England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM HIS HONOUR JUDGE PEPPITT QC
(SITTING AS A HIGH COURT JUDGE)

		Royal Courts of Justice Strand, London, WC2A 2LL
		Tuesday, 16 November 1999

Tuesday, 16 November 1999

JUDGMENT

LORD WOOLF MR :

1. This is a judgment of the Court. This appeal arises out of three actions brought by Sandra Penney, Helen Palmer and Lesley Cannon. The actions were tried in Canterbury before His Honour Judge Peppitt QC sitting as a Deputy High Court Judge. After a hearing lasting two weeks, between 18 – 29 January 1999, the judge gave judgment in favour of the claimants on the issue of liability. With the agreement of the parties, he did not deal with the question of causation. The actions arose out of four cervical smears taken from the three claimants in the years 1989, 1990 and 1992 as part of the national cervical screening programme (CSP). Each of the four smears were reported by the primary screeners as being negative.
2. The consequence of the negative reports was that there was no timely follow up or diagnostic or therapeutic intervention as a result of the screening and each of the subjects went on to develop invasive adenocarcinoma of the cervix. This meant the claimants had to undergo surgery which included a hysterectomy. Prior to the hearing in the court below, Judge Peppitt had declined to transfer the cases to London on the grounds that some of the relevant plaintiffs were very ill, with "time not on their side".

Background to the Appeal

3. The Health Authority regards this as a test case. The Authority considers that this is a very important case not only for the claimants but also to those responsible for carrying out the CSP nationally. The Authority is concerned that the case could damage the CSP by requiring inappropriately high standards of care by those responsible for primary screening. It is suggested the decision of the judge could also unnecessarily undermine the confidence of the public in the CSP. While we understand the reasons why the Authority is so concerned, we do have reservations as to whether the implications of this case are as great as the Authority believes.
4. Although the screening programme in its present form began in 1988, screening has a substantially longer history. In the early years there was controversy as to the benefits which CSP was likely to achieve. However, the CSP is now recognised to have been remarkably successful in detecting pre-cancerous changes in squamous cell carcinomas which are known as cervical intra-epithelial neoplasia or "CIN". This is a cancer found in the ectocervix or outer cervix. It was only later that it was appreciated that the screening programme could also be used to identify pre-cancerous changes which can result in adenocarcinoma, endocervical carcinomas. This is possible as a result of the examination of cells taken from the inner cervix. This condition is known as cervical glandular intra-epithelial neoplasia or "CGIN". CGIN is less common and less easy to detect in cervical smears than CIN. However its incidence has for some time been increasing. CGIN is the condition which the claimants allege that screening in their cases should have revealed.
5. If detected at an early stage both these pre-cancerous or cancerous conditions can usually be effectively dealt with by relatively minor surgical treatment. The position can, however, be very different if this does not happen.
6. The cells required for a cervical smear are obtained by scraping the cervix with a spatula. The spatula is shaped so that it can collect cells from both the ectocervix and the endocervix. However if necessary a brush can be used to obtain cells from the endocervix.
7. After the cells have been harvested by the use of the spatula and, if necessary, the brush, they are smeared onto a slide and this smear is then stained with a special dye and fixed. The dye results in the different structures showing up in different colours. The squamous cells ordinarily have an appearance rather like a fried egg, the yoke being known as the nucleus and the white as the cytoplasm. The cells from the endocervix are said to be columnar or glandular because they are more oblong in shape though their shape when examined on a slide will depend on the direction from which they are being viewed.
8. After a smear has been taken it is flooded with a fixative solution to preserve the cells. When the slides have been stained a thin glass cover slip is applied.
9. Before the judge there was an agreed amended general or generic report (the generic report) which provided background information for the purposes of the case. The report had been prepared by Dr. Elizabeth Hudson in collaboration with Dr Boon and Dr Hughes, who are the experts called on behalf of the defendants. It was then amended so that it included agreed additions by the claimants' experts, Professor Cotton and Professor Krausz. This report makes the important point that screening should be conducted "so that the best interests of patients are protected" while at the same time attempting to achieve a balance between detecting as many abnormalities as possible without unnecessarily subjecting the individual from whom the smear is taken to unnecessary anxiety.

The Screening

10. The generic report makes it clear that cervical screening does not provide a fault proof test. Even if the manner in which test is taken and interpreted is exemplary, not all cases of cervical pre-cancer or invasive cancer will be detected. This is especially true of adenocarcinoma which is more difficult to detect by cervical smear. As the generic report points out, the abnormality may be high up the endocervical canal out of reach of the usual smear taking instrument. In addition in the case of CGIN the distinction between normal and abnormal results is less well defined. However, reporting needs to be adjusted to take this into consideration.
11. The screening is done by qualified biomedical scientists or qualified cytology screeners. They sit at their microscopes for periods of up to two hours at a time examining methodically the whole of the area of the slide. If screeners consider that a smear is in the normal range (that is negative) or of such poor quality that it should be reported as inadequate and repeated, they report this. If the screener detects or suspects that a smear is abnormal they are required to pass it to a supervisory checker who either confirms the opinion of the primary screener or if it is still considered abnormal, passes it on to a pathologist for examination and report. The report will usually be for a repeat test or referral to a gynaecologist for further investigation.
12. Primary screeners usually screen on average six to eight slides per hour but they should not be subject to pressure to keep to a particular timescale because some slides are more difficult to examine than others. The task of examining the slides can be made more difficult because of the presence of red blood cells or inflammatory cells (white blood cells) which are common contaminants. Cells may also be spread unevenly, and distorted and interpretation is made more difficult when epithelial cells show severe reactive changes.
13. The classification of cells by screeners should be uniform in accordance with the requirement of the CSP. The relevant classes for present purposes are :

Inadequate	Insufficient or poorly visualised cells
Negative	No abnormal cells seen
Borderline changes	Changes of uncertain significance
Mild dyskaryosis	Mild neoplastic squamous changes usually CIN 1.
Glandular neoplasia	Severe glandular cell changes possible adenocarcinoma

The classification also deals with three more categories of increasingly severe dyskaryosis, namely CIN2 and 3 and invasive squamous carcinoma. The Greek word dyskaryosis refers to an abnormal nucleus.

14. Inadequate smears should result in the test being repeated within three months. Borderline smears should be repeated after three, six or 12 months. The borderline category is used when after examination by the screener, checker and pathologist there is uncertainty whether the cells in a smear are within the normal range. There are "grey areas" between the cells which are clearly normal and cells which are equally clearly abnormal. Borderline is a holding category until the uncertainty is removed as a result of the test being repeated.
15. The generic report makes it clear that even in the best laboratories there will be some false positive and some false negative cervical smear results. A mistake is not necessarily a result of negligence. National standards published in 1996 indicated that primary screeners should detect 85-95% of abnormal smears so that false negative reports do not exceed 5-15%. More recently introduced quality control is intended to reduce false negative reports to 5% or less.
16. The generic report also describes the training of screeners. Screeners, based on their training and experience, have to make a number of judgments, including deciding whether to examine the slides at high magnification. Where a patient who has been tested has developed cervical cancer notwithstanding a negative report, there is a retrospective review. The results of the review vary. There can be no abnormal cells disclosed by the review or there can be a few dispersed dyskaryotic cells which would be unlikely to be detected. In other cases there can be cells which should have been detected by screening but which were wrongly interpreted.

The Wells Report

17. In relation to the claimants, the screening service was provided by the Kent and Canterbury Hospitals NHS Trust ("the hospital"). The screening service provided by the hospital became the subject of adverse publicity in 1996. The Hospital subsequently admitted that an independent check by another laboratory disclosed sufficient screening errors to justify re-screening all of the most recent smears taken from women in the area. Because of the publicity and a report by the hospital that its cytology laboratory failed to meet good practice guidelines between 1993 and 1995 due to understaffing and lack of adequate and effective management, Sir William Wells, Chief Executive of the South Thames Region of the NHS, was appointed to conduct an independent inquiry. His report which was published on 21 October 1997 contained a catalogue of serious criticisms of the hospital. These included the fact that the laboratory was poorly managed, relations between staff were poor, there was poor laboratory performance from at least 1990 until 1994, including under-diagnosis of abnormal smears, and several cases of serial misreporting of smears extending back into the 1980s.
18. The review was admitted by the Authority for the purpose of the proceedings but its relevance was denied. On this appeal Mr Faulks QC on behalf of the Authority challenges the way the report was treated by the judge. In his judgment he indicated that in a sense the denial of the Authority that the report was relevant is well founded for, as he stated :

"I cannot simply rely upon the inquiry findings as evidence specifically referable to the cases which I have to consider. But just as evidence of a well-ordered and efficient laboratory could properly have been adduced by the hospital as the background against which the plaintiffs' allegations should be judged so to the same extent (but no more) I am entitled to have regard to the very different findings of the committee."

19. Having set out this quotation it is convenient to indicate that we consider that the approach which the judge states he adopted to the report was wholly appropriate. The report was not irrelevant but it was of limited significance to the decisions to which the judge had to come on the issues which he had to determine. It was not more than part of the background against which he had to judge the evidence.

The Evidence

20. The other evidence which was before the court was limited. It was confined to the evidence and reports of the five experts to which we have already referred, the generic report and a number of journals and other publications which were agreed to be a comprehensive selection of the relevant literature. No screener had made a statement or was called to give evidence. Nor was any witness called who could give direct evidence as to the training and management of the screeners who examined the slides which are the subject of the claim. As oral evidence, the claimants were content to rely on the evidence of their experts and the Authority was content to rely on its experts. There were of course available to the court the relevant slides and the report on the slides made by the individual screener who was involved. It was suggested that the members of this court might wish to examine the slides under a microscope, but bearing in mind that we are an appellate court, we did not consider that such an examination would help. We were however prepared to view a video on the screening process. This was a helpful introduction to the evidence.
21. The judge found that all five pathologists who gave evidence before him were doctors of distinction. Having regard to their qualifications and experience, the judge could not have come to any other conclusion. He also found that they "gave their evidence with complete integrity and in a genuine attempt to assist the court". Both Mr Faulks and Mr Badenoch QC, on behalf of the claimants, accepted that this approach to their evidence was appropriate. Having had the opportunity to read extensively from the transcripts of their evidence, we adopt the same approach. However, Mr Faulks and Mr Badenoch did repeat the submissions which they had made to the judge as to why their side's experts' evidence should be preferred. We are unimpressed by submissions related to the age or experience of the experts. This is subject to an exception in relation to Dr Hudson. Her experience was especially relevant since she was regularly involved in training programmes for cytoscreeners, has written extensively on the questions which the claims raise and has been closely involved in the evolution of the standards which cytoscreeners should adopt.

The Law

22. In his judgment, the judge having set out the background facts with admirable clarity, including a detailed account of the task of the cytoscreeners, turned his attention to the law. He began by indicating that both parties agreed that :

"The standard which I have to apply is that of a reasonably competent screener exercising reasonable care at the time when the screening took place. I must ignore any advances in screening practice which have occurred since the relevant events. I must also put out of my mind when considering the extent of the screeners' duty of care the fact that all three [claimants] subsequently developed carcinoma.

Equally importantly I must bear constantly in mind that in cases where an exercise of judgment is called for, the fact that with the benefit of hindsight that judgment was exercised wrongly is not itself proof of negligence."

23. The judge then went on to cite from Bolam v Friern Hospital Management Committee [1957] 1 WLR 583, Maynard v West Midland Regional Health Authority [1984] 1WLR 634 and Bolitho v City and Hackney Health Authority [1998] AC 232. The judgments in these cases are very well known and it is not necessary to cite extensively from them. However we would draw particular attention to the words of Lord President Clyde in Hunter v Hanley [1955] SLT 213 which the judge cited from the speech of Lord Scarman in Maynard :

"In the realm of diagnosis and treatment there is ample scope for genuine difference of opinion and one man clearly is not negligent merely because his conclusion differs from that of other professional men ... The true test for establishing negligence in diagnosis or treatment on the part of the doctor is whether he has been proved to be guilty of such failure as no doctor of ordinary skill would be guilty of, if acting with ordinary care."

24. To this passage Lord Scarman added :

"... that a doctor who professes to exercise a special skill must exercise the ordinary skill of his speciality. Differences of opinion and practice exist, and will always exist, in the medical as in other professions. There is seldom any one answer exclusive of all others to problems of professional judgment. A court may prefer one body of opinion to another: but that is no basis for a conclusion of negligence." (at p.638)

25. To those passages it is only necessary to make one short additional citation from the speech of Lord Browne-Wilkinson in Bolitho at p.1158 H. Lord Browne-Wilkinson stated :

"In my view the court is not bound to hold that a defendant doctor escapes liability for negligent treatment or diagnosis just because he leads evidence from a number of medical experts who are genuinely of the opinion that the defendants' treatment or diagnosis accorded with sound medical practice(.. the court has to be satisfied that the exponents of the body of opinion relied on can demonstrate that such opinion has a logical basis. In particular in cases involving, as they often do, the weighing of risks against benefits, the judge before accepting a body of opinion as being responsible, reasonable or respectable will need to be satisfied that, in forming their views, the experts have directed their minds to the question of comparative risks and benefits and have reached a defensible conclusion on the matter."

26. Both before the judge and before this court counsel were agreed that the approach indicated in the passages which have been cited should be applied to these cases. We agree. The screeners were exercising skill and judgment in determining what report they should make and in that respect the Bolam test was generally applicable. Later authorities make clear that this it is the appropriate standard to apply. However, as we will explain, the fact that two sets of competent experts genuinely hold differing opinions as to whether or not at the relevant date, which is the date of the examination, the screeners could without being negligent have diagnosed the smears as negative does not necessarily provide the solution to the dispute on liability in these cases.
27. There is the qualification which Lord Browne-Wilkinson identified in the passage already cited from his opinion in Bolitho. In addition the Bolam test has no application where what the judge is required to do is to make findings of fact. This is so, even where those findings of fact are the subject of conflicting expert evidence. Thus in this case there were three questions which the judge had to answer :

What was to be seen in the slides?

At the relevant time could a screener exercising reasonable care fail to see what was on the slide?

Could a reasonably competent screener, aware of what a screener exercising reasonable care would observe on the slide, treat the slide as negative?

28. Thus, logically the starting point for the experts' reasoning was what was on the slides. Except in relation to the slide known as Palmer 2, as to which there was a striking conflict, as a result of a meeting which took place between the experts they were in substantial but by no means total agreement. In so far as they were not in agreement, the judge had the unenviable task of deciding as a matter of fact which of the experts were correct as to what the slides showed. This was a task which required expert evidence. However the evidence having been given, the judge had to make his own finding on the balance of probabilities on this issue of fact in order to proceed to the next step in answering the question of negligence or no negligence. Having come to his own conclusion as to what the slides showed, the judge had, therefore, then to answer the 2^nd and 3^rd questions in order to decide whether the screener was in breach of duty in giving a negative report. Whether the screener was in breach of duty would depend on the training and the amount of knowledge a screener should have had in order to properly perform his or her task at that time and how easy it was to discern what the judge had found was on the slide. These issues involved both questions of fact and questions of opinion as to the standards of care which the screeners should have exercised. As already indicated, there was virtually no evidence of the actual training provided to the primary screeners. The approach of the experts was to give their opinion, based on their respective interpretations of what was on the slide, on the general question of whether a reasonably competent screener, exercising the appropriate standard of care, could treat the slide as negative.
29. The distinction between issues of fact and issues as to what is or is not an appropriate response to facts when the facts have been ascertained is illustrated by the case of Loveday v Renton and Welcome Foundation Ltd [1990] 1 MEDLR 117. In summarising his conclusions in that case Lord Justice Stuart-Smith said (at p.182) :

"(1) The preliminary issue to be determined by the court is, can pertussis vaccine cause permanent brain damage in young children? ... The burden of proof rests on the plaintiff and the standard of proof is that of the balance of probability. It must be shown that it is more likely than not that the vaccine can cause permanent brain damage.

(2) Medical and expert opinion is deeply divided on the issue. The question has to be determined on all the evidence in the case, which is primarily the oral evidence of the witnesses tested in cross examination. The court cannot simply accept the opinion or belief of a witness, however eminent, that such is or is not the case. The basis for the opinion must be examined, tested against other evidence, for consistency and logic and the validity of the reasoning.

(3) The question is not answered by showing that there is a respectable and responsible body of medical opinion that the vaccine can, albeit rarely, cause permanent brain damage, or that this view may be more widely held than the contrary. The opinion of others not called to give evidence is not admissible to prove the truth of the opinion. The works of learned and qualified authors form part of the general corpus of medical and scientific learning on the subject and can be relied upon and adopted by suitably qualified experts. These experts may have their opinions tested in the light of the literature."

The Appellants' Case

30. The appellants challenge the judge's decision on the central question of the relevant state of knowledge at the time when each of the slides was reported upon. However, it is accepted that the judge asked the right question, that is "How a reasonably competent cytoscreener at the relevant time should have reported her slide?". Mr Faulks submits that the opinions of the experts called on behalf of the appellants were that abnormalities to be seen on the slides (other than Palmer II, which requires separate treatment) would not have been recognised as such by a reasonably competent cytoscreener at the time. In those circumstances, the classifications of the slides as "negative" could properly have been made by a reasonably competent cytoscreener. It is submitted that these opinions of the appellants' experts were based on a thorough knowledge of the standards of cytoscreeners, the evolution of knowledge, practice in laboratories, examination questions, proficiency tests and all the available literature. They amounted to a good defence upon a proper application of Bolam principles. The judge is criticised for holding that Bolam did not apply and, in so far as he did, for deciding what were really matters of opinion as if they were questions of fact. He is also criticised for not giving detailed reasons in his judgment for rejecting the appellants' case as to the state of knowledge. It is submitted that at the very least there should be a retrial to permit a judge to come to a reasoned decision on this central issue. At the same time, it is accepted as being highly undesirable that there should be a retrial and Mr Faulks said that the appellants would not wish to put the individual respondents through one. But his instructions did not extend to being able to give any assurance that this would not happen.
31. In addition to what appeared on the slide, Mr Badenoch submitted that the question of what was known at the time of the examination of the slides was a question of fact to which Bolam could have no application. This submission we regard as being only partially correct. The state of knowledge may be objectively discernible and therefore be a matter of fact. However, there would be room for differences of opinion as to the extent that screeners at a particular time should be required to be aware of the latest learning on a particular subject. This would be a subject on which respectable opinion could differ as to what is the appropriate standard. Thus, for example, one view might be that the already demanding cytoscreeners' task should not be over-complicated by training them to spot the more subtle and complex changes the relevance of which was not yet fully understood; another view might be that they should be trained to do the best they can to spot and evaluate anything which might be relevant. The evidence did not suggest any divergence of view such as this. Similarly, there might be a difference of opinion as to how much judgment a screener should personally exercise once a potential abnormality was spotted; but again the evidence did not suggest any such divergence of view.

Mrs Penney's Slide

32. The judge's approach to this subject can be illustrated by the slide relating to Mrs Penney dated 31 December 1992. We therefore take her case first. The judge commenced by recording that her slide was by common consent "difficult to interpret though none of the experts argued that the cytoscreener might reasonably have failed to observe what it contained." He then set out the views of the four experts namely the claimants' two experts and Dr Hudson and Dr Boon. Their descriptions of what the slides showed differed in detail. He then set out the fact that a meeting took place on 5 December 1998 when all the disputed smears were reviewed and it was agreed that the slides showed :

"Four or five inflamed groups of endocervical cells showing changes in nuclear and cytoplasmic morphology. It was not a normal slide."

33. The judge went on to state that the experts were agreed that at the time they gave evidence (ie not at the time when the screener examined the slide) a reasonably competent cytoscreener would have reported the smear at least as "borderline". At the relevant time that is 1993, a cytopathologist (in other words someone more skilled than a cytoscreener) would at least have reached the conclusion that this smear was "probably borderline". He then records the fact that the claimants' experts considered that the smear should have been classified at the time as borderline whereas the other experts took a different view.
34. On this evidence the judge said :

"The issue which I have to decide therefore is whether in the light of what he saw the cytoscreener was negligent in failing to classify the smear as at least borderline. On this issue I prefer the views of Professors Cotton and Krausz. The slide was difficult to interpret even by a Consultant Pathologist. But the abnormality was there to be seen. Whether that abnormality was precancerous or reactive and thus benign it was not for the cytoscreener to decide. His function was to observe and to record. He had neither the knowledge nor the experience to diagnose. In my judgment no competent cytoscreener would have dismissed the possibility that the abnormality on this smear was precancerous. Accordingly if one applies the "absolute confidence" test propounded by Dr Hudson this smear had to be classified at least borderline. Anything less would place the patient in danger.

I have already set out the conflicting views of Doctors Hudson and Boon. In so far as the import of those views is that no cytoscreener could have been expected in 1993 or later to say affirmatively whether the abnormality on the smear was reactive or precancerous I accept them. But that is not the test which both parties agree that I should apply. I have to consider whether the reasonably competent cytoscreener in 1993 could properly have dismissed the possibility that the abnormality was precancerous. I have already indicated that in my judgment he could not. On the smear were four or five abnormal groups of endocervical cells which I think on the balance of probability were precancerous. No-one who gave evidence before me sought to suggest how a cytoscreener in 1993 might have been able competently to distinguish between precancerous and reactive changes. Indeed the evidence was to the contrary - that the two are often indistinguishable. In those circumstances I consider that the cytoscreener was wrong to classify this smear as anything less than borderline."

35. Pausing there it is important to note that the judge is indicating that his understanding of Drs Hudson and Boon's evidence is that they were saying that a cytoscreener could not in 1993 have been expected to say one way or the other whether the smear was abnormal or not.
36. The judge then went on to say :

"I should say at the outset that I find the Bolam principle ill-fitting to the facts of Mrs Penney's case. In Bolam and the cases which followed the court was concerned with an aspect of professional conduct of which some members of the profession, but not others, disapproved. In other words in those cases the defendants' experts sought to justify as an acceptable professional practice what the defendant did or did not do. Here the position is different. All the experts agree that the cytoscreener was wrong. No question of acceptable practice was involved. The issue here to which the experts' evidence was directed was whether the cytoscreeners conduct though wrong, was excusable. This seems to me to fall outside the Bolam Principle.

But if I am wrong about this I remain of the view that Bolam does not assist the defendants. For I do not consider that the evidence of Drs Hudson and Boon stands up to the logical analysis as that phrase was used by Lord Browne-Wilkinson in Bolitho at p.1160 c-d. This is not to disparage the evidence of either. It is rather that in my judgment their opinions cannot stand with 'the absolute confidence' test which Dr Hudson herself propounded with the agreement of the other experts. Here were admitted abnormalities which, to put if most favourably to the cytoscreener, he could not positively have said were not pre-cancerous. Neither Dr Hudson nor Dr Boon suggested that the cytoscreener had the ability to draw any such distinction, still less how they should apply it. It seems to me therefore that having regard to the potentially disastrous consequences of a mistaken classification a reasonably competent cytoscreener should have classified the smear as borderline even though that classification might have caused the patient short-term distress and perhaps the discomfort and embarrassment of a further smear. I cannot believe that any woman would not be prepared to put up with both if the alternative was an undiagnosed potential carcinoma.

It is for these reasons that I have preferred the evidence of Professors Cotton and Krausz to Drs Hudson and Boon in finding as I do that the defendants were negligent and in breach of the duty which they owed to Mrs Penney in failing by their cytoscreener to classify her slide in January 1993 as at least borderline."

37. In these passages the judge clearly sets out his reasons. In two different passages he sets out the correct standard of care. He also makes it clear by his references to 1993, that contrary to Mr Faulks' submission he appreciates that he has to focus upon that year. He makes findings about what the slide show. He concluded there was an abnormality which even a consultant pathologist would find difficult to interpret although the abnormality was there to be seen. Because the slide was difficult to interpret in order to protect the patient, in accordance with Dr Hudson's own approach, the slide should have been classified as at least borderline.
38. The judge describes the Bolam principle as being "ill fitting to the facts of Mrs Penney's case". On an analysis of the judge's reasoning this is an understandable statement for him to make. On the judge's approach, this was not a case where there were two acceptable standards of professional conduct involved. The only comment of the judge which is difficult to understand is his questioning "whether the cytoscreeners conduct though wrong, was excusable". However, this does not matter because the judge also, in the alternative, applied the Bolitho test. The judge's reasoning was that, if the cytoscreener could not say with a reasonable degree of confidence that the abnormalities were not precancerous, then it was inconsistent with what Dr Hudson and Dr Boon were agreed was the right approach to report her slide as negative.
39. In his submissions, Mr Faulks makes a number of criticisms of the reasoning of the judge. He describes it as simplistic. He submits it amounted to no more than saying that if a slide contains features which might be abnormal then a negative classification should not be given. In relation to Mrs Penney's slide, we do not consider this criticism is appropriate. The judge was basing himself on the difficulty of interpreting that particular slide. We do not regard the reasoning of the judge as inconsistent with the statistics which show that a false negative rate would not be less than 5% to 15%.
40. The judge's reliance on the absolute confidence test is understandable. The phrase 'absolute confidence' was no more than shorthand for the approach which on examination of the transcripts it seems to us all the experts endorsed. To take single passages from the experts' evidence can be misleading, but as an example supportive of the judge's approach reference can be made to the transcript of day 6 at p.79. Dr Hudson was asked "I wonder if you could help the learned judge in terms of the concept of doubt in what the screener, in what is expected of the screener in terms of their response to a relatively difficult smear?".
41. She gave the answer :

"Certainly if they have doubt they must pass it on to a checker, and when they are training they have doubts all the time and discuss this with their trainer but once they have passed their exam they are expected to make a certain number of decisions themselves and indeed quite a lot of decisions (".

42. The judge also asked :

"Is that the test, anything short of absolute confidence within the normal range to pass on?"

Answer :

"Yes, my Lord".

43. What the judge regarded as being an illogical approach on the part of Dr Hudson was explored during her cross examination. For example, (day 6 p.115/116) during her cross examination, she argued that a screener need not refer a particular smear, although earlier she had accepted that in the case of this type of smear a high degree of expertise would be required before it could be confidently stated that the slide was negative.
44. In the case of Dr Boon, he makes it absolutely clear that the cytoscreener should only treat a slide as negative if he or she is satisfied beyond reasonable doubt that that is the position. (Day 7 p.72/73) Dr Hughes, the Authority's other expert, also adopted the same approach as to "the absolute confidence test".
45. The generic report supports the judge's approach to the absolute confidence test. At page 4 it makes the point that "competent cytology screeners should not miss smears with significant numbers of abnormal cells. More subtle abnormalities or small numbers of abnormal cells can inevitably be missed by competent screeners under normal screening conditions."
46. At page 9 it states "The primary screener's job is to make a judgment based on training and experience about the normality or possible abnormality of a smear. The primary screener is encouraged to refer to the checker and pathologist if in any doubt (but population screening by current methodology would not be possible if all cervical smears had to be reported by checkers or pathologists)."
47. Then the Wells Report states that staff at the primary screening level:

"... may make the final decision on and report negative and inadequate smears. If there is any doubt or if abnormalities are detected, the slide will be passed on to a technical "checker" ..."

48. The critical question is whether the judge was entitled to conclude that in the case of Mrs Penney's slide a reasonably competent cytoscreener would have at least been aware that the slide was difficult to interpret. On the judge's findings of fact, this was what the cytoscreener should have concluded. There were abnormalities to be seen on the slide and the reasonably competent cytoscreener could not with confidence have concluded that they were not pre-cancerous. The cytoscreener should then have referred the slide for further examination. Either because of lack of training or because of the way in which the slide was examined, the cytoscreener did not do this.
49. In evidence and in argument before the judge, there was a considerable amount of discussion of the need to have a balance between sensitivity and specificity. There needed to be a balance on the part of the screeners because otherwise there would be an excessive risk of their reaching "false negative" or "false positive" conclusions. However, as the judge made clear, from the point of view of the patient, a false negative could have very adverse consequences. A false positive would have nothing like this disadvantage to the patient. This is because a false negative could have even fatal results for the patient, while as long as the checker or pathologist corrected the mistake made by the screener, a false positive annotation of the slide would have no adverse consequence to the patient. If they repeated the screener's mistake the patient could be caused anxiety, but this is a small price to pay for the protection against the adverse consequences that could result from a slide wrongly being classified as negative. We have no doubt on the evidence that the screening programme should aim by enhanced training of those who operate it to reduce so far as possible the number of false positives and false negatives. On the undisputed evidence before the judge however this cannot logically compromise the need for a primary screener to apply the absolute confidence test.
50. Dr Hudson, while adhering to the approach which we have indicated of placing the patient's safety first, did stress that screeners were required to exercise judgment and it was on the basis of this requirement that the screeners should exercise judgment that she would not criticise the screener who examined Mrs Penney's slide. While this approach may be justified in relation to appropriate smears the judge was entitled to conclude that it could not be justified with regard to the "4 or 5 abnormal groups of endocervical cells" which he found to be present on Mrs Penney's slide. It is correct that the groups were agreed by the experts to be inflamed but this made the screener's task more difficult and to that extent supported the judge's conclusion. Despite the experts' agreement as to what the slide showed Dr Hudson placed less emphasis on what the slide revealed. She considered this to be "slightly atypical endocervical epithelial cells", while Dr Boon based his opinion that it was wholly unreasonable to expect a competent screener to have consistently detected such material on his conclusion that there was no suspicion of glandular neoplasia.
51. The judge did not accept Dr Hudson's and Dr Boon's descriptions of what the slides showed. He was entitled to prefer the evidence of the claimants' experts that there were abnormalities on Mrs Penney's slide which were there to be seen. Applying the agreed absolute confidence test to these abnormalities, the judge was entitled to come to the conclusion which he did. Mrs Penny's case can also provide the answer to Mr Faulks contention that the judge failed to deal adequately with the evidence which was given as to the extent of knowledge of which screeners should have been aware at the relevant time. The judge's approach was to ask himself could a reasonably competent screener pass the slide as negative. Basing himself on his findings as to the abnormalities the slides showed, he considered it was not possible for the reasonably competent screener to do otherwise than take the safe course and not trust the slide as negative. To come to such a decision it was not necessary to make a detailed analysis of the state of knowledge of screeners in 1993. All the judge had to be satisfied was that the reasonably competent screener should be able to recognise abnormalities which were present, and be unable to conclude with confidence that there was an innocent explanation for their presence.

Mrs Cannon

52. Having dealt in some detail with Mrs Penney's slide, it is possible to approach the three remaining slides which have to be considered more briefly. In Mrs Cannon's case the relevant slide was taken on 14 July 1992. It was labelled as "negative. Endocervical cells present". When the four experts, who gave evidence as to this slide, (who did not include Dr Hudson) met on 5 December 1998, they could not agree as to what the slide showed. The claimant's doctors found severe dyskaryosis. The Authority's experts thought what was present was not so obvious and referred to borderline changes. Dr Boon observed reactive changes consistent with active inflammation and no frankly dyskaryotic cells. He considered that it was impossible to distinguish inflammatory changes from low grade dyskaryosis. Dr Hughes observed varying degrees of abnormality of group architecture and nuclear morphology which with hindsight was probably the precursor of adenocarcinoma, but he considered that as initially viewed they could still represent reactive changes. The Authority's experts basing themselves on their assessment of what this slide showed thought the screener could be excused for regarding them as negative. However, the judge agreed with Professors Cotton and Krausz that there were abnormalities present which should have been clearly apparent to a cytoscreener in that year. If that was the factual position, then the failure to refer was a breach of duty. In our judgment the judge on the evidence was entitled to come to that factual conclusion.
53. In dealing with Mrs Cannon's case, the judge specifically dealt with the argument which had been advanced before him and was repeated before us that if cytoscreeners referred all smears about which they entertained doubt, the scheme would be over clogged with referrals and in consequence a number of women would suffer unnecessary distress. Rightly in our judgment the judge did not find that argument convincing, partly because of what we have indicated already as to the relative disadvantages of a false negative diagnosis as compared with a false positive diagnosis and also because the only effect of a false positive diagnosis would be to require a repeat smear in 6-12 months.
54. Mrs Cannon's slide was reviewed at a Birmingham laboratory in April 1996 and again reported as negative. The judge considered this fact but as he pointed out he had no evidence as to the circumstances in which the review was conducted. He thought it was probably no more than a rapid rescreening. Again a detailed analysis of the state of knowledge in 1992 was unnecessary.

Mrs Palmer

55. We turn to Mrs Palmer. In her case there is a need to consider two smears. One taken on 25 April 1989 and the second on 27 March 1990. The judge referred to them as Palmer I and Palmer II and we will do likewise.
56. In Mrs Palmer's case, Palmer I was reported "negative. Endocervical cells not present. Suggest repeat". The report on Palmer II read "Negative. Endocervical cells present". In December 1995 Mrs Palmer was unfortunately diagnosed as suffering from adenocarcinoma.
57. As the report on Palmer I suggested a repeat, even if there was a misdescription, it is not clear that there could be any adverse consequences to Mrs Palmer. However, as the issue of causation is not before us, we should not assume that if there was negligence, Mrs Palmer is not entitled to have the finding in her favour upheld. Mr Badenoch submitted the description "negative", even after the report in relation to Palmer II, could have had an adverse affect upon her treatment.
58. In relation to what the slides showed, the views of the experts were again divided with the claimant's experts of the opinion that the slide showed a more serious situation than that indicated by the defendant's specialist. Professor Krausz found 8 clumps of cells of which some were dyskaryotic. Professor Cotton regarded the cells as readily visible under low magnification. Dr Boon, although he may have been persuaded there were eight clumps, considered the size of the cells was such that it was possible to miss them altogether. Dr Hughes agreed with Dr Boon and regarded the abnormalities as borderline.
59. The judge accepted the evidence of Professors Cotton and Krausz that the cytoscreener should have spotted the endocervical cells on Palmer I. He regarded their evidence as being more logical and more consistent than that of the other experts. He referred to Dr Boon having shifted his ground. He did not however accept Professors Cotton's and Krausz's evidence that the cytoscreener should have identified the endocervical cells as being possibly dyskaryotic. He formed his own judgment as to the facts. The judge considered that the cytoscreener having identified the endocervical cells should have referred the slide on, since a cytoscreener was not equipped to assess the true significance of the cells which were present.
60. In relation to Palmer II there was a remarkable difference of opinion between the experts. The views differed as to what the cytoscreener should have reported from Professor Cotton's "urgent referral absolutely mandatory" to negative by Dr Boon and Dr Hughes. Part of the explanation for the striking difference of opinion was that Drs Boon and Hughes attributed the changes which the slide showed as being indicative of or commensurate with inflammation. It was common ground that a screener would (as happened) have identified the relevant cells.
61. Again the judge preferred the evidence of Professors Cotton and Krausz. If he was entitled to conclude they were more likely to be correct as to what the slide showed, then he was entitled to reach the conclusion which he did. What the slide showed was a matter which had to be determined by the judge on the experts' evidence. The experts' views could not be reconciled. Mr Faulks submits that Professors Cotton and Krausz were almost certainly wrong. He makes a number of specific criticisms of their evidence. However, none are critical and they are not sufficient to show the judge was wrong to accept the evidence of Professors Cotton and Krausz. The evidence justified a finding that the slide showed clear abnormalities and the very disagreement between the experts underlined the fact that no primary cytoscreener could treat the abnormalities as innocuous.

The Court of Appeal's approach to the judge's findings as to expert evidence

62. Mr Faulks on this issue and on other expert issues was inviting us on the basis of the material which we had before us, namely the transcripts of the oral evidence, the written evidence and the literature to take a different view on the factual issues from that of the judge. However we find that it is not possible to do this. It appears to us that the approach of the judge was logical. He was clearly not impressed by the way Dr Boon changed his ground in relation to Mrs Cannon. The significance of this can only be properly assessed by a judge who heard him give evidence which we have not. Although we have read the evidence and the literature (which we found of little assistance) as Mr Faulks requested, we do feel that even in a case of this sort, the judge was in a better position than we are to assess the expert evidence. He was also entitled in our judgement to take into account the background of the Wells Report to which we have referred. In relation to Palmer II, he was entitled to take into account the fact that Mrs Palmer unfortunately developed cancer. He could do so, not for the purposes of performing the inappropriate exercise of saying as she developed cancer and the screener gave a negative report that report must be negligent; but because the fact that Mrs Palmer developed cancer is a factor to which Mr Badenoch can point as being consistent with the evidence of his experts.
63. The judge in his judgment pointed out passages in the evidence of Dr Boon and Dr Hughes which might indicate that they were departing from the absolute confidence test. However while we understand the judge's reasons for doing so, we regard this as unnecessary for his conclusion. If what Professors Cotton and Krausz said appeared on the slides was in fact there to be seen, the slides should have been referred for further examination.
64. Appropriately Mr Faulks cited the judgment of Bingham LJ in Eckersley v Binnie (1988) p.1. In a dissenting judgment Bingham LJ having been referred to the relevant authorities accurately stated the role of the Court of Appeal. He pointed out that while in relation to expert evidence an appeal is by way of a rehearing "the Court of Appeal is a court of review". "It does not approach the resolution of factual issues as if the sheet before it were blank. It has to be persuaded that the trial judge, who is a primary judge of fact, has plainly erred." Bingham LJ then indicated that it would be rare to interfere with a trial judge's conclusion where they rest on the credibility and demeanour of witnesses. And more important for this case, he made the following statement with regard to expert evidence :

"In resolving conflicts of expert evidence, the judge remains the judge; he is not obliged to accept evidence simply because it comes from an illustrious source; he can take account of demonstrated partisanship and lack of objectivity. But, save where an expert is guilty of a deliberate attempt to mislead (as happens only vary rarely), a coherent reasoned opinion expressed by a suitably qualified expert should be the subject of a coherent reasoned rebuttal, unless it can be discounted for other good reason. The advantages enjoyed by the trial judge are great indeed, but they do not absolve the Court of Appeal from weighing, considering and comparing the evidence in the light of his findings, a task made longer but easier by possession of a verbatim transcript usually (as here) denied to the trial judge." (p.77/78)

65. In examining the evidence which the experts gave in this case, we have tried to faithfully apply this guidance of Bingham LJ. Having done so, we have come to the conclusion that the judge was entitled to come to the conclusions which he did. Although the evidence was substantially confined to that of the experts and although there are transcripts available, even in a case of this nature, the judge still has an advantage over an appellate court of having heard the oral evidence. In addition it is always important for this Court to remember that, as Bingham LJ said, the Court of Appeal is a court of review even though the appeal is by way of rehearing. This court's task is to rectify an injustice arising out of the decision of the court of first instance. Unless it is satisfied that there has been an injustice, the decision of the lower court should stand.

The examination of the slides at the Krausz Laboratory

66. After Mr Faulks had finished his submissions on behalf of the Authority, Mr Pittaway made submissions on a separate point. This was in relation to the judge's decision to allow evidence to be given of the results of examination of the slides which Professor Krausz had arranged to be carried out on the slides by cytoscreeners at his laboratory. As the Professor acknowledged, there were a number of deficiencies in this exercise. However the judge in making a ruling that the evidence could be admitted, accepted that its evidential value was very seriously diminished. He thought it was of "fringe relevance". He in fact made no reference to this evidence in his judgment. However although careful attention had to be paid to its deficiencies, the evidence could not be said to be inadmissible. It was of some weight and of some support of the claimants' case. The approach of the judge was the correct one.

Conclusions

67. Before leaving this case, we should stress the following points so that the significance of the decisions of the court in these cases can be properly appreciated :

This judgment only relates to the four slides which were the subject of the claims. It is critical to the findings of the judge and the conclusions of this court that the judge decided that the slides contained obvious abnormalities which meant that their assessment as negative was negligent.

If the abnormalities observable on the slides had been different, the decision of the judge could have been different. This case does not decide that negligence by a cytoscreener can be established by showing that someone who has had a slide labelled negative unfortunately develops cervical cancer. It was not in dispute in this case that cervical cancer can develop even though a relatively recent slide is properly labelled negative. The fact that in the majority of cases this does not happen does not mean that it cannot happen even though a high degree of care is exercised.

The judge was not rejecting the general approach of the Authority's experts and in particular that of Dr Hudson. He was finding that because of the observable abnormalities on the slides the slides should not have been labelled negative in order to comply with the approach (the absolute confidence approach) that those experts supported.

We dismiss the appeal.

Order: Appeal dismissed with costs. Leave to appeal to the House of Lords refused.