In a recent post I discussed the Department of Health’s electroconvulsive (ECT) statistics and how only a fraction of the ECT actually used in England finds its way into the statistics. Reasons suggested by the Department of Health for the under-reporting of ECT have included:
i) That many trusts do not realise that ECT should be recorded as an operative procedure
ii) Mental Health Trusts which do not usually carry out operations sometimes do not appreciate that there may be ECT codes relevant to them, and therefore leave them out.
That was written in 2003 (Electro Convulsive Therapy: Survey covering the period from January 2002 to March 2002, England, published by the Department of Health, page 2). Since 2006 however the Department of Health has changed the terminology: what they now count is “procedures and interventions” and by no stretch of the imagination can psychiatric hospitals say they do not carry out procedures and interventions.
The HES website (Hospital Episodes Statistics) now publishes provider level statistics, that is statistics from the individual trusts. (A trust is the governing body of a hospital or group of hospitals. There are not vast numbers of mental health trusts in England – fewer than 100).
It is possible to see if the figures for ECT use are reasonable or not by comparing them to the figures produced by the Care Quality Commission (CQC) for requests to use ECT on non-consenting patients. On most occasions when a psychiatrist in England wants to give ECT to someone who cannot consent, they have to apply for permission from the CQC in the form of a request for a Second Opinion Appointed Doctor (SOAD) to visit. The CQC keeps count of these requests. Apart from anything else they have to pay the SOADs who are paid per visit, and that is an incentive to accurate record-keeping. Not all requests result in ECT being administered, but most do (some requests are cancelled and on very rare occasions SOADs say no). People being treated without their consent still form a minority of ECT patients, albeit an increasingly large one – in the 2002 survey it was nearly a quarter. So it is possible to estimate – very roughly – how much ECT a hospital uses by how many requests for SOAD visits they make.
In the provider level analysis for 2009/2010, over a quarter of trusts had no reports of ECT, although they had made requests to the CQC for a SOAD visit to approve ECT for non-consenting patients. Amongst the trusts that reported no ECT were some with the highest rates of SOAD requests:
- Greater Manchester West MH NHS Foundation Trust (41 SOAD requests – no ECT reported)
- Leicestershire Partnership NHS Trust (33 SOAD requests – no ECT reported)
- North Essex Partnership NHS Trust (59 SOAD requests – no ECT reported)
- Northumberland, Tyne and Wear NHS Trust (39 SOAD requests – no ECT reported)
- South London and Maudsley NHS Foundation Trust ( 45 SOAD requests – no ECT reported)
- West London MH NHS Trust (38 SOAD requests – no ECT reported)
Only one or two trusts were reporting figures that at least looked probable:
- Cornwall Partnership NHS Trust (12 SOAD requests – 451 ECT administrations)
- Kent and Medway NHS and Social Care Partnership Trust (43 SOAD requests – 547 ECT administrations, which is just about possible but would suggest a high use of ECT on non-consenting patients)
The rest of the trusts are reporting some ECT use but the numbers are improbably small. That means that the problem isn’t just that hospitals don’t realize that the coding system applies to them. For example, Lincolnshire Partnership NHS Foundation Trust made 12 requests for SOAD visits but reported only 18 administrations of ECT.
Are some hospitals reporting patients, rather than individual administrations of ECT (which is what they should be reporting)? Is there confusion about what counts as a main intervention or procedure? Or just totally random under-reporting?