Lawfulness of earlier first statutory referrals by the MOJ

Law and practice Mental Health Tribunal
,
1

In Ministry of Justice, ‘Mental Health Casework Section stakeholder update’ (July 2025) it is explained that s37/41 restricted hospital order patients (or s45A hybrid order patients) who had previously been transferred under s48/49 before sentencing would begin to have their first referrals earlier: three years from the start of the s48/49 rather than the s37/41. This policy change is based on CS v Elysium Healthcare [2021] UKUT 186 (AAC), in which UTJ Mitchell decided that the tribunal retained jurisdiction in relation to a s47/49 application when the Court of Appeal replaced the original prison sentence with a s37/41 restricted hospital order.

Section 71(2) states:

The Secretary of State shall refer to the appropriate tribunal the case of any restricted patient detained in a hospital whose case has not been considered by such a tribunal, whether on his own application or otherwise, within the last three years.

Two solicitors told me last week that the new policy is definitely wrong because of the “statutory bar” on applying for the first six months of s37/41, and cited as authority AC v Southern Health NHS Foundation Trust [2024] UKUT 297 (AAC), in which UTJ Jacobs decided that the tribunal lost jurisdiction in relation to a conditionally discharged patient’s application when the patient was recalled.

I don’t think the situation is clear cut, and the cases are such a muddle that it seems impossible succinctly to describe their combined effect, but here are some thoughts.

  1. To be honest it had never occurred to me whether the patient’s “case” in s71 included earlier contiguous periods of detention as a restricted patient (though I’d wondered something similar about the tribunal rules in Potential reduction in paper reviews for CTO patients). The new policy is based on the conclusion that it does, though I’m not sure that a decision on continuing tribunal jurisdiction is necessary to reach that conclusion, or that the conclusion necessarily follows from a decision that the tribunal retains jurisdiction. UTJ Jacobs said in AC that the “statutory bar” on applying also covered “pursuing the application to its conclusion”. I can see the argument that whatever effect it has should apply equally to pursuing a reference, but it’s not clear that this has any effect on when the MOJ should make a reference.

  2. The decisive factor for UTJ Mitchell in CS was that that the patient had “remained throughout a restricted patient”. UTJ Jacobs in AC didn’t address that reasoning at all, but simply said that CS shows that “the precise analysis depends on the provisions involved”. He didn’t explicitly distinguish the cases, despite each remaining throughout a restricted patient. A difference between the two cases is that only CS remained a detained restricted patient, so arguably the CS approach still applies at least to this type of status change (which would support the new MOJ policy).

  3. When UTJ Jacobs mentioned the “statutory bar” in AC it was in the context of GM v Dorset Healthcare University NHS Foundation Trust [2020] UKUT 152 (AAC) in which he had decided that the tribunal lost jurisdiction in relation to the patient’s s3 application when a s37 hospital order was made. That case seems to have been decided on the importance of “judicial oversight” – which was provided by the judge when imposing s37 – with the the “statutory bar” being evidence of a policy that there should be no judicial oversight during the first six months (something which wasn’t directly relevant to AC, where there was no judicial decision and should be a mandatory recall tribunal).

  4. In DD v Sussex Partnership NHS Foundation Trust [2022] UKUT 166 (AAC) UTJ Jacobs himself subsequently decided that the tribunal retained jurisdiction on a s37/41 application after conditional discharge. If the “statutory bar” had been decisive then his decision would have been different (as a 12-month “statutory bar” applied to DD) but UTJ Jacobs said he could “find nothing inconsistent” with his decision in GM. Again, “judicial oversight” seemed decisive – and the “statutory bar” in this case was only relevant because it reduced “judicial oversight” (there would be no mandatory tribunal reference).

  5. In relation to the new MOJ policy there are arguments in both directions: either a s48/49 to s37/41 patient remains throughout a restricted patient, and the relevant sections are much more closely analogous to the s47/49 to s37/41 change in CS, so the approach in CS should be followed; or the s37/41 sentencing provided sufficient judicial oversight, and the statutory bar prevents further judicial oversight, for six months, so the approach in AC should be followed.

