Lord Moylan to Ed Humpherson: Abortion Statistics for England and Wales (6 May 2025)

Dear Mr Humpherson,

I apologise for not replying before to your letter of 20th February send in response to mine about statistics on complications arising from abortion. I am glad that you have agreed to undertake a compliance check of the statistics on abortion complications published by DHSC to ensure they meet the standards in the Code of Practice for Statistics. I am only disappointed that the check is to be delayed until the first quarter of 2026/7.

One of the reasons for the delay, you say, is to allow time for DHSC to work with digital experts and system providers “to improve the design of the Abortion Notification System”. I find this rather alarming, as it seems to me to put the cart before the horse. Given the manifest inadequacy of the Abortion Notification System to capture more that a small number of complications, would it not be better for the compliance check to be undertaken first so that the inadequacies, and remedies for them, could be identified and this analysis then be used as a basis for improving the design of the ANS? I should be grateful to know what your reflections on this might be.

Perhaps I could comment on a few other points on your letter, but as background may I just be clear these are “national statistics” which are required to be collected under statute (Abortion Regulations 1991, regulation 4 and schedule 2, which specifies the complications to be reported). It is not therefore wholly relevant whether there has been “user demand” for the more comprehensive data set produced by the Office for Health Improvements and Disparities in November 2023 and in any event, it is not clear who the relevant “user” should be, since the availability of reliable statistics on complications goes to the question of informed patient consent, not merely to the interest shown by clinicians.

There are always of course certain advantages in maintaining a statistical series over time and you state that this is also true in the case of ANS data as currently recorded. But there is no suggestion, certainly not from me, that the collection of ANS returns on complications should be abandoned. Information on complications drawn from ANS data is an indispensable part of the picture, but one of dwindling relevance given the changes in the abortion clinical landscape. Having more robust statistics relevant to the real world would not mean a rupture in the ANS data series. And in any case, ANS data over a much wider field that simply complications. Whilst insufficient for the reporting of complications, the ANS is necessary for those other data.

You refer to the resource constraints faced by analytical teams at DHSC and I can appreciate that the compilation for the first time of relevant Hospital Episode Statistics by OHID will have been somewhat onerous. But each of the complications arising from abortion in a hospital setting is assigned an ICD-10 code and a new system would make routine the central reporting of these codes and the number of incidents pertaining to each. (The ICD019 codes are set out in the Annex to the OHID report.) Indeed this seems to me to illustrate the compelling reasons for you to carry out your compliance check first and then allow the DHSC to spend resources digitalising their collection system once it is clear what data you are looking to collect and whether they should include the relevant ICD-10 codes recorded by hospitals.

Finally, you say that the ANS data capture complications not requiring hospitalisation. I would make the point rather differently. As I said in my earlier letter to you, even with the huge enhancement in robustness achieved by the OHID, there is still a lacuna in regard to complications dealt with by a GP (or for that matter resolved with telephone advice from 111). I do not have a solution to suggest for making good this deficiency except to say that I suspect the numbers are small and that in any case, a GP or 111, presented with the symptoms of the principal recordable complications (haemorrhage, uterine perforation, sepsis) would in any event refer the patient to A&E and thus the case would be captured in the Hospital Episode Statistics.

As you can see, I am concerned that, by deferring the compliance check until after the developments the DHSC envisage are carried out, opportunities will be lost and excuses created for further deferring the improvement in statistics for which the OHID has blazed a path. I should be keen to hear your considered view.

I should be happy to meet if you felt that would be useful.

Yours sincerely,

Moylan

Lord Moylan to Ed Humpherson: Abortion Statistics for England and Wales (6 May 2025)

Related Links: