IQ Legal TrainingBerkeley Lifford Hall Accountancy Services

Home > Judgments > 2006 archive

Meadow v General Medical Council [2006] EWHC 146 (Admin)

Ruling concerning immunity from further proceeedings arising from evidence given in court by an expert witness.

Neutral Citation Number: [2006] EWHC 146 (Admin)

Case No: CO/5763/2005




Royal Courts of Justice

Strand, London, WC2A 2LL

Date: 17 February 2006


Mr Justice Collins


Meadow Appellant

- and -

General Medical Council Respondent

Nicola Davies Q.C. & Ian Winter (instructed by Hempsons) for the Appellant

Roger Henderson Q.C. & Adam Heppinstall (instructed by The GMC) for the Respondent

Hearing dates: 23 – 25; 27 January 2006


Mr Justice COLLINS :

1. Professor Sir Roy Meadow is an eminent paediatrician. He is now some 73 years old and has retired from the clinical practice of medicine. He qualified in 1960 and in 1970 was appointed to the post of senior lecturer and consultant in paediatrics and child health at Leeds University. Following his observations in two cases in which the only explanation for the child's illness was that it had been fabricated by a parent, in 1977 he wrote and had published in the Lancet an article entitled "Munchausen's Syndrome by Proxy. The Hinterland of Child Abuse".

2. The article led to awareness in the medical profession of the possibility that false illnesses in children were being created and so child abuse was being practised. He continued to research into this area and in the 1980s began to be used to make reports and give evidence in family proceedings. In 1991 he was asked for the first time to produce a report for criminal proceedings and was involved in the notorious prosecution of Beverley Allitt. He continued to be used in court, mainly in family proceedings. He gave reports in all in some 10 criminal cases. He maintained a general paediatric practice and, as the many testimonials which were provided to the GMC and to this court show, he was regarded as a superb practitioner and teacher. Many families have cause to be grateful to him for what he did for their children. In 1980 he became Foundation Professor and Head of the Department of Paediatrics and Child Health at St James' Hospital in Leeds. On his retirement from clinical practice in 1998, he was appointed Emeritus Professor in Child Health at Leeds University.

3. Professor Meadow prepared at about the time of his retirement a paper for publication describing his clinical experience relating to children who had died because of what was regarded as child abuse. He sought in that paper to identify clinical features common to those cases. This paper was in due course published in the Archives of Disease in Childhood under the title 'Unnatural Sudden Infant Death'. As is usual in such publications, the paper was sent out in draft to be reviewed by his peers. Its relevance will become apparent when I refer to the evidence which was before the Fitness to Practise Committee of the GMC against whose decision Professor Meadow appeals in these proceedings.

4. In 1998 Professor Meadow, whom I shall henceforth refer to as the appellant, was approached by the Cheshire Constabulary and asked to provide a medical opinion on the deaths of Christopher and Harry Clark, sons of Sally Clark, both of whom had died when a few weeks old. The appellant reviewed all the material provided to him, in particular the medical records and the findings of the pathologist, and in June 1999 provided a statement of 10 pages in which he concluded that the deaths were not natural. He was called to give evidence both at the committal proceedings and at the trial in the Chester Crown Court. Sally Clark was convicted of murder of both children. She appealed to the Court of Appeal which upheld the convictions on 2 October 2000. In 2002 it was discovered that the results of important and relevant microbiological tests had not been disclosed by the pathologist. This led to a referral to the Court of Appeal which allowed Sally Clark's appeal on 29 January 2003. No retrial was ordered.

5. Complaint was made to the GMC by Sally Clark's father against the appellant. This alleged, broadly speaking, that the evidence he had given to the criminal courts had been badly flawed, particularly in the misuse of statistics, and so he deserved to be found guilty of serious professional misconduct and dealt with accordingly. The complainant did not suggest that he desired to have him erased from the register but that he should be prevented from acting as an expert in child protection cases. In due course, following a hearing lasting some 16 days between 21 June and 15 July 2005, the Fitness to Practise Panel (FPP) found serious professional misconduct proved and ordered that his name be erased from the register. He appeals both against the finding of serious professional misconduct and the sanction of erasure.

6. The decision of the FPP in this case has concerned medical practitioners who are asked to prepare reports for or to give evidence in court. Those concerns are the more acute since the FPP specifically found that the appellant had not intended to mislead the court and that there was no evidence of any calculated or wilful failure to use his best endeavours to provide evidence. He had acted in good faith. There can be no doubt that the decision has had a damaging effect in that it has increased the reluctance of medical practitioners to involve themselves in court proceedings, particularly in cases before the Family Court. In evidence before the FPP, Professor Sir Alan Craft, the President of the Royal College of Paediatrics and Child Heath, identified the concerns. They have been reiterated and reinforced in correspondence sent to the appellant's solicitors and to the GMC and I am aware of the real difficulties experienced in the Family Division because of the reluctance of doctors to produce reports and give evidence. Professor Craft said this:-

"[The campaign against paediatricians in the field of child abuse] has had an absolutely enormous effect on paediatricians. Paediatricians are frightened of getting involved in child protection work … I do not think you can actually underestimate what being reported to the GMC actually does to you – and paediatricians, I think, are pretty sensitive people, that is probably why they are paediatricians – and they do take it incredibly personally when the letter drops through the door saying that they have been reported to the GMC. It has a huge effect on them and on their families and on their children, particularly if there is a press campaign associated with it which there often is. Children have been excluded from school, people have had their car tyres slashed – all sorts of things that are really quite horrible have happened to paediatricians, so it is not surprising that they are fearful of being involved in child protection."

7. This appeal of course concerns medical practitioners. But the possibility that they may be disciplined even to the extent of losing their livelihood will apply to other professionals who give expert evidence to courts. It is particularly worrying that disciplinary action may result even if reports have been prepared and evidence given in good faith and with no intent to mislead. Accordingly, I received a request from the Expert Witness Institute to allow it to intervene in order to "indicate some issues of principle relating to the duties of expert witnesses in both the criminal and civil courts". I indicated that I was prepared to accept written submissions, provided that both parties to the appeal had no objection. Unfortunately, there was a failure by the EWI to notify the parties and the court assumed from the terms of the letter of request that the parties had been notified. In the result, I received and read the submissions, which I have taken into account only insofar as they make submissions on the general duties of expert witnesses and the jurisdiction of regulatory bodies in disciplining them in respect of evidence given by them. Mr Henderson Q.C., on behalf of the GMC was unhappy at what had happened. He suggested that, although such interventions are more common in the higher courts, they should not normally be allowed at the first instance level. However, he recognised that I had read them and he was able, so far as necessary, to deal with them and so he did not maintain his objections. I do not need, therefore, to consider whether they ought to have been admitted. I merely observe that the court should be careful not to allow costs to be increased by unnecessary interventions and that they should be limited to such interventions as genuinely assist the court, for example by identifying matters which go beyond the circumstances of the case in question and raise arguments which might otherwise not be properly considered.

