Projecting excess mortality due to infectious diseases during a crisis: methods and application to the Gaza Strip, 2023-2024

BackgroundExcess mortality due to infectious diseases is a key consequence of crises due to armed conflicts and natural disasters, but is measured retrospectively if at all. The Israeli operation in the Gaza Strip has exposed the population to known risk factors for increased infectious disease burden (overcrowding, malnutrition, inadequate water, sanitation and hygiene, disrupted health services). We developed a novel method for forward-projecting mortality due to both endemic and epidemic-prone infections in Gaza between 7 October 2023 and 6 August 2024 (ten months). MethodsAfter defining alternative crisis scenarios, we used expert elicitation to quantify the probability of specific epidemics and increases in transmissibility and case-fatality. We analysed past vaccination coverage and natural exposure information to estimate population susceptibility. For endemic infections, we took pre-war deaths, adjusted for COVID-19, as the counterfactual baseline, multiplied by the above parameters to compute projections. For epidemics, we projected deaths by simulating a susceptible-exposed-infected-recovered process. All analyses were by disease and age and implemented stochastically to propagate parameter uncertainty. ResultsOver the total period, we estimated that about 4000 endemic infection deaths would occur under a status quo scenario, ranging from 3400 to 4600 under ceasefire and escalation scenarios and comprising mainly COVID-19, influenza and pneumococcal disease. For epidemics, we projected 8500 deaths under the status quo but with a wide 95% confidence interval (0 to 100,300, rising to 128,900 under military escalation), reflecting high uncertainty about whether epidemics would indeed occur; cholera, measles and polio were the most likely epidemic threats. ConclusionThe model generated actionable projections for the Gaza Strip under a single framework covering all infectious diseases of public health interest and accounting for the effects of crisis on immunity, transmissibility and case-fatality. The method needs to be improved, e.g. by explicitly featuring the effect of malnutrition and strengthening expert elicitation techniques, before it can be generalised to other crises.