Why has the UN revised its Gaza death toll for women and children?

The United Nations has revised down the figure it publishes for the proportion of women and children killed in Gaza, leading to claims fewer Palestinians have been killed since Israel launched its offensive following the Hamas attack of 7 October.

On 6 May, the UN said that 69% of reported fatalities were women and children. Two days later, it said this figure was 52%.

The overall number of reported deaths in Gaza - which currently stands at more than 35,000 - has not changed, but the UN now says incomplete information has led to the revision.

The UN says it is now relying on figures from the Hamas-run health ministry in Gaza, rather than from the Hamas-run Government Media Office (GMO).

The Israeli foreign minister highlighted the UN's change, external, calling the lower reported death tolls for women and children "the miraculous resurrection of the dead in Gaza" and accusing the UN of relying on "fake data from a terrorist organisation".

The figures have been released in different ways by the two separate authorities - the GMO and the health ministry - and the UN has now switched from quoting the former to the latter to break down demographic data.

At the start of the war the health ministry only reported deaths for which details had been registered in hospitals, whereas from November, the GMO included an additional category of deaths recorded in "reliable media reports".

More recently the health ministry has included these media reports compiled by the GMO, along with an additional category of deaths recorded by family members by filling in an online form.

Gaza's health ministry is now therefore reporting these figures:

Some of these for which information is missing (such as an ID number or date of birth) are included in the overall number of deaths in the health ministry figures, but are not broken down demographically.

The number of women and children killed is deeply contested because it is often interpreted as an indication of the number of civilian deaths.

In a recent interview, external, Israeli Prime Minister Benjamin Netanyahu said about 30,000 people had been killed in Gaza, with 14,000 of them being "terrorists" and 16,000 civilians. He did not provide any evidence for those figures.

Keeping precise records of casualties during wartime is challenging, so it is therefore not unusual for the numbers reported to vary, according to Prof Erica Charters who specialises in the study of war.

The initial figure for how many people were killed in Hamas's 7 October attacks was originally put at 1,400 by Israeli authorities, but was later revised down to about 1,200.

At the time, the Israeli authorities said the misidentification of some of the burnt bodies was the reason for the discrepancy.

The GMO has consistently given a higher figure for the proportion of women and children in all fatalities than has the health ministry.

On 6 May, the UN reported 34,735 deaths - of which there were 9,500 women and 14,500 children, citing the GMO as its source.

The two days later, the UN released a further report, switching its sourcing to the health ministry.

The result of this was that although the overall recorded death toll was almost unchanged (34,844), the number of registered deaths of women (4.959) and children (7,797) had both fallen significantly.

This difference was because those individuals with incomplete information were not included in the demographic breakdown.

The GMO says the proportion of women and children killed in the conflict is around 70%.

The UN's latest report, citing health ministry data, says out of 24,686 fully identified fatalities, 52% were women and children, 40% were men, and 8% were "elderly", without specifying whether they were male or female.

It isn't clear what age the report considers to be "elderly" and how it categorises children.

We have done our own analysis of detailed data released by the health ministry, and also found that 52% of the fatalities were listed as women and children (anyone under 18). In addition, 43% were men and another 5% were "unknown" (missing information such as an age or gender).

For the GMO's figure to be compatible with the health ministry's data, almost all of the 10,000 deaths not fully identified by the ministry would have to have been women and children.

"It's not logically impossible... but it really strains credibility," says Prof Michael Spagat, who specialises in examining death tolls in conflicts around the world.

We asked the GMO why the proportion of women and children they have recorded as killed is so much higher than in the health ministry's data, but they did not provide a direct response to the discrepancy.

We also approached the health ministry for comment.

Prof Spagat also told the BBC that the death toll in Gaza was "staggeringly high" given the war has only been going on for six months, noting that the verified hospital deaths still represent about 1% of Gaza's pre-war population of 2.4 million.

Asked by BBC Verify why it made the change, the UN said it used the GMO data for women and children killed because the health ministry had not been publishing these figures.

"Once the ministry of health in Gaza published breakdowns that were more comprehensive, backed up by a list of names, the UN reports reflected that data instead," a UN spokesman said.

He added that in both cases, the figures were attributed to their source and a "clear note" was added making clear the UN's "inability to presently provide an independent confirmation of the data".

The UN has previously said that it has a "long-standing co-operation" with the Gaza health ministry, and that its previous reporting has been considered credible and "well developed".

Data analysis by Rob England and Daniel Wainwright