8. Before going to the circumstances in more detail, I should deal at the outset with a point that I raised but which was not taken either before the FPP or in the grounds of appeal. However, since it goes to the jurisdiction of the FPP to deal with a complaint such as that made against the appellant, it seemed to me that it was a point which ought to be considered, particularly as it might be determinative of this appeal.

9. The point is based on the immunity from suit of a witness in respect of evidence he gives in a court of law. That immunity applies as much to an expert as to any other witness: see X (Minors) v Bedfordshire CC [1995] 2 A.C.633 approving Evans v London Hospital Medical College [1981] 1 W.L.R. 184. The immunity extends to any civil proceedings brought against a defendant which are based on the evidence which he gives to a court. It extends to any statement which the witness makes for the purpose of giving evidence.

10. The immunity has not been extended to prevent the bringing of disciplinary proceedings. That seems to be because the argument has not hitherto been deployed that the rationale that lies behind the grant of immunity from suit should apply equally to such disciplinary proceedings. It is clear that proceedings before the GMC have been brought in the past which have certainly included heads of charge based on evidence given, but most seem to have been based on other conduct as well. And it is clear that to produce a report or to give information which is sufficiently flawed as properly to be regarded as serious professional misconduct will not attract immunity even though it is used as the basis for evidence given subsequently. The report must have been prepared with a view to its being used or in the knowledge that it will probably be used in evidence in court. This distinction is important and is recognised in the authorities, in particular in the recent decision of the House of Lords in Darker v Chief Constable of the West Midlands Police [2001] 1 A.C.435, a case to which I will have to return. That case was concerned with a claim that police officers had fabricated evidence against the claimants. At page 449B-C, Lord Hope of Craighead referred to:-

"… the distinction … between the act itself and the evidence that may be given about the act or its consequences. The distinction rests upon the fact that acts which are calculated to create or produce false evidence or to destroy evidence have an independent existence from, and are extraneous to, the evidence that may be given as to the consequences of those acts. It is unlikely that those who have fabricated or destroyed evidence would wish to enter the witness box for the purposes of admitting to those acts of fabrication or destruction. Their acts were done with a view to the giving of evidence not about the acts themselves but about their consequences. The position is different where the allegation relates to the content of the evidence or the content of statements made with a view to giving evidence, and not to the doing of an act such as the creation or the fabrication of evidence."

11. The distinction may, as Lord Hutton observed at p.469G, appear in practice to be a fine one. He gives the example of the officer who falsely states in evidence that the defendant made a confession, who would be immune, and the officer who fabricates a note containing an admission which the defendant never made, who would not be immune in respect of proceedings based on the fabricated notes. While Darker's case was concerned with misfeasance in public office and conspiring to injure and so bad faith, there seems no reason why the principle should not apply in an appropriate case to negligent acts as well as to dishonest ones.

12. It is accepted by Mr Henderson Q.C. that the appellant would be immune from civil suit in respect of the matters alleged against him in the disciplinary proceedings. The lengthy and detailed heads of charge were based on and substantially limited to the statements he had made for the purpose of giving expert evidence and the evidence he had given in the committal proceedings and the trial of Sally Clark for the murder of two of her children. This case therefore raises directly the question whether immunity from suit should be extended to provide immunity from disciplinary proceedings and, if so, whether there are any qualifications which are appropriate if that extension is in principle justified.

13. Immunity from suit extends to the honest as well as the dishonest witness. It is based on public policy which requires that witnesses should not be deterred from giving evidence by the fear of litigation at the suit of those who may feel that the evidence has damaged them unjustifiably. It is not the result of the litigation that matters but the burden upon the witness of having to defend himself whether or not he may be confident of the outcome. Stanton v Callaghan [2000] 1 Q.B. 75 concerned the immunity of a structural engineer who had produced a report, said to have been prepared negligently, in respect of a claim which had had to be withdrawn when its deficiencies were identified following a joint meeting of experts. The Court of Appeal upheld the striking out of the claim. At p.107A, Otton LJ said this:-

"I pause here to note that immunity is not granted primarily for the benefit of the individuals who seek it. They themselves are beneficiaries of the overarching public interest, which can be expressed as the need to ensure that the administration of justice is not impeded. This is the consideration which should be paramount. And it is not only the conduct of the immediate hearing which we should consider to be the "administration of justice". This is not a narrowly drawn phrase; it is best served by a purposive construction. In this I agree with Lord Wilberforce who said in Roy v Prior [1971]AC 470, 480: "immunities conferred by the law in respect of legal proceedings need always to be checked against a broad view of the public interest.

Each party who comes, or is about to come, before a court is participating in an event which supervenes individual concerns and interests. When we are concerned with the proper and smooth administration of justice through our legal system we should not seek to place burdens on those who participate in it at any stage. Thus I do not think it necessary to make distinctions between the various reasons which have been given to justify the granting of immunity and approach this situation in an algorithmic fashion and say that some reasons should apply to some cases but not to others: the case is best approached by asking the simple question would it serve the interests of the administration of justice to grant immunity? To answer this question we need to examine the role and place of an expert in the legal system."

14. Otton LJ went on to point out that an expert witness owes a duty to the court which overrides that to his client. Accordingly, he should not be vulnerable to claims from disgruntled clients. That was the main consideration in Stanton v Callaghan, but Otton LJ's observations are, as he himself recognised, applicable generally. Nourse LJ, after noting that the extent of an expert witness's immunity from suit was still in course of development and would and should be developed on a case by case basis, said this (at p.109C):-

"… I see no justification for distinguishing between an expert and a lay witness, either on the ground that the expert is usually remunerated for his services or on the ground that he may be less likely than a lay witness to be deterred from giving evidence. Nor would I make any distinction between civil and criminal proceedings. An immunity founded on requirement of public policy that witnesses should not be inhibited from giving frank and fearless evidence cannot afford to make distinctions such as these. If they were allowed, it would never be certain that the public policy would not sometimes be put at risk."

15. Immunity, as I have said, extends to the dishonest as well as the honest witness. The dishonest may be guilty of the criminal offence of perjury and can be prosecuted if sufficient evidence exists. But, if such evidence is not available (for example, because there is no independent corroboration), the immunity exists because of the requirement that a witness should be able to give evidence free from fear of any reprisal. The public policy states that the need to protect the honest witness may result in immunity for the dishonest, but the balance between the right of an individual to make a claim and the need in the interest of the administration of justice to ensure that witnesses give evidence in the knowledge that they cannot be subjected to action which may seek to penalise them is struck by giving priority to the latter. In Darker v Chief Constable of the West Midlands at p.464D, Lord Hutton put it thus:-

"The reason for the rule is grounded in public policy: it is to protect a witness who has given evidence in good faith in court from being harassed and vexed by an action for defamation brought against him in respect of the words which he has spoken in the witness box. If this protection were not given persons required to give evidence in other cases might be deterred from doing so by the fear of an action for defamation. And in order to shield honest witnesses from the vexation of having to defend actions against them and to rebut an allegation that they were activated by malice the courts have decided that it is necessary to grant absolute immunity to witnesses in respect of their words in court though this means that the shield covers the malicious and dishonest witness as well as the honest one."

