preventive medicine policy, resulting in healthier and protected populations. This analysis employed a mathematical model for the transmission of polio within deployed military populations interacting with local populations in an endemic setting. Results from model simulations described the potential benefits of protecting these troops via routine blanket immunization, predeployment booster immunization, and residual protection resulting from childhood vaccination.
In the absence of blanket immunization on recruitment, immunity to polio disease among
Though the increased percentage in transmission resulting from dropping blanket immunization was nonzero, the overall risk of new infections among both deployed and nondeployed service members was extremely low, resulting from the combination of high US childhood immunization coverage rates, conferment of partial protection against polio transmission by IPV, and low disease incidence levels globally. At this range of risk, the likelihood of importation of polio cases among deployed soldiers, and subsequent spread to their nondeployed counterparts, is exceptionally small even in the absence of blanket immunization.
Given preexisting protection resulting from routine childhood vaccination, predeployment booster of service members driven by travel to polio-endemic regions is sufficient to prevent additional transmission among both deployed and nondeployed populations based on these results. Blanket mandatory polio vaccination of Department of Defense service members appears to be epidemiologically redundant, and dropping this routine immunization will not adversely affect troop readiness or mission objectives.
Kellie McMullen is a military service member (or employee of the US Government). This work was prepared as part of the official duties. Title 17, USC
The authors declare they have no conflicts of interest.
This work was funded by the Military Vaccine Agency, US Government Work (17 USC 105), and NHRC Contract no. N62645-15-F-1002. The authors wish to thank LCDR Lori N. Perry for her assistance with this contract and manuscript.