If you’ve followed the international campaign to brand Israel as a criminal country over the past two years, you’ve likely encountered statistics. Some float through U.N. briefings, appear in medical journals and law reviews, and are cited by activists as proof that Israel has waged a systematic campaign against Palestinian hospitals and medical workers. Those numbers come from the World Health Organization’s database of “attacks on health care.” It is treated, almost universally, as recording verified facts. Not so.

We spent months examining that database and how it actually works. What we found should concern anyone who cares about integrity around these issues. The data underlying some of the most serious accusations Israel faces does not hold up to scrutiny. Almost no one knows this. Few have studied the WHO dataset before.

First, the WHO’s definition of “attack” bears little resemblance to what the term means under the actual laws of war. It includes verbal threats and what the WHO calls “psychological violence,” a category broad enough to include intimidation or embarrassment. It also includes arrests, detentions, and security searches. By the WHO’s own written methodology, a doctor delayed at an IDF checkpoint and arriving late to a shift can be logged as an “attack on health care.” An ambulance searched for smuggled weapons counts the same way, even though the Red Cross considers this practice lawful. And Israel has good reason to search ambulances, given Hamas’ documented history of using ambulances to move fighters and weapons.

This means that when government officials, U.N. rapporteurs, or news organizations cite “hundreds of WHO-verified attacks” in Gaza or the West Bank, they are citing a number that conflates real violence with routine security operations and administrative friction. Our analysis found that only about a quarter of logged events in Gaza, the West Bank, and East Jerusalem meet even a loose threshold for violence, and only 6% involve a fatality, among the lowest severity rates the WHO tracks in armed conflicts anywhere in the world. The aggregate figure, repeated without that context in nearly every WHO citation, tells a story the underlying data doesn’t support.

The second problem goes to the heart of how Israel’s military conduct in Gaza gets judged. International law is clear that a hospital forfeits its special protection the moment it’s used for military purposes, such as a command center, a weapons cache, or a tunnel network. This is settled law, not an Israeli talking point. When the IDF moved into Al-Shifa Hospital in November 2023, following extensive evidence that a command-and-control complex sat beneath it, it was acting on a legal basis international law has long recognized. Yet the WHO’s data appears to log the episode as more than a hundred separate “attacks,” with no methodology that accounts for, or even acknowledges, the loss of the hospital’s protected status. Every citation of that figure treats a military operation with a recognized legal basis as an atrocity statistic.

Then there is Al-Ahli Arab Hospital — the case that should trouble anyone who cares about how quickly false accusations take hold and how rarely they’re corrected. When the hospital was struck on October 17, 2023, reportedly killing and wounding hundreds, Israel was blamed instantly, everywhere. Days later, forensic evidence (including from Western intelligence services) traced the explosion to a Palestinian rocket that misfired on launch. The correction reached a fraction of the audience that the original accusation did. And the event remains logged today, permanently, as the single deadliest entry in the WHO’s entire global database — still cited by activists and U.N. officials as evidence of Israeli war crimes and even genocide, years after the evidence pointed elsewhere. The WHO’s system has no mechanism by which Israel, or any state, can formally contest a mistaken entry.

Who is actually filing these reports? In Gaza, the primary reporting channel is a health ministry that answers to Hamas, an organization with every incentive to document Israeli actions in granular detail and zero incentive to disclose its own use of hospitals and ambulances for military purposes. The database doesn’t account for that asymmetry, and neither do the international bodies that cite it. The result is a dataset structurally tilted before a single event is even logged, built on reporting from an organization that the United States, the European Union, and Israel all designate as a terrorist group.

These numbers are being fed into prosecution dossiers, U.N. Human Rights Council reports, and academic literature building the legal case against Israeli officials. Diaspora Jewish organizations fighting these accusations in op-eds, in Congress, and on campus are often arguing against a statistic their opponents treat as beyond dispute, with few people knowing how poorly it was assembled.

We are not arguing that no legitimate criticism of Israeli conduct exists, or that health care workers in conflict zones don’t deserve real protection — they do, and better data would serve that goal far better than what currently exists. But everyone has a direct stake in knowing that one of the most-cited evidentiary pillars in the international campaign against Israel is far shakier than its footnotes suggest. They should demand that the WHO fix its methodology by adopting clearer definitions, independent verification, and a right of reply for accused states. The WHO should act like the credible institution it claims to be, especially when the cost of its sloppiness falls so heavily, and undermines the entire credibility of its database on attacks on healthcare. 

Leslie Johns is a professor of political science and law at UCLA. Gregory Rose is a professor of law at the University of Wollongong. They are the authors of a report on the WHO’s health care attack data published by the Henry Jackson Society and The Hague Initiative for International Cooperation.