16. Although Lord Hutton was referring specifically to actions for defamation, it is clear that the public policy which grants immunity extends for the same reason to any action brought, whether or not it alleges malice, bad faith or dishonesty. The rule has a long history. In R v Skinner (1772) Lofft 54, Lord Mansfield observed:-

"Neither party, witness, counsel jury or judge can be put to answer, orally or criminally, for words spoken in office."

In respect of witnesses, 'in office' can only refer to giving evidence. The only qualification to this is a prosecution for perjury or, possibly, an attempt to pervert the course of justice. In Watson v McEwen [1905] AC480 at p.482, Lord Halbury, L.C. observed:-

"The broad proposition I entertain no doubt about, and it seems to me to be the only question that properly arises here; as to the immunity of a witness for evidence given in a court of justice, it is too late to argue that as if it were doubtful. By complete authority, including the authority of this House [see Dawkins v Lord Rokeby (1875) LR 7HL 744] it has been decided that the privilege of a witness, the immunity from responsibility in an action where evidence has been given by him in a court of justice, is too well established now to be shaken. Practically I may say that in my view it is absolutely unarguable – it is settled law and cannot be doubted. The remedy against a witness who has given evidence which is false and injurious to another is to indict him for perjury; but for very obvious reasons, the conduct of legal procedure by courts of justice, with the necessity of compelling witnesses to attend, involves as one of the necessities of the administration of justice the immunity of witnesses from actions brought against them in respect of evidence they have given. So far the matter, I think, is too plain for argument."

17. There is clear evidence before me, and common sense points in the same direction, that the possibility of disciplinary proceedings based on a complaint by someone affected by the evidence given has a serious deterrent effect. It is in those circumstances difficult to follow why the public policy based on the need to protect the administration of justice should not prevent disciplinary proceedings. They can result in penalties which are more serious than an award of damages and have the effect on the practitioner which was described by Professor Craft.

18. It is, however, to be noted that in Darker's case the House of Lords was at pains to state that the protection should not be given any wider application than was absolutely necessary in the interests of the administration of justice – see per Lord Cooke of Thorndon at p.453, citing observations of Sir Thaddeus McCarthy, P in Rees v Sinclair [1974] 1 NZLR 180 at 187. Lord Clyde dealt with the matter more extensively at pp 456H to 457H, saying:-

"It is temptingly easy to talk of the application of immunities from civil liability in general terms. But since the immunity may cut across the rights of others to a legal remedy and so runs counter to the policy that no wrong should be without a remedy, it should be only allowed with reluctance, and should not readily be extended. It should only be allowed where it is necessary to do so. As McCarthy P observed in Rees v Sinclair [1974] 1 NZLR 180, 187: "The protection should not be given any wider application than is absolutely necessary in the interests of the administration of justice …". Furthermore the idea of a universal immunity attaching to a person in the performance of some particular function requires to be entertained with some caution. As Lord Wilberforce observed in Roy v Prior [1971] AC470, 480: "Immunities conferred by the law in respect of legal proceedings need always to be checked against a broad view of the public interest". Once a situation has been identified as deserving of immunity it may readily be accepted that the immunity is in its quality absolute. But the process of identification may require to be undertaken with a particular eye to an evaluation of the public interests involved. The quality of an immunity may be absolute, but its application may not be invariable.

On the other hand there has to be some degree of certainty about the existence of an immunity for it to be effective. The matter cannot be entirely left as one to be determined on each and every occasion. For the immunity of a witness to be effective it is necessary that the person concerned should know in advance with some certainty that what he or she says will be protected. So even although the matter may depend in any case upon a balancing of interests it ought to be possible to predict with some confidence whether or not an immunity will apply. The law has sought to achieve this by making it clear that the substance of the evidence presented to the court in judicial proceedings will be immune from attack. But a more difficult question arises with regard to the preparation of material and the investigation of a case before the matter comes before the court.

Two reasons can be identified for the justification for granting an immunity to witnesses from civil process. They were expressed by Lord Wilberforce in Roy v Prior in these terms, at p 480:

"the reasons why immunity is traditionally (and for this purpose I accept the tradition) conferred upon witnesses in respect of evidence given in court, are in order that they may give their evidence fearlessly and to avoid a multiplicity of actions in which the value or the truth of their evidence would be tried over again"

So far as the first of those reasons is concerned it may be considered necessary that witnesses should be granted an immunity so as to secure that they may enjoy a freedom to express themselves without fear of any consequences to themselves. In the interests of the judicial process a witness should not be exposed to the risk of having his or her evidence challenged in another process. Those engaged in the judicial process should be under no restraint from saying what has to be said and doing what has to be done for the proper conduct of that process. As Salmon J observed in Marrinan v Vibart [1963] 1 Q.B. 234, 237:-

"This immunity exists for the benefit of the public, since the administration of justice would be greatly impeded if witnesses were to be in fear that any disgruntled and possibly impecunious persons against whom they gave evidence might subsequently involve them in costly litigation""

19. I would emphasise the sentence (at p 489G):-

"In the interests of the judicial process a witness should not be exposed to the risk of having his or her evidence challenged in another process.

While all the authorities have concerned immunity from suit, the rationale behind the rule which is recognised by that observation leads me to the view that not only is there no reason in principle why it should not apply to disciplinary proceedings such as those with which this appeal is concerned but every reason why it should so apply. There can be no doubt that the administration of justice has been seriously damaged by the decision of the FPP in this case and the damage will continue unless it is made clear that such proceedings need not be feared by the expert witness.

20. I recognise that there is a public interest in play here which goes beyond the rights of others to a legal remedy. That is the need to ensure that public confidence in the profession is maintained and that no practitioners who have shown themselves to have fallen below the standards required should be able to continue to practise uncontrolled. Any expert witness will know that he has a duty to the court and must bear in mind his obligations in that regard. They are helpfully set out in a passage from the judgment of Cresswell J in The Ikarian Reefer [1993] 2 Lloyd's Rep 68, at 81-82:-

"The duties and responsibilities of expert witnesses in civil cases include the following:

1. Expert evidence presented to the Court should be, and should be seen to be, the independent product of the expert uninfluenced as to form or content by the exigencies of litigation [Whitehouse v Jordan [1981] 1 W.L.R. 246 at p 256 per Lord Wilberforce).

2. An expert witness should provide independent assistance to the Court by way of objective unbiased opinion in relation to matters within his expertise (see Polivite Ltd v Commercial Union Assurance Co Plc [1987] 1 Lloyds Rep 279 at p.386 per Mr Justice Garland and Re J [1990] F.C.R. 193 per Mr Justice Cazalet). An expert witness in the High Court should never assume the role of an advocate.

