Scottish Court of Session Decisions

OUTER HOUSE, COURT OF SESSION [2011] CSOH 63
A137/09	OPINION OF LORD TYRE in the cause HELEN McGLONE Pursuer; against GREATER GLASGOW HEALTH BOARD Defenders: ­­­­­­­­­­­­­­­­­________________

Pursuer: A Smith, Q.C., Hodge; Balfour + Manson LLP

Defender: McLean, Q.C., Davidson; Scottish Health Service Central Legal Office

6 April 2011

1. Introduction

[1] On 28 January 2008, while working in Geneva, the pursuer had a cervical smear test which reported the presence of abnormal features. After further investigation, she was diagnosed as having invasive adenocarcinoma of the cervix. On 30 May 2008 she underwent a radical hysterectomy operation, all conservative options for treatment having been excluded by the size of the tumour. At that time she was 28 years of age.

[2] In this action, the pursuer seeks reparation from the defenders for loss and damage caused by the misinterpretation and misreporting of two cervical smear tests carried out on 15 December 2005 and 9 March 2006. The pursuer avers that if either of these tests had been correctly interpreted and reported, she would have been referred for further tests and treatment that would have revealed the presence of cancer which, at that time, was either pre-invasive or at least at a sufficiently early stage of invasion that it could have been conservatively treated by a method which preserved her fertility. The defenders admit that the test carried out on 15 December 2005 was misinterpreted though not that it was misreported. They admit that the test carried out on 9 March 2006 was both misinterpreted and misreported. They aver, however, that the pursuer already had significant invasive cancer when the two smears were taken and that if her cancer had been diagnosed in 2006 the treatment would have been materially the same as it was in 2008, with the same impact upon inter alia her fertility.

[3] The action came before me for proof restricted to certain issues of causation. In order to focus the issues for the diet of proof, a draft note listing the following issues was tendered by senior counsel for the defenders shortly after the proof commenced:

"On a balance of probabilities:

1. What report should correctly have been given of the cervical smear that the pursuer underwent on 15 December 2005 [n.b. it is agreed that the report should have been given by 6 January 2006]?

2. What medical management of the pursuer (including any further investigation and/or treatment) would have taken place as a result of such correct report being given, and when?

3. What report should correctly have been given of the cervical smear that the pursuer underwent on 9 March 2006 [n.b. it is agreed that the report should have been given by 23 March 2006]?

4. What medical management of the pursuer (including any further investigation and/or treatment) would have taken place as a result of such correct report being given, and when?"

Although no formal interlocutor was pronounced, these questions provided the framework for the scope of the proof. However, it became apparent that the principal issue between the parties was implicit in questions 2 and 4: namely, what would have been discovered in the course of the further investigation that would have followed a correct report of either smear test? There was a consequential issue regarding the appropriate medical management of the pursuer's condition as disclosed by that investigation. I deal with each of these matters, as well as the first and third questions listed in the note, in the course of this opinion.

[4] For ease of reference later, it is convenient to record here that in the course of the proof I heard evidence from the following witnesses.

For the pursuer:

• Professor Dulcie Coleman, Honorary Consultant Cytopathologist and Professor Emeritus, Imperial College, Hammersmith Hospitals NHS Trust, London;
• Dr Roberto Dino, Consultant Cyto-Histopathologist and Honorary Senior Lecturer, Imperial College, Hammersmith Hospitals NHS Trust, London;
• Mr W Patrick Soutter, Honorary Consultant Gynaecologist and Honorary Reader in Gynaecological Oncology, Imperial College, Hammersmith Hospitals NHS Trust, London;
• Dr Graham McKillop, Consultant Radiologist, Royal Infirmary of Edinburgh;
• Dr Urszula Bankowska, retired Consultant in Family Planning & Reproductive Health, Sandyford Initiative, Glasgow;
• The pursuer.

For the defenders:

• Dr Mary Stephen, retired Consultant Cytopathologist, Glasgow Royal Infirmary;
• Dr Audrey Brown, Consultant in Sexual and Reproductive Healthcare, Sandyford Initiative, Glasgow;
• Dr Alistair Williams, Reader in Pathology, University of Edinburgh, and Honorary Consultant Gynaecological Pathologist, Royal Infirmary of Edinburgh;
• Dr David Farquharson, Consultant in Obstetrics & Gynaecology, Royal Infirmary of Edinburgh;
• Mr Raj Naik, Consultant Gynaecological Oncologist, Northern Gynaecological Oncology Centre, Queen Elizabeth Hospital, Gateshead;
• Dr Edward Duvall, Senior Lecturer in Cytopathology, University of Edinburgh, and Honorary Consultant, Directorate of Pathology, Lothian Health NHS Trust.

In addition, the parties agreed (i) that a precognition of Dr Jonathan Weintraub, Consultant Pathologist, Viollier Weintraub Laboratory, Geneva, Switzerland, was to be accepted as the equivalent of his evidence; and (ii) that an affidavit of Dr Alison Bigrigg, Consultant in Obstetrics & Gynaecology, Sandyford Initiative, Glasgow, who was unable due to illness to attend the proof, was to be accepted as the equivalent of her evidence.

[5] At the conclusion of the hearing of evidence, senior counsel for each of the parties provided a written note of his submission, for which I am grateful. Counsel for the pursuer also provided an Appendix to his submission entitled "The Medical Background" consisting of (i) a glossary of terminology based largely upon a glossary in a report by Mr Soutter; (ii) a description of the national cervical screening service based largely upon a report by Professor Coleman; (iii) a resume of the pursuer's medical history; and (iv) a summary of the oral evidence of certain witnesses regarding the two smear tests on 15 December 2005 and 9 March 2006. Counsel for the defenders provided a response which accepted parts (i) and (ii) of the pursuer's Appendix as correct subject to certain comments and qualifications. As will be apparent, I have made use in this opinion of these parts of the Appendix, and in so doing have had regard to the defenders' comments and qualifications. The correctness of the material in parts (iii) and (v) of the pursuers' Appendix was not accepted by the defenders and I have therefore treated it as part of the submission on behalf of the pursuer.

2. Factual background

Cervical cancer

[6] The cervix is the neck or narrow, lower part of the uterus and protrudes into the vagina. It consists of the endocervical canal, which is continuous with the endometrium of the uterus, and the ectocervix which protrudes into and is continuous with the vagina. The surface of the cervix is covered by a mucous membrane known as epithelium. The endocervical canal is lined with a single layer of glandular (or columnar) epithelium. The ectocervix is lined with layers of squamous epithelium. The squamous and glandular epithelia meet at the squamo-columnar junction. The position of the squamo-columnar junction changes during a woman's lifetime. It is initially located at the external os where the endocervical canal opens into the vagina. In young women, especially those using oral contraceptives, the squamo-columnar junction is likely to move outwards towards the vagina; in older women it is more likely to be found further up the endocervical canal. When the squamo-columnar junction is located on the ectocervix, the area of the ectocervix which is lined with glandular epithelium is known as an ectopy. On examination it appears to have a reddish colour due to the visibility of blood vessels through the single layer of glandular epithelium.

[7] The area around the squamo-columnar junction is known as the transition zone or transformation zone. New cells in this zone may develop into either glandular or squamous epithelium, depending upon whether the squamo-columnar junction is in the endocervix or ectocervix. If the squamo-columnar junction is in the ectocervix, glandular cells in the transformation zone change to squamous cells, a process known as squamous metaplasia.

[8] Cervical cancer is caused by human papillomavirus (HPV) infection. Cells which are developing or transforming in the transformation zone are particularly vulnerable to attack by carcinogens, and most cervical cancers begin in this zone. There are two common varieties of cervical cancer. The more common of the two is squamous cell carcinoma which arises in squamous epithelial cells and accounts for around 80-85% of cases. The less common is adenocarcinoma which arises in glandular epithelial cells and accounts for most of the remainder. The most common type of adenocarcinoma is the endocervical cell type, so called because the cells look like endocervical (glandular) epithelium: the name does not necessarily imply that the cancer is located in the endocervix. Cervical cancer is described as invasive when it ceases to be confined to the epithelium and encroaches into the neighbouring tissue known as the stroma.

Cervical screening

[9] Most cases of invasive cervical cancer are preceded by the development of a pre-invasive lesion of the cervix. The pre-invasive stage of squamous cell carcinoma is known as cervical intraepithelial neoplasia (CIN). CIN is a term given to a spectrum of squamous epithelial lesions in the cervix ranging from low grade lesions (CIN1) to high grade lesions (CIN3). CIN1 lesions are characterised by minor changes in the cervical epithelial cells whereas the cellular changes in CIN3 are more severe. The rate of progression of CIN lesions is estimated to be within 5-15 years (mean 8 years). The pre-invasive stage of invasive adenocarcinoma is known as cervical glandular intraepithelial neoplasia (CGIN) or adenocarcinoma in situ (AIS). If left untreated, CGIN can progress to invasive cancer. The mean rate of progression of CGIN to invasive adenocarcinoma was estimated by Professor Coleman as being up to ten years. CIN3 and CGIN may both be found in the same cervix suggesting a common aetiology for the two types of cancer.

[10] Both pre-invasive and invasive cancers can be detected by the Papanicolaou (or "Pap") smear test. The test is used worldwide as a method of screening for pre-invasive or early invasive cervical cancer. In its pre-invasive and early invasive stage cervical cancer is usually silent: i.e. most women will not have any symptoms or signs of the disease, which is why the test is so valuable. The aim of the cervical screening programme is to detect women with pre-invasive disease when treatment is confined to local ablation of the lesion and the risk of recurrence is very low indeed. A cervical screening programme was introduced in the United Kingdom in 1966 in a semi-organised fashion but screening was only carried out every five years. The rate of cervical cancer did not decrease and, following a review in 1988, three-yearly screening was introduced on a systemised basis for all women aged between 20 and 65.

