Crisis in Gaza: Scenario-Based Health Impact Projections

This project aims to provide projections of excess mortality for different causes of excess mortality (mortality modules) in Gaza under three scenarios: 1) ceasefire, 2) status quo (continuation of conditions observed during the period of the war from October 2023 to 15 January 2024), and 3) escalation. These estimations will provide decision-makers and responders information to plan for and respond to different scenarios. The data may also be used to inform diplomatic efforts addressing this war. The current projection is for six months from 7 February- 6 August 2024. The estimates will be periodically updated and will take into account the evolving crisis, availability of new data, and improved methods until May 2024.

We will prepare scenario-based projections of overall mortality and four cause-specific mortality modules: 1) traumatic injury, 2) infectious diseases, 3) maternal and neonatal health, and 4) non-communicable diseases that, taken together, encompass nearly all direct and indirect causes of mortality. A methods annex technical document detailing the methods underpinning the estimates, and an openly accessible repository of data and analysis code are available. Presentations targeted for specific stakeholders will occur.

The project is implemented by researchers from the Health in Humanitarian Crises Centre at the London School of Hygiene & Tropical Medicine , and the Johns Hopkins Center for Humanitarian Health at Johns Hopkins University.

The research is being conducted with the support of the UK Humanitarian Innovation Hub (UKHIH) and the donor, the UK Foreign, Commonwealth & Development Office (FCDO). The work UKHIH supports aims to harness UK and international capabilities to support problem- led, systems innovation in the humanitarian sector.

Scenario-based projections will be released in a publicly available report alongside an open repository of data and analysis code, together with a methods annex technical document detailing the methodology used. An Arabic summary of each report will also be provided. The report will be periodically updated over the next five months and consider the evolving situation, availability of new data and improved methodologies.

All public facing project resources will be uploaded periodically on the website. If you wish to receive a notification when updated projections are published, please sign up to our mailing list.

Three different scenarios were chosen to be sufficiently distinct in terms of their likely impact, and, taken together, encompass a foreseeable range of trajectories for the crisis. The scenarios include a ceasefire (optimistic) scenario, a status quo (middle) case scenario (reflecting continuation of conditions seen between 7 October 2023 and 15 January), and an escalation (pessimistic) scenario. Projections for each scenario have a time horizon of 6 months. Each scenario is mainly specified in terms of three elements: 1) intensity and typology of the military activity in Gaza; 2) occurrence and duration of any pauses or ceasefires; and 3) extent of humanitarian space and operational adaptation.

For each scenario, we specified ranges for the values of specific risk factors for disease transmission or progression, including displacement, inadequate shelter and overcrowding, exposure to war injuries, food security/nutritional status, and insufficient water, sanitation and hygiene.

Finally, for each scenario we specified percent access to key health services, including vaccination, outpatient (primary) care and specific inpatient services such as intravenous antibiotics, rehydration and respiratory support. This was based on analysis of disruptions to these services, and realistic prospects for their rehabilitation.

We reviewed and extracted a range of previously collected data from the period before the war, including previous wars in Gaza, and from the period since the crisis’ start. We also sourced publicly available data and peer-reviewed articles from similar crisis contexts. Reports were curated by extracting key meta-data and vetted for quality based on set criteria. Finally, we used expert consultations and structured elicitation questionnaires to help us estimate key quantities for each scenario, including the transmissibility and fatality of infections and the availability and quality of trauma surgery care. As the project progresses, we will continue to source and update data and improve expert elicitation processes. An open repository of information will be published alongside each report.

These terms can be used interchangeably. In the Gaza Projections Report, excess mortality means the projected number of deaths above what one would have been expected without the conflict, during the same time period.

Israel’s public health infrastructure is intact and functioning, Gaza’s is not. Additionally, there is a better understanding of the impacts of the ongoing war in Israel compared to Gaza. In Gaza the war has disrupted health services and resulted in overcrowding, inadequate water, sanitation and hygiene conditions, and insufficient food intake. To better address the humanitarian crisis in Gaza both during and after the war, it is important to understand how the conflict is contributing to disease and death there.

We are seeking to estimate excess mortality due to all causes. As such, following the broad categories of the International Classification of Diseases, we have developed estimates of excess mortality due to 1) traumatic injuries, 2) infectious diseases, 3) maternal and neonatal causes, including deaths and stillbirths, and 4) non-communicable diseases. Furthermore, malnutrition is analysed as an underlying cause of mortality. Morbidity due to mental health may be included in future iterations of the projections. Taken together, these broad disease areas encompass most causes of mortality. These estimates do not distinguish between civilians and combatants.

Excess maternal, neonatal, and stillbirth mortality was estimated using the Lives Saved Tool or LiST; this extensively documented model uses published estimates of the relationships between changes in service coverage (e.g. antenatal care) and mortality or stillbirth risk. LiST models have been widely used and were built using a hierarchy of evidence on the effect of interventions on mortality.

For other disease areas, we developed computational models that, where possible, incorporated known ranges of uncertainty in various parameters such as exposure-outcome effects and the coverage of specific public health services. Briefly,

• For traumatic injuries, we used publicly available casualty data, data from similar conflicts, studies on survival rates of trauma patients, and expert-elicited values for case fatality of the most prevalent trauma types in conflict settings to estimate our projection in trauma injuries and deaths. Our estimates include the proportion of the injured who die of wounds due to lack of adequate medical services.
• For malnutrition, we estimated caloric intake and translated this into percent weight loss based on a published model; these weight losses were then applied to baseline data to project prevalences of acute malnutrition;
• For infections, we divided work into epidemic and endemic infections. For endemic infections, we multiplied baseline mortality data by relative risks representing higher transmissibility and case-fatality. For epidemics, we simulated a large number of epidemics based on expert-elicited values of epidemic risk, susceptibility to infection and disease based on vaccination coverage and natural exposure, and epidemiological parameters sourced from the literature. Susceptibility to infection and disease for different vaccine- preventable infections was estimated in a separate model.
• For non-communicable diseases, we multiplied baseline mortality data by relative risks representing changes in access to treatment among prevalent and incident cases, with relative survival with and without treatment quantified through literature review.

All our methods are extensively documented in a Methods Annex Technical Document can be reproduced using data and code provided.

No. The results are projections that use historical and present data and models to estimate the future under scenarios that we as researchers specifically set. Scenario-based projections are a ‘what-if’ exercise: we do not claim that any of the three scenarios will in fact materialise, but instead attempt to quantify the expected health consequences of these distinct alternative trajectories. The results should not be interpreted as a prediction or forecast.

All data sources come with their own biases and limitations, and the research team has attempted to adjust for these within the modelling approach. The best-available data have been used but, as is the case in all crisis situations, the available data are incomplete or may feature inaccuracies. The Methods Annex Technical Document discusses the limitations and biases of the estimates.

The range of scenario-based projections is not a prediction of what will happen but rather indicates what could happen under different scenarios, so as to enable appropriate planning, response and decision-making.

The project was reviewed and received approval from the Johns Hopkins Bloomberg School of Public Health Institutional Review Board and the London School of Hygiene & Tropical Medicine Research Ethics Committee.

All queries relating to the project should be directed to info@gaza-projections.org.