3. An expert witness should state the facts or assumption upon which his opinion is based. He should not omit to consider material facts which could detract from his concluded opinion (Re J sup)

4. An expert witness should make it clear when a particular question or issue falls outside his expertise.

5. If an expert's opinion is not properly researched because he considers that insufficient data is available, then this must be stated with an indication that the opinion is no more than a provisional one (Re J sup). In cases where an expert witness who has prepared a report could not assert that the report contained the truth, the whole truth and nothing but the truth without some qualification, that qualification should be stated in the report (Derby & Co Ltd and Others v Weldon and Others, The Times, Nov9, 1990 per Lord Justice Staughton).

6. If, after exchange of reports, an expert witness changes his view on a material matter having read the other side's expert's report or for any other reason, such change of view should be communicated (through legal representatives) to the other side without delay and when appropriate to the Court.

7. Where expert evidence refers to photographs, plans, calculations, analyses, measurements, survey reports or other similar documents, these must be provided to the opposite party at the same time as the exchange of reports (see 15.5 of the Guide to Commercial Court Practice)."

In addition, he will know that he must give evidence honestly and in good faith and must not deliberately mislead the court. He will not expect to receive protection if he is dishonest or malicious or deliberately misleading.

21. Since I am applying a principle based on public policy to grant an immunity which has not hitherto been explicitly recognised, I can, I think, consider whether public policy requires that an absolute immunity should be granted. The approach of their Lordships in Darker's case indicates that immunity from suit, in respect of which the law has granted absolute immunity, should be confined as narrowly as reasonably possible. That approach and the need to balance the countervailing public interests persuades me that a blanket immunity is not necessary. Barristers and solicitors owe duties to the court and may be subjected to disciplinary action in respect of their conduct in litigation. That does not inhibit them because they know that they must maintain the necessary standards before the court and will be liable to action if they do not. But witnesses are in a somewhat different position, particularly when they become involved in litigation fortuitously, perhaps because as a doctor they treated a particular child and abuse is suspected.

22. In my judgment, the immunity has to cover proceedings based on a complaint (whether or not it alleges bad faith or dishonesty) made by a party or any other person who may have been upset by the evidence given. Public policy, as reflected in the observations of the various judges which I have cited, requires at least that. But I see no reason why the judge before whom the expert gives evidence (or the Court of Appeal when that may be appropriate) should not refer his conduct to the relevant disciplinary body if satisfied that his conduct has fallen so far below what is expected of him as to merit some disciplinary action. I note that such referrals have been made, although I do not think the immunity point has been argued: see Hussein v William Hill Group [2004] EWHC 208 (Q.B.) per Hallett J and Pearce v Ove Arup (Case No HC 1996 06040 (2 November 2001) per Jacob J. In Paragraph 62 of that case, Jacob J said this:-

"I see no reason why a judge who has formed an opinion that an expert had seriously broken his Part 35 duty should not, in an appropriate case, refer the matter to the expert's professional body if he or she has one. Whether there is a breach of the expert's professional rules and if so what sanction is appropriate would be a matter for the body concerned."

The witness should, as Jacob J stated, be given an opportunity to make representations before any referral took place.

23. Such a referral would not be justified unless the witness's shortcomings were sufficiently serious for the judge to believe that he might need to be removed from practice or at least to be subjected to conditions regulating his practice such as a prohibition on acting as an expert witness. Normally, evidence given honestly and in good faith would not merit a referral. Mr Henderson was concerned that to draw the line at dishonesty or recklessness could mean that a practitioner who gave seriously defective evidence which was honestly given but resulted from for example ill health was able to continue in practice to the danger of the public. I recognise that possibility: the judge is likely also to recognise it if it arises in any given case.

24. No system can be perfect. It is, as Mr Henderson submitted, at least in theory possible that a practitioner whose shortcomings are not recognised by the court may escape deserved sanctions. This would particularly be so if the practitioner did not give evidence because court proceedings were, as in Stanton v Callaghan, never pursued. However, I think that this problem is more theoretical than real. It is unlikely that a single case involving a poor report or evidence would on its own show that the practitioner was unfit to practise and so a danger to the public. His report would become known and he would not be invited to give evidence in the future. Further, if he was so poor, he would be likely to show his defects in a subsequent case. Mr Henderson raised the issue of accreditation which, for example, would affect a pathologist. Could he be removed from the list of those entitled to act for the Home Office on the basis of poor evidence in a particular case? The answer must be that he could. Just as a private client is entitled to cease to instruct an expert if dissatisfied with his performance so can the Home Office. If that is done, he has a right of appeal.

25. The precise boundaries of the immunity will have to be established on a case by case basis. For example, where serious defects in the expert's evidence only came to light after a court hearing, it may be possible to go back to the judge to ask him to consider a referral. If there is an appeal, the Court of Appeal can take the necessary action. But what is of fundamental importance is that a witness can be assured that if he gives his evidence honestly and in good faith, he will not be involved in any proceedings brought against him seeking to penalise him. The risk of a judge deciding that there should be a referral in such circumstances is so remote as to be virtually non-existent.

26. I have no doubt that the complaint against the appellant should not have been pursued. Since it was based upon his evidence in court, he had immunity. Mr Henderson submitted that this was to run counter to the requirement imposed by Parliament that the GMC investigate any complaint made to it. But I see no problem; a complaint can only be pursued if the law permits it. In the case of a complaint made against an expert which arises from evidence he has given in court, the law prohibits, save in the circumstances I have indicated, the matter being pursued. If the complaint does relate to the evidence or the preparation of evidence, the GMC must bring it to an end. A similar procedure must be applied by other disciplinary bodies. The result will be that experts can give evidence free from the fear of subsequent disciplinary action unless they act so contrary to their obligations to their profession and to the court that the court decides to make a complaint. Only in such circumstances will disciplinary action be permissible.

27. It follows that the appeal against the finding of serious professional misconduct must be allowed since the FPP should not have considered the complaint.

28. Since this case may go further, I must deal with the matter on the assumption that the FPP was entitled to consider the complaint. The allegations did not touch on the appellant's skills as a doctor nor did they impugn his conclusions which were based on the pathological findings. He was found guilty of serious professional misconduct because he had relied on statistical material in a way which was flawed and had, it was alleged, not disclosed that he was not a statistician and had no expertise in the understanding or application of statistics. His evidence had been badly wrong in misusing statistics to conclude that the chances of a second natural death occurring in a family were miniscule, amounting to 1 in 73,000,000. This evidence may have influenced the jury to convict of murder.

29. That conduct which is not directly connected with a doctor's practice as such can amount to serious professional misconduct cannot be doubted. Any conduct which brings the profession into disrepute is capable of justifying disciplinary action. In Roylance v GMC [2000] 1AC 311, Lord Clyde, giving the judgment of the Privy Council, observed that misconduct involved acts or omissions which fell short of what was proper in the circumstances and that the standard of propriety might often be found by reference to the rules and standards normally required to be followed. It must be serious and must be linked to the profession of medicine. He pointed out that conduct removed from the practice of medicine might qualify if it was of a sufficiently immoral or outrageous or disgraceful character. This was because the public reputation of and public confidence in the profession could be adversely affected. Dr Roylance was the chief executive of a hospital in which there had been excessive mortality rates of children who underwent cardiac surgery and had failed to take steps to deal with the problem. His conduct was not in the class of moral turpitude or of so outrageous a nature as to bring the profession into disrepute. Nevertheless, he was properly convicted since he could not divorce his duties as a medical practitioner from those as an administrator.