[11] Until the early years of the 21^st century, the conventional laboratory method was to spread the smear on a glass slide. This resulted in a high number samples being categorised as "inadequate", meaning that they were poorly visualised due to contamination by blood and other materials. The test now involves the removal of a sample of epithelial cells from the surface of the cervix using a specialist sampling device which in modern practice is a cervical broom or brush. Because the device collects cell material largely from the ectocervix, it is designed primarily to sample squamous cells, although abnormal glandular changes can also be detected. The sample of cells are prepared by the "thin prep" method as a suspension and sent to a cytology laboratory, where a special automated "thin prep" machine transfers them to a glass slide. They are thereafter stained and examined in the light microscope. The initial microscopy is carried out by a cytotechnologist (primary screener) who is trained to recognise dyskaryotic (abnormal) cells among the many thousands of normal cells in the smear. The primary screener is required to refer the smear to a more experienced cytotechnologist or pathologist for confirmation if he or she detects abnormality suggestive of pre-cancerous (CIN or CGIN) or cancerous changes in the cervix, or if he or she is not absolutely confident that the smear is negative. Use of "thin prep" liquid-based cytology has greatly reduced the number of smears categorised as inadequate due to the presence of blood or inflammatory cells. The screeners and checkers issue reports on negative or inadequate smears. A pathologist issues the report in the event of an abnormal smear.

[12] There are certain generally-recognised reporting categories, although there may be variations from one laboratory to another. At the time when the pursuer had her smears tested by the defenders, the following categories were used by them: 0 unsatisfactory; 1 negative; 2 borderline; 3 mild dyskaryosis; 4 moderate dyskaryosis; 5 severe dyskaryosis; 6 ?invasion; 7 glandular abnormality; 8 adenocarcinoma; 9 other. The category "?invasion" refers to invasion by squamous cell carcinoma. Each category is associated with a recommended management strategy. An inadequate or unsatisfactory smear must be repeated within 3 months. At the material time, women in Scotland who had a smear reported as borderline or mild dyskaryosis (CIN1) for the first time would normally be advised to have a repeat smear in 6 or 12 months (depending upon the practice of the laboratory in question), since it is not uncommon for low grade lesions to regress. Women whose smear was reported to contain moderately or severely dyskaryotic cells (CIN2 or CIN3) or glandular abnormality (or worse) would normally be referred for colposcopy.

[13] Colposcopy is a non-invasive procedure carried out by a gynaecologist or general practitioner trained in the technique. It involves the examination of the cervix under low power magnification. CIN usually stains white (aceto white) when dilute acetic acid is applied to the cervix and fails to stain when Schiller's iodine is applied. Changes in the vascular pattern can be recognised by the colposcopist after the application of acetic acid which provides him or her with information about the exact location, size and severity of the CIN lesion. A biopsy is taken for histological confirmation prior to commencement of any treatment.

Treatment options following diagnosis of pre-invasive or invasive cancer
[14] If colposcopic examination and biopsy confirm the presence of a high grade pre-invasive lesion or of early stage invasive cancer, the next step is likely to be an operation to excise from the cervix that part of the tissue containing the transformation zone where the lesion is most likely to be located. This may be done with a scalpel (cone biopsy) or by laser excision using a fine loop of hot wire or needle through which an electric current runs to cut the tissue in a cone shape (LLETZ). These procedures do not compromise fertility, although they do create some risk of miscarriage or early delivery in future pregnancies. The excised tissue will then be subjected to histological examination to assess the nature and extent of the lesion, including the extent of invasion, if any, of neighbouring tissue. The expectation is that in most cases the whole affected area will have been removed by the cone biopsy or LLETZ, as the case may be, with excellent prognosis.

[15] In cases where examination and biopsy disclose frankly invasive cancer which has not spread beyond the uterus, or where it is apparent from histological examination that conservative measures have failed to remove the whole tumour, the standard treatment has been a radical hysterectomy, i.e. removal of the uterus, cervix and surrounding tissue, usually combined with removal of the pelvic lymph nodes to which the cancer might have spread. Pregnancy is not possible after this operation. In recent years, however, an alternative surgical procedure known as radical trachelectomy has been developed. In this procedure the cervix, the upper vagina and the tissues on either side of the cervix are removed. Because the uterus is preserved, pregnancy remains a possibility and research indicates that around 60% of pregnancies following this treatment have resulted in live births. The procedure was pioneered in France by Dargent in the late 1980s and was available in the United Kingdom in 2006. For a consultant advising a patient on whether radical trachelectomy would be an appropriate treatment, the principal consideration is risk of failure to eradicate the cancer. In a paper first published in 1994 and referred to during the hearing, Dargent noted three risk factors: tumour diameter, capillary-like space involvement and adenocarcinoma, of which the first was regarded as the most important. It was a matter of agreement among the witnesses who expressed a view on this matter at the hearing that radical trachelectomy would not normally be an appropriate treatment for a tumour with a diameter greater than 2cm.

The pursuer's medical history
[16] So far as material to the matters within the scope of the restricted proof, the pursuer's medical history may be summarised as follows. Before May 2005 her medical records are unremarkable. She began taking the oral contraceptive pill Brevinor in 1997. A cervical smear taken on 30 October 2003 was reported as negative, with advice that it should be repeated in three years. On 11 May 2005 the pursuer attended without prior appointment the Sandyford Initiative, a clinic providing family planning and reproductive health services in Sandyford Place, Glasgow. She requested tests for chlamydia and gonorrhoea. She reported significant post-coital bleeding and dyspareunia after intercourse with her most recent sexual partner. She is noted as having declined a blood test due to "acute needle phobia". She returned to the clinic on 6 June 2005 when, on examination, she was found to have a large ectopy and a slightly raised pH level. She reported having had post-coital bleeding on one occasion and also having had a vaginal discharge which was clear with no smell or itch and which was almost clearing up.

[17] On 18 November 2005 the pursuer returned to the Sandyford clinic complaining of abnormal bleeding over the past few weeks while on the Brevinor contraceptive pill. She was noted as being "very anxious ++ re amt of blood being passed". She also complained of a vaginal discharge described as yellow with no odour which had increased during the last two weeks. A cervical smear was taken but was reported as unsatisfactory due to heavy blood staining and inflammatory exudates.

[18] On 15 December 2005 the pursuer attended the colposcopy clinic at the Sandyford Initiative. She was seen by Dr Bigrigg who carried out a colposcopic examination after prescription for the pursuer of certain anti-anxiolytic drugs. The examination disclosed the presence of an ectopy. The pursuer complained of heavy intermenstrual bleeding for 1 week in May and in November and of a constant vaginal discharge. She is noted as having reported that the symptoms were so bad that she could not go to work, give talks, travel, have sex or go to the gym. (I discuss below the pursuer's account of this consultation.) Dr Bigrigg referred the pursuer for an ultrasound scan, noting in the internal request form that the pursuer "may not accept vaginal probe".

[19] The first of the two cervical smears with which this action is concerned was also taken on 15 December 2005. The test result was initially reported as negative with a repeat advised in three years. However, on 6 January 2006, a letter was sent by Dr Stephen to Dr Bankowska which stated as follows:

"In the course of doing an internal audit this smear was reviewed by the Lab Manager. It was originally reported as negative repeat at follow up but on review the smear consists almost entirely of endocervical cells and acute inflammatory cells. There are virtually no squamous cells present. One or two groups of endocervical cells show mild nuclear atypia although there are no definite features of a glandular abnormality. Nevertheless the smear should have been reported as inadequate and not as negative, and we feel this patient should have a repeat smear."

[20] In the meantime, the pursuer had had an ultrasound scan at the Sandyford clinic on 4 January 2006, carried out by Dr Rosie Cochrane. In the report of the scan results, the following entry was made under "Additional Comments":

"No cysts no polyp in uterus but * ?? debris ?? polyp within cervical canal *"

Following the scan, Dr Cochrane sent a note to Dr Bankowska in the following terms:

"This lady was seen at colp 15.12.05 - she has discharge + IMB + ectopy. She's very anxious and was not happy with consultation. She has requested f/u by yourself (you have spoken to her on the phone previously).

Scan today showed normal endometrium but possibly an endocervical polyp ? significance.

I have added her to a free space on your list 19.1.06 and given her a double appt as I think she requires prolonged discussion. I hope this is appropriate."

In fact Dr Bankowska saw the pursuer on 12 January 2006 and discussed treatment options with her for "persistent discharge and ectopy". The pursuer agreed to treatment of the area of ectopy by cold coagulation but at a later date. Her contraceptive pill was changed from Brevinor to Norimin. Dr Stephen's letter of 6 January had not been received by Dr Bankowska in time for this consultation. At about this time the pursuer moved to Geneva to live and work, but returned to Scotland from time to time and continued to receive treatment at the Sandyford clinic.

[21] Dr Bankowska next saw the pursuer on 9 March 2006. She carried out a colposcopic examination, noting the presence of an ectopy and that the position of the squamo-columnar junction was on the ectocervix. She carried out cold coagulation treatment (which is not "cold" but requires the placing of a hot probe within the cervix) of the ectopy without anaesthetic, although the pursuer is noted as having been "very anxious". The second of the cervical smears with which this action is concerned was also taken on 9 March 2006. By letter dictated on 10 April 2006 Dr Bankowska advised the pursuer that the smear had been reported as negative and continued:

"If you are still having symptoms, in terms of discharge and irregular bleeding after 3-4 months, please arrange to come back and see us. Otherwise, if your symptoms are settling, I do not think we need to re-examine you. You should have a further smear test checked in three years time in the normal screening programme (March 2009)."

[22] The pursuer's vaginal discharge continued to be problematic for her and she arranged a further appointment at the Sandyford clinic on 29 June 2006 while home from Geneva. At this consultation the pursuer is noted as having complained of "Very heavy persist vag discharge - soaks through clothes and deterrent to [sexual intercourse]". Once again Dr Bankowska carried out a colposcopic examination. She noted the presence of a small ectopy and again observed the presence of the squamo-columnar junction on the ectocervix. She also noted "Clear discharge minimal". Microscopy disclosed the presence of "clue cells" which may be indicative of bacterial vaginosis ("BV"). The pursuer was advised to stop her contraceptive pill and to have a trial of metronidazole, an antibiotic prescribed to treat BV. She also had a second ultrasound scan on the same date. In the report of the scan results, the following entry was made under "Additional Comments":

"Cervical lining is thick but regular not polyp - quite often seen in normal scans No [illegible word] non-tender"

[23] The pursuer's GP records include an entry for 25 October 2006 which states "Discharge had settled but not fully. Try another course + Dalacin gel. If no better back to GUM [i.e. genito-urinary medicine] clinic." She next attended the Sandyford clinic on 30 November 2006 when she was seen by Dr Brown whose note on the computer records reads as follows:

"Long history noted. Only potential cause for discharge is BV. Has had metronidazole previously and felt that it improved discharge briefly but problems returned. Is now back on Brevinor and bleeding pattern is fine. Long discussion about the problems and how to approach this. Empirical suggestion: metronidazole one week, for three cycles. Happy to try. Asked to keep diary of discharge symptoms."

