Introduction

Never in the recent history of armed conflicts – with the possible exception of Iraq and Darfur (Sudan) in the early 2000s – has the assessment of civilian death tolls held as much significance as in Gaza, after Hamas’s massacre in Israel on October 7, 2023. This is reflected in the media coverage, political statements, and the interest shown by researchers.

Several studies have been conducted to estimate mortality in Gaza in the current conflict. This editorial provides a background to these studies, with a focus on the recent study1 published in The Lancet (henceforth, the Lancet study) and discusses the challenges of communicating findings with significant political implications. The Lancet study estimates that the health impact of the war in Gaza, in terms of excess traumatic deaths, has substantially exceeded the official figures reported by the Gaza Ministry of Health (GMoH). The study has attracted widespread media attention, but, to my knowledge, it has yet to prompt a public response from the scientific community.  

The companion article in this issue of Epidemiologia&Prevenzione (pp. 84-90) explores the risks of data from conflict settings – particularly numbers of civilian deaths, refugees, and internally displaced persons – being instrumentalised for political agendas, potentially fuelling hate in an increasingly polarised world. Estimates of civilian death tolls in conflicts are often taken at face value and widely circulated by journalists and can stir up public opinions. They are readily embraced and leveraged by activists and politicians when they align with their agendas. Conversely, political actors may dismiss the estimates outright, without proper scrutiny or counterarguments, when they expose the harmful consequences of their policies. Former U.S. President Biden expressed doubts about the GMoH reported death toll, stating that he had “no confidence in the number that the Palestinians are using”.2 Israeli Prime Minister dismissed the GMoH death figures, labelling them as “fake data from a terrorist organisation”.3 Neither the U.S. nor the Israeli authorities provided evidence to substantiate their statements.

GMoH mortality data and previous studies

The track record of the GMoH mortality figures from past wars is reasonably good. For instance, for the 2014 war in Gaza, the discrepancy between GMoH data and those reported by Israeli authorities was 8%,4 while the difference with the United Nations Commission of Enquiry’s figures was 2.6%.5 

In response to scepticism from some political parties regarding the GMoH reported fatalities in the current war, studies have sought to evaluate the accuracy of these figures. A study compared the GMoH mortality rates of the first month of the war with corresponding rates among United Nations Relief and Works Agency (UNRWA) staff, independently reported, as well as with rates among healthcare workers, published by the GMoH and WHO.6 The analysis showed consistencies among the rates and their age distribution, concluding that data fabrication was unlikely. A similar analysis by other researchers, covering a different period, showed consistent mortality trends between the GMoH and UNRWA data.7 The authors of the second study argue that the lower mortality rates reported by the GMoH, compared to those of UNRWA, support the claim that the GMoH figures have not been inflated. Furthermore, an unknown number of bodies – possibly as many as 10,000 as of May 2024 – were feared to remain buried beneath the rubble of bombarded buildings.8

At the time of going to press, a new study was published.9 The study, based on demographic methods, confirms the overall accuracy of the GMoH data. It also estimates a life expectancy loss of 31.1 years for both sexes combined, during the first 12 months of the war. This estimate nearly halves prewar life expectancy levels.

Another study estimated that, as of July 2024, “186,000 or even more deaths could be attributable to the current conflict in Gaza”.10 Estimating total excess deaths, i.e., the number of deaths that would not have occurred in the absence of the conflict, is challenging, as discussed in the companion E&P article. The authors used a ‘conservative’ ratio of four indirect deaths per one direct death,11 based on a range of 1 to 15 ratios estimated in conflicts over recent decades, and applied it to the GMoH mortality data.12,13 However, the choice of this ratio is questionable, given the assorted methods used to generate these estimates across different contexts, each with a unique pre-war burden of disease, conflict intensity, health system performance, and levels of humanitarian assistance.

