The Jameel Institute Crisis Vaccination Planner (JICVP)

JICVP is an interactive modelling tool designed to forecast outbreak risks for key vaccine-preventable diseases (VPDs) and inform the prioritization of reactive vaccination campaigns in humanitarian emergencies.

In conflict-affected settings, disruptions to routine healthcare and immunization services increase the risk of VPD outbreaks. JICVP addresses this challenge by integrating a stochastic, mathematical model of disease transmission with pre-conflict demographic, immunization and case data, flexible impacts on health system functioning and outbreak occurrence, and the ability to model a variety of settings and diseases.

The tool generates time-dependent outbreak probabilities and estimated case burdens, allowing users to simulate how vaccination disruptions increase susceptibility over time, quantify outbreak risks under varying vaccination coverage scenarios, and assess outbreak severity and the impact of delayed or proactive vaccination campaigns. By enabling users to explore different intervention scenarios, JICVP supports rapid decision-making to mitigate the public health impacts of immunization disruptions.

Intended Use

JICVP is intended as a decision-support tool for planning vaccination responses in crisis settings. It does not provide real-time outbreak forecasts or directly predict current epidemiological outcomes. Instead, it helps users understand outbreak risks under different disruption scenarios, estimate the value of restoring vaccination, and support strategic decision-making around timing, targeting, and disease prioritisation in resource-limited settings.

JICVP is currently in a prototype phase and is not intended for real-time outbreak prediction. Instead, it serves as a planning tool to illustrate the potential consequences of immunization disruptions and the benefits of various vaccination strategies in crisis settings.

We are actively engaging with stakeholders, including humanitarian partners, to guide the future development of JICVP and ensure it aligns with operational needs in the field.

Using JICVP

Disease Comparison Dashboard

The Disease Comparison tab helps answer the critical question: if vaccination stops, which disease is most urgent for proritising vaccination?

For the selected country, the tool runs stochastic outbreak simulations for Measles, Diphtheria, and Pertussis simultaneously, assuming routine vaccination has completely stopped. Each disease receives a composite risk score (0-100) based on susceptibility gap (40% weight), transmissibility (30% weight), and time to outbreak (30% weight). Diseases are then classified as Critical (≥70), High (40-69), Moderate (20-39), or Low (<20) risk.

The simulation uses 10 stochastic runs per disease with 10 initial seeded infections, projecting outcomes over a 3-year period with no routine vaccination. The population data comes from UN World Population Prospects.

Key outputs include outbreak risk ranking, expected case burden with 95% confidence intervals, time to outbreak milestones (100 and 1,000 cases), and a comparison of current susceptibility against herd immunity thresholds.

Single Disease Analysis

The Single Disease Analysis tab allows detailed exploration of a specific disease scenario. The Model Setup panel shows historical context including demographics, vaccination coverage trends, reported cases, and current immunity profile by age. The Model Outputs panel compares outbreak trajectories under two scenarios: pre-conflict vaccination continues versus routine immunisation stopped. Users can specify custom scenarios with different seeded infections and partial vaccination recovery for each year.

PREVAIL: The Underlying Model

JICVP is powered by PREVAIL (Projecting Responses to Epidemics with Vaccine-preventable disease Analysis and Integrated Learning), an R package developed by the Jameel Institute. PREVAIL provides the epidemiological modelling engine that drives all simulations in this dashboard.

The PREVAIL model builds on a generalisable, stochastic SEIR (Susceptible–Exposed–Infectious–Recovered) framework, extended to include compartments for severe infection and recovery with complications. The model is stratified by age, vaccination status, and risk group, and incorporates dynamic processes such as aging, vaccination, waning immunity, and movement between risk strata.

Epidemiological dynamics are informed by a combination of empirical and synthetic data, including:

• UN World Population Prospects for demographic structure, fertility, mortality and migration
• WHO data for historical vaccine coverage and case data
• POLYMOD study for age-specific contact matrices
• Disease-specific parameters including R₀, incubation and infectious periods, and vaccine efficacy

To estimate prior exposure and immunity, the model supports several approaches: constant seeding of infections, direct use of WHO-reported case data, or hybrid methods that combine empirical and simulated data depending on disease and country context.

For more details on PREVAIL, including the full model documentation and source code, visit the PREVAIL GitHub repository.

Contact Us

For inquiries or to provide feedback on JICVP, please contact: oliver.watson15@imperial.ac.uk

Contribution Acknowledgements

JICVP is developed by the Jameel Institute at Imperial College London, with contributions from:

Arran Hamlet¹, Paula Christen¹, James Hay^2,3, Bhargavi Rao^4,5, and Oliver Watson¹

Affiliations

1. MRC Centre for Global Infectious Disease Analysis, WHO Collaborating Centre for Infectious Disease Modelling, Jameel Institute, Imperial College London, United Kingdom
2. Big Data Institute, University of Oxford, United Kingdom
3. Pandemic Sciences Institute, University of Oxford, United Kingdom
4. Médecins Sans Frontières, United Kingdom
5. London School of Hygiene and Tropical Medicine, United Kingdom

Funding Acknowledgements

This project is supported by Community Jameel as part of the Jameel Institute – Realtime Intelligent Support for Emergencies (JI-RISE) initiative.

Disclaimer and Licensing

The JICVP model code is provided as is, without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages or other liability, whether in an action of contract, tort or otherwise, arising from, out of or in connection with the software or the use or other dealings in the software.