[24] On 29 May 2007 the pursuer was once again reviewed by Dr Brown. On this occasion the pursuer reported that she had been using empirical metronidazole for what was presumed to be recurrent BV. It was noted that two months previously the pursuer had taken a higher dosage than normal of metronidazole which appeared to improve her symptoms but made her feel nauseated. Dr Brown advised her to try a monthly course of vaginal metronidazole gel rather than oral doses. At a further review by Dr Brown on 19 July 2007, the pursuer reported that she had noticed a symptomatic improvement whilst using metronidazole gel but that the discharge returned after she stopped. The pursuer declined an examination but was reported to be "upset regarding impact on personal life and relationship, or lack of". She was given a further prescription for metronidazole gel to use daily for six weeks before returning for review. There is no record of any further visit by the pursuer to the Sandyford clinic or of her receiving treatment elsewhere for either vaginal bleeding or discharge.

[25] As I noted at the outset of this opinion, the pursuer had a cervical smear test in Geneva on 28 January 2008. It was reported to show squamous and glandular endocervical cells with atypical nuclear features suspicious of CIN3 with an atypical glandular component. Two small biopsies taken on 12 February 2008 were reported to show at least high grade glandular abnormality. The subsequent history of the pursuer's treatment in Switzerland is narrated by Mr Soutter, in his report dated 11 November 2010, in the following terms which I do not understand to be controversial:

• On 5 March 2008 Dr Sabeti performed a loop biopsy of the cervix in three portions. All of these samples were massively infiltrated by an adenocarcinoma. There was also some high grade CIN3. It was thought possible that this might have been a villoglandular carcinoma and so slides were sent to Dr Debra Bell in the Mayo Clinic, USA. It was important to do so because such a variant is associated with a better outcome, and would affect treatment. She concluded that this was a moderate to well differentiated adenocarcinoma of the endocervix. Although the depth of infiltration could not be assessed exactly from the biopsy it seemed likely that it was more than 3mm. She also reported the presence of CIN3.
• Dr Sabeti referred the pursuer to a Dr Megevand. On 18 March 2008 Dr Megevand performed a laparoscopic sampling procedure of the pelvic lymph nodes. None of the three nodes identified contained any tumour. Dr Megevand had a discussion with the pursuer who was not prepared to agree to the idea of a hysterectomy. He explained to her that the conservative procedure of trachelectomy was not recommended for tumours more than 2cm in diameter because of the risk of recurrence. It seemed in her case that the tumour was at least 2cm in diameter.
• On 11 April 2008 Dr Megevand reluctantly undertook a trachelectomy. It became apparent during this procedure that there was an extensive cervical cancer and so the procedure was limited to one intended to be for diagnostic purposes only. The histology of the specimen he removed described a moderately differentiated adenocarcinoma measuring 4.3 x 2.8 x 2.8cm. The tumour extended to the margins of the specimen.
• Dr Megevand explained the situation to the pursuer, advising her that radiotherapy possibly in association with chemotherapy or radical surgery should be considered. However given the size of the tumour additional treatment with radiotherapy would probably also be needed. There was some discussion as to how the ovarian tissue might be preserved.
• On 30 May 2008 Dr Megevand performed a radical hysterectomy using a standard technique. Dr De Candolle assisted by transposing the ovaries high into the paracolic gutters in the hope that they would not be affected by postoperative radiotherapy.

·       The histology of the uterus showed residual moderate to well differentiated adenocarcinoma of the endocervix that infiltrated the isthmus and lower part of the body of the uterus. The closest surgical margin was 1mm. The total tumour volume was estimated by Dr Weintraub to be 29cm³ based upon the previous cone biopsy, trachelectomy and the hysterectomy specimen.

• The pursuer was seen for follow up on 16 June 2008 by Dr Popowski, consultant radiotherapist. He advised that she should have additional radiotherapy to the pelvic lymph nodes and the vaginal vault. Radiotherapy was carried out over several months.

3. Issues of law

Onus of proof

[26] Before turning to the arguments of the parties and my assessment of the evidence, it is necessary to address the issue of onus of proof, as the parties each maintained that the onus lay on the other. For the pursuer it was submitted that the onus lay on the defenders to prove, on balance of probability, that the outcome of hysterectomy would have been the same even if the negligence had not occurred. The defenders had a system designed to protect participants such as the pursuer from the development of cancer. Where such a system fails, as it did in the pursuer's case, there is at least a prima facie presumption of causation. In support of this proposition, reference was made to McGhee v National Coal Board 1973 SC (HL) 37 and Fairchild v Glenhaven Funeral Services Ltd [2003] 1 AC 32. Counsel founded in particular upon the following passage from the speech of Lord Wilberforce in McGhee at page 55:

"First, it is a sound principle that where a person has, by breach of a duty of care, created a risk, and injury occurs within the area of that risk, the loss should be borne by him unless he shows that it had some other cause. Secondly, from the evidential point of view, one may ask, why should a man who is able to show that his employer should have taken certain precautions, because without them there is a risk, or an added risk, of injury or disease, and who in fact sustains exactly that injury or disease, have to assume the burden of proving more: namely, that it was the addition to the risk, caused by the breach of duty, which caused or materially contributed to the injury? In many cases, of which the present is typical, this is impossible to prove, just because honest medical opinion cannot segregate the causes of an illness between compound causes. And if one asks which of the parties, the workman or the employers should suffer from this inherent evidential difficulty, the answer as a matter in policy or justice should be that it is the creator of the risk who, ex hypothesi must be taken to have foreseen the possibility of damage, who should bear its consequences."

This approach was not restricted to employer's liability. Authorities such as Chester v Afshar [2005] 1 AC 134 demonstrated that strict rules of "but for" causation should not be applied where the defenders' duty would be rendered devoid of content by expecting proof of the impossible. In the present case it was unrealistic to impose an onus on the pursuer to prove the impossible, i.e. that at the time when the defenders' negligent acts occurred she did not have a tumour of a size for which radical hysterectomy was the only appropriate treatment.

[27] For the defenders it was submitted that even where, as in this case, breach of duty was not in dispute, the onus lay upon the pursuer to prove that the defenders' reporting failures had caused her appreciable disadvantage that she would not, in any event, have suffered. For a summary of the current law, counsel referred to the following excerpt from Jackson & Powell, Professional Liability (6^th ed, 2007, with Third Supplement, 2010) at para 13-112:

"Whether the claim is brought in contract or tort, it is first necessary to determine whether the medical practitioner's breach of duty caused the injury complained of. The burden of proof of causation is as a matter of principle always on the claimant. Causation can be a simple matter of fact or a combined matter of fact and law. This leads to the need to consider different 'tests' or rules for causation of the injury complained of, namely 'but for' causation (also referred to as 'orthodox' or 'threshold' causation), material contribution, material increase in risk. The latter two tests might equally be considered to be modifications of the 'but for' test, which are applied when certain circumstances arise. Those cases which might be described as exceptions to the usual need to prove causation (such as the House of Lords' decisions in Fairchild v Glenhaven Funeral Services and Chester v Afshar) are simply instances where a new rule on causation was applied or the orthodox rules were modified. These cases are not an indication of a general relaxation in the requirement on a claimant to prove causation. They are, and were always intended to be, of limited and exceptional application. Subsequent decisions have confirmed that this is correct."

It was submitted that all of the cases relied upon by the pursuer had special circumstances, none of which applied in the present case. In McGhee, the issue was whether the defenders' breach of statutory duty in failing to provide adequate washing facilities had been the cause of the pursuer's dermatitis and it was held that a material increase in risk was sufficient to establish liability. Fairchild was a case dealing with the particular situation where the claimant had suffered injury from asbestos exposure and it could not be established scientifically during which of two or more employments the injurious exposure had occurred. In these circumstances it was again held that a material increase in risk was sufficient to satisfy the causal requirements for liability. Chester v Afshar was also an extension of normal causation principles in a "failure to warn" case, which this was not. The present case was more akin to Hotson v East Berkshire Area Health Authority
[1987] AC 750, discussed by Jackson & Powell at para 13-117, in which the plaintiff's claim failed on application of ordinary principles of causation. Here the pursuer required, according to normal and well-established rules to satisfy the court as to what would have happened had the correct report of her smear test been made in January 2006 or March 2006, as the case may be, so that this could be contrasted with what is known to have happened.

[28] In my opinion the submissions of the defenders on this point are to be preferred. It does appear to me to be clear from the speeches delivered in Fairchild that the cases founded upon by the pursuer are exceptional ones in which policy considerations have been held to necessitate departure from what Lord Bingham of Cornhill described (para 35) as "the ordinary approach to proof of causation". Lord Nicholls of Birkenhead warned at para 43 of the need for restraint in any relaxation of the threshold "but for" test of causal connection, and noted that policy questions would loom large when a court has to decide whether the difficulties of proof confronting a claimant justify taking this exceptional course. The present case is not one concerning material increase in risk, material contribution or failure to warn, and does not therefore appear to me to fall within the circumstances in which the traditional approach to causation has been varied. It would, in my view, be inappropriate for a judge sitting at first instance to extend an exception founded upon policy considerations to circumstances in which it has not previously been applied.

[29] In any event I consider that the cases founded upon by the pursuer are distinguishable. In both McGhee and Fairchild, the contracting of the disease by the claimant was caused by the defender's breach of duty. Similarly, in Wilsher v Essex Area Health Authority
[1988] AC 1074, a medical negligence case, the act which the claimant alleged (unsuccessfully) had caused his blindness, namely the administration of excess oxygen, was something done by an employee of the defendant health authority. In the present case it is not contended that the pursuer contracted a glandular abnormality capable of developing into a cancerous tumour because of any act or omission by the defenders. As I have noted, cervical cancer originates from the attack of cells in the cervix by the HPV virus, and it is not suggested that the defenders were responsible for this in the pursuer's case. The claim is rather that the defenders were negligent in failing to report the presence of a cancerous or pre-cancerous lesion at a time when conservative treatment would have been appropriate. In these circumstances it seems to me to be quite clear that the onus rests upon the pursuer to prove that such treatment would indeed have been appropriate had the problem been identified and reported when it ought to have been. I consider that the present case is closer in principle to Hotson v East Berkshire Area Health Authority. In that case the plaintiff, who was aged 13, fell from a tree and sustained a fracture of the left femoral epiphysis. His injury was not correctly diagnosed or treated for five days and ultimately he developed avascular necrosis of the epiphysis involving disability of the hip joint. After trial, the judge (Simon Brown J) held that there was a 75% probability that avascular necrosis would have developed even if the claimant had received prompt diagnosis and treatment. His award of damages for the loss of a 25% chance of recovery was overturned by the House of Lords, but for present purposes it is sufficient to note the following observation of Lord Ackner at 793:

"...I have sought to stress that this case was a relatively simple case concerned with the proof of causation, upon which the plaintiff failed, because he was unable to prove, on the balance of probabilities, that his deformed hip was caused by the defendants' breach of duty in delaying over a period of five days a proper diagnosis and treatment."

The pursuer's claim in the present case is focused in her averment at page 48C of the Record that

"Had the pursuer's cervical cancer been diagnosed at an earlier stage, the losses hereinbefore condescended upon would not have been incurred..."

In my opinion this is a case in which the traditional approach to causation falls to be applied. The onus rests, in my view, upon the pursuer to prove, on balance of probability, the averment which I have just quoted as an essential ingredient of her claim.

Legal proof and scientific proof

[30] A related issue that was raised in submissions is the correct approach to the concept of "proof" in a case such as this which depends to a material extent upon expert scientific evidence. I was reminded by counsel for the pursuer of, and have attempted to bear in mind while formulating my opinion, the following passage from the opinion of Lord Prosser in Dingley v Chief Constable, Strathclyde Police 1998 SC 548 at 603:

"In ordinary (non-lawyers') language, to say that one regards something as 'probable' is by no means to say that one regards it as 'established' or 'proved'. Yet in the civil courts, where we say that a pursuer must prove his case on a balance of probabilities, what is held to be probable is treated as 'proved'. I do not suggest that any lawyer will be confused by this rather special meaning of the word 'proved'. But speaking very generally, I think that the civil requirement of a pursuer - that he satisfy the court that upon the evidence his case is probably sound - would in ordinary language be regarded as very different from, and less stringent than, a requirement that his case be established or proved. More importantly in the context of a case such as the present, the fact that the two concepts are distinct in ordinary language, but the same in this legal context, seems to me to give rise to a risk of ambiguity or misunderstanding in the expressed opinions of expert witnesses. And this risk will be increased if the expert in question would normally, in the exercise of his profession, adopt an approach to such issues starkly different from that incumbent upon a court. Whether one uses the word 'scientific' or not, no hypothesis or proposition would be seen as 'proved' or 'established' by anyone with any form of medical expertise merely upon the basis that he had come to regard it as probably sound. (Indeed, I think even the word 'probable' would be reserved for situations where the likelihood is thought to be much more than marginal). And even if, in relation to any possible proposition or hypothesis, such an expert even troubled to notice that he had come to the point of regarding it as not merely possible but on balance 'probable', then I think he would regard that point as one from which he must set off on further enquiry, and by no means as being (as it is in the courts) a point of arrival. Mere marginal probability will not much interest him. But it must satisfy a court."

The decision of the First Division in Dingley was affirmed by the House of Lords (2000 SC (HL) 77). In the course of delivering a speech with which the other members of the Judicial Committee agreed, Lord Hope of Craighead made the following observations at page 89:

"...It is right to bear in mind that there is an important difference between the exacting standards of thought and analysis which the academic will expect of medical scientists and the task of a judge when he is considering whether the essential elements in a pursuer's case have been established on a balance of probabilities.

This is not to say that the judge need not examine the detail of the evidence. Of course he must, as his task is to identify the real issues in the case and then determine where the balance lies between the competing positions revealed by the evidence on each side. In a case such as this, an important part of his task is to assimilate and understand the oral and written evidence and to penetrate the arguments which have been developed by the expert witnesses. But when it comes to the point of exercising his judgment on these issues, he must be careful to avoid applying the standard of proof which the expert would apply to them. As Lord President Cooper said in Davie v Magistrates of Edinburgh [1953 SC 34], at p 40:

'Expert witnesses, however skilled or eminent, can give no more than evidence. They cannot usurp the functions of the jury or judge sitting as a jury, any more than a technical assessor can substitute his advice for the judgment of the court.'

The function of the judge in a civil case is to decide where the truth lies or whether the case has been made out, on a balance of probabilities. One cannot entirely discount the risk that, by immersing himself in every detail and by looking deeply into the minds of the experts, a judge may be seduced into a position where he applies to the expert evidence the standards which the expert himself will apply to the question whether a particular thesis has been proved or disproved - instead of assessing, as a judge must do, where the balance of probabilities lies on a review of the whole of the evidence."

[31] These observations appear to me to be relevant to any complaint by the pursuer in the present case that to impose the onus of proof upon her would be to expect her to prove the impossible. According to my understanding of the dicta which I have quoted from Dingley, it is not my task in the present case to decide whether I am satisfied, to the standard which a scientist would apply in assessing whether something has been proved, the location, nature or size of whatever lesion or tumour was present in the pursuer's cervix in December 2005 and March 2006. My task is, rather, to determine, on balance of probabilities, after review of the whole of the evidence, whether the lesion or tumour present at that time was such that, had it been correctly identified and reported, the pursuer would have undergone appropriate treatment that was more conservative than the radical hysterectomy which was found necessary in 2008. Clearly that is not a straightforward issue, but discharge of the onus of proving it to the civil standard seems to me to be the type of task commonly faced by pursuers in medical negligence claims.

4. Assessment of the evidence: interpretation and reporting of cervical smears
Introduction
[32] In paragraph 3 above, I set out the questions which provided the framework for the restricted proof. As it turned out, the issues raised by questions 2 and 4 were more controversial, and occupied considerably more time during the proof, than those raised by questions 1 and 3. For reasons which will become apparent, my conclusions in relation to those aspects of questions 1 and 3 which remained in dispute by the end of the proof are not critical to my overall determination on causation. I therefore find it convenient to deal with questions 1 and 3 first, before turning to consider the evidence relating to the various lines of argument which the parties developed with regard to questions 2 and 4.

Interpretation and reporting of smear taken on 15 December 2005

[33] It will be recalled (see paragraph 19 above) that the report on this sample that was ultimately made by the defenders to Dr Bankowska at the Sandyford clinic, following an internal audit, was as set out in Dr Stephen's letter of 6 January 2006, namely that the smear should have been reported as inadequate and not as negative, with a recommendation that the pursuer have a repeat smear. The cervical cytopathology slides for the various samples, including that taken on 15 December 2005, were made available to and examined by expert cytology witnesses for the pursuer (Professor Coleman and Dr Dina) and for the defenders (Dr Duvall). Dr Dina stated in his report that he had disseminated the slides to the cytoscreeners in his laboratory. He noted that "all the screeners thought that this slide was diagnostic of ?glandular neoplasia" and expressed the view that "In essence this is a slide containing atypical cells of glandular type and request for urgent referral to colposcopy should have been issued". In oral evidence Dr Dina explained that "glandular neoplasia" is a general term used to describe a tumour arising from glandular cells when it is not known whether the tumour is pre-invasive or invasive. On receipt of such a report, the colposcopist would, in accordance with national guidelines, carry out an examination and then proceed to biopsy or, if colposcopy was negative, bring the case to a multi-disciplinary meeting. For his part, Dr Duvall noted changes which amounted "to what are, at least, changes bordering on dyskaryosis in endocervical glandular cells". His opinion was that the smear should have been reported as showing borderline changes in endocervical glandular cells with very occasional squamous cells showing borderline nuclear changes. The appropriate recommendation would have been a repeat smear (as was in fact recommended, for different reasons). For the sake of completeness it should be noted that Dr Coleman noted a pattern "suggestive of CIN3" (i.e. of squamous rather than glandular cell abnormality) and considered that the correct report would have been "Numerous severely dyskariotic cells present suggestive of CIN3. Urgent referral for gynaecological opinion and colposcopy advised".

[34] It is therefore a matter of agreement among the three expert witnesses that the slide was misinterpreted to some extent. With the benefit of hindsight it would appear that the view of Dr Dina and Dr Duvall that the sample displayed abnormality of glandular cells should be preferred to that of Professor Coleman. In her evidence at the proof, Dr Stephen disagreed with Professor Coleman's report (which she had seen) and also with the view of Dr Dina that there was glandular abnormality suggestive of adenocarcinoma. She regarded Dr Duvall's assessment as reasonable. The significant difference between the views of Dr Dina and Dr Duvall lies in what they regarded as the appropriate recommendation: Dr Dina envisaged an urgent referral for colposcopy whereas Dr Duvall would have recommended a repeat smear which, depending on the practice of the laboratory, would have been carried out in either 6 or 12 months. My impression was that Dr Duvall was exercising great care not to be influenced by hindsight. He fairly accepted that there was room for debate on interpretation of this slide and that the decision on the appropriate course of action would be taken by the gynaecologist to whom the report was made rather than by the reporting cytologist. That being so, it is of importance that the evidence of Dr Bankowska (to whom the report would have been made) was that it would have made little difference to her whether she had received a report in the terms favoured by Dr Dina or by Dr Duvall: in either case, she would have regarded a biopsy as necessary. As the pursuer had shown herself to be a difficult patient with regard to examination and sampling, Dr Bankowska considered it likely that following discussion with the pursuer she would have referred her to a gynaecological oncologist for follow-up, which would have included biopsy. Other witnesses who gave evidence on this point (Mr Naik and Dr Farquharson) agreed that investigation by colposcopy and biopsy would have been appropriate where a glandular abnormality was reported; Mr Naik regarded even a "borderline" report as extremely worrying. In the end, therefore, it is unnecessary for me to make a choice between the opinions expressed by Dr Dina and Dr Duvall on this point: I find that the outcome would have been likely to be the same in each case, and different from what actually happened.

Interpretation and reporting of smear taken on 9 March 2006

[35] As previously noted (paragraph 21 above), this smear was reported as negative. There is no material disagreement between the parties with regard to the correct interpretation of this slide and the recommendation which ought to have been made. Dr Dina considered that a report of "?glandular neoplasia" should have been issued with a recommendation for urgent referral to colposcopy. Dr Duvall considered that the smear showed the presence of groups of cells clearly showing endocervical glandular abnormality and that the appropriate recommended management was referral for colposcopy. (Professor Coleman again regarded the pattern of abnormality as suggestive of CIN3; with the benefit of hindsight it again appears that the view of Dr Dina and Dr Duvall that there was glandular abnormality should be preferred.) On the evidence of Dr Bankowska already mentioned, I am satisfied that a report in March 2006 of glandular abnormality would have resulted in referral of the pursuer for the carrying out of a biopsy.

Conclusion

[36] I therefore find that the consequence in terms of further investigation of the correct interpretation and reporting of the 15 December 2005 smear and of the 9 March 2006 smear would have been the same: on balance of probabilities, the abnormal test result would have been reported as a matter of urgency by Dr Bankowska to the pursuer, who would have been referred to a gynaecological oncologist for further investigation which may or may not have included colposcopy but would have included biopsy. I also find, in the light of my reasoning set out below in relation to the other issues canvassed at the proof, that nothing turns on whether such further investigation would have followed upon the reporting of the 15 December 2005 smear or of the 9 March 2006 smear: the outcome, so far as material to these proceedings, would have been the same. It is common ground that if colposcopy and biopsy had been carried out in about either January or April 2006, the pursuer would have been found to have a glandular abnormality that required urgent treatment.

5. Assessment of the evidence: appropriate medical management
Introduction

[37] I now turn to consider the issues raised in questions 2 and 4 of the framework for the restricted proof. In order to make a finding as to what would have been the appropriate medical management of the pursuer in January or March 2006, as the case may be, it is necessary for me first to make a finding as to the stage of progression of the lesion or tumour at that time. This was one of the most controversial issues at the proof. In this chapter of my opinion I begin by setting out briefly the arguments of the parties with regard to this issue, in order to give context to the discussion which follows. I then address in turn the various lines of evidence upon which reliance was placed by one or other of the parties.

Argument for the pursuer

[38] For the pursuer it was not accepted that by early 2006 the pursuer had a tumour that had passed from the pre-invasive to the invasive stage. Although Professor Coleman had stated in her report that it was highly likely that the pursuer's cancer would have been at the pre-invasive stage in March 2006, her position in oral evidence appeared to be that although there was no definite evidence of invasion, it was not possible on the basis of the cytology alone to express a view one way or the other. Dr Dina would not have expected a cytologist's report to express an opinion as between pre-invasive and invasive cancer. Dr Duvall similarly did not consider that one could tell from cytology alone. On the whole evidence, Mr Soutter considered that this was a very small tumour which would probably have been reported as AIS by a pathologist, but the size was more important than the description. Even if the tumour had become invasive, it was submitted on behalf of the pursuer that the evidence demonstrated that it did not, at either of the critical dates in 2006, have a diameter of 2 cm or greater. The pursuer's case was based to a large extent upon the evidence of Mr Soutter and had two principal strands. First, Mr Soutter had estimated the size of the pursuer's tumour in early 2006 by using (i) information from the pursuer's Swiss medical records concerning the estimated size of the tumour when it was diagnosed and removed in 2008 and (ii) published data regarding the growth rates of tumours of various types. This latter information was contained in a monograph by Gordon Steel entitled "Growth Kinetics of Tumours" (to which I shall refer as "Steel") published in 1977 and collating data from a number of published papers containing research into the volume doubling times of primary and metastatic tumours. Although published over 30 years ago, Steel had stood the test of time and was regarded as a reliable source of information. Using the data in Steel, and using a geometric mean doubling time of 77 days, the volume of the tumour in mid-January 2006 could be calculated to be approximately 14mm³. Even if large margins of error were built into the calculation, the result produced was still a tumour whose diameter was measured in millimetres, and certainly not one with a diameter greater than the 2cm which witnesses agreed was normally the upper limit for performance of a radical trachelectomy. The figures derived from Steel were cross-checked by Mr Soutter using data from a separate study by Jacobs in 1986 of women who had died as a consequence of squamous cell carcinoma of the cervix. The doubling times derived were similar to those measured in lung metastases according to Steel, and supported the estimate based on those measurements.

[39] Secondly, the size of the tumour could be estimated by reference to the presence - or absence - of relevant symptoms at the material time. It was recognised that both post-coital bleeding and vaginal discharge could be symptoms of invasive or perhaps even pre-invasive disease. However, in the pursuer's case, there had only been one episode of post-coital bleeding and a further episode of inter-menstrual bleeding in November 2005. It was not, therefore, a persistent problem and the single episode of post-coital bleeding could be explained by a mechanical cause. The most likely explanation for the persistence of the pursuer's clear, watery vaginal discharge was either bacterial vaginosis or an increased physiological discharge of unknown aetiology. It was extremely unlikely that either the bleeding or the discharge was caused by the presence of a cancerous or pre-cancerous tumour. The ultrasound scans carried out in January and June 2006 disclosed nothing indicative of the presence of a cancerous abnormality. Most significantly, no invasive cancer was detected by any of the three colposcopic examinations carried out in December 2005 and in March and June 2006. On each of these occasions the squamo-columnar junction was clearly visible. The vast majority of tumours originate in the transformation zone around the squamo-columnar junction. If a tumour of significant size had been present in 2005 and 2006, it is therefore likely that it would have been detected by colposcopy. The appropriate inference to be drawn was that the tumour present at that time was not large enough to be seen clearly and was of a size of the order calculated by Mr Soutter.

Argument for the defenders

[40] For the defenders it was submitted that investigation by colposcopy and biopsy carried out within a few weeks after receipt of a correct smear test report of glandular abnormality would have disclosed that the pursuer had an invasive and infiltrative adenocarcinoma of the cervix, located within the endocervical canal and extending at least 2cm along at least one dimension. The evidence for this conclusion lay in the contemporaneous observations and in the pursuer's symptoms. The pursuer's smears, correctly interpreted, were consistent with the pursuer having invasive adenocarcinoma from November 2005. Histology indicated that the adenocarcinoma would be of an infiltrative type and would be likely to grow up the endocervix in a tube-like shape. The ultrasound scan reports indicated that there was something unusual about the pursuer's cervical canal in January - June 2006. A tumour in the endocervix would not be visible on colposcopy. In any event the colposcopists who examined the pursuer in 2005 and 2006 had not been alerted by smear test results to the presence of an abnormality and may therefore have failed to notice the tumour, or mistaken it for an ectopy, even if it was visible. With the benefit of hindsight, the pursuer's symptoms of vaginal bleeding and discharge can be seen as indicative of the presence of a well-established cancer. They emerged in about 2005 and were regarded by the pursuer as severe. It was more likely that a group or pattern of symptoms had a single cause than that there were different causes for separate symptoms. A tumour causing symptoms, certainly of bleeding, is likely to have a diameter of at least 3cm and may be much larger. By 2008 it was moderately large, of a size more often seen in elderly patients or in those who have not availed themselves of the screening programme. There was evidence which indicated that this was an indolent rather than an aggressive tumour which would have grown more slowly than as calculated by Mr Soutter.

[41] So far as Mr Soutter's calculations were concerned, the defenders' objection was not to the substance of Steel but rather to the use made of it by Mr Soutter. For various reasons, the conclusions which he drew from the figures in Steel were not legitimate. None of the metastases in Steel's table of figures was from a primary cervical carcinoma. In any event metastases were likely to behave differently from primary tumours and in particular were likely to be more aggressive, especially in a favourable environment such as the lung. Figures published elsewhere in a textbook by Hall (2000) appeared to show a doubling time for primary adenocarcinomas of 166 days, i.e. more than double the figure of 77 days used by Mr Soutter. They also indicated that metastatic adenocarcinomas grow faster than primary tumours. Steel made the point that tumours were unpredictable and his own figures demonstrated that there was a wide range of doubling times. No allowance had been made for the indolent nature of this particular tumour. There were therefore too many scientific weaknesses for Mr Soutter's estimate to be regarded as reliable. Nor did the Jacobs paper afford a scientific cross-check. For present purposes, this was an inappropriate group of cases which ex hypothesi had had the most intractable and aggressive cancers. In any event Mr Soutter's use of the data was based upon arbitrary assumptions as to the size of tumours at various times. No allowance was made for the possibility that tumour growth actually accelerates after unsuccessful radiotherapy treatment. For these reasons the pursuer's estimate of tumour size in 2006 should be rejected.

