Hexpress Healthcare Ltd v CQC  EWHC 2943 (Admin) – Is 0.01% a fair sample?
Many healthcare providers will be all too aware of how difficult it can be to obtain an interim injunction against the CQC. Hexpress Healthcare will present few surprises in this regard. However, two points of interest emerge. The court reconsiders the CQC’s procedure for considering factual accuracy comments (“FAC”) from a provider prior to publication. In addition, the court makes brief observations on the appropriateness, or otherwise, of making generalised criticisms about a provider from a tiny sample of patient records.
Hexpress operates an online service to provide medication for over 60 conditions. Patients fill out an online set of questions, the information is reviewed by a doctor, and the prescription issued – all without any face to face involvement with the patient. Between April and August 2021 five concerning incidents arose regarding overdoses of medication, two of which were fatal. The CQC conducted an announced inspection in May 2022. The team selected a sample of nine patient records, of which three were unreviewable for technical reasons. The sample of six records which could be reviewed represented just 0.01% of all patient records available. Out of these all but one raised concerns.
On 17 May 2022 the CQC issued a draft report giving Hexpress an overall assessment of “requires improvement”. Hexpress submitted its FAC in response. On 20 October 2022 the CQC sent Hexpress a final inspection report and confirmed that this report and its response to the FAC would be published on 25 October 2022. Its assessments remained unchanged. Before it could do so, Hexpress issued judicial review proceedings and obtained an interim injunction until the permission hearing. This was heard before Mostyn J.
The majority of his judgment dealt with the suggestion raised by Hexpress that the CQC’s approach to the FAC was procedurally unfair.
Under the Health and Social Care Act 2008 (“the Act”), the CQC is empowered to register and review health and social care services. Whilst its reports have the power to “wreak serious economic and reputational damage”, it is striking the Act allows no right of appeal or challenge.
The CQC’s Handbook mitigates the severity of this position to a limited extent in that it allows a provider to make representations both before and after publication. Andrews J in R (On the Application of SSP Health Ltd,) v Care Quality Commission  EWHC 2086 (Admin) summarised it in this way:
“i) prior to publication, service providers can challenge the factual accuracy and completeness of the evidence and findings on which the ratings are based, as well as the proposed ratings themselves…
ii) after publication, service providers can seek a review of ratings. The Handbook makes it plain that the only grounds for requesting such a review is that CQC did not follow the process of making ratings decisions and aggregating them (i.e. the process set out in the Handbook). Service providers cannot request reviews on the basis that they disagree with the judgments made by CQC, “as such disagreements would have been dealt with through the factual accuracy checks…”.”
This case concerned the first of these avenues. This right to provide comments is limited. SSP makes clear this factual accuracy check process applies only to primary facts which can be objectively shown by reference to other hard evidence to be demonstrably wrong. It does not extend to “abstract fact” like a person’s state of mind or a finding based on the evaluation of these primary facts.
SSP discussed the appropriate approach to the FAC process. This judgment was interpreted – not least by Holgate J in R (Babylon Healthcare Ltd) v Care Quality Commission  EWHC 3436 (Admin) – to mean the following steps should be taken by the CQC:
“i) the CQC inspection team sends the draft report to the provider;
ii) the provider sends to the CQC its list of alleged factual inaccuracies with reasons (“the FAC list”);
iii) the CQC inspection team prepares its response to the FAC list (“the CQC response”);
iv) the CQC then sends its response to the provider;
v) if the CQC response is not accepted by the provider then it spells out its reasons in a reply (“the reply”) and at the same time seeks an independent internal review of the CQC inspection team’s response in the light of the reply;
vi) an independent reviewer considers the draft report, the FAC list, the CQC response, and the reply and makes a recommendation;
vii) in the light of the recommendation the CQC inspection team then makes a final decision; and
viii) the report is then finalised and published.”
Prior to the instant case, however, the CQC decided to condense the process as follows:
“i) the CQC inspection team sends the draft report to the provider;
ii) the provider sends to the CQC its FAC list;
iii) the CQC inspection team prepares its response to the FAC list;
iv) an independent reviewer considers the draft report, the FAC list, and the CQC response, and makes a recommendation;
v) in the light of the recommendation the CQC inspection team then makes a final decision; and
vi) the report is then finalised and published.”
Hexpress challenged this innovation as one of its grounds of review. However, declining permission on this point, Mostyn J commented that he considered this new condensed approach to be compatible with SSP. If it was not, he declined to follow SSP. In addition, he expressly disagreed with Babylon. In his view, the CQC’s amended procedure was significantly quicker and removed “time-wasting work of supererogation”. The condensed approach did not result in any unfairness:
“I remind myself that we are talking here about errors of primary concrete fact, and not about value judgments on those primary facts. Those errors will have been clearly pointed out in the initial FAC request by the provider. They will either be right or wrong. If the alleged errors are rejected by the inspection team they will have had to have done so for good reasons, referable to hard objective evidence. It is hard to see what the reply can contribute further to the debate.”
On the other hand:
“time is of the essence when preparing the consequential report. Given the nature of the main objective of the CQC, namely to protect the public… This key factor strongly militates in favour of the condensed process”.
Having rejected this procedural challenge, Mostyn J gave permission to appeal on one ground. Was the CQC justified drawing general conclusions from such a small sample? As he put it:
“can a generalised criticism that the provider’s standard of safety is “inadequate”, or that its standards of effectiveness and leadership “require improvement”, logically and fairly be extrapolated from six medical records out of 60,000?”
He granted permission for expect statistical evidence, and for the CQC to adduce evidence relating to resourcing issues. As he observed, unhelpfully perhaps for Hexpress:
“I do not doubt that there would be resource implications if the CQC’s inspection teams had to review hundreds of medical records on every inspection such as this. The answer to the questions might be, therefore, that given the resources available to the CQC, and the duty imposed on it to protect the public, there is no other way of forming an evidence base from which to extrapolate conclusions, notwithstanding that the result of the ratings and of the report could have very serious implications for the provider.”
The last issue for consideration was the question of an injunction. Of course, this question was not governed simply by American Cynamid  AC 396 – something more is required to prevent the publication of a statutory body’s report. The bar is high. The requirement has been expressed as requiring “most compelling reasons“, “exceptional circumstances“, “pressing grounds“, or “extreme circumstances“. Examples where it may be appropriate include if there was evidence the public body was engaged in a vendetta against the provider, or the report was manifestly false and littered with errors. Mostyn J considered,
“I cannot say that the facts of this case, even taken at their highest, come close to demonstrating the exceptionally strong grounds needed to prevent a public body from publishing a report that Parliament has required it to publish.”
We await the full hearing on this issue of using such a narrow sample to found general criticisms. Whilst there may well be resourcing issues to justify dip sampling, it is difficult to see how this could justify such a small fraction being used to cause such potential damage to a business. If 0.01% is not too small a sample, what is?