  6. I think a three-judge Upper Tribunal panel, or the Court of Appeal, should want to find a way to say that no application or reference survives a statutory bar on applications and no reference may be made during one, and that there had been sufficient judicial oversight when the statutory bar commenced. That seems to have been the Parliamentary intention, but it doesn’t seem to be the current interpretation provided by the Upper Tribunal. Though for as long as the MOJ continue with the early references, and the tribunal continue to accept them, this particular policy is unlikely to be challenged.

What do you think?

2

I’m firmly of the view that the imposition of a s.37/41 necessarily restarts the reference clock. DD was really an anomaly, driven by the unsatisfactory hypothetical of endless cycles of conditional discharge and recall preventing any tribunal scrutiny. I think the most prominent issue is that of “judicial oversight” and if any court/tribunal decides a section is appropriate (by imposing or reviewing it) the clock should reset.

When we argued AC, one of our submissions was that recall provides no “judicial oversight” - it’s an administrative act, but the statute treats it as if a fresh s.37/41 had been imposed by the court. Had that point been accepted, I think the law here would be far clearer.

1 Like
3

I’ve slept on it and I’m afraid have some more thoughts:

  1. The policy says it’s based on the decision about tribunal jurisdiction in CS, but really it’s based on the underlying logic of the reason for that decision. CS was about the fact that s70 doesn’t differentiate between different types of restricted patients who are detained in hospital when making applications, and the policy is really based on the fact that s71(2) similarly doesn’t differentiate when making references. That helps me understand something in point 1 above.

  2. There are two separate questions, one about whether the tribunal retains jurisdiction (on either applications or references), and the other about the timing of the “reference clock” as you call it, but they depend on similar considerations.

    (a) In relation to tribunal jurisdiction, I suppose if you treat s47/49 to s37/41 (CS) and recall (AC) as being effectively the same things (focussing only on the statutory bar on applications) we then would have two conflicting cases and the question would be whether the later overrules the earlier (I doubt it, especially as the later one is so sparsely reasoned and doesn’t expressly disagree, but I don’t know which should then be followed). If they aren’t about effectively the same things then they can both co-exist in relation to the different “provisions involved” (and I think that’s what most people would have thought).

    (b) In relation to the reference clock and the MOJ policy, I think CS with its focus on statutory construction (the generic use of “restricted patient”) is more relevant than AC with its focus on judicial oversight (and it being provided by the mandatory recall reference rather than the previous tribunal application). Eventually it might be that “s.37/41 necessarily restarts the reference clock” (see point 6 above) but I’m not sure it’s the law as it stands.

  3. I had a look at Anselm Eldergill, Mental Health Review Tribunals: Law and Procedure (Sweet and Maxwell, London 1997) and in relation to tribunal jurisdiction he was quite clear in chapter 10 that the statutory bar on applications meant that previous tribunal applications would cease to have effect. I wonder what he would say about this now, and about the reference clock. @medicolegal

  4. The MOJ supported the patient’s arguments in CS on tribunal jurisdiction, and it seems that the MOJ and MHT have been allowing s48/49 to s37/41 status change patients to continue with their applications for a while (see Tribunal application 48/49-37/41), so their new policy on the “reference clock” is just a continuation of this approach.

1 Like
4

I think a large part of the decision in CS (although I may be misremembering) was because the custodial sentence was replaced by a hospital order, so it took effect as if the hospital order was originally imposed when the custodial sentence was. I recall a point being that the patient would be disadvantaged if they were unable to pursue their application because they were successful in their appeal to the CrCoA.

There’s a clear semantic difference between:

a) P applying whilst on a s.47/49 or s.48/49, but then being sentenced to a s.37/41 (judicial oversight of the crown court breaking the chain); and

b) P applying whilst on a s.47/49, but their custodial sentence being retroactively replaced by the Court of Appeal (no judicial oversight about whether the patient should presently be under s.37/41).

Recognising I’m perhaps commenting on what I think the law should be than what it is, but in my eyes judicial oversight appears determinative (if we ignore AC).

5

The crux of it really is judicial oversight / supervision. That’s the underlying purpose of all of Part V

Why? First, most of the compuslory measures are achieved administratively, not judicially. But those measures have a heavy impact on a bunch of rights, so the Act wants to ensure that it’s all endorsed by a judicial body. No compulsory measure is ever left to just the doctor’s or secretary of state’s discretion (except the really short term powers and, weirdly in my view, CTO recalls).

Second, in relation to restricted patients, because this was one of the big innovations (or maybe evolutions) from 1959 to 1983, following Winterwerp and the 1982 X v UK. Those developments bothered the government, which led to the whole thing being expressly drafted to comply with art5.

Third, we know this because they literally said it. It’s in Jones in the notes to s69(2) - p412 of the 27th ed. Lord Belstead explained that the reason for s70(a) was that the detention had just been endorsed by a court (i.e. the sentecing judge) but some other patients whose cases hadn’t been before a court for a while were lumped together with those whose cases had (i.e. when transferred from prison to hospital). So 69(2) was enacted to fill that gap.

Upshot - the Act always wants every patient’s case to be considered by a judicial body, and never wants that not to happen (hence the backstop reference provisions).

And the Act wants that consideration to keep happening, because its purpose is to cause a change for the better in the mental state of the patient. So nature & degree etc can be expected to change, & so the justification for detention needs to be kept continually under review. It’s in the DNA of the statute.

Hence s.70: judicial consideration’s recently been done by a Crown Court judge so the Act basically says ‘ok, that’ll do for the time being but ideally I want it back before a judge after 6 months’. But if judicial supervision hasn’t been done recently, because the patient is a different species of ‘restricted patient’, the Act makes a correction to get the case back into its overarching scheme. Thus s.69(2).

The anomaly

The big anomaly is s.75(1)(b) - the Act applies s.70 to recalled CD patients but there’s no equivalent to s.69(2) for them, even though you’d expect there to be. You’d expect it because transferred prisoners and patients recalled from CD are both detained in hospital as a result of an adminstrative and not judicial decision (the SSJ in both cases). And it’s even worse for recalled patients because they’re going from liberty to detention, whereas prisoners are going from one form of detention to another.

So why enact s.69(2) for prisoners but not recalled patients? i.e. why disapply s.70(a) for one but not the other? This was the problem in AC. It’s a conundrum. I mean, you’ve got the manadatory reference in s.75(1)(a), but is that a complete answer to why the patient can’t apply where other patients can? The CA in Rayner said it wasn’t, but that it was ok because habeas/JR filled the gap.

Thing is, this takes us full circle back to X v UK and the genesis of the Act - the whole point was to enact a complete (statutory tribunal) scheme for judicial supervision that didn’t require patients to go to the High Court. That’s not to mention the entitlement to funding differences between the MHT and the Admin Court. And I think I’m right in saying that, eslewhere in the Act, the existence of a reference never precludes an applicaton.

One answer to this might be that the real purpose of s.75(1)(b) is to reset the clock for recalled patents’ right to apply to the FTT. This really is needed because otherwise the eligibility periods following recall would all be different depending on (a) the date of the original 37/41 and (b) the date of recall. So if eligiblity to apply continued to run from the original 37/41 date, some recalled patients would be apply to apply as soon as they were recalled, but others would have to wait for months. I haven’t yet figured out how the separate s.75(2) rights and timings fit into all this - it’s v tricky.

And that only relates to 70(b), not 70(a).

Anyway, sorry this is a bit of a bleuurghh post - the whole thing bothers me because I can’t yet see how the really clear logic of the rest of the Act fits with s75(1)(b). I’d welcome any insights about where I’ve gone wrong in my analysis (I fear I might have), and about what the anwer might be.