Estimation of tumour size by use of volume doubling time

[42] In order to attempt to assess the reliability of Mr Soutter's calculation of the growth rate of the pursuer's tumour, it is necessary to look at Steel in more detail. Mr Soutter made use principally of a table (Table 1.2, at pages 48-49) summarising data from published measurements of 780 tumours. Primary tumours had been studied in four anatomical sites: lung, breast, colon-rectum and bone. Metastatic tumours had been studied in the lung, lymph nodes and in superficial sites. Lung tumours were subdivided by histological type (i.e. squamous cell carcinoma, adenocarcinoma or other) and metastases by the tissue of origin. It is apparent from the table that different types of tumour have widely differing mean volume doubling times. For example, primary bronchial adenocarcinoma has a mean volume doubling time of 148 days (with 95% confidence limits of 121-181), whereas the figure for lung metastases of all adenocarcinoma within the data - the figure used by Mr Soutter in his calculations - is 77 days (with 95% confidence limits of 69-87). (By far the longest doubling time is a time of 632 days for primary adenocarcinoma of colon and rectum. Doubts are, however, expressed in the Steel paper regarding the possibility of errors in measurement and I accept Mr Soutter's opinion that this figure should not be treated as reliable.) There are few instances in the Steel figures where the doubling time of a metastatic adenocarcinoma of a particular primary location can be compared with that of a primary adenocarcinoma of the same location. Mr Soutter did, however, place weight on one instance where this can be done to some extent: the figure for primary breast carcinoma is 95 (with 95% confidence limits of 68-134) and the figure for metastases observed in the lung of breast adenocarcinoma is 73 (with 95% confidence limits of 55-97). Mr Soutter noted that Steel (at page 51) did not consider the difference between these figures to be statistically significant. It did not support the proposition that the doubling time of metastatic tumours was faster than that of primary tumours.

[43] In order to counter the conclusions drawn by Mr Soutter from Steel, he was referred by counsel for the defenders to a table appearing in a 2000 publication by Hall entitled Radiobiology for the Radiologist (which had in turn been referenced in a 2003 article by Wyatt et al). This table shows an average doubling time of 166 days for primary adenocarcinoma compared to 82 days for lung metastases of adenocarcinoma. The point arose in cross-examination shortly before a weekend adjournment. When Mr Soutter returned to the witness box he indicated that having had an opportunity to consider the figures in the Hall publication he was of the view that the conclusion drawn from them that lung metastases grew significantly more quickly than primary tumours was unfounded. He noted that the source of the figures in the table in Hall was a 1971 paper by Charbit et al which was also among those used by Steel. Perusal of the Charbit paper disclosed that whereas the primary adenocarcinomas included in the analysis were all either bronchial or breast cancer, none of the lung metastases was of a bronchial adenocarcinoma. As primary bronchial adenocarcinoma is known from the Steel data to be slow-growing, one was not comparing like with like: a probable explanation for the difference that was noted in the Charbit paper itself. It appears to me that Mr Soutter's criticism of the basis upon which a conclusion was drawn by Hall is well founded, and I place no weight on that conclusion. Counsel for the defenders suggested in his submissions that Mr Soutter's keenness to assemble an argument against the Hall figures when they were challenged demonstrated that he had a personal interest in defending his own system based on Steel which had previously been presented in court in England and which he might wish to use again in future. I reject that submission. In my view it was entirely reasonable for Mr Soutter to investigate a challenge to his methodology and I would have found it much less helpful if he had not been able to offer his view in response to the challenge.

[44] I return, therefore, to my assessment of the use made by Mr Soutter of the Steel data to estimate the growth rate of the pursuer's tumour. Having considered the matter carefully, I am not persuaded that the figures in the Steel table can legitimately be used to support a conclusion that the pursuer's tumour would have been approximately 14mm³ in volume in mid-January 2006 and approximately 28mm³ in volume in early April 2006, or even within the range which Mr Soutter was willing to concede of up to four times greater or less than those figures. In the first place, it appears to me that the criticisms which Mr Soutter made of use of the Charbit data by Hall can be made of his own use of Steel. As the defenders submitted, the doubling time of 77 days which Mr Soutter has used is for lung metastases of adenocarcinoma of locations throughout the body, with no positive indication that any of them was from a primary tumour in the cervix. Given the variety of mean doubling times disclosed for lung metastases of adenocarcinoma of different locations (ranging from 33 to 94), it does not appear to me that any reliable conclusion can be drawn from these figures regarding the mean doubling time of lung metastases of cervical adenocarcinoma.

[45] In the second place, I am not satisfied that use of a geometrical mean gives a result in which any confidence can be placed for the purposes of the issue arising in the present case. After the conclusion of the proof, I was provided with a copy of the whole of chapter 1 of Steel. The following passage concerning the mean doubling time of lung metastases of adenocarcinomas appears at page 47:

"...This example illustrates the very considerable range of growth rates that is found within any particular tumour type. The sample of 159 adenocarcinomas has a geometrical mean doubling time of 77.8 days, but 36 per cent of the tumours have a doubling time outside the range 37-163 days (± 1 s.d.) and 4 per cent of tumours lie outside the range 18-343 days (± 2 s.d.). For the benefit of those who are not familiar with statistics, it should be pointed out that this great range is not inconsistent with the claim that there are significant differences between some of the groups of tumours that are listed in Table 1.2, nor with the rather narrow confidence limits indicated in Fig. 1.19. The confidence limits indicate the range of doubling times that has a 95 per cent probability of containing the mean."

I refer to this passage merely because it contains factual information regarding the range of doubling times which is not found in Table 1.2. In the context of the present case, it seems to me (in agreement with an observation made by Mr Naik in the course of his evidence) that the range is of greater significance than either the mean or the 95% confidence interval. The range of growth rates underlying the mean which was used by Mr Soutter in his calculation appears to me to underline the point made in the Steel monograph that unpredictability is often the most noticeable characteristic of primary growths. Rates of growth vary from one patient to another and may also vary over time with regard to the same patient. In the case of the pursuer, the defenders' expert witnesses, Dr Williams, Dr Farquharson and Mr Naik, were of the opinion that this was a slow-growing, indolent tumour. This view was based upon histological and other investigations carried out at the time of removal of the tumour, when it was described as moderate- to well-differentiated, with no involvement of pelvic lymph nodes or lymphovascular space invasion. These features, it was said, demonstrated that the tumour was not aggressive and therefore likely to be slow growing. In contrast, Mr Soutter's view was that characterisation of a tumour as aggressive or indolent was not related to growth rate (but was related rather to propensity to metastasise) and that there was no evidence that the pursuer's tumour was growing more quickly or more slowly than average. I do not find it necessary to choose between these points of view because it seems to me that even if Mr Soutter's view were to be preferred, given the wide range of growth rates disclosed by Steel, absence of evidence as to the growth rate of the pursuer's tumour is not a sound reason for adopting the mean doubling time as the growth rate likely to have occurred in her case.

[46] In the third place, even if Steel does not demonstrate that metastases grow faster than primary tumours, I am not satisfied that there is anything in the paper which supports the converse proposition, i.e. that they may be presumed to grow at a similar rate. Steel's own conclusion on this point is cautious. At page 51, conclusion (vi), it is noted (as I have already observed) that the difference between primary and lung metastatic tumours of breast carcinoma is "not significant", but it is also noted that one study indicates that superficial soft-tissue metastases from breast carcinoma grow faster than either lung metastases or primary tumours. No further conclusion is stated by Steel regarding comparison of primary tumours and metastases.

[47] For all of these reasons, I am not persuaded that Mr Soutter's conclusions based upon the data in Steel, and in particular upon the figure therein for mean doubling time of lung metastases of all adenocarcinomas, afford a reliable estimate of the size of the pursuer's tumour in early 2006, even within the broader range of possibilities which Mr Soutter was willing to concede. Nor, in my opinion, does the Jacobs paper provide a cross-check of any real value. Although an excerpt from Jacobs was among the productions, it was not specifically referred to in the course of the proof. It is not directly concerned with tumour growth rate but Mr Soutter explained that his purpose in referring to it was to show that cervical tumours can grow quickly. He required to make certain assumptions regarding the rate of progression of the tumour between the time when it was confined to the cervix and the massive involvement of vital organs. Having done so, he increased the safety margin of his calculation by reducing the growth rate by two doublings and making no allowance for deceleration in growth. This produced a range of possible average doubling times of 5 to 10 weeks (35 -70 days), which was similar to the measurements in Steel for lung metastases. It respectfully appears to me that there are too many assumptions made in Mr Soutter's calculation for it to be regarded as affording a cross-check with scientific credibility. He accepted that he had used round figures which seemed appropriate on the basis of his clinical experience. Once again, it seems to me that the use of averages tells us nothing upon which reliance can be placed in assessing the growth rate of the pursuer's tumour in particular. I also consider that there is force in the defenders' criticism that the "cross-check" uses data from a group of women with those cancers which proved to be the most aggressive (according to Mr Soutter's usage of the word) and that this ought not to be relied upon as affording a sound basis for calculation of growth rates of cervical carcinoma more generally.

[48] I accept that there may come a point when it can be said in a particular case, at least on balance of probabilities, that use of doubling times demonstrates that a tumour would not have been greater than a particular size at a given time in the past. In the course of his cross-examination of Mr Naik, counsel for the pursuer put it to him that if one assumed for ease of arithmetic that the volume of the pursuer's tumour in 2008 was 32cm³ (instead of the 29cm³ calculated by Dr Weintraub) and that the volume doubling time was 200 days (i.e. a figure well beyond the 95% confidence limits for any of the tumours in the Steel table other than the unreliable colon-rectum adenocarcinoma figure), then after allowing for the four doublings which would have occurred since January 2006 the volume of the tumour at that time would have been 2cm³, giving a diameter - according to counsel's calculation, and bearing in mind that the tumour was likely to be ellipsoid and not spherical in shape - of 1.56cm. On the basis of this calculation Mr Naik was content to accept that the diameter of the tumour in 2006 was less than 2cm. It will be recalled that the significance of a diameter of 2cm or less is that according to the evidence treatment by trachelectomy instead of radical hysterectomy was not ruled out. For my part, given the reservations that I have expressed regarding the use of the Steel figures to estimate the doubling time of the pursuer's cervical adenocarcinoma, I am reluctant to place any weight even upon a calculation which, on the face of it, appears to use figures at the extremity of likely doubling times. In the end, however, this is of little practical significance because, for the reasons given below, I have not found it necessary to base my decision upon the likelihood of the pursuer having been appropriately treated by trachelectomy in 2006.