Projections of excess mortality were developed for the period of February-August 2024, covering the scenarios of ceasefire, status quo, and war escalation.14 The study estimated that, with no epidemics occurring, the projection for the ceasefire scenario would be 6,550 excess deaths, 58,260 excess deaths for the status quo, and 74,290 excess deaths for the war escalation. In a separate study, covering May to August 2024, an estimated 3,509 traumatic injury deaths, or 39 deaths per day, were projected for Gaza’s Rafah governorate, where most Gazans sought refuge.15

The Lancet study

The study used a capture-recapture analysis to estimate deaths due to traumatic injuries in Gaza between October 2023 and June 2024. Three lists were used: from the GMoH morgues, a GMoH online survey, and social media obituaries. Duplicates were removed using weighted variables. Records were then matched across the three lists and a capture-recapture analysis was performed, using statistical models, whose outcomes were averaged. The study estimated 64,260 deaths (95%CI 55,298-78,525) due to traumatic injuries during the period, 41% more than the GMoH reported mortality data. The rate ratio with all-cause mortality in 2022 was estimated at 14.0 (95%CI 12.8-17.6). Women aged 18-64 years, children under 18 years of age, and people of more than 65 years – groups least likely to be combatants – represented respectively 20.2%, 33.3%, and 5.8% of the deaths for which age and sex were available. The total estimate of 64,260 deaths, or about 3% of pre-war population, is a much higher mortality figure than that recorded in previous conflicts in Gaza. The analysis excluded deaths from indirect, non-violence-related causes, such as those resulting from the collapse of the health system, starvation,16 or other risk factors, which would have raised the total death toll.

The Capture-Recapture method, originally developed to estimate wildlife population sizes, has been adapted for other purposes, including epidemiological studies. It was first applied to estimate civilian mortality during the Biafra War of 1967-1970.17 Since then, its use during conflicts has been limited to Kosovo,18 Colombia,19 and Sudan,20,21 arguably due to the advanced statistical expertise required. In simple terms, the method integrates multiple information sources by analyzing the overlap among independent lists of decedents, each with varying degrees of incompleteness. It is reasonable to assume that the lists of deaths in Gaza miss some cases, with the level of incompleteness increasing over time. Indeed, in humanitarian crises, health information systems often deteriorate as health workers and record-keeping officials are killed, data transmission is disrupted, hospitals are destroyed or overwhelmed, and lifesaving tasks take precedence over data collection. Merging incomplete lists inevitably results in an undercount of deaths. The capture-recapture method attempts to adjust for the number of deaths missing from all lists by fitting statistical models to the overlap among lists. The estimate of missed deaths is then added to the deaths recorded in the de-duplicated lists to calculate the total estimated death toll. The method implies that matching errors among lists are minimised by removing duplicates and that some statistical assumptions, such as the independence of lists, are satisfied. 

“All models are wrong, but some are useful”22

The article’s authors deserve commendation for their efforts in tackling the formidable methodological challenges involved in addressing the gaps in the GMoH data. The integration of multiple data sources has, however, involved the use of advanced statistical methods. To the average reader of The Lancet, even more to the journalists who have reported on the study’s findings, the methods may have appeared as a black box, challenging to unpack in terms of complexity, underlying assumptions, and limitations. 

The Lancet article has understandably and rightfully received prominent press coverage in the current political context, with major news outlets featuring the new death estimates as headlines. The media, however, are often drawn to the apparent and seductive simplicity of a few numbers to represent complex phenomena, such as the impact of war. The fact that the study’s researchers are world-renowned, affiliated with top-tier academic institutions, and the medical journal itself is highly respected, can also be intimidating. As George Orwell once remarked, “To say ‘I heard it on the BBC’ is almost the same as saying ‘I know it to be true’”.23 Furthermore, the estimated 41% undercount of traumatic deaths, widely circulated, could create a strong “anchoring effect”:24 our tendency to fixate on the first number we have heard, especially if it is strikingly high or low.

Conclusions

The Lancet study may stimulate a broader discussion among public health experts on various aspects, including the methods for assessing civilian mortality in extreme crisis settings and the importance of effectively communicating findings to audiences with limited understanding of the complexities involved. Data do not speak for themselves; they must be interpreted within the broader humanitarian crisis context, acknowledging the inevitable limitations in the evidence. Researchers and public health experts share the responsibility to translate study findings – especially those with potentially significant political and policy implications – into narratives and messages that minimise the risk of being misinterpreted or manipulated.  