30. There is equally no doubt that bad faith is not necessary for a finding of serious professional misconduct. Thus, for example, negligence may suffice, but it must be negligence to a high degree. The test has been described as conduct which would be regarded as deplorable by fellow practitioners or properly informed members of the public. The same test is likely to remain despite the amendments to the legislation, but I do not need to go into that.

31. An appeal under s.40 of the Medical Act 1983 is not limited to a review – see CPR 52.11 and the Practice Direction at Paragraph 22. However, this court will not interfere unless persuaded that a decision, whether in respect of a finding of misconduct or the sanction imposed, is clearly wrong. I need not cite the various authorities to which I have been referred. Suffice it to say that that is the test which is applied. I prefer to place no further gloss upon it.

32. The appellant's statement in the criminal proceedings contained the following paragraphs under the heading: Two Infant Deaths in One Family:-

"Even when an infant dies suddenly and unexpectedly in early life and no cause is found at autopsy, and the reason for death is thought to be an unidentified natural cause (Sudden Infant Death Syndrome) [SIDS], it is extremely rare for that to happen again within a family. For example, such a happening may occur 1:1000 infants, therefore the chance of it happening twice within a family is 1:1,000,000.

Neither of these two deaths can be classified as SIDS. Each of the deaths was unusual and had the circumstances of a death caused by a parent."

It will therefore be apparent that the statistic was not of itself material to the appellant's opinion that the deaths were not natural. But it obviously tended to negative a defence that the deaths resulted from SIDS and may well have persuaded a layman that the risk of two natural deaths of unknown cause was very small indeed and so it supported the view that these were not natural deaths. The statistical error was obvious from the squaring of the 1:1000 to make 1:1,000,000. Such squaring is only permissible if the two events are truly independent. Since in families there are bound to be other matters to be taken into account, such as genetic or environmental factors, squaring is inappropriate. An example of a truly independent event (assuming, I suppose, no variation in the way in which it is tossed) is the tossing of a coin. Each time, the chances of landing the same side will be squared. Indeed, the evidence given before the FPP by a statistician indicated that dependence should always be assumed and that it would have to be disproved before squaring of the risk of a second event was a valid exercise.

33. That statement was made on 5 June 1998 and was read out in the Committal proceedings as the appellant's evidence in chief. This was on 25 May 1999. In the course of his evidence, he referred to the paper I have mentioned in Paragraph 3 of this judgment, which was published in January 1999. It was a study of 81 children who were initially thought to have died of natural causes but were subsequently adjudged by either criminal or family courts to have been killed by a parent.

34. He was cross examined by Mr Kelsey-Fry (now Q.C.), junior counsel representing Sally Clark. He put to the appellant various research papers which he suggested showed that the risk was higher. Those the appellant did not believe were reliable since they did not deal with true cases of SIDS because in many there were other possible causes, both natural or unnatural. But he was not challenged on his use of statistics. He was asked:-

"Am I right in thinking that, in the event of a family suffering a cot death, a SIDS, an unexplained death, research shows that the chances of a repeat occurrence once the first has happened, of course, the chances of a repeat occurrence are effectively the same? In other words, the fact that there is one, does not enhance the chance of another?"

The appellant agreed. It was then suggested that research showed that such a risk was less than 1:1,000,000, but the appellant did not accept that that was so. Mr Kelsey-Fry did not challenge the squaring exercise or suggest that it was flawed.

35. In August 1999, Professor Fleming, Professor of Infant Health and Development Physiology at the Institute of Child Health at Bristol University, asked the appellant to write the preface to a report on a three year study, which had been commissioned by the Department of Health investigating factors contributing to sudden unexpected deaths in infancy. It was known as the CESDI study. The appellant agreed to Professor Fleming's request and on 14 August 1999 was sent a pre-publication draft of the report. It was a most ambitious and extensive project and had examined some 472,823 live births over a 3 year period.

36. It contained an assessment of infants and families at risk of SIDS. I should, I think, set the relevant passage out in full:

"Overall in the population included in this study the SIDS rate was 0.768 per 1000 live births – i.e. approximately 1 baby in 1300 died as SIDS. From our data it is possible to identify within the population a number of factors which are associated with an increased risk of SIDS. The identification of families at higher risk of SIDS is of importance in allowing the appropriate deployment of scarce health care resources, and in attempting to achieve changes in life style or patterns of childcare that might reduce this risk. For families already at low risk, knowledge of the factors influencing risk may help to provide reassurance and encouragement in continuing appropriate patterns of care.

Table 3.6.1 shows three factors that are associated with an increased risk of SIDS in both univariate and all multivariate models, and the likely effect of the presence or absence of each factor on the incidence of SIDS, along with effect when combining these factors.

Table 3.6.1 SIDS rates for different factors based on the data from the CESDI SUDI study

SIDS Rate per 1000 livebiths* SIDS incidence in this group*

Overall rate in the study population 0.768 1 in 1303

Rate for groups with different factors

Anybody smokes in the household 1.357 1 on 737

nobody smokes in the household 0.199 1 in 5041

No waged income in household 2.057 1 in 486

At least one waged income in household 0.479 1 in 2088

Mother ,<27 years and parity>.1 1.762 1 in 567

Mother>26 years or parity=1 0.531 1 in 1882

None of these factors 0.117 1 in 8543

One of these factors 0.619 1 in 1616

Two of these factors 1.678 1 in 596

All three of these factors 4.674 1 in 214

*Based on the number of livebirths in each study region from 1993 to 1995 inclusive (OPCS)

Thus an infant living in a household in which nobody smoked had a risk of SIDS of around 1 in 5000, whilst if anyone in the household smoked this risk rose to around 1 in 700. Similarly for an infant in a household in which there was no waged income, the risk was around 1 in 500, compared with 1 in 2000 if there was a waged income.

The correlation between the factors was taken into account when more than one factor was used to calculate the rate, but, because all three factors are independently significant in the multivariate analyses, the presence of more than one will have an increased effect.

Thus it can be seen that for infants in families in which all three factors are present the risk of SIDS was 1 in 214, compared with a risk of 1 in 8543 for infants in families with none of the factors – i.e. a 40 fold difference in risk.

Since the factors will generally remain the same (with the possible exception of maternal age below 27 years) for a subsequent child, the risk of SIDS to a subsequent child in a family in which one infant has already died will range from 1 in 214 to 1 in 8543. This does not take account of possible familial incidence of factors other than those included in the above table.

For a family with none of these three factors, the risk of two infants dying as SIDS by chance alone will thus be 1 in (8543x8543) i.e. approximately 1 in 73 million. For a family with all three factors the risk will be 1 in (214x214) i.e. approximately 1 in 46,000. Thus, for families with several known risk factors for SIDS, a second SIDS death, whilst uncommon, is 1600 times more likely than for families with no such factors. Where additional adverse factors are present, the recurrence risk would correspondingly be greater still.