Symptoms - general observations

[49] The expert witnesses for both parties placed considerable weight upon symptoms displayed by the pursuer during 2005 and 2006, but for opposite reasons. The defenders' witnesses considered the pursuer's symptoms, when taken together with reports of investigation of them, to indicate the presence of an invasive tumour of significant size. Mr Soutter, who was the principal expert witness called by the pursuer on this matter, considered that the symptoms which the pursuer reported were not indicative of the presence of a tumour and that the absence of indicative symptoms or of findings on examination suggestive of the presence of invasive carcinoma was an indication that no such tumour was then present. Since there is a clear conflict between the expert witnesses on these matters, it is necessary for me to explain the basis upon which I have decided which evidence to prefer. Additionally, since the pursuer gave evidence that was not in all respects consistent with her medical records, it is appropriate also for me to make findings in relation to her credibility and reliability.

[50] Both parties in their submissions attacked the impartiality of the experts led by the other side. For the defenders, it was argued that Mr Soutter had developed a "tag team" arrangement with Professor Coleman for cases of this type, and had an interest to maintain and defend the system which he had developed for assessing the presence and size of cancerous lesions. It was readily apparent to me that Mr Soutter is an experienced witness who is well accustomed to participating in medical negligence claims. While bearing this in mind, I found him to be an impressive and reliable witness. His expertise in the field in which he gave evidence is beyond question. Although I have rejected his argument based on the Steel data, I find no reason to conclude that it was maintained out of self-interest. His scrutiny of the pursuer's medical records had been thorough and he was willing to correct himself on at least one occasion (concerning the recording by the colposcopists of the location in the pursuer's cervix of the squamo-columnar junction) when this was necessary. He was willing to give careful consideration to hypotheses put to him in cross-examination and was able, as a consequence of prior preparation, to provide a reasoned response to them. For these reasons I regard him as a witness in whose opinion on clinical matters I can place confidence.

[51] For the pursuer it was submitted that Mr Naik, Dr Williams and Dr Farquharson gave an appearance of bias and failed to acknowledge or address in their reports any factors or assumptions that might be regarded as favourable to the pursuer. They had all been willing to support, without proper foundation, the view (ultimately discarded) that the pursuer's tumour was a villoglandular adenocarcinoma, which would have suited the defenders' case because such tumours were regarded as slow-growing. They each had an incomplete or erroneous understanding of the pursuer's symptoms. Dr Williams had given what purported to be expert opinion on matters outwith his expertise and qualifications. As with Mr Soutter, I am entirely satisfied that each of these witnesses was doing his best to assist the court with regard to this case. I am less convinced that all of them had based their expressed opinions on an accurate reading of the pursuer's medical records. By way of examples, Mr Naik and Dr Williams appear to have placed emphasis on the pursuer's bleeding having been more persistent than is disclosed by the records. Mr Naik had failed to notice various entries in the records which stated that the pursuer was a non-smoker. Dr Williams' description in oral evidence of her vaginal discharge as "initially clear, subsequently smelly and yellow" was not in accordance with the records. In these circumstances, I consider that I have to approach the conclusions of these two witnesses based on the pursuer's symptoms with some caution. In the case of Dr Williams, it was apparent that in certain respects, such as his calculation of the tumour volume, his evidence strayed beyond areas in which he had either personal experience or had carried out research in order to enable him to express a view. Under cross-examination he retreated very significantly from the opinions expressed in his report and was often unable to provide a satisfactory answer to questions which challenged the views that he had expressed. I regret that I do not feel able to place reliance on his evidence. I do not make the same criticisms of Dr Farquharson who was, in my view, a careful and convincing witness. He was criticised in the pursuer's submissions for expressing the view "simply because he thinks so" that the diameter of the tumour in 2006 was about 2cm, but my understanding is that the basis for this figure is that the tumour was what was seen by the first ultrasound scan. The criticism accordingly appears to me to be unfounded.

[52] Unusually, in this case, the pursuer gave her evidence after having heard almost all of the evidence of witnesses called to give factual and expert evidence on her behalf. For the defenders it was submitted that I should assume that the pursuer would have worked out where her best interests lay in the evidence that she gave, and to treat with scepticism any gloss that she attempted to place on the medical records. I am in no doubt that the pursuer is a person of acute intelligence who has taken a close interest in this litigation and who is indeed capable of working out where her best interests lie. However, it would be another matter altogether to conclude that the pursuer's evidence was, intentionally or otherwise, misleading. My overall impression of the pursuer was that she was a credible and reliable witness doing her best to give an accurate account of events which have preoccupied her since the 2008 diagnosis. Having said that, there is, on occasion, a notable discrepancy between the description of the severity of her symptoms as given by her to the medical staff at the Sandyford clinic between 2005 and 2007 and the account which she gave in evidence to the court. It is apparent from the medical records and from the evidence of Drs Bankowska and Brown that the pursuer was a difficult and anxious patient, resistant to physical examination while demanding lengthy explanations. On at least one occasion it appears that her subjective perception of her symptoms was exaggerated: at the consultation with Dr Bankowska on 29 June 2006, she is reported to have complained of a "very heavy persistent vaginal discharge" soaking through her clothes, whereas on examination (as the pursuer acknowledged in her evidence) Dr Bankowska observed a "minimal" clear discharge. Dr Brown, too, wished to carry out an examination to make an objective assessment but desisted when the pursuer became upset. The principal occasion of conflict between the medical records and the pursuer's evidence is in relation to the consultation with Dr Bigrigg on 15 December 2005. Dr Bigrigg stated in her affidavit that she has no recollection of this meeting but I have set out the relevant passage from her notes at paragraph 18 above. The pursuer's account of the consultation was that she did not say to Dr Bigrigg that her symptoms were so bad that she could not go to work, give talks, travel, have sex or go to the gym. As a matter of fact she was not unable to do these things at that time. She explained that she had felt that Dr Bigrigg was rushing her and failing to give her the attention she required, and that she had wanted to emphasise to Dr Bigrigg how her symptoms might affect her everyday life. I am content to accept the pursuer's account of what she said. It appears to me to be consistent with her admitted propensity, putting it at its lowest, not to play down her symptoms when seeking medical attention. I do not consider that anything in the medical records casts doubt upon the credibility or reliability of the pursuer's evidence regarding the settling of her symptoms of bleeding and, ultimately, of vaginal discharge.

Symptoms - bleeding

[53] The pursuer suffered one episode of post-coital bleeding in May 2005. She also had an episode of heavy inter-menstrual bleeding in November 2005. No further episodes occurred thereafter. In January 2006 the pursuer's oral contraceptive was changed although she later reverted to Brevinor which she had previously been taking. Dr Williams, Dr Farquharson and Mr Naik considered that irregular bleeding could be a symptom of invasive cancer (though not of pre-invasive cancer, which was asymptomatic). All were however persuaded during cross-examination to agree that these symptoms on their own would not necessarily give cause for concern. The episode of post-coital bleeding could reasonably be explained by the pursuer having sexual intercourse with a particular partner at a time when she had the ectopy that was observed on examination in May 2005. Mr Naik considered that the likelihood of it being symptomatic of cancer was less than 1%. Dr Bankowska noted that inter-menstrual bleeding was fairly common among young women using oral contraception and that the change of pill appeared to have resolved the problem. Mr Soutter's view was that if the bleeding in 2005 had been indicative of the presence of a cancerous tumour it would not have settled. Because it did settle, the theory was "untenable". In my opinion, having regard to the existence of an obvious mechanical cause for the single episode of post-coital bleeding and the fact that the inter-menstrual bleeding settled without recurrence following a change of contraceptive pill at the beginning of 2006, neither episode of bleeding should be regarded as indicative of the presence in early 2006 of a tumour of significant size.

Symptoms - discharge

[54] The pursuer's vaginal discharge was undoubtedly more persistent than her bleeding episodes. There was general agreement among the witnesses that a persistent discharge could be indicative of the presence of a cancerous tumour. According to Mr Soutter, however, vaginal discharge associated with cervical cancer is almost invariably bloodstained, discoloured, or purulent if the tumour is infected. Dr Farquharson agreed that a discharge related to cancer tends to be bloodstained. Mr Naik estimated that a clear, odourless discharge would be indicative of cancer in only 1-2% of cases; Mr Soutter, under reference to published material, noted that discharge was the presenting complaint in only 4% of women with cervical cancer and described the likelihood of a clear discharge being due to cervical cancer as "minute". Most of the witnesses who expressed a view on this matter regarded bacterial vaginosis as the most likely cause. A raised pH value, as the pursuer had in May 2005, was suggestive of infection. The presence of "clue cells" in June 2006 was indicative of BV; Dr Farquharson described it as a "hallmark" of BV. For the defenders it was submitted that the diagnosis of BV was unconvincing because the discharge did not have the typical "fishy" smell and was not cured by metronidazole. These submissions are not, however, supported by the evidence: Dr Brown observed that odour was found in some cases but not in others; Dr Farquharson agreed. Mr Soutter noted under reference to published material that the treatment success of metronidazole in BV cases is only approximately 60%. The suggestion that persistence of the symptoms was attributable to treatment failure was also criticised by the defenders when there was an "obvious" alternative explanation. For my part, I do not regard an explanation which applies in only 1 or 2% of cases as an obvious alternative. In any event it appears that metronidazole did have some effect on the pursuer's symptoms. The medical notes of her consultations with Dr Brown on 30 November 2006, 29 May 2007 and 19 July 2007 indicate that the pursuer had noticed a symptomatic improvement while taking metronidazole but that the discharge had returned when she stopped. The pursuer herself gave evidence, which I accept, that over time the discharge settled at a level that was manageable and no longer causing her concern. That too appears inconsistent with it having been a symptom of the cancer which was developing during the same period.

[55] Taking all of these factors together, I find that it is extremely unlikely that the vaginal discharge of which the pursuer complained during the period from May 2005 to January 2008 is indicative of the presence of a tumour which had become invasive and sufficiently large to be symptomatic by December 2005 or by March 2006. The most likely explanation appears to me to be that the cause was BV, although the possibility remains that it was caused by another infection which was never specifically identified.

Ultrasound scan reports

[56] The pursuer underwent ultrasound scans on 4 January and 29 June 2006. As mentioned above, the operator who carried out the scan on 4 January 2006 noted the presence of "??debris ??polyp within cervical canal". Dr Farquharson and Mr Naik placed considerable emphasis on this report as indicating the presence of an abnormality within the pursuer's endocervix which, with the benefit of hindsight, was probably a tumour of sufficient size to show up during the scan. According to Dr Farquharson, it was rare for an ultrasound scan operator to comment on the cervical canal. The abnormality would have to be around 2-3cm in size for the operator to see anything. Dr Williams also expressed the view in his report that the comment in the ultrasound scan report was an indication of invasive adenocarcinoma but was persuaded in cross-examination to agree that the comment was neutral. Mr Soutter's view was that if the feature noted in January had been a tumour in the cervical canal it would have been much more prominent in June. Dr McKillop, a consultant radiologist, described these findings as "relatively innocuous". His view was that debris within the cervical canal could be explained by the pursuer's recent symptoms of heavy bleeding. While acknowledging that it was difficult to review images from a test carried out by another person, he saw nothing which appeared to him to be sinister. Dr McKillop interpreted the report of the June scan as indicating that the operator had had available the note of the January scan and for that reason had looked closely at the cervical canal, which would not normally be done. The finding in June of a "thick but regular" cervical lining was, like the January finding, entirely neutral.

[57] I accept the view of Dr McKillop that the ultrasound scan reports should be regarded as neutral and not supportive of the presence of a tumour in the pursuer's endocervix in 2006. I also consider that Mr Soutter was correct in his view that a cancerous abnormality observed in January would have been more prominent six months later. There remains, however, the question whether, in the light of knowledge that in 2006 the pursuer did have either pre-invasive AIS or invasive adenocarcinoma, the ultrasound findings can be seen with the benefit of hindsight to have been indicative of the presence of adenocarcinoma in the cervical canal. This in turn raises the question whether, as the defenders submit, the various symptoms and observations that I have discussed amount to more in sum than they do if considered separately. A single explanation for a pattern of symptoms should, as Dr Farquharson had observed, be preferred to a multiplicity of explanations for each symptom viewed individually. For the pursuer, on the other hand, it was submitted that if as in the present case each of the separate strands contained nothing indicative of the presence of a tumour of significant size, then adding them together still produced nothing. I agree with this latter submission. Summarising my findings thus far, I have found that it is very unlikely that the pursuer's bleeding episodes were symptomatic of cervical cancer, that it is equally unlikely that her persistent vaginal discharge was symptomatic of cervical cancer, and that the ultrasound scan reports are at best for the defenders neutral as to the presence of a tumour of significant size. Taking these strands of evidence together they do not, in my opinion, amount to a body of evidence which would persuade me that in January or April 2006 the pursuer had an invasive and infiltrative adenocarcinoma, 2cm in diameter along at least one dimension, located in her cervical canal.

Colposcopy findings

[58] It remains, however, to determine whether the pursuer has proved, on balance of probabilities, that the lesion or tumour which all accept was present in early 2006 was of a size and nature such that she would have received a more conservative treatment than radical hysterectomy. For the pursuer it was submitted that the necessary evidence can be found in the contemporaneous evidence provided by the reports of the three colposcopic examinations carried out in December 2005, March 2006 and June 2006. All three reports mention the presence of an ectopy. None mentions any abnormal lesion. Two of the examinations were carried out by Dr Bankowska. Her position in evidence was that in the course of her career she had twice seen what turned out to be adenocarcinoma and that it had looked quite striking at colposcopy. She could not believe that if there had been a tumour of significant size present on the ectocervix she would not have seen it. She also denied that she could, as suggested by Mr Naik in his report, have mistaken an adenocarcinoma for an ectopy. Mr Soutter thought it highly unlikely that two different colposcopists would have missed evidence of an invasive tumour greater than 1cm in diameter. Such a tumour would distort the surface of the cervix and might appear as a nodule protruding above the surface or as a small ulcerated area with large blood vessels running to, over and around it. Even if a colposcopist did not know what it was, he or she would be alerted to the need for further investigation by biopsy.

[59] I have already stated that I feel able to place confidence in the evidence of Mr Soutter. I was also impressed by Dr Bankowska who appeared to me to be a careful and thorough clinician who gave her evidence in a balanced and impartial manner. I am satisfied that when carrying out her colposcopic examinations she would not have missed a lesion of significant size on the pursuer's ectocervix nor mistaken an adenocarcinoma for an ectopy. Having regard to her evidence and to the supporting evidence of Mr Soutter, I find on balance of probability that if a tumour of diameter greater than 1cm had been present in the pursuer's ectocervix at the time of colposcopy carried out by Dr Bankowska on 9 March and 29 June 2006, it would have been observed and noted by her. On that basis I find that no tumour of that diameter was present in the ectocervix at those dates, from which it follows that no such tumour was present in early January 2006 or mid-March 2006 when the misinterpreted smear test reports were made.

[60] These findings do, of course, depend upon it being established that the glandular abnormality which the pursuer undoubtedly had in 2006 was located on the ectocervix and not in the cervical canal, where it would not have been visible to the colposcopists. The positive case in favour of location in the cervical canal is that, firstly, adenocarcinoma is a disease of the glandular cells most of which are in the endocervix, and secondly, an abnormality in the canal was detected by the January ultrasound scan. Against that case is the statistical evidence, upon which the expert witnesses were broadly agreed, that the vast majority of tumours, including adenocarcinoma, originate in the transformation zone. Dr Dina stated that 98% of tumours arise in this area; Dr Farquharson agreed both with this figure and that it applied to adenocarcinoma in particular. Mr Soutter stated that his impression was that virtually all squamous cell carcinomas and the vast majority of adenocarcinomas begin in the transformation zone. These views bear out the explanation offered by Professor Coleman at the outset of the proof that dividing cells are a target for genetic damage. It should also be noted, however, that in a published work (a chapter contributed to Lower Genital Tract Neoplasia, edited by MacLean et al, 2003), Mr Soutter referred to studies indicating that AIS was not located in the transformation zone in 29% of cases, although in women under 36 years the lesion was usually less than 10mm from the squamo-columnar junction. It is also relevant to note in this connection an observation by Dr Duvall that the area sampled in a smear test is the ectocervix, as it is very difficult to sample the endocervix; it should be recalled that, properly interpreted, the smear samples taken from the pursuer in December 2005 and March 2006 showed abnormalities in both.

[61] I remind myself that the standard of proof which I should apply in deciding this issue is that of balance of probabilities. Having regard to the fact that the colposcopic examinations show that the squamo-columnar junction was at all material times located on the pursuer's ectocervix, the statistical evidence, which was broadly undisputed, appears to me to afford a strong indication that the glandular abnormality which was present at that time and which subsequently developed into adenocarcinoma was located on the ectocervix and not out of sight in the cervical canal. Such statistical evidence might, of course, require to yield to contrary evidence specific to a particular case, but in the present case, taking into account Dr McKillop's view that the ultrasound scan findings should be regarded as neutral, I do not consider that there is such contrary evidence. Conversely, the observation by Dr Duvall which I have mentioned appears to me to support the statistical likelihood that the abnormality was located on the ectocervix. For these reasons, I hold that it is of probable that the pursuer's adenocarcinoma originated in the transformation zone around the squamo-columnar junction which was visible on the ectocervix in 2006, rather than further up the cervical canal.

[62] It was Mr Soutter's opinion that a tumour on the ectocervix with a diameter not greater than 1cm - and with a consequent volume of around 500mm³ or less - could be treated by means of LLETZ or cone biopsy with a strong likelihood of complete eradication. The defenders' witnesses did not express a view directly on the appropriate treatment of a tumour of this size because they were of the view that it was larger and probably on the endocervix. I accept Mr Soutter's evidence on this point. I therefore find, in response to the second and fourth questions set out in paragraph 3 above, that if a correct report of the pursuer's cervical smear had been given either in January 2006 or in March 2006, the medical management which would have taken place would have been treatment by LLETZ or cone biopsy. Such treatment would probably have been carried out within a short period of time after receipt by the Sandyford clinic of smear test results reporting a glandular abnormality such as to cause concern. I accept Dr Bankowska's evidence that the pursuer would have been seen urgently within about two weeks at the colposcopy clinic and probably referred on to a gynaecological oncologist for the biopsy treatment to be carried out.

[63] In the light of these findings it is unnecessary for me to consider whether the pursuer would have received treatment by radical trachelectomy instead of radical hysterectomy. I should, however, indicate briefly for the sake of completeness that if I had not found that in 2006 the pursuer's tumour was of a size and location which rendered treatment by biopsy appropriate, I would have had some difficulty in holding that the pursuer had discharged the onus of proving that trachelectomy would have been the appropriate procedure. I would be addressing a factual situation (which I have found not to be the case) in which the pursuer's tumour was an adenocarcinoma of up to but not more than 2cm in diameter, probably located in the endocervix. In these circumstances the likely sequence of events was described by Dr Farquharson as follows. The pursuer would have had a colposcopic examination and treatment by LLETZ. Depending upon the results of histological examination of the excised tissue, there might have had to be a second loop excision further up the cervical canal. If this contained abnormalities, the pathology would have been discussed and an MRI scan arranged. In the event that the presence of a 2cm lesion within the canal was indicated, there would have been discussion at a multi-disciplinary meeting of whether a fertility-sparing option could be considered, but radical hysterectomy would have been the treatment of choice. Mr Naik's view was to similar effect. He emphasised that in the case of adenocarcinoma it was less easy to be confident that there were not further lesions within the canal. Mr Soutter did not express any view on a choice between trachelectomy and hysterectomy as he was clearly of the view that the appropriate treatment was biopsy. On that state of the evidence, I consider that it would have been too speculative for me to find, on the hypothesis of fact discussed in this paragraph, that the pursuer would have been appropriately treated by trachelectomy, even if she herself had been aware of the possibility and had tried to persuade those managing her to carry it out instead of hysterectomy. On the view that I have taken, however, this issue does not arise.

Disposal

[64] As counsel for the defenders accepted, it is appropriate, in the light of the defenders' admission of liability, to sustain the pursuer's first plea in law and I would propose to do so. However, in accordance with counsel's suggestion, I shall put the case out By Order to hear submissions on further procedure.