In Miller & Another v Health Service Commissioner for England [2018] EWCA Civ 144 the Court of Appeal quashed both the ombudsman’s decision to investigate and the conclusions of that investigation. The court found that multiple forms of unlawfulness were compounded by a variety of poor adjudicative practices.

The Health Service Commissioner for England (the “ombudsman”) is a statutory public authority. It is empowered to investigate under the Health Service Commissioners Act 1993 (the “1993 Act”), including where a complaint is made that a family health service provider’s actions have given rise to injustice or hardship. In the Miller case, the ombudsman upheld a complaint about the medical treatment provided by GPs to a patient, Mr Pollard, and found that, with appropriate care, his subsequent death would have been avoided. The GPs who had treated Mr Pollard brought a judicial review challenge. After losing in the High Court they were triumphant on appeal.

The value in this judgment lies both in the Court of Appeal’s firm reiteration that proving unlawful conduct on the part of the ombudsman is very difficult, and its equally uncompromising identification of a spectrum of unlawful conduct on the facts before it. Serious failures were identified in relation to four main areas:

  1. the application of the test for commencing an investigation;
  2. the way the investigation was conducted;
  3. the standard of review; and
  4. the way conclusions were reached.

1. Test for commencing an investigation

The court determined that the 1993 Act requires two questions to be answered before the ombudsman can commence an investigation: (1) is there a legal remedy, and, if so (2) is the ombudsman satisfied that in the particular circumstances is it not reasonable to expect the complainant to resort to it? The court will not interfere with the ombudsman’s approach to the question of reasonableness unless that approach discloses an error of law or the conclusion is irrational. In Miller, the court found that the ombudsman had failed to make any decision in accordance with its statutory duty. There was only a record of the assessor’s opinion that there was no reasonable alternative remedy, and no information had been obtained or analysis done to support that position. The lack of such a decision was “fatal to the decision to investigate which was accordingly unlawful” and therefore quashed.

2. Investigation process

Before setting out its analysis, the court emphasised that the legislative scheme in the 1993 Act makes the ombudsman’s process an informal and inquisitorial, non-judicial, complaints mechanism. The procedure is “that which the ombudsman considers appropriate in the circumstances of the case”.

That said, basic requirements are laid down for how that mechanism should function, including the section 11(1A) obligations on the ombudsman to give the person complained of the opportunity to comment on the allegations contained in the complaint before deciding to conduct an investigation. Such statutory requirements are complemented by the context dependent requirements of common law procedural fairness and the obligation to act rationally. Having regard to all this, the court emphasised the following essential safeguards:

  • The requirements of fairness are to be judged in the context of “the ombudsman’s statutory purpose and the fact that serious criticism of the [GPs’] professional practice might ruin their reputation, professional standing and ability to earn a livelihood…”
  • Administrative and adjudicative decision making bodies should not consider relevant material (supportive or adverse to their case) without giving the affected person the right to comment upon it. This is not a procedural requirement to disclose the entirety of the evidence the ombudsman obtains but material that is relied upon must be disclosed, as must material contrary to the ombudsman’s provisional or final conclusions.
  • Further, the plain language of the 1993 Act requires that an opportunity to comment be given before a decision to investigate is made. This is a real protection, not a technicality.
  • It is necessary for the allegations to be set out with the particularity with which the ombudsman intends to investigate them, which must be sufficient for them to be answered in a meaningful and focussed way. The ombudsman cannot investigate what it wants, or expand beyond the content of the complaint, so good practice would be to disclose the complaint that has been made and identify that which is to be investigated.

The ombudsman’s failure to give an opportunity to comment until after the investigation was well underway (and a draft report with recommendations had been prepared) was a breach of section 11 of the 1993 Act which was enough to render the purported decision to investigate unlawful. However, for completeness the court highlighted several departures from good adjudicative practice. These included, for example, failures to disclose (a) the complaint letter or the record of a conversation enlarging on it; (b) the assessment of the complaint leading to the decision to investigate; or (c) the clinical reports on which the ombudsman relied.

3. Standard of review

The court determined that two questions must be answered in relation to the standard of review: (1) does the ombudsman have to adopt a particular standard; and (2) by what standard did the ombudsman measure the GPs’ actions in this case? Reviewing earlier authority, the court confirmed that the ombudsman has a broad discretion to: “decide and explain what standard he or she is going to apply…that standard will not be interfered with by the court unless it reflects an unreasonable approach”. In other words, the court will only intervene in extreme situations. For example, it might do so if the standard adopted is “incapable of being readily discerned or tends to produce inconsistent decisions”. The ombudsman is also entitled to change its policy on the standard of review.

The court’s reluctance to interfere did not, however, prevent it from doing so in this case. It identified that on this occasion the ombudsman had decided to choose a statement of good practice and measure the doctors against it. This standard was incoherent, with “no yardstick of reasonable or responsible practice” and the “risk of being a lottery dependent on the professional opinion of the advisor that is chosen.” It met the high legal threshold of being ‘unreasonable’ or ‘irrational’ and was therefore unlawful.

4. Pre-determined conclusions

The final main area for criticism was the adversarial attitude of the ombudsman to the investigation process. The test the court applied was whether a fair minded and informed observer, having considered the facts, would conclude that there was a real possibility that the investigation was biased by pre-determination. It considered that the ombudsman’s officers had used “firm concluded and adverse” language which gave no hint of doubt that the GPs were culpable. Adverse opinions were inappropriately contained in the file made available to the investigators. The draft report was shared with the GPs to give them an opportunity to “dispute” or “overturn” provisional findings and recommendations rather than inform them. Furthermore, the final report made no reference to the expert opinions submitted to the ombudsman on behalf of the GPs and no explanation was given for their rejection.

The court found that, overall, the investigation file gave “every appearance of pre-determination and almost none of a fair handed approach.” Due process and fairness may be respected in a draft report procedure, but what should happen, the court made clear, is that “the style of provisional evaluation should clearly admit of the possibility that there may be other explanations and opinions that the ombudsman has yet to consider.”

Conclusion

Historically, judicial reviews of the decisions of statutory ombudsmen have been distinguished by the court’s reluctance to interfere with what is self-evidently a widely drawn inquisitorial jurisdiction whose very purpose is to avoid litigation. The approach taken by the Court of Appeal in this case suggests that the intensity of review of these types of decisions may be shifting although it is too early to tell whether this case will fall into a wider pattern. In any event the decision does provide a helpful indication of the minimum standards that any inquisitorial process should adopt and therefore has wider application beyond the Parliamentary and Health Service Ombudsman.