From a public health point of view, the precision of estimates is less important for guiding the humanitarian response than the assessment of the magnitude of the crisis’s impact on health. Even if the GMoH latest report of 47,161 cumulative deaths25 were correct and not undercounted, the death toll would remain tragically high and could not be justified by any political or military motive: a single civilian death in a conflict is one too many.

At the time of writing, a fragile ceasefire between Israel and Hamas remains in effect. However, even if, hopefully, the violence was to finally end, the impact of the war on the health and wellbeing of Gazans will remain for a long time. Many more civilians could die in the coming months due to the public health crisis and the collapse of the health system. Additionally, the reverberations of the war will result in many more people suffering from disabilities, the hazard of mines and unexploded ordnance, and the consequences of delayed treatment of diseases, as well as mental and psychological problems. Recovery will require a political solution to secure peace, substantial aid and investment to rebuild thousands of homes and critical infrastructure, the creation of sustainable livelihood opportunities, and – perhaps most challenging of all – the restoration of Gazans’ trust in the future.

Behind the statistics and their controversies lie individual tragedies, as Nesrine Malik reminds us about Gaza: “Too often, individual losses during this war have been subsumed by a broader tussle over whether the numbers of the dead were correct, whether they were justified and even necessary”.26

Conflicts of interest: none declared

References and notes

  1. Jamaluddine Z, Abukmail H, Aly S, Campbell OMR, Checchi F. Traumatic injury mortality in the Gaza Strip from Oct 7, 2023, to June 30, 2024: a capture-recapture analysis. Lancet 2025:405(10477):469-77. doi: 10.1016/S0140-6736(24)02678-3
  2. McGreal C. Can we trust casualty figures from the Hamas-run Gaza health ministry? The Guardian, 27.102023. Available from: https://www.theguardian.com/world/2023/oct/26/can-we-trust-casualty-figures-from-the-hamas-run-gaza-health-ministry (last accessed: 25 January 2025).
  3. Horton, J, Sardarizadeh, S, Durbin, A. Gaza war: Why is the UN citing lower death toll for women and children? BBC, 6 May 2024. Available from: https://www.bbc.com/news/world-middle-east-69014893 (last accessed: 21.01.2025).
  4. Debre I. What is Gaza’s Ministry of Health and how does Gaza’s Ministry of Health calculate the death toll? AP News, 07.11.2023. Available from: https://apnews.com/article/israel-hamas-war-gaza-health-ministry-health-death-toll-59470820308b31f1faf73c703400b033 (last accessed: 26.01.2025).
  5. United Nations. Report of the independent commission of inquiry established pursuant to Human Rights Council resolution S/21. 25.06.2015. Available from: https://www.un.org/unispal/document/auto-insert-185919/ (last accessed: 25.01.2025).
  6. Jamaluddine Z, Checchi F, Campbell OMR. Excess mortality in Gaza: Oct 7-26, 2023. Lancet 2023;402(10418):2189-90. doi: 10.1016/S0140-6736(23)02640-5.
  7. Huynh BQ, Chin ET, Spiegel PB. No evidence of inflated mortality reporting from the Gaza Ministry of Health. Lancet 2024;403(10421):23-24. doi: 10.1016/S0140-6736(23)02713-7
  8. United Nations News. 10,000 people feared buried under the rubble in Gaza. 02.05.2024. Available from: https://news.un.org/en/story/2024/05/1149256 (last accessed 21 January 2025)
  9. Guillot M, Draidi M, Cetorelli V, Monteiro Da Silva JHC, Lubbad I. Life expectancy losses in the Gaza Strip during the period October 2023, to September, 2024. Lancet 2025:405(10477):478-85. doi: 10.1016/S0140-6736(24)02810-1
  10. Khatib R, McKee M, Yusuf S. Counting the dead in Gaza: difficult but essential. Lancet 2024;404(10449):237-38. doi: 10.1016/S0140-6736(24)01169-3
  11. Indirect deaths are those that are not caused by the direct consequences of intentional violence (traumas and injuries).
  12. Geneva Declaration Secretariat. Global burden of armed violence. Geneva: Geneva Declaration Secretariat; 2008. Available from: https://www.unodc.org/documents/data-and-analysis/Crime-statistics/Global-Burden-of-Armed-Violence-full-report.pdf (last accessed: 20.01.2025).
  13. Wise PH. The Epidemiologic Challenge to the Conduct of Just War: Confronting Indirect Civilian Casualties of War. Daedalus 2017;146(1):139-54.
  14. Jamaluddine Z, Chen Z, Abukmail H, et al. (2024). Crisis in Gaza: Scenario-based health impact projections. Report One: 7 February to 6 August 2024. London, Baltimore: London School of Hygiene and Tropical Medicine, Johns Hopkins University; 2024. Available from: https://gaza-projections.org/gaza_projections_report.pdf (last accessed: 23.01.2025).
  15. Igusa T, Chen Z, Gnaedinger, A, Checchi F, Spiegel P. Crisis in Gaza: Projected Deaths due to Traumatic Injuries in the Rafah Governorate. Report 2: 20 May to 17 August 2024. London, Baltimore: London School of Hygiene and Tropical Medicine, Johns Hopkins University; 2024. Available from: https://gaza-projections.org/docs/report2/traumatic_injuries_rafah.pdf (Last accessed: 08.01.2025).
  16. De Waal A. How Many People Have Died of Starvation in Gaza? World Peace Founfation; 17.12.2024. Available from: https://worldpeacefoundation.org/blog/how-many-people-have-died-of-starvation-in-gaza/ (last accessed: 28.01.2025).
  17. Roberts L. Advances in monitoring have not translated into improvements in humanitarian health services. Prehosp Disaster Med 2007;22(5):384-89. doi: 10.1017/s1049023x00005094
  18. Ball P, Betts W, Scheuren F, Dudukovich J, Asher J. Killings and refugee flow in Kosovo, March-June 1999: A report to the International Criminal Tribunal for the FormerYugosla-via. American Association for the Advancement of Science 2002. Available from: https://www.icty.org/x/file/About/OTP/War_Demographics/en/s_milosevic_kosovo_020103.pdf (last accessed: 20.01.2025).
  19. Lum K, Price M, Guberek T, Ball P. Measuring Elusive Populations with Bayesian Model Averaging for Multiple Systems Estimation: A Case Study on Lethal Violations in Casanare, 1998-2007. Stat Politics Policy 2010. doi: 10.2202/2151-7509.1005
  20. Dahab M, Abdelmagid N, Kodouda A, Checch F. Deaths, injuries and detentions during civil demonstrations in Sudan: a secondary data analysis. Confl Health 2019;13:16. doi: 10.1186/s13031-019-0199-8
  21. Dahab M, AbuKoura R, Checchi F, et al. War-Time Mortality in Sudan: A Capture-Recapture Analysis. Lancet Preprint, 12.11.2024. Available from: http://dx.doi.org/10.2139/ssrn.5016438 (last accessed: 28.01.2025).
  22. Box GEP. Robustness in the strategy of scientific model building. Technical Summary Report #1954. Mathematics Research Center, University of Wisconsin-Madison, 1979. Available from: https://apps.dtic.mil/sti/pdfs/ADA070213.pdf (last accessed: 28.01.2025).
  23. Hudson M, Stanier J. War and the media: A random searchlight. New York: New York University Press; 1998.
  24. Kahneman, D. Thinking, fast and slow. London: Penguin Books; 2011.
  25. UN OCHA. Reported impact snapshot | Gaza Strip (22 January 2025). Available from: https://www.ochaopt.org/content/reported-impact-snapshot-gaza-strip-22-january-2025 (last accessed: 27.01.2025).
  26. Malik N. Goodbye to the lost children of Gaza. You were loved, you are remembered, you did not deserve it. The Guardian, 27.01.2025. Available from: https://www.theguardian.com/commentisfree/2025/jan/27/children-gaza-loved-remembered-innocent-mourn (last accessed: 27.01.2025).

 

          Visite