Whilst child abuse and non-accidental injury are associated with many of the same factors as an increased risk of SIDS, the increased risk in the above calculation is derived from a population in which careful attempts have been made to exclude those deaths for which abuse by a parent or carer was identified as a probable casual factor. When a second SIDS death occurs in the same family, in addition to careful search for inherited disorder there must always be a very thorough investigation of the circumstances – though it would be inappropriate to assume maltreatment was always the cause."

In due course, the table, but not, it seems, the text, was put before the jury at the Crown Court. It was his evidence based on his understanding of this that was largely the source of the complaint which led to the finding of serious professional misconduct.

37. On 4 October 1999, shortly before the trial was due to commence, the appellant was requested to and did attend a meeting resulting from the disclosure of the defence medical reports. The meeting was largely concerned with pathology, but there was a discussion about SIDS because it was believed that the defence might rely on SIDS. In fact, they did not. However, on 5 October 1999, the appellant produced a short handwritten supplementary statement which read:-

"Since writing my report, I have read the reports of other medical experts.

Apart from non-accidental injury, no likely specific medical cause of death has been proposed. Thus it is suggested that the deaths of both children should be considered as examples of SIDS.

The likelihood of SIDS rises with social circumstances. The most recent estimation of the incidence in England, is that for a family in which the parents do not smoke, in which at least one has a waged income, and in which the mother is over the age of 26 years, the risk is 1 in 8543 live births.

Thus the chance of 2 infant deaths within such a family being SIDS is 1 in 73,000,000."

As will be obvious, this was based on the extract from the CESDI study which I have already cited. It was an statement based on a misunderstanding of the significance of the squaring. The squaring was not intended to be a guide to the risk of recurrence. The figures given were estimates based on a mathematical modelling and were not observed rates. Since independence could not be assumed, the squaring was a statistically invalid assumption and was intended to do no more than show that it produced in truth an underestimate of the real risk. I am bound to say, having read Professor Fleming's evidence (he was a witness before the FPP), I am far from clear why the squaring exercise was included at all.

38. The appellant tried without success to contact Professor Fleming to ensure that he had correctly understood the significance of the table. But on 19 October 1999, at 8.48am Professor Fleming faxed to the defence solicitors and to Mr Kelsey-Fry a letter which followed contacts from the defence solicitors about the appellant's supplementary statement and the extract from the CESDI report. So far as material, it read as follows:-

"The purpose of including these calculations within the CESDI SUDI (Sudden Unexpected Death in Infancy) report, was to point out that, for families with infants at high risk (and these are particularly families living in socio-economically deprived circumstances and those in which one or more adults smoke heavily) the risk of a second death occurring purely by chance, without the need to adduce deliberate or other actions by a parent, or the need to suggest the presence of a familial or genetically determined condition, would be approximately 1 in 45,000. Thus, whilst rare, such an event would not be of such rarity as to require the assumption of harm by a parent or carer.

The question of second and subsequent deaths in families without risk factors is, however, very much more difficult to deal with and the statistics upon which these calculations are based, whilst coming from the largest study of sudden infant death ever conducted, must be seen as having a large confidence interval, that is to say that, whilst the risk is approximately 1 in 8,500 for a baby to die in such a family, the extreme rarity of such an event makes this statistic potentially somewhat unreliable and open to the effect of other (unmeasured) parameters which may influence the risk.

It is also important to point out that, in a family in which a single baby died suddenly and unexpectedly as a cot death, the risk to a subsequent baby could vary between 1 in 214 and 1 in 8500 depending upon the presence, or absence, of the various risk factors mentioned above. If this second event were indeed truly independent of the first, then the assessment of the difference in probability of a second death occurring, in relation to the presence, or absence, of one or more of the risk factors, would be determined by the risk to that second baby i.e. the risk would lie somewhere between 1 in 200 and 1 in 8,500.

A further complicating factor in assessing the risk of occurrence of sudden infant death within a family is the potential importance of factors not included in the simplified risk scoring system noted, but which may have a major impact upon the risk of a baby dying, e.g. birth weight, gestation, post-natal growth pattern, sex of the infant, sleeping position, heating, heavy wrapping, the presence of recent illness. Whilst overall, none of these factors have as big an effect on the risk of babies dying as the four factors listed in the above risk score, for infants with none of the risk factors included in the risk score, the presence of one, or more, of these secondary factors may have a substantial effect upon the risk of death.

In summary, therefore, the risk scoring system which we have developed is primarily aimed at trying to identify families for whom the risk of a subsequent baby dying is substantially increased compared with the general population. Because of the extreme rarity of sudden death in families with none of these risk factors, the use of this risk score for such families is potentially much less reliable."

39. The defence thus had the necessary ammunition to question the appellant's use of the statistics. He gave his evidence on 20 October, but could have been recalled if Professor Fleming's letter had arrived too late to be properly absorbed. But his use of the statistics was not challenged. Further, the appellant was aware that the defence had as one of their experts Professor Berry who had been a joint author of the CESDI report. The material questions and answers in cross-examination by experienced leading counsel for the defence, Mr Julian Bevan, Q.C., were as follows:-

"Q. On your own table when Christopher was born his chances in relation to a cot death were, taking your own figure, 1 in 8543? – A. Around there, yes. I say around because as this paper mentioned, this figure analyses the three biggest risk factors and there are other things that can modify it, but I think for practical purposes 1 in 8,500 is a starting point.

Q. He died. When Harry came into this world, yes? – A. Yes.

Q. When he was born the chances of Harry dying, the chances of him dying of a cot death were exactly the same, were they not, 1 in 8,543? – A. Yes, that is correct.

Q. It's a bit like a coin, isn't it? If you flip a coin, heads or tails, yes? – A.Yes.

Q. It's the same odds each time, isn't it, one to one? – A. Yes, and that's why you don't just look at … This is why you take what's happened to all the children into account, and that is why you end up saying the chance of two children dying naturally in these circumstances is very, very long odds indeed, one in 73 million. You know, I mean …

Q. That's a double death every hundred years – A. I know, but I mean, you know, I know Mr Kelsey-Fry is interested in betting odds and you know, it's the chance …

Q. I don't know how you knew that – A. At a previous hearing; but it's the chance of backing that long-odd outsider at the Grand National, you know; let's say it's an 8o to 1 chance, you back the winner last year, then the next year there's another horse at 80 to 1 and it is still 80 to 1 and you back it again and it wins. Now here we're in a situation that, you know, to get to these odds of 73 million you've got to back that 1 in 80 chance four years running, so yes, you might be very, very lucky because each time it's just been a 1 in 80 chance and you know, you've happened to have won it, but the chance of it happening four years running we all know is extraordinarily unlikely. So it's the same with these deaths. You have to say two unlikely events have happened and together it's very, very, very unlikely.

Q. Have you ever heard – I hope it's not too frivolous a remark to make but have you heard the expression "Lies, damned lies and statistics"? – A. I don't like statistics but I'm forced into accepting their usefulness."

The appellant accepted that the illustration he gave based on the odds of winning the Grand National was insensitive. But it was no more than an illustration which would bring home to the jury the risk represented by 1 in 73 million.

40. In dismissing Sally Clark's appeal, the first Court of Appeal accepted that the appellant had been entitled to speak to what was in the literature including statistics. Absent any objection – and none was made at the trial – 'the expert can rely on an up-to-date reputable study such as the CESDI' – see paragraph 112. They recorded that Professor Berry had been called by the defence and that he had said that the 1 in 8543 statistic was an observed figure thereby, it seems, falling into the same error as the appellant. They said this (in Paragraph 144):-

"In our judgment … Professor Meadow's opinion was based on his expert assessment of the medical and circumstantial evidence, not on the statistical material."

They rejected the contention that he had been responsible for the 'prosecutor's fallacy' (see R v Doheny & Adams [1997] 1 CR App R.369). The case was never put to the jury on the basis that the chances of the defendant being innocent were 73 million to one and the court said in terms that they were satisfied that the appellant had not contributed to the danger of misinterpretation of the evidence of the risk – see paragraph 155. They said:-

"If Mr Bevan Q.C., for the defence, had understood him to be saying that the odds against both of those deaths being a SIDS death were 73 million to 1 that is a point which would certainly have been brought out in cross-examination and not left where it was, with the remark 'Lies, damned lies and statistics'."

At paragraph 163, the court said this:-

"Professor Meadow did not misuse the figures in his evidence, though he did not help to explain their limited significance."

41. The second appeal was allowed largely because of the failure to disclose significant results of microbiological tests. The statistical evidence was a second ground. The Court expressed concern that no steps had been taken to seek to exclude the evidence. While juries might well be aware that cot deaths are rare and that 'two deaths in a family are much rarer still' they said (Paragraph 175):-

"Putting the evidence of 1 in 73 million before the jury with its related statistics that it was the equivalent of a single occurrence of two such deaths in the same family once in a century was tantamount to saying that without consideration of the rest of the evidence one could be just about sure that this was a case of murder."

The difference of approach of the two courts is obvious, but it does not mean that the first was acting unreasonably. The second court dealt with the evidence in Paragraph 178 as follows:-

"The argument before us would have addressed the question whether the 1 in 73 million figure was misleading in itself quite apart from the use made of it at trial. On the material before us, we think it very likely that it grossly overstates the chance of two sudden deaths within the same family from unexplained but natural causes. There is evidence to suggest that it may happen much more frequently than suggested by that figure although happily the risk remains a relatively unlikely one. The figure of 1 in 73 million was disputed by Professor Berry in his evidence who pointed to the obvious dangers of simply multiplying the risk of one such recurrence by the same figure to obtain the chance of two such deaths. Quite what impact all this evidence will have had on the jury will never be known but we rather suspect that with the graphic reference by Professor Meadow to the chances of backing long odds winners of the Grand National year after year it may have had a major effect on their thinking notwithstanding the efforts of the trial judge to play it down."

They indicated that, if the matter had been fully argued, they would probably have allowed the appeal on this ground too.

42. The appellant faced lengthy heads of charge which contained a number of statements of fact, most of which were admitted, interspersed with allegations which could be relied on to establish misconduct. The appellant is undoubtedly a person who requires precision of language and was reluctant to accept any allegation which in his view was open to misinterpretation. For example, he would not accept the accuracy of the allegations that he held himself out as being expert in matters relating to deaths classified as SIDS and the statistics concerning them. He did not challenge that he was ready, willing and considered himself able to give such evidence. The distinction, while narrow, is obvious (indeed, the allegation of holding out was deleted). He was represented by solicitor and counsel and no doubt would have received advice on what should and should not be admitted. It must be borne in mind that, although bad faith was not specifically alleged, it was hovering in the background. The lengthy and hostile cross-examination (he gave evidence for some 5 days) certainly did nothing to reassure him that a finding of bad faith would not be sought or made.

43. The important allegations can be summarised as follows:-

"(a) the appellant failed to provide a fair context for the limited relevance (if any) of SIDS deaths and the statistics were misleading and irrelevant;

(b) the squaring exercise was erroneous and failed to have regard to dependence and ignored common environmental, genetic or biological components and their interaction;

(c) he wrongly implied that two deaths were independent; and

(d) the giving of such evidence was outwith his experience. He failed to disclose that he was not a statistician and so was in breach of his duties as an expert."

44. I do not propose to go through the evidence in any detail. Professor Fleming stated what the CESDI study had not been intended to do and that it had been misunderstood by the appellant. But he accepted that the information had been misinterpreted. He accepted too that the author was not the best person to judge the clarity of what he had written. He said:-

"I accept that there has now been considerable discussion and misinterpretation of a lot of the information, largely because people are looking only at a few paragraphs in isolation rather than the sort of multiple pages before where we describe in technical detail the processes. But yes clearly it is possible to misinterpret it. It has been misinterpreted .."

45. Professor Cox, an eminent Statistician, accepted that the squaring exercise was often done but was very often wrong in tending to make a small probability much too small. It would be unreasonable or at least incautious to assume that one death had no influence on the probability of a subsequent death in the same family. He said:-

"The information that one event has already occurred would almost invariably increase the probability of a second event in a similar situation."

Professor Aitkin supported Sir David Cox's views but agreed that the mistake to assume independence was common among statistical lay men. Finally, Professor Golding, a paediatric epidemiologist, was called to try to demonstrate that the true risk of a subsequent SIDS death was 1 in 75. Her attempt failed: that allegation was not proved. But she did accept that the appellant's quotation of the 1 in 73 million risk was a mistake which was easily made. She observed, in answer to a question asking for her view about someone giving expert evidence and using statistics wrongly:-

"I think it is always difficult to present information on something you are not fully versed in. I am sure I do that too, but in this case it had very grave consequences."

46. Professor Golding was cross examined about the studies she had relied on to establish an increased risk and some which pointed in the other direction were put to her. It is not entirely clear whether the upshot of all the evidence was that she believed the risk was increased in the sense that one death tended to show that factors were in play in addition to those which produced the figure of 1 in 8543 so that that figure was too high and the risk of a subsequent death, although smaller was itself to be heightened or whether the risk of a second death was always likely to be lower than that of the first, whatever the risk of the first may have been. As the second Court of Appeal indicated, the general view is that the risk of a second death is lower and there is research material which supports that view. Certainly the experience of those such as the appellant who have worked in the field for much of their working live is that the risk of a second death is indeed smaller. Furthermore, as the appellant pointed out in his evidence, many of the studies, particularly the earlier ones, suffered from a failure properly to define SIDS and included reference to deaths which should not have been classified as SIDS.

47. One of the matters he was questioned about at length was his reference in his first statement to the 1:1000 risk rate. Where did this come from? It appeared in his 1999 paper in the following paragraph:-

"The reason that more than half the reported families included more than one dead child is likely to be because the courts were impressed by evidence that it was highly improbable for two or more children to die in infancy of undiagnosable natural causes: "if there is a 1/1000 chance of a child dying suddenly and unexpectantly of natural causes in the first year of life, the chance of two children within a family so dying is 1/1,000,000". A parent who kills only one child is much likely to be incriminated than one who kills or abuses two or more. Nevertheless, the finding of 26 serial killers is worrying."

The passage in quotation marks has no attribution. In his evidence, the appellant said he could not recall where those figures came from and he recalled writing them on a blackboard in a lecture and reference from a member of the audience. But he said that the figure was, as he put it, a ball-park figure. In reality, it seems that it was based on his general experience and was used as an average. That it was properly so regarded became apparent from the CESDI report, which gave an average of 1 in 1300 odd. It may well be that the appellant did not explain things as clearly as he should have done. The hostility of the cross-examination (which was not conducted by leading counsel for the Respondent) cannot have helped.

48. There has been no challenge to the FPP's findings of fact. Mr Henderson asserts that for an expert witness to go beyond his competence and to profess to be an expert on matters on which he is not an expert amounts to misconduct. It is worse when he volunteers the evidence rather than being led into it in cross-examination. To do this without disclosing that he is not an expert in the matters is misconduct. The fact that it happened in a double murder trial made it worse. The fact that a competent expert in the relevant area would not have given such misleading evidence underlines the seriousness of the misconduct. It produced the prosecutor's fallacy, which again increased its seriousness.

49. Mr Henderson was compelled to accept that if the appellant had said that he was not an expert in statistics but believed that his interpretation of the figures in the CESDI report was correct, he might have had difficulty in seeking to uphold the finding of serious professional misconduct. If the appellant had been asked whether he was an expert in statistics, he would have admitted that he was not. Mr Henderson suggested that if he had volunteered that information, the evidence should have been excluded. Since the defence did not take that point, I cannot accept that that would necessarily have happened. He also ridiculed the evidence of the appellant that his reliance on the statistics was akin to reporting on radiological or other expert medical data in forming his view. I do not accept that criticism.

50. The FPP gave relatively lengthy reasons for concluding that serious professional misconduct was proved. It criticised the appellant for failing to meet his responsibility to use statistics 'in accordance with good statistical principles and practice in relation to matters within your expertise' and continued:-

"You owed a duty to identify relevant matters (including assumptions) on which your statistical evidence was based. You failed in this duty. You should have refrained from giving expert evidence upon matters beyond your competence, but this, again, you failed to do."

This ties in with the alleged failure to comply with the fourth principle in The Ikarian Reefer, namely that an expert should make it clear where a particular issue lies outside his expertise. Mr Henderson supported this approach by pointing out that the appellant had in his evidence in the Crown Court frequently refused to deal with matters based on, for example, pathological findings which were outside his expertise. He had not done the same in relation to the statistics.

51. Nevertheless, the FPP acted too harshly in concluding as it did. The appellant gave evidence of his concerns at giving evidence and the difference between criminal and family courts. He had honestly and as he believed correctly relied on his understanding of the statistics. He had not concealed their source and he was aware that the defence had access to experts. He expected his evidence to be challenged and the adversarial process to establish any errors. He never put himself forward as an expert in statistics. While I accept that he can properly be criticised for not making it clear that he was not an expert in that field, I do not accept that his failure was as heinous as the FPP indicated.

52. The FPP then went on to deal with the 1:1000 and 1:1,000,000 references. It said this:-

"The Panel has heard expert statistical evidence (which it accepts) that the squaring of the 1:1000 ratio to conclude that there was 1 in a million incidence of double SIDS deaths within a family was incorrect. Furthermore you were unable to explain from where you derived these figures. You said in evidence before this Panel that you thought someone in the audience of a lecture you were giving had said this, and that you had remembered putting the figures "on a blackboard somewhere", although you could not recall when and where. The Panel considered this explanation to be unacceptable, and the members were of the opinion that this highlighted your less than rigorous use of statistics and your inability to adhere to strict scientific principles in so doing."

I accept the criticism made by Ms Davies that this was unfair and did not properly reflect his evidence. I have already referred to that part of the evidence.

53. In dealing with the CESDI study, the FPP said that it produced evidence that 'there is an elevated risk of a second SIDS death in one family after there has been one such death'. I am far from sure that that reflects the evidence; it may depend on what is meant by elevated risk. Elevated above what? Their criticism based on the prosecutor's fallacy was also unfair and might well not have been made if they had seen the judgment of the first Court of Appeal. The appellant did not produce the prosecutor's fallacy. He merely gave what he believed to be accurate evidence based on the CESDI study. It was not for him to decide what use was made of that evidence. The FPP stated that his eminence meant that he had a unique responsibility to take meticulous care in such a grave case. I do not think that eminence imposes a greater burden. The FPP said that:-

"Your misguided belief in the truth of your arguments, maintained throughout the period in question and indeed throughout the inquiry is both disturbing and serious."

That in my judgement was hardly fair. In truth, until he had the criticisms put to him, he made one mistake and had no reason to believe he was wrong. His evidence at the inquiry was given to try to show that he had honestly believed that he had not made any mistake.

54. Finally, the FPP decided:-

"The Panel, having considered all those matters, has concluded that your errors, compounded by repetition, over a considerable period of time, constitutes such a serious departure from, and falling short of, the standards expected of a registered medical practitioner, that it finds you guilty of serious professional misconduct."

I have no doubt that that conclusion is not justified by the evidence before the FPP. As I have said, he made one mistake, which was to misunderstand and misinterpret the statistics. It was a mistake, as the Panel accepted, that was easily and widely made. It may be proper to have criticised him for not disclosing his lack of expertise, but that does not justify a finding of serious professional misconduct.

55. Ms Davies submits that the conclusion that he had acted in good faith and that there was no evidence of calculated or wilful failure to use his best endeavours to provide evidence precluded a finding of serious professional misconduct. I accept that such a finding can be made even though there has been no bad faith or recklessness. But it will only be in a very rare case that such a finding will be justified. The lapses in question must be serious indeed to lead to such a finding in the absence of bad faith. I am satisfied that the lapses in this case did not justify the finding.

56. It follows that I would allow the appeal against the finding of serious professional misconduct. It is difficult to think that the giving of honest albeit mistaken evidence could save in an exceptional case properly lead to such a finding.

57. I need say little about the sanction. It was unnecessary since the appellant had retired from clinical practice. It was imposed in the teeth of the many testimonials to him and the knowledge that he had made a real contribution to paediatric medicine. The FPP referred to the seriousness of his 'undermining of public confidence in doctors who have this pivotal role in the Criminal Justice System'. If the full facts are taken into account and the media campaign based on a lack of knowledge of all the circumstances is ignored that comment is unjustified. And to say, as the FPP did, that his conduct was 'fundamentally incompatible with what is expected by the public from a registered medical practitioner' approaches the irrational.

58. I am satisfied that no more than the imposition of a condition not to engage in medico-legal work would have been appropriate. In truth, the finding itself was sufficient. But, in the light of my conclusions on the finding of serious professional misconduct, I will say no more.

59. The appeal is allowed on all grounds.