High Court dismisses negligence claim against HSE for failing to clinically recall patient after clear mammogram.

[2020] IEHC 286 THE HIGH COURT [2017 No. 5298P] BETWEEN SIOBHAN FREENEY PLAINTIFF

– AND – HEALTH SERVICE EXECUTIVE DEFENDANT JUDGMENT of Ms. Justice Niamh Hyland delivered on 29 May 2020

Introduction

1. The plaintiff, Mrs. Siobhan Freeney, asserts that the national screening service, BreastCheck, run by the HSE, have been negligent in the way in which her breast screening was carried out in June 2015. At that time, she was called in the usual way for her two-yearly mammogram (as she was over 50 years of age and under 65). Following a mammogram that took place in the mobile screening unit in Gorey, Co. Wexford, her images sent for review by radiologists. A note from the radiographer went to the two radiologists reading her mammogram, in the following terms: “Hx of cysts. Lady feels a cyst in the R breast. Couldn’t tolerate compression on R” (the “Note”). Both radiologists considered the plaintiff’s images and Note separately (as is normal practice) and neither thought she required clinical recall for further examination.

2. In November 2015, the plaintiff noticed symptoms of concern in her right breast and she went to her GP who immediately referred her for triple assessment at St. Vincent’s Hospital, Dublin 4. She was diagnosed in December 2015 as having invasive lobular carcinoma. She underwent neo-adjuvant chemotherapy, a mastectomy of her right breast, axillary clearance whereby all her lymph nodes in her right armpit were removed, and radiotherapy following the mastectomy. Happily, the treatment was successful and the plaintiff is, at present, free of cancer.

3. In short, the plaintiff says that her screening by BreastCheck was carried out negligently, that she should have been clinically recalled and sent for further examination in June 2015, and, had that been done, the cancer diagnosed in December 2015 would have been detected in June 2015. She accepts that she would nonetheless have received treatment for cancer including a mastectomy but says that she would likely have avoided chemotherapy and radiotherapy to the chest wall and would have required only a sentinel node biopsy. Had that been the case, it is said she would have avoided certain side effects of her actual treatment, including lymphoedema. Based on her experts’ estimate of the size of the tumour in June 2015, her life expectancy would have been shortened by between 3.4 years and 0.9 years (depending on the doubling time of the tumour used)

instead of the current 6.9 years. The plaintiff asserts she should have been clinically recalled where:

- There were suspicious irregularities in the mammographic images warranting further investigation;

- She informed the radiographer carrying out the mammogram, Ms. Mathews, of clinical symptoms that ought to have warranted a recall.

4. A further issue in the case is whether negligence, if present, in fact caused harm to the plaintiff. This requires me to consider two separate questions – first, whether cancer was in fact present in June 2015 and second, if so, whether it would have been diagnosed had the plaintiff been clinically recalled. It is agreed between the parties that if the answer to those questions is yes, then some harm was done to the plaintiff by the diagnosis taking place in December 2015 rather than June 2015, the nature of which is to be determined at a separate quantum hearing, (as ruled on the first day of the trial following an application for a staggered trial by the defendant).

Legal Standard

5. The law in this jurisdiction in respect of professional negligence, including in the context of screening decisions, has been recently restated by the Supreme Court in the case of Morrissey v. HSE [2020 IESC 6]. Clarke C.J., giving the decision on behalf of the Supreme Court, summarised the legal standard of care in a clinical negligence claim, or indeed any professional negligence claim, as requiring “the court to assess whether no reasonable professional of the type concerned could have carried out their task in the manner which occurred in the case in question. That overall test requires a court to determine what standard a reasonable professional would apply” (paragraph 16.2). He observed that the test in Dunne (an Infant) v. National Maternity Hospital [1989] I.R. 91 remains the basis for identifying the legal standard of care by reference to which a claim in clinical negligence is to be assessed.

6. Helpfully, the case of Morrissey sets out a clear road map in a case of alleged professional negligence. The starting point is the identification by the court of the standard of approach that would have been applied by a professional of the appropriate standing or skill as the person against whom the allegation of negligence is made (paragraph 6.6).

Or, as restated at paragraph 6.12 of the judgment of the Chief Justice, “what would an ordinary competent professional of the type and skill of the individual concerned have done, and did the professional who is sued meet that standard”?

7. Here, as in Morrissey, there is no issue concerning the adoption of a common practice with inherent defects that ought to have been obvious. Insofar as the allegation of negligence with respect to the reading of the mammographic images is concerned, there is no dispute about the standard of approach: rather the dispute centres on whether there was any cause for concern on the images. In other words, what I must consider is whether the two consultant radiologists who read the plaintiff’s mammogram, Dr Pender and Dr Coffey, fell below the accepted standard of approach.

8. Insofar as the clinical presentation allegation is concerned, it is necessary to identify the appropriate standard of approach as there is a dispute between the parties on this point.

In this respect, the Guidelines for Clinical Recall 2014 (referred to within BreastCheck as the standard operating procedures or “SOP” ) as developed by BreastCheck are highly relevant. The Guidelines themselves are not challenged in these proceedings. The Guidelines were adopted pursuant to the Guidelines for Quality Assurance in Mammography Screening (4th edition 2015) (“Guidelines for Quality Assurance”), developed by BreastCheck. Under the Guidelines for Clinical Recall, clinical recall for reasons unrelated to the results of the mammogram will only be made for significant signs and symptoms that are known risk factors for cancer, where those signs and symptoms are either observed by the radiographer or made known to the radiographer by the woman being screened. Those known risk factors are: (a) new lump (b) skin dimpling/tethering/inflammation (c) recent nipple inversion. Where those risk factors are not present, but the radiographer obtains information they consider potentially relevant, under the guidelines they may write a note to alert the radiologists to that information.

The radiologist must read the note and use personal clinical judgement in deciding what action if any should be taken.

9. I must decide upon the correct standard of approach for a competent consultant radiologist, operating in a national screening service such as BreastCheck, with equivalent clinical recall guidelines, when deciding whether or not to recall a woman for reasons of clinical presentation. I must then consider whether Dr Pender and Dr Coffey met that standard in deciding not to send the plaintiff for clinical recall having regard to the content of the Note.

10. A separate case is made against the radiographer, Ms. Mathews, who carried out the mammogram on the plaintiff on 17 June 2015. The allegation against her is a straightforward one; that in drawing up the Note for the radiologists to consider, she failed to fully and accurately record her conversation with the plaintiff. I must decide both the correct standard of approach of a radiographer when recording a conversation with a client while carrying out a mammogram and whether that standard was met.

11. The two aspects of the plaintiff’s case stand alone: in other words, when considering each aspect, I must test the conduct of the radiologists and (where relevant) the radiographer irrespective of my conclusions on the other aspect.

12. The correct standard of approach is one of fact in each case as identified at paragraph

6.13 of Morrissey. As observed there and in the following paragraphs, it is not for me to determine the standard of approach or impose my own views as to the requisite standard of approach. Rather it is the standards of the profession itself as demonstrated by the evidence that impose the standard required. As noted at para 6.37 of Morrissey, it is necessary to tender expert evidence as to the appropriate standard of approach for the relevant professional. Both parties have tendered a consultant radiologist as to the relevant standard of approach. There is no dispute about the appropriateness of those witnesses to give evidence on these matters. In respect of the standard of approach in respect of the clinical presentation allegation, the plaintiff has also sought to tender evidence from a consultant breast surgeon and a consultant histopathologist and to introduce into evidence material obtained on cross examination of two of the defendant’s witnesses, being a consultant oncologist and consultant breast surgeon. I address the admissibility and weight of that evidence below. Pleadings

13. The proceedings were issued on 9 June 2017 against a range of defendants and were pleaded at a very high level of generality initially. On 5 June 2018, particulars were delivered that essentially alleged that the defendant was negligent in reviewing, interpreting and reporting on the plaintiff’s mammogram, in failing to identify suspicious features on the mammogram, and in failing to follow up on those abnormal features, On 6 June 2018, the proceedings were discontinued against all defendants except the HSE.

Further particulars of personal injury were provided but no further particulars of negligence and breach of duty until 14 February 2020, some 4 days before the case commenced on 18 February 2020. Those particulars identified the clinical presentation aspect of the case as follows:

“Failed to refer the Plaintiff for further assessment, including ultrasound examination, having regard to her clinical presentation and/or symptoms on 17 June 2015”.

This extremely late addition of a very substantive aspect of the case had unfortunate implications for the conduct of the trial. It meant that the report of Professor Whitehouse (consultant radiologist, expert witness for the plaintiff) on this aspect of the case was only provided on 18 February 2020 i.e. the day the trial commenced. Professor Bundred, consultant breast surgeon, expert witness for the plaintiff, provided his report on this aspect of the case on 17 February 2020. Dr Given-Wilson, consultant radiologist, expert witness for the defendant, did not do a written report on this aspect of the case, but dealt with the matter in evidence, with her views being put to Professor Whitehouse in cross examination. There were various time-consuming disputes during the trial about the provision of documents. Had the plea been raised in a timely fashion, those disputes would have been resolved by the time the case came on for hearing in the normal fashion. It was never made clear why the plaintiff raised this aspect of the case so late in the day. Professor Whitehouse received the Note on 21 June 2017, along with the other records from BreastCheck. I was told that on 12 February 2020 the plaintiff’s solicitors had a pre-trial consultation with Professor Whitehouse and asked him to consider the Note and whether it was of any significance in the case. Professor Whitehouse in his evidence said he had not received the Note until close to the trial. I think that is unlikely to be correct as it was specifically confirmed that he had been provided with the Note on

21 June 2017.

The Plaintiff’s medical history

14. The plaintiff was born on 13 August 1960 and lives in Gorey, Co. Wexford. She is married to Paul Freeney. They have two adult children. She worked as a special needs assistant and was made permanent in 2015. She has not worked for some time and had to retire in 2019 due to the ongoing effects of her treatment for cancer.

15. In May 2010, the plaintiff felt a lump in her left breast and attended her GP. She was referred for assessment to the breast unit at St. Vincent’s Hospital, with the area of clinical concern identified as a lump in the lower outer quadrant of her right breast.

Ultrasound showed a complex cystic mass in the lower outer quadrant of the right breast which was biopsied and confirmed to be benign. A 12mm simple cyst in her left breast was aspirated and was not of clinical concern.

16. Thereafter, the plaintiff came under the care of BreastCheck. Her initial visit and mammogram took place on 26 May 2011 and she received a routine recall i.e. recall after two years. She was called for her second mammogram on 2 July 2013 and again received a routine recall after two years.

17. Her third visit took place on 17 June 2015. The mammogram was carried out at the mobile unit of the Merrion Unit of BreastCheck, which was stationed at that time in Gorey, Co. Wexford. The plaintiff was seen by Ms. Nithin Mathews, a senior radiographer with BreastCheck. Two views were taken of each of the plaintiff’s breasts – a mediolateral oblique (“MLO”) view and a cranio-caudal (“CC”) view. The images produced from the mammogram were sent to the BreastCheck Merrion Unit in Dublin 4 and were read by Dr Pender, consultant radiologist with BreastCheck, as first radiological evaluation on 18 June 2015 and by Dr Coffey, consultant radiologist with BreastCheck, as second radiological evaluation on 20 June 2015.

18. During the mammogram, the plaintiff had a discussion with Ms. Mathews. Ms. Mathews did not identify that the plaintiff should be clinically recalled. However, she made a note of her conversation with the plaintiff and annotated one of the images from the mammogram with a sticker with the word “N” on it, as per the Guidelines on Clinical Recall, to alert the radiologists reading the mammogram to the existence of the Note.

19. Both radiologists gave evidence that they read the images and did not identify any suspicious features and did not believe a clinical recall was necessary. The extracts from the IT system recording the reading were put in evidence. Both radiologists also gave evidence that they read the Note and did not think a clinical recall was necessary. Both radiologists acted independently of each other and did not confer, as is the practice in BreastCheck unless there is a difference of opinion as to whether a client should be recalled.

20. In November 2015, the plaintiff noticed nipple retraction on her right breast. She attended her GP, Dr Clare, on 25 November 2015 and was referred for assessment to the breast clinic at St. Vincent’s Hospital. She attended the triple assessment clinic on 11 December 2015 where clinical examination showed a lump in her right breast. Ultrasound showed an irregular mixed echogenicity area involving the upper right breast spanning greater than 5cm. Following a core biopsy, the histopathology showed invasive lobular cancer, grade 2, oestrogen positive, HER2 negative. A mammogram showed increased parenchymal density throughout the right breast centred in the upper central breast. The ultrasound right breast showed an irregular mixed echogenicity area involving the upper right breast spanning greater than 5cm and a pathological lymph node. CT examination of the chest, abdomen and pelvis on 30 December showed a 6mm internal mammary lymph node. An MRI of the right breast on 12 January 2016 showed a 5.8 x 5.5 x 6.4cm area of non-mass enhancement, centred in the upper central breast, corresponding to the known tumour, with enlarged enhancing right axillary lymph nodes. It concluded there was a 6.4cm area of non-mass enhancement upper central right breast corresponding to the known tumour. It was decided the plaintiff would undergo MRI, staging, neoadjuvant chemotherapy and mastectomy clearance. The plaintiff underwent neoadjuvant chemotherapy. On 25 April 2016 an MRI identified a tumour bed with with a non-mass enhancement of 4cm. The residual enhancing lymph node was much smaller.

21. On 28 May 2016 the plaintiff underwent a mastectomy and axillary clearance i.e. lymph node removal. Histopathology of the specimen was reported as showing, inter alia, that there was residual invasive lobular carcinoma grade 1, that the largest invasive focus measured 18mm and that residual invasive carcinoma extended over 70mm at least, that invasive tumour cellularity was 10%, that 42 axillary lymph nodes were negative for metastases and that the post-treatment biomarkers were oestrogen receptor positive and HER2 negative.

Invasive Lobular Cancer

22. The nature and presentation of invasive lobular cancer (“ILC”) are particularly relevant to the first issue to be determined, i.e. whether there was negligence on the reading of the mammogram. There is in fact little dispute between the parties as to the core features of ILC. Between 8.5% and 10% of breast cancers are ILC. It grows in a single file distribution, known as “Indian file”, which does not initially disturb the background structure of the breast but rather leaves the tissue intact, thus making it difficult to detect on a mammogram. It spreads in a more diffuse, infiltrative way than other cancers. ILC has an infiltrative pattern rather than forming a mass. This is because of the genetic alteration of the e-cadherin gene in ILC, resulting in the cells separating rather than sticking together as is normal in other cancers. Classical ILC, which forms 80% to 85% of cases, generally grows slowly and is usually given a mitotic rate of 1, being the lowest rate. This refers to how quickly the mitochondria in the cells are dividing. The fastest growing cancers would have a mitotic rate of 3.

23. Brem comments that the unique histology of ILC contributes to the difficulty in early detection, both clinically and radiographically. As per Professor Whitehouse’s evidence, it is notoriously difficult to identify it in practice, resulting in (anecdotal) evidence of significant litigation in the U.S. for failure to diagnose it on mammography. Professor Ellis, consultant pathologist giving evidence on behalf of the plaintiff, also identified the difficulty in recognising lobular cancer.

Breast density

24. The question of breast density is also relevant to the alleged negligence on the reading of the plaintiff’s mammogram. Breast density refers to the mammographic appearance of the breast, with density reflecting varying amounts of fat (dark areas on mammograms)

and stromal and epithelial tissues (white areas on mammograms). Breast density is measured as the absolute amount of dense or white areas in the breast or a proportion of the mammogram that is composed of dense tissue. Increased breast density can make it more difficult to detect smaller cancers with mammography as it masks the cancer. The sensitivity of mammography detecting cancer in women with almost entirely fatty breasts is 88% as compared to 69% for women with heterogeneously dense breasts and 62% for women with extremely dense breasts. There is also evidence that high breast density increases breast cancer risk apart altogether from the risk of masking.

25. It is accepted that the plaintiff has a history of breast density. Indeed, as identified by the plaintiff in her evidence, she has established a blog to inform women of the consequences of having dense breasts from the point of view of detection of breast cancer. She asserts that she should have been told to avail of ultrasound rather than mammography as a diagnostic tool to detect cancer due to her dense breasts. However, this complaint did not form part of her case. When being cross examined, she accepted that the right upper area of her breast had always felt lumpy. The radiology report from 2010 identifies both breasts as dense and glandular. Dr Coffey noted that her mammogram of 2015 showed lumpy tissue, Professor Whitehouse and Professor Bundred identified her breasts as dense and Dr Given-Wilson identified them as very dense. I do not have to decide the precise characterisation of the plaintiff’s breasts since whether they are characterised as dense or very dense, both sides agree that density makes the detection of breast cancer on a mammogram more difficult due to the masking effect of density. In those circumstances, I conclude that the plaintiff had either dense breasts or very dense breasts, thus making it more difficult to detect the presence of cancer on the plaintiff’s mammogram in June 2015.

26. There was some controversy about a minute written by Professor Ann O’Doherty, director of BreastCheck, recording a meeting she had with the plaintiff on 6 January 2016. In the minute, it is recorded that Professor O’Doherty told the plaintiff she did not have dense breasts, while later she records she told the plaintiff that her breasts were moderately dense not high density while explaining to the plaintiff why she had not been advised to have ultrasound. There is no other reference in the minute to breast density. Professor O’Doherty was not called to give evidence. The plaintiff has raised the failure by the defendant to make that note available to her legal team until Day 2 of the hearing. It was said that the late provision of that note was detrimental to the plaintiff as her experts ought to have seen it in advance given the reference to density identified above. I deal with that issue below but, in short, I do not consider that the passing reference in the note to density was required to allow the plaintiff’s experts to arrive at their own conclusions on density. Professor Whitehouse, Professor Bundred and Professor Ellis, all experts for the plaintiffs, based their findings on density on the primary sources before them, i.e. the plaintiff’s mammograms, the reports on mammograms, ultrasound and MRI, and the histopathology reports. Professor O’Doherty’s views, as recorded in the minute, would not, in my view, have altered their conclusions on density. Moreover, there is no controversy on this issue: every expert in the case who considered density, irrespective of whom they were acting for, concluded that the plaintiff had either dense or very dense breasts.

Breast screening at BreastCheck

27. BreastCheck, the National Breast Screening Programme, was established in 1998. It is a free health service that offers eligible women i.e. women aged 50-69, a mammogram every two years. A mammogram is an x-ray of the breasts to look for breast cancer. It aims to reduce the number of deaths from breast cancer in the relevant age group. With over 2,700 women diagnosed every year, breast cancer remains the most commonly diagnosed cancer in women in Ireland and has the second highest mortality rate.

28. The breast screening is carried out by taking two images of each breast – a craniocaudal (head to toe) image (CC), from the top of the breast down through the breast and a mediolateral image being the lateral image of the breast, which includes the axillary tail and the upper part of the breast (MLO). The breast is compressed by the machine taking the images to (a) spread out the tissues to get better detail (b) reduce the radiation dose to the breast and (c) to keep the breasts steady so there is less movement artefact. That compression may cause discomfort or pain. It is a short procedure, usually taking 5-7 minutes in total. In BreastCheck, all mammograms are carried out by a radiographer.

There is no clinical examination of the client. A woman undergoing a mammogram at no stage sees a radiologist.

29. The images are then examined separately by two consultant radiologists, i.e. double screening. The room in which the images are screened is completely dark and quiet.

There are three screens, two of which show the images. The screens are about 21 inches in cross sectional diameter. They do not confer when reaching their initial conclusion. Dr Pender explained that having radiologists read images from a mammogram separately means that 5% more cancers get picked up. If there are differing views between the radiologists as to whether a woman should be clinically recalled for further investigation, the view of the highest reader prevails, i.e. the radiologist who wishes to send the woman for recall. If clinical recall is necessary, the woman is sent a letter and is seen at BreastCheck usually within two to three weeks of the screening where she will undergo triple assessment i.e. assessment by a radiologist using mammogram and ultrasound, biopsy if appropriate and assessment by a breast surgeon.

30. As with all screening programmes, the service is not a diagnostic one. The women called for screening are not called because they have presented with symptoms to their doctor but because they fall within a particular age group. If a person presents with breast symptoms to their GP, their doctor cannot refer them to BreastCheck.

31. Mammograms cannot detect all cancers. Cancers developed between screening intervals (in Ireland a two-year interval) are known as interval cancers. Dr Given-Wilson explained that if 1,000 women are screened, the programme will find cancer in about 8 of them.

The remaining 992 will be screened in two years. Any women in that group who present with cancer prior to the next screening are described as having an interval cancer. The standard figure in European countries, including Ireland, is that of those 1,000 women, 2 will develop an interval cancer.

32. Interval cancers may be distinguished into four categories – true interval cancers i.e.

those that showed normal or benign features in the previous screening mammograms, false-negative cancers i.e. when signs suspicious for malignancy are retrospectively seen on a mammogram, minimal signs i.e. cancers showing detectable but non-specific signs at the latest screening and occult tumours i.e. cancers that show clinical signs of the disease despite a lack of mammographic abnormalities either at screening or diagnosis .

Quality Assurance Guidelines/ Guidelines for Clinical Recall (SOP)

33. The Quality Assurance Guidelines note in the Preface that the primary aim of breast screening is to reduce the number of breast cancer mortalities in the eligible population through early detection. The Guidelines set out quality assurance standards for each relevant area, being client care, radiography, radiology, pathology, surgery, physics, epidemiology and breast care nursing. Section 3.9 provides for clinical recall as documented in the standard operating procedure (SOP). As noted above, the SOP refers to the Guidelines for Clinical Recall. Given the importance of these Guidelines in the determination of the clinical presentation limb of this case, they merit being quoted in extenso. The first section is entitled

“1. Purpose and Scope (including Quality Standards)” and provides:

To ensure that significant signs and symptoms of breast problems are identified and recorded for clients who attend screening.

To ensure information regarding a Clinical Recall is available to the radiologist when reading and the radiologist is aware of any client falling into the Clinical Recall category.

To ensure the radiographer recognises and makes note of relevant clinical symptoms and that recall to assessment may be required.

The fourth section is entitled “4. Method” and provides:

“During screening mammography the radiographer may identify or be made/aware of one of the following significant signs or symptoms:

New lump

Skin dimpling/tethering/thickening/inflammation

Recent nipple inversion

The radiographer may identify this client a CLINICAL RECALL so that she may be called back to an assessment clinic.

Procedure

4.1 The radiographer takes note of any clinical signs or symptoms she may observe either before or during the mammogram. NB: The radiographer should not perform a clinical examination of the client.

4.2 If a significant sign or symptom is identified by either the radiographer or the client (above list) and requires follow-up then the client is identified as a CLINICAL RECALL by the radiographer. …

4.4 When a radiographer wishes to alert the Radiologist to relevant information that is not covered by the category clinical recall, the radiographer annotates the clients images with the word “NOTE” and documents the relevant information on the clients datasheet as per departmental protocol.

The radiologist has responsibility for reading any note and will use personal clinical judgement in deciding whether or not any action should be taken in relation to the content of any note.”

On the front of the SOP, there is a box entitled “Document Revision History”. This identifies that paragraph 4.4 (above) was added on 2 April 2014 and that a previously existing point regarding the radiographer asking client if she has noticed any changes in her breasts is to be removed. The previous list of signs classed as clinical recall reasons are palpable lump, skin thickening/tethering, bloody nipple discharge and recent nipple inversion . Thus “palpable lump” has been replaced with “new lump” in the new version of the Guidelines on Clinical Recall.

34. The core features of BreastCheck screening may be summarised as follows:

- It provides a binary result i.e. should the client be recalled for clinical assessment or not.

- No report is produced describing the findings from the mammogram.

- The only interaction the client has is with a radiographer, whose expertise is in the taking of images.

- There is no breast examination.

- The radiographer is expressly instructed not to ask the client questions about her breast history. However, if the client volunteers information about her breasts identifying any of the signs and symptoms in the Guidelines for Clinical Recall, or the radiographer observes them, the radiographer should designate the client for clinical recall. If information provided or observation by the radiographer does not fall within those signs or symptoms but may potentially be relevant, then the radiographer makes a note of that information.

- Screening is carried out on all women who respond to the invitation by BreastCheck

– about 45,000 of 55,000 invited annually – and the population being screened are therefore healthy, in the sense that they are not attending for a mammogram because of symptoms.

35. Those features of the screening service may be contrasted with those present when a woman visits a GP due to symptoms of concern to her, where:

- The woman will be symptomatic;

- The woman will be seen by the GP;

- The woman will be physically examined by the GP, who may be able to palpate any abnormalities of concern, depending on the type of abnormalities;

- The GP will take a clinical history and very likely ask the patient questions, including in relation to any previous palpable abnormalities, their designation and their treatment, if any;

- If the woman displays symptoms of concern, she will likely be referred to a hospital for assessment, where, if she undergoes triple assessment, she will likely be seen by a radiographer, a radiologist and a surgeon;

- She will have a mammogram and a report will be done on the resulting images that will analyse the results of the mammograph (see for example the report on the plaintiff’s mammograph of 2010 or December 2015).

Evidence adduced on behalf of the Plaintiff in respect of alleged negligence

36. Factual evidence was given by the plaintiff. Professor Whitehouse, consultant radiologist, was called as an expert witness. The plaintiff also sought to call Professor Bundred, consultant breast surgeon and Professor Ellis, consultant pathologist, as expert witnesses in respect of the clinical presentation aspect of the case and I explain below why I give their evidence in that respect no significant weight .

The plaintiff

37. The plaintiff, Mrs. Freeney gave evidence in person over the course of three days. She gave slightly different versions of her experience when having the mammograph but in my view the core components of her evidence did not change. She gave evidence that she told Ms. Mathews of her history of cysts, specifically from 2010 when she had two cysts investigated, one of which turned out to be a simple cyst in her left breast and the other which turned out to be a complex cyst in the lower outer quadrant in her right breast.

38. During her examination in chief, in response to a question whether she had any symptoms or any cause of concern when she went for the mammogram in June 2015, the plaintiff said nothing that she was aware of that was any different. She then referred to the pain she felt during compression of the right breast, that she told Ms. Mathews that she had lumpy breast tissue and she was always conscious of the upper right breast area, that she told Ms. Mathews she was very concerned about a lump in her upper right area (it is clear from the evidence that this was after the painful compression), and said:

A. And then we discussed the fact that there had been a lump there, there had been, you know, a look at it some years earlier.

Q. Yes.

A. So for me the lump wasn’t necessarily new, but I had a renewed concern over it because of the pain.”

39. Then, a short time later, counsel for the plaintiff read out the handwritten Note and asked:

“Q. Does that reflect what you said?

A. Absolutely, yeah.”

40. In my view, this exchange is of considerable significance. Had the plaintiff believed that she had told Ms. Mathews of a lump rather than a cyst, or that she had told her about something new, she would have identified that the Note was inaccurate or incomplete.

Instead, she agreed without any caveats or reservations that the Note reflected what she said to Ms. Mathews. Later in cross examination, while the plaintiff was being asked about her concerns, she said that she had talked about “these cysts” and was happy that the mammogram really wasn’t showing anything that she needed to be concerned about.

Shortly afterwards, the plaintiff said:

“A. My concern would always have been there about a palpable area in my upper right which I now know to be a quadrant and I don’t believe I have ever gone anywhere in relation to my breast health without mentioning it and I definitely mentioned it on the day, yes. And you know, that’s where the comment came from me in relation to a history of cysts because that is what had been explained to me in relation to that, you know, area. They were cysts, yes.” ….

“A. Okay. So she asked me did I have a history of anything in relation to it and I said I had been told that I had a lot of cysts and I pointed to the upper right area…. As I said, I had gone to have it checked on many occasions.”

41. Further on in the cross examination, she said in respect of the upper right area that she had been told that they were cysts but that this was the lump that she felt was causing the pain in the compression. She went on to observe that this was not the first time she had shown a concern over her right breast and wasn’t the first conversation she had in relation to it.

42. Discussing the plaintiff’s concerns on going into the mammography appointment and whether she had any cause for concern at that point, she said:

“A. Nothing that I hadn’t already, you know, brought to the attention of – you know, the lump was there, it was palpable, it had been looked at, I had been reassured so …

Q. Right. So when you went in in June 2015, you thought – you understood you had a lump at that stage; is that right?

A. Yes, yes.

Q. Okay. And then when you had the mammogram and you had the pain that you talked about, you then thought that the pain was coming from the existing lump; is that right.

A. It certainly felt like that. The compression certainly felt like that – yes yes.”

43. This evidence is somewhat confusing since the cyst in the plaintiff’s right breast in 2010 was in the lower right quadrant, whereas the plaintiff gave evidence that her area of constant concern was the upper right quadrant. Nonetheless, the totality of the plaintiff’s evidence demonstrates that following the pain she experienced on compression, she did not interpret this as evidence of a new lump/cyst but rather this provoked a new concern about an existing area that had been previously been brought to the attention of various medical personnel. Following the mammogram, the plaintiff initially said that the pain “felt even more painful after the compression”, but on cross examination said that once the compression was lifted, she really didn’t feel there was a huge issue there.

44. When the plaintiff went to see her GP in November 2015, the GP found a palpable lump in the upper right quadrant. The plaintiff was cross examined about the fact that the GP’s referral letter referred to the fact that the duration of symptoms in relation to the breast was 3 weeks. The plaintiff contested the accuracy of that note and said that the threeweek period might refer to the time from which she first noticed the nipple retraction, which had caused her to go to the GP and that at no point did she ever say her symptoms in respect of the lump were there for three weeks. Given the evidence of the plaintiff to the effect that she had a long standing presence of a lump/cyst in her right breast, and the emphasis she placed on finding that her nipple had retracted, I consider that it is likely that the GP referral letter mistakenly referred to the fact that a lump in her right breast had first been identified by the plaintiff three weeks earlier. In fact, it seems more likely that the symptom of nipple retraction was first identified by the plaintiff three weeks earlier.

Professor Whitehouse

45. Professor Graham Whitehouse, consultant radiologist, gave evidence for the plaintiff in relation to the reading of the mammographic images and the contents of the Note. He was the professor of diagnostic radiology at the University of Liverpool from 1976 to 2001 while also working as a consultant radiologist at the Royal Liverpool University Hospital dealing with breast radiology (among other matters) and at the Liverpool Women’s Hospital where his special area of expertise was gynaecological radiology. In 1978 he set up a research unit looking into the feasibility of screening mammography on a mobile basis. In the late 1980’s the UK Department of Health set up a breast screening programme and he was one of a group of pathologists, along with radiologists and surgeons, that looked into the feasibility and practicality of setting up breast screening.

He also assisted setting up the Australian Breast Screening Programme. He has various publications on breast radiology.

46. After he retired in 2001 from the University of Liverpool, he continued to do locums which involved screening mammograms and taking part in assessment clinics. He is now an emeritus professor of the University of Liverpool. In 2010 he retired and since then has written litigation reports, for both plaintiffs and defendants, about 58-85 per year with about one third of them dealing with breast problems. On his evidence, Professor Whitehouse does not appear to have ever been involved in running a screening programme.

47. He accepted under cross examination that he does not now meet the professional standards required of radiologists involved in the NHS Breast Screening Programme (“BSP”) and other breast screening as identified in the “Guidance on screening and symptomatic breast imaging” 4th edition, November 2019, published by the Royal College of Radiologists. These standards require that to screen mammograms, the radiologists must undertake a minimum of 5,000 screening and/or symptomatic cases per annum.

This is to be contrasted with the requirement for symptomatic breast imaging, where radiologists are required to undertake a minimum of 500 symptomatic cases per annum.

They also require a certain minimum participation in multidisciplinary meetings.

Evidence on the reading of the mammographic images

48. Professor Whitehouse provided a written report dated 14 August 2017. He was provided with, and reviewed the images from 2010, as well as those from 2011, 2013, June 2015 and December 2015 (the cancer being diagnosed in December 2015). However, because the only mammograms reviewed by Dr Pender and Dr Coffey were those of 2011, 2013 and June 2015 (and there is no criticism of that approach) I will focus on his reviews of those mammograms. In his report, Professor Whitehouse observes that there is no contemporaneous report found in respect of the 2011, 2013 and June 2015 mammograms. Professor Whitehouse concludes that in 2011, the appearances show no change from 2010 where the breasts were dense with bilateral symmetry, save that nodularity is less of a feature. In 2013, he concludes again there is no significant change from 2011, the breasts remaining quite dense with bilateral symmetry, overall nodularity less of a conspicuous feature and no dominant mass and no suspicion of malignancy.

49. In June 2015, he finds there is a significant change to the right breast appearance since 2013, whereby on the CC projection there is an area of distortion and spiculation just lateral to the midline, and on the MLO projection there is now an ill-defined area of increased density within the upper half of the breast, with the diameter of the mass being approximately 2 cm. He asserts that those new features are highly suspicious of cancer.

In respect of the December 2015 mammogram (where cancer was detected), he notes that architectural distortion in the mid line is visible on the CC projection and that it is difficult to measure the exact size of mass in the upper half of the right breast but is approx. 2.5 cm in diameter and has a spiculated margin. He concludes that the June 2015 mammogram shows features highly suspicious of malignancy in the upper half of the right breast and that not to have reported and followed up on the abnormal appearances was negligent.

50. He also did a written report on the clinical presentation aspect of the case that touched on the mammographic images and included the observation that if the mammogram had been read by radiologists, attention would have been drawn to the obvious asymmetry and possibility of a mass in the upper part of the right breast seen on MLO image and the distortion on the CC image. This is the first time Professor Whitehouse identifies his belief that the images had not been read at all by radiologists.

51. In his oral evidence, Professor Whitehouse explained he had received the images on a disc and indicated there was a slight degradation on compact disc since they were recording original digital images but it was a small factor generally in respect of the quality of the images. In a hospital setting the original digital images are being viewed.

Professor Whitehouse presented a small book of four images to the Court, consisting of the right MLO 2013 (image 1), right CC 2015 (image 2), right MLO 2015 (image 3) and left MLO 2015 (image 4). A book of images was also produced by the Defendant, which consisted of 19 slides in total, of which 12 were various mammographic images of the plaintiff’s breasts in 2007, 2011, 2013, June 2015 and December 2015. He defined spiculation as linear extensions coming out from a mass lesion, with the mass being generally rounded, lobulated, and where the lines extend from the margin of a mass to the surrounding breast tissue, observed that was a feature highly suspicious of cancer.

Right MLO image 2015

52. Considering the right MLO (no. 3 on his booklet and Slide 4 on BreastCheck booklet), Professor Whitehouse described a fairly ill-defined area of increased density within the upper half of the breast, describing it as “rather diffuse but it has got some nodularity to its margin”. He went on to describe the increased density in the 2015 image (MLO) as a patchy increased density, ill-defined. As compared to 2007, he indicated the 2015 density was becoming more focal, more localised. He identified at the top part of the breast a “lobulated density there which didn’t exist two years before” and indicated that the margins of it were slightly spiculated. He indicated that a similar mass could not be seen for the equivalent image in 2013 (no. 1 on his book). He said he measured the visible mass on the mammogram at 2cm and formed the impression that it extended beyond that. On being asked did he agree that a diffuse abnormality was not the same thing as a mass he said

“A. Well again, this is so often a matter of semantics, really. You are looking at the diffuse mass and when you identify it as a diffuse mass and you look at previous mammograms you can see that there is an increased density, ergo it is a mass, which just happens to be ill-defined.

Q. Well, a diffuse abnormality is different from a mass, we can agree with that?

A. Well, a diffuse abnormality is, yes.

Q. All right.

A. But it’s not in this case because we are looking at a tumour which is diffuse in its margins. It has appeared and it wasn’t there before. So this equates it as being a mass.”

53. I found this evidence confusing and it did not assist me in understanding how Professor Whitehouse defined a mass.

Right CC image 2015

54. On the right CC image (no. 2 in his book of images) he referred to the fact that there was “some distortion of the architecture, well normal architecture, but particularly there is some spiculation, and there is sort of pointing of little needles. It is not needle like extensions. It is not actually – it is difficult to tie in with the mass …”. He later said you could see “probably some architectural distortion” also. When asked to point out the architectural distortion, he said “well, it’s sort of going with it really. It’s less specific but it’s a spiculation which is dominantly here” . On being asked in cross examination to compare the CC view on the 2011 mammograms with the 2015 mammograms (CC view, slide 12) Professor Whitehouse reversed his view, stating in respect of the 2015 CC view, “… there is architectural distortion of here and there is, subtly, spiculation, but certainly architectural distortion, in comparison to this in 2011, which is just an overlaying of normal tissues. But I think here that the arrangement of the tissues around there does indicate this combination of architectural distortion and spiculation”. He also said that in respect of the 2011 mammogram, it was just an overlaying of normal tissues.

55. In relation to spiculation, on having it put to him there was no drawing together in a star shaped fashion, he said there was a vague star shape but that the radiations were not complete in a circular fashion, being a variable shape, not a regular, star shape or not complete in a star configuration. He said with ILC, “it is difficult to make out, perhaps, a main mass from which, a main nidus from which you are seeing spiculations appearing … They are not entirely in a circle… It reflects the rather sort of invasive ill-defined nature of the basic tumour. You have got the spiculation coming from various edges of the infiltrative carcinoma” .

56. Responding to the proposition put to him in cross examination that on the CC view there were similar lines to those he was identifying in 2015 in all the other mammograms (i.e.

2007, 2011, 2013) and that none of them were spiculation properly so called, he responded that there was a sort of linear appearance in places but it was not in the same linear configuration as in 2015. In respect of the reports from December 2015 he observes that “the spiculation … is far more marked. The mass is appearing denser. It is more irregular. It is bigger”. He was asked to consider the MRI of the breast taken in December 2015 (page 48 of the core book of medical records) and notes that it is described there as a “non-mass enhancement” as it doesn’t look like a mass, because it is so diffuse and spreading it has been called a non-mass enhancement but the way it has been taken up by the dye indicates that it is a carcinoma.

57. On re-examination, discussing what he believed to have been present in June 2015, he said “Obviously it did look like a cyst to the radiologist, they passed it as normal”. There is no evidence that either radiologist identified a cyst. He confirmed that as per his report, he could not find a contemporaneous radiological report from the mammogram and not to have reported it or followed up was negligent. On cross examination, he gave evidence that he was aware of hindsight bias and tried to avoid it at all costs and accepted that when he was reviewing the mammograms, he knew it was a lady who had developed cancer. He accepted that for 10 years he had had no experience in looking prospectively at screening mammograms and that in the vast majority of the cases he had looked at in the last 10 years in the context of his medico-legal work, the woman had cancer. He gave evidence that cysts were very unusual after the menopause.

Evidence on clinical presentation

58. Professor Whitehouse produced a short two-page report exclusively on the clinical presentation issue on 14 February 2020. He identified that he had recently received a copy of the breast screening document pertaining to the plaintiff’s visit to the screening unit on 17 June 2015. The totality of his criticisms is contained in the following paragraph: The Radiographer’s comment on Ms. Freeney feeling a lump and intolerance of compression is sufficient to require attendance at an Assessment Clinic …The Radiographer’s comment, even with the mammogram being of diagnostic quality and even in the unlikely event of the Radiological opinion being “Normal” should also result in Assessment referral. Not to have done so was a breach of duty.”

59. In his direct evidence, he was asked to comment on Ms. Mathew’s Note and he said it gave a steer to look particularly carefully at the right breast, noting that cysts in postmenopausal women were very unusual indeed, and that with that record, a red sticker ought to have been put on the image. Referencing the letter “N” placed on the right MLO June 2015, he said that was “the equivalent of the “red dot” used in the UK, which, when placed on the record of a mammogram, dictates the woman be referred to the assessment clinic. He also said the Note fell within the comments, made in the Quality Assurance Guidelines i.e. that if a radiographer has a concern it should be referred for assessment. However, he did not address the Guidelines for Clinical Recall in his direct evidence or the distinction between a new and existing lump. On cross examination, he accepted he was not criticising those Guidelines.

60. When told by counsel that the plaintiff had told the court that she was sore and that she had a lump, and asked to comment, he indicated that the lump was something that would need to be investigated by triple assessment but that the pain was a secondary symptom.

He concluded there was a breach of duty “on the fact that neither the radiographer took it any further with the red dot and the radiologist, if there was a radiologist, didn’t do anything about it either”. However, when cross examined, having been shown the Guidelines on Clinical Recall, Professor Whitehouse accepted that the radiographer did the right thing by making a note of it. Importantly, when considering 4.4 of the Guidelines, he stated the judgment of the radiologist to refer or not “would be based on their interpretation of the mammogram and they would’ve had a steer to the right breast and if they thought there was no cause for alarm, there are no worrisome features, then they would not take any further action.” In response to a question on cross-examination as to what should happen where there was no cause for concern in respect of the mammogram, he said “Yes I think that would be enough. They rely on the mammogram, which after all would be the first line of investigation if it went to assessment”. However, on re-examination, he reverted to the approach taken in his written report to the effect that irrespective of the mammogram, it was a lump that should be evaluated.

Evidence adduced on behalf of the Defendant in respect of alleged negligence

61. Factual evidence was given by Dr Coffey, Dr Pender and Ms. Mathews. Dr Given-Wilson, consultant radiologist, was called as an expert witness. At the trial, objection was taken to Dr Coffey giving evidence at all on the reading of the images, on the basis that (a) she would be doing so as expert witnesses and Order 39, Rule 58(3) (as introduced by S.I.

254/2016) only entitled a party to one expert on each topic (b) she would be using her expert status to self-corroborate her own actions and (c) she could not recall reading the plaintiff’s mammographic images. (Neither radiologist recalled reading the plaintiff’s images or the Note. This is unsurprising given that each of them read 20,000 mammograms per year). On 3 March 2020, I gave a ruling (Freeney v. HSE [2020 IEHC 115]), to the effect that Dr Coffey was not being put forward as an expert witness but rather as a professional witness as to fact and that she was entitled to give evidence on the facts the subject of this case. In relation to recall, I indicated that I would decide upon the weight to be given to her evidence at the close of the trial.

62. This objection was renewed in the plaintiff’s written submissions, where it was said that no weight should be placed on the evidence of Dr Coffey and Dr Pender in respect of how they would have interpreted the Note or read the mammograms because of the factors identified above. Curiously, despite this submission, heavy reliance was placed on certain aspects of Dr Pender’s evidence in respect of the interpretation of the Note. In any case, I was satisfied that both Dr Coffey and Dr Pender, while not recalling reading the plaintiff’s images and Note, were able to provide a clear account of their normal practice faced with mammographic images and a note such as those in evidence. The fact that both radiologists continue to work at BreastCheck up to the present day was no doubt of assistance in this respect. In respect of the complaint that they were giving expert evidence, as per the ruling on 3 March, Dr Pender and Dr Coffey did not give evidence in respect of their practice as experts but as factual witnesses. Their evidence was central to understanding the events the subject matter of the proceedings. Every person charged with professional negligence, whether they are named as defendants or not, are entitled to have their version of events heard if they decide to give evidence. For those reasons, I consider it appropriate to treat their evidence in the same way as that of any other factual witness.

63. It is useful at this point to engage with the obligation of the plaintiff to establish negligence separately on the part of each radiologist. The defendant asserted there was a high onus on the plaintiff as she was obliged to show that Dr Pender and Dr Coffey separately “got it wrong”. In response the plaintiff observed that in Morrissey, four cytotechnologists had made two separate errors, and noted that if something was missed that ought to have been seen, it was irrelevant how many radiologists missed it. In my view, the fact that two professionals acted in the same way in the same situation does not absolve the court from applying the usual Dunne/Morrissey test as to whether those acts were negligent and I have proceeded on that basis. Dr Coffey

64. Dr Coffey is a consultant radiologist who works full time in BreastCheck. She qualified in 1980, became a registrar in the Mater Hospital in radiology, and then went to the Mayo Clinic in the US for a two year fellowship including body and breast imaging. She stayed on there as a consultant for two years, then came back to Dublin and worked in the Blackrock Clinic where she established their breast imaging service. When the breast screening programme began in 1999/2000 she joined BreastCheck and worked half the time there and the other half in St. Michael’s hospital in Dun Laoghaire, where she did general radiology. In 2016, due to an increasing workload when the BreastCheck programme was extended to age 69, she joined BreastCheck full time.

65. Her special interest is breast imaging. She reads about 20,000 mammograms per year and she works in the assessment clinics for women who have been recalled back following a screening mammogram. When women are recalled they have further imaging with dedicated mammography (since 2017 using digital breast tomosynthesis) and focused ultrasound trying to further evaluate the problem seen on the initial mammogram. Some of those women will require ultrasound-guided core biopsies and she does those also.

66. She was asked to look at the images from the plaintiff’s mammograms when being examined in chief. Dr Coffey explained that one of the first things she looks for is whether it is a technically good mammogram i.e. whether the whole breast is included, whether there is no blurring, no skin folds overlying the tissue etc. There were no technical issues with the plaintiff’s images from 2015. On the MLO view, there was some asymmetry in the right breast, in that there is a sort of white patch there which is not symmetrical to the left, what would be described as an asymmetrical density. There was no corresponding density on the CC view. Therefore, there is no mass as for a mass to be labelled as a three-dimensional abnormality, it must be there on both views. Accordingly, she compared the asymmetry with earlier mammograms, being 2011 and 2013. She says the asymmetric density has remained unchanged over four years, that worrying features do not persist over four years unchanged and therefore she would have no cause for concern. Professor Whitehouse’s concerns in respect of spiculation and distortion on the CC images were put to her and she said she saw no spiculation or distortion on those images. She compared them with the 2013 images and saw no change. She said she thought it was unlikely she would have called for the 2007 images as she was reading a 2015 mammogram and had a 2011 mammogram with no change.

67. She gave evidence of the codes the radiologist can enter after reading the mammogram, going from R1 for no abnormality to R5 for an abnormality that is thought to be definitely malignant. A woman who is R1 or R2 will not be recalled but R3 or over will be recalled.

Dr Coffey noted that when a woman is first invited to participate in BreastCheck she will receive a leaflet identifying what breast screening is, what happens during a mammogram, her BreastCheck appointment etc. The leaflet identifies that not all breast cancers can be found by a mammogram. When a woman is invited to a specific appointment, she will receive a leaflet that again gives information about the process and identifies that a small number of cancers are not found by mammogram and that even after a screening, women should stay breast aware and look at and feel their breasts for changes. After a mammogram, if a woman receives a R1 or R2 reading, she will receive a letter stating that the mammogram was considered to show no abnormality, that she will be recalled in another two years if she is still within the age range, that in a very small number of instances breast cancers are not detected by mammography and that she should go to her GP should she notice any change or symptoms.

68. Dr Coffey then explained the circumstances in which a radiographer who carries out a mammogram makes a note on the patient’s records and she explained that it was the radiographer’s way of communicating an issue to the radiologists, usually explaining why the mammogram may be technically imperfect, either necessitating a technical recall so another mammogram can be taken or indicating why there is no point in initiating a technical recall, for example if the patient is unable to cooperate. She also explained the notes can also tell the radiologist anything else the radiographer considers relevant. Dr Coffey gave evidence that the changes to point 4 in the Guidelines for Clinical Recall (identified above) updating the signs classed as clinical recall were intended to simplify matters and make it easier for radiographers to know what a clinical recall is, since their tendency was to want every patient back for fear of missing something. She went on to indicate that a “new palpable lump” was really a more critical finding – that a lot of patients would have a palpable lump they know about, they have had it previously assessed, and that is not what they wanted in a clinical recall. Dr Coffey expanded on this in cross examination .

69. She indicated that a radiologist can override a clinical recall but that would be relatively unusual. Dr Coffey explained that if she felt a note by a radiographer was telling her something that should have been a clinical recall, then she could upgrade from a note to clinical recall, including where the mammogram was normal. She observed that in the vast majority of cases where there is clinical recall there is also a radiological abnormality.

Dr Coffey then explained the procedure she would follow if she was reading a mammogram and the letter N appeared on it, stating that she would assess whether it was of importance in the technical aspect of the image, would consider whether it should be upgraded to a clinical recall and would always read the note and make a judgment on it. Turning then to the actual Note made by Ms. Mathews, she identified that she interpreted it as follows:

“A understand this note to be the radiographer telling me that she feels the compression may be suboptimal in that breast and her reasons for that are that the patient is telling her that she has pain and tenderness which the patient is putting down to a cyst.”

70. In fact, there was general agreement that the images in this case were of good quality and there was no necessity for a technical recall. When asked about the statement “lady feels a cyst on the right breast” in cross examination, and whether it meant the plaintiff felt with her fingers what she describes as a cyst in her right breast, she answered “No, I don’t because that would have been a palpable abnormality which, if new, the radiographer would have … made his clinical recall”. Dr Coffey observed that it would be unusual to see a note that constitutes a clinical recall as the radiographers are very good at following the SOP (Guidelines on Clinical Recall) and are diligent about it and interested in it.

71. In relation to the feeling of tenderness, Dr Coffey gave evidence that on a mammogram, the feeling of tenderness typically waxes and wanes so it would not be unusual for somebody to have two or three mammograms without a problem and then find another one particularly uncomfortable and then find the next one comfortable. There was some considerable debate during the evidence as to whether a patient could be expected to know whether a mass was a cyst or a lump, and the significance of a patient’s designation of an abnormality as a “cyst”. Dr Coffey explained that if the radiographer formed the impression the abnormality was new, whether it was described as a lump or a cyst, the woman would get a clinical recall. “

72. In relation to pain, Dr Coffey gave evidence that she did not regard a woman complaining of pain, such that she couldn’t tolerate compression, as a significant sign or symptom.

Dr Pender

73. Dr Pender is a consultant radiologist who works in BreastCheck and St. Columcille’s Hospital. After obtaining her basic radiology qualification in Beaumont Hospital and the Royal College of Surgeons, she became a fellow of the Faculty of Radiologists in RCSI in 1996 and a fellow of the Royal College of Radiologists UK in 1997. In the same year, she went to Ontario and did a two-year fellowship to acquire specialist breast training as well as general radiology. As part of that she did symptomatic work and worked in the Ontario Breast Screening Programme, which provided for a total referral basis of 20,000 mammograms and 3,500 breast ultrasound studies per year. She returned to Ireland as a consultant radiologist in 1999, dividing her time between BreastCheck and general radiology work in St. Columcille’s Hospital in Loughlinstown. She reports on approximately 20,000 mammograms per year in BreastCheck, as well as participating in weekly triple assessment clinics, weekly multi-disciplinary meetings, teaching of rotating radiology registrars, radiographers and breast care nurses and quality assurance. She was chair of the BreastCheck QA committee 2006-2009. In 2003 she co-authored the “Quality Assurance in Radiology” chapter of the BreastCheck “Guidelines for Quality Assurance in Mammography Screening” 2nd ed. She participated in the drafting of the Guidelines on Clinical Recall (SOP), which she described as a collaborative undertaking between the radiographic service manager, herself, and the radiologist and the Clinical Director to formulate the policy for the unit. The SOP has been reviewed in an external QA accreditation process by EUREF, being the European Reference Organisation for Quality Assured Breast Screening and Diagnostic Services

74. She confirmed that at the time of the plaintiff’s mammogram in June 2015, the SOP was the applicable document. As with Dr Coffey, Dr Pender did not recall reviewing Ms.

Freeney’s mammograms from June 2015. However, she was able to give evidence of her practice when reviewing mammograms and confirmed that she would have followed that practice in relation to the plaintiff’s mammogram. Her practice is to look at the four images from the current mammogram first and then compare each of the four images with the oldest mammogram available for the patient (in the plaintiff’s case, the 2011 mammogram). She then looks at the two current craniocaudal projections, then the two current mediolateral projections, then all four images together. She then pulls up any intervening mammograms and compares them with the previous images and with the current images. If there is a circle on the mammogram which says either “clinical recall”

or “note”, in the old practice (which I understand was the applicable practice in 2015) she takes the data sheet out and makes any decision on that and then goes back to the original eight hanging i.e. the four images from the current mammogram and the four images from the oldest mammogram available and then makes her decision and either recalls the patient to clinical assessment or assigns them to routine recall.

75. Given the available images in this case, Dr Pender gave evidence that she would have looked at the 2015 mammogram, compared it with the 2011 mammogram and then compared it with the 2013 mammogram. She thought it unlikely she would have looked at the 2007 mammogram. Dr Whitehouse’s views on the 2015 mammogram were put to her and she indicated that she did not agree with them. Specifically, she said she saw asymmetry of the parenchyma but it was stable and hadn’t changed. She did not see a discernible mass on any of the projections and she did not see spiculation or architectural distortion on the MLO or the CC projections for 2015. She said there was no significant change in the mammograms between 2011 or 2015 that concerned her.

76. She confirmed by reference to the records that she had read the mammogram on 18 June 2015. She identified what would happen where there was a dispute over a recall and indicated that in BreastCheck, it is done by “highest reader”, so if, after a discussion between the two radiologists who had read it and possibly a third radiologist, there is no consensus, then if any one person feels strongly that they would like the patient recalled to assessment, then that is what happens. She gave evidence that she looks at every clinical recall that comes through the unit and reviews the indications for same and feed back to the radiographers as part of the teaching programme in terms of the indications they have made for clinical recall if they fall into the SOP or not and look at the outcomes for those women.

77. When asked whether had she or Dr Coffey or Professor O’Doherty expressed a concern that radiographers were unnecessarily recalling women, she indicated that this was the reason for having the SOP and constantly reviewing the clinical recalls. She noted that a negative effect of screening that they are criticised on and rightly so, is recalling women to assessment who don’t have cancer either clinically or on the basis of their mammograms as they have a responsibility not to recall someone if it is not warranted.

She notes that sometimes radiographers err on the side of caution and if in doubt they will assign a clinical recall. She noted that this was the purpose of reviewing every clinical recall and doing the audits, to look at the indications, see if they fall within the Guidelines and feed back to the radiographers to refine what they do. But she indicated that the radiographers and radiologists err on the side of caution and if there is any doubt for the clinical recall, unless there is a clear reason why it shouldn’t be a clinical recall, the woman is brought back for assessment. She was extensively cross examined about the issue of erring on the side of caution and indicated that if a woman meets the indication for clinical recall in the SOP, then she should be recalled as a clinical recall and that where a radiographer has any doubt, they tend to assign a clinical recall.

78. In respect of the programme, she noted the programme was a mammographic screening programme and all guidelines, including those of BreastCheck, the NHS and European would suggest that although it is primarily a mammographic screening service, there must be a mechanism for the radiographer to alert the radiologist of the most significant findings. She said:

“A. The information that has been proven to give the best benefit in terms of increasing or picking up cancers is whether the patient has a new lump and that is what we want to focus on and not any other information, and women who have a new lump tend to volunteer that spontaneously to the radiographer and that is what we are looking for and that is what the SOP is focused towards finding”.

79. She added that breast symptoms are extremely common, that the screening programme is a mammographic screening programme and the BreastCheck literature tells women if they have any symptoms that they are concerned about, they should go to their GP and that is for the symptomatic service. She said BreastCheck do want to pick up the most significant clinical findings of a new lump and recall those patients and that is what the Guidelines recommend they do.

80. In response to cross examination about the lack of a similar note to that of 17 June 2015 on the plaintiff’s previous mammograms, Dr Pender indicated it was very common to have a note on one mammogram and not on any previous or other mammograms. She indicated that if a woman says she has a lump in her breast for a significant amount of time or one that has been investigated previously, and that lump has got bigger, then if a note was made about that by the radiographer, she would convert it to a clinical recall.

Equally she would accept that as a clinical recall if the radiographer had assigned it as such. She said she would use clinical judgment on it as it didn’t fall into a neat category.

81. In respect of the wording of the Note, she interpreted it as follows:

“A. Up to 20% of our screening mammograms will have a note attached to them. My interpretation of this is the radiographer is explaining to me that the woman is finding it difficult to tolerate the compression due to pain or tenderness, that she is feeling pain or a sensation which she is attributing to a cyst or something similar to what she has experienced before with a history of cysts. But in itself, it doesn’t give me cause for concern. I feel the radiographer is saying that the woman found it very difficult or painful to tolerate the compression. Therefore, the mammogram might not be of optimal or best quality, but that is the best that she can achieve.

So then I look at the mammogram and again confirm it is of diagnostic quality. ”

82. She accepted that the Note could be open to other interpretations and that someone else might find a doubt in what the Note means but she stated she was happy with her interpretation of the Note. She indicated that the clinical recall is designed to catch women who have a significant symptom who have not already attended the symptomatic services.

83. In respect of the reference in the Note to the plaintiff feeling a cyst in her right breast, Dr Pender observed she may be feeling a symptom rather than a lump, possibly pain or tenderness or a sensation similar to what she may have noted previously with cysts. She noted that word “feels” doesn’t mean it is palpable. Equally the sensation of pain can be more widespread, thus it is not necessarily focal. She said that when a woman comes into a mobile unit, and if she feels a lump, a new lump, she is usually very anxious to communicate that as early as possible to the radiographer. She said that in breast screening they don’t want to recall cysts as they are not usually cancerous and there is a very low incident of cancer. If the mammogram is normal, then it would be their practice not to recall such a patient. If a patient has a symptom that is bothering them or causing them significant symptoms, including a large fluid filled cyst, then the route is through the symptomatic services. The screening mammogram is to look for cancer on the mammogram. The appropriate route is through the symptomatic service.

Ms. Mathews

84. Ms. Mathews is a radiographer who did a Batchelor of Radiography and a postgraduate mammography with BreastCheck in Dublin and who worked in mammography exclusively from 2007 to 2016 in BreastCheck. In 2016 she went to Perth and is working as a radiographer in mammography there. She gave evidence by video link from Australia. In common with the radiologists, she could not remember her specific interaction with the plaintiff and thus gave evidence of what her usual practice would be, as prompted by considering the records in respect of the plaintiff. First, she was clearly aware of the SOP and said she always followed it, so if a patient told her there was a new lump or skin dimpling, inflammation or a recent nipple inversion, she would bring that to clinical recall.

She confirmed she followed the SOP. When asked if a patient didn’t tell her of those situations, but there was information she wanted to pass onto the radiologist, she said she would write the details and make it a note. She characterised the matters she would write down as follows: “If the patient tells me something they want the radiologist to know, something significant, history, then I would write that down so the radiologist is made aware”. She indicated she goes on the basis of what the patient tells her and that does not know a patient’s history unless the patient tells her. When asked how she would treat a lump reported by the plaintiff that didn’t fall within the SOP, she indicated, by reference to the Note she had made, as follows: “Yes if it’s an existing lump or if there has been a history, I would put that, document it exactly like that so that the radiologist can see it, yes” .

85. In cross examination, she indicated the difference in approach she takes depending on whether what is identified is a new lump/cyst or not. For a new lump she indicated that she would ask when the patient felt it so there is a timeline and that “For lumps that is not new, you just write history of lump or cyst or whatever, and then make a note of what the lady has told you, or the patient has told you, so the radiologist can decide once they see the mammogram” .

86. In relation to a possible different approach depending on whether a lump or a cyst is identified by a patient, Ms. Mathews, when asked whether she saw a difference between a woman saying to her that she had a lump or a cyst, explained that she was not a doctor, not a physician, just an imaging professional, so that when a patient tells her something, that is what she writes down. An issue arose between the parties as to whether the plaintiff had in fact identified that she “felt” a lump or cyst in her right breast at the time of her interaction with Ms. Mathews. Contrary to the defendant’s approach that the plaintiff had not identified this, the terms of the Note itself make it clear that the plaintiff said that she “felt” a cyst in her right breast. When asked why she wrote that down, her evidence was as follows:

“A. Because that is what she has told me, and during the mammogram, I don’t know, maybe she must have felt some tenderness or pain.

Q. Yes

A. In the right. So maybe she associated it with the cyst, meaning that she was feeling, which she had mentioned to me, and I have written that down. ”

87. Ms. Mathews went on to explain, when asked again why she had written down “lady feels a cyst in her R breast”, that the patient had told her this, that she had to document it for the radiologists so that the radiologists would know it, and that whatever the patient tells her she has to put it down for the radiologist. In relation to her approach when writing a note, under cross examination, on being asked whether she had to exercise a certain amount of professional judgment in what she writes, she replied that she can only write what the patient tells her and can’t interpret it or make it her own way and that she writes down all relevant information.

Dr Given-Wilson

88. Dr Given-Wilson, consultant radiologist, qualified as a doctor in 1980 and spent nine years doing specialist radiology training in King’s College Hospital in London, the last year and a half doing a breast fellowship. In 1989, she was appointed a consultant at St. George’s Hospital to set up their breast screening unit covering the whole of southwest London, with about 150,000 women being regularly screened on a three-year cycle. Dr Given- Wilson directed that unit for ten years and was then clinical director of St. George’s Hospital for ten years, until 2015. She then stepped back from management but has maintained her clinical practice in breast screening, spending two full days a week in the breast screening unit i.e. four sessions, including overseeing an assessment clinic and the screening multidisciplinary weekly team meetings. She is heavily involved in organising the screening programme in England, chairing the Adult Reference Group which looks at the evidence for all adult screening programmes and reports on that to the UK National Screening Committee, the group advising England, Scotland, Wales and N. Ireland on what screening programmes to undertake and how to do them. She is on the Clinical Professional Group and the Breast Advisory Screening Committee, two of the major committees overseeing the running of the English national screening programme. She is a member of the National Radiology Quality Assurance Committee for breast screening which oversees the production of the QA guidelines ensuring the quality of the screening programme. She is a member of the National Advisory Committee on breast cancer screening, which advises on more major modifications to the UK National Screening Committee. She has published very extensively on matters relating to breast radiology.

89. Dr Given Wilson confirmed in evidence that she meets the Guidance on screening and symptomatic breast imaging referred to above, i.e. she undertakes a minimum of 5,000 screening and/or symptomatic cases per annum.

Evidence on reading of mammographic images

90. In her report, Dr Given-Wilson reviewed the plaintiff’s mammograms from 2007, 2010, 2011, 2013, June 2015 and December 2015. However, as with Professor Whitehouse, I will focus on her review of the 2011, 2013 and June 2015 mammograms. She concludes that each of the three mammograms is of good quality and show dense glandular tissue, no mass or suspicious features and no change from previous films. In the mammogram of December 2015, she considers there is some diffuse increase in the overall density of the upper right breast compared to June 2015 but no obvious mass, distortion or suspicious calcification. She notes the appearance of increased density on the right is subtle and non-specific but is likely to represent the palpable mass and area of abnormality better demonstrated on the ultra sound and MRI.

91. She expresses the view that there were no suspicious features on the plaintiff’s screening mammograms from 2011-2015 and only a subtle increase in density to suggest cancer on the December 2015. She concluded it was still mammographically subtle at the time of the plaintiff’s diagnosis in December 2015, when other imaging and biopsy confirmed the diagnosis.

92. Dr Given-Wilson noted that the plaintiff showed very dense glandular tissue on all her mammograms and that cancers not seen on mammography are more common when the breast tissue is dense with glandular tissue difficult to see through, as this reduces the ability of the reader to distinguish masses from tissue. She noted that mammography shows reduced sensitivity for diagnosis of cancer in women with dense breasts from 88% of cancers being detected in women with fatty breasts to 62% in women with very dense breast tissue.

93. She concluded that the cancer was not visible on the screening mammography in June 2015, therefore could not be diagnosed on screening and that the screening mammogram was accurately reported as there were no diagnostic features of malignancy to be seen.

94. In her oral evidence, she identified that in the process of compressing the breast for a mammogram, often there will be minor differences between one film and the next, because the way that the tissue is arranged and the way that it is squashed means that it will never be absolutely identical. She noted that sometimes in different views there are bits that are denser but that one must go to the second view of the mammogram to see whether the density persists or whether it is transient i.e. positional or composite shadows that is a reflection of the way it was squashed.

95. She observed there are multiple lines right throughout the breast tissue, being ligaments.

She noted a spicular mass is a central density with radiating lines from it and that it is necessary to consider if lines on a mammogram are actually radiating from a central point or are simply the sort of overlapping lines that often do cross over that one would expect to see as part of the normal effects of compressing a breast from mammography. She observed that when she read the films, she did not think there was any significant difference between the two series of films being 2011 and 2015. She returned to this theme in cross examination, having had put to her Slide 8 (MLO 2015 and 2011) and Professor Whitehouse’s view that it showed an increased density, a mass and was irregular in shape:

“A. If the whole of the rest of the breast was fatty and this was the only bit you had, then you would be much more focused on it. But the point with this set of mammograms is that this lady has very similar islands of glandular tissue throughout … part of the problem with looking at something like this is you get hindsight. You say “That was in the area where the cancer developed, so that dense bit must be the cancer. Actually, when you look at these films over the series and even when you look at an individual one, that is the normal pattern of her glandular tissue. She has very dense islands. They have little kind of linear bits around the edge because of the ligaments that come into them and it makes it very difficult to pick up cancers in tissue like this because it may not look that different from what the normal tissue looks like” .

96. When asked about what one would expect to see if ILC was present, Dr Given-Wilson explained one would expect to see a focal density and to call something a mass it needs to be seen on the MLO and CC views. She would expect to see spiculations coming in a star-like pattern to the centre of it. Where a density is only seen in one view, it is called a density or asymmetric density and not a mass. Later, when looking at the MLO 2013 (no. 3 on Professor Whitehouse’s book) and having had Professor Whitehouse’s view that there is light speculation on same, she says she does not spiculation, she sees overlapping lines, consistent with the ligaments that hold the breast tissue together .

Commenting on slide 7, being MLO 2015 compared to 2013, she notes the breast tissue on the area identified by Professor does appear a little denser than 2013. But then going to slide 8, 2015 as against 2011, she notes the density looks very similar to how it did in 2011 and indeed 2007 (previously looked at). She concludes there is a slightly subtly varying density that kind of goes up and down from examination to examination as the breast is compressed and arranged position slightly differently. Referring then to slide 6, where views for 2007, 2011, 2013 and 2015 are seen, she says: “A. the 2015 one, slightly denser and on the 2013, slightly less dense, all the tissue is a little bit better spread out. On the 2011, it is slightly denser again and on the 2007, even denser and probably closest to the 2015. So I don’t see a progressive change to a developing mass… ”

97. In relation to the CC view, comparing the 2011 and 2015 view, she says there are lines in the area of the breast where the cancer was found but that they are no different from the lines seen back in 2011 and that the overall arrangement of dense tissue standing is no different, with the general pattern staying relatively similar between those two films. She observed that if the court concludes there was a tumour in June 2015 that did not alter her view about the acceptability of the mammograms, as in screening you can only go on the evidence in front of you i.e. the mammogram.

Evidence on clinical presentation

98. In her evidence on clinical recalls, she identified that because some cancers don’t show up on mammograms but are producing a clinical lump, if a woman comes for screening and identifies a very significant symptom, such as a new palpable lump, then there is an option for that to be brought to the attention of the radiologist for a clinical recall to be initiated. She observes that it is a very small number of the recalls, the vast majority being based on mammographic abnormalities. She gave evidence that one would not normally recall for a palpable lump that was not new, as a number of women have lumps that are longstanding.

99. In relation to the Note, when asked what the radiologist should do, Dr Given-Wilson gave the following evidence:

“A. They would consider it. It’s clearly a piece of information that is being brought to their attention but I think if it doesn’t fall within the SOP then they would not initiate a clinical recall on that basis. It is very common that one gets feedback from radiographers at the time of mammography that the women has pain or has a history of previous problems, or various things which do not have the same sort of trigger value as a new lump. So they have got an SOP, I think they would judge whether it fell in under the remit of the SOP and if it did recall and if it didn’t not recall” .

100. In cross examination she explained that it happens very frequently that women make comments about things like pain or lumpiness or some history when they come for screening and there must be a process as one should not recall those women and therefore there must be focused on the real trigger symptoms. She noted that if plaintiff had used the word “lump” in the context of “I have a new lump”, then it should have triggered a clinical recall as per the SOP.

101. Dr Given-Wilson was cross-examined about the necessity for a clinical recall arising out of the Note and was consistent in her answers – that if the plaintiff had referred to a new lump there should be a clinical recall. She differentiated between the diagnostic process and the screening process, which as she noted was focused on people who are not symptomatic, where the mammogram is used to look for evidence of cancer but noted there was also a kind of safety net, whereby if the patient mentions a significant symptom they can get a clinical recall on that basis. However, she observed that it was not the same sort of route as the diagnostic route where the patient is complaining, and the doctor refers them on to the diagnostic clinic.

Conclusions on the reading of mammographic images

102. Before addressing the specific evidence given, I outline my views on the qualifications and approach of both expert witnesses in respect of the alleged negligent misreading of the mammographic images.

103. Starting with Professor Whitehouse, he proceeded on the basis that (a) no radiologist had read the images and (b) there had been a failure to provide a report on the plaintiff’s June 2015 mammograms, which failure was negligent. I find that neither of those assumptions were correct.

104. In respect of the reading of the images, he gave his report and his direct evidence while under the impression that no radiologist had read the mammogram. Had this been the case, it would have represented a very serious systemic failure on the part of BreastCheck. But in fact, as is clear from the summary of facts above, Dr Pender gave the first radiological evaluation on 18 June 2015 and Dr Coffey the second on 20 June 2015.

Professor Whitehouse was criticised for this assumption by the defendant. In fairness to Professor Whitehouse, the document at page 216 of the core book of medical records entitled “radiology data sheet” might suggest that the mammograms had not been read, as there is a space for a signature from both radiologists but no signature. However, that document itself calls for further investigation by anyone seeking to understand the plaintiff’s review, since although the plaintiff’s name, address and the date of her visit appear on it, nothing else is filled in, including the section for signature by the radiologists or the outcome section. The document does not permit of a conclusion that no radiologist viewed the images; because it is blank in the material respects, it ought to have prompted inquiry. (The explanation for the lack of entries was given in evidence: the paper form is no longer filled out and the information is now entered onto the computer).

When one goes to the print-out of the relevant computer records, produced in the supplemental book of medical records produced by the defendant, one can see that Dr Pender and Dr Coffey both signed off on the images taken in June 2015 and both classified them as R1 (meaning no concern) (pages 205 to 208). Apparently, those computer records were not provided to Professor Whitehouse until 24 February 2020, the day before he gave evidence. In any case, counsel for the plaintiff tells me that it was not until cross examination that a legible copy of the IT records showing a sign off by the radiologists was provided to Professor Whitehouse.

105. At no point did Professor Whitehouse inquire into this stage of affairs. He prepared his report on the basis that the mammographic images had not been read and explicitly stated this was the case in his second report. A significant amount of his evidence was given while he was under a misapprehension as to what had happened. I find it difficult to understand how Professor Whitehouse could have concluded there was negligence without forming a complete picture of the screening of the plaintiff, including confirming that the images had not been read by radiologists. In fact, Professor Whitehouse only became aware that two radiologists had read the mammogram while he was being crossexamined. Yet strangely he did not identify it as negligent for a screening service to fail to ensure the mammographic images were read.

106. As identified by Clarke J in Morrissey, expert evidence may assist the court in deciding on the facts of the case whether the conduct of the relevant professional has lived up to whatever standard of approach has been established (paragraph 6.37). When, as in this case, the expert witness has not correctly identified what that conduct was, it inevitably gives rise to a loss of confidence in the thoroughness of that expert’s preparation.

107. Second, I have serious concerns about Professor Whitehouse’s view that it was negligent not to have produced a radiological report following the plaintiff’s screening. In both his report and his oral evidence, he identified the lack of a report. But both Dr Coffey and Dr Given-Wilson gave evidence that screening is a binary process and no report is produced, Dr Coffey explaining:

“A. It’s a binary decision. So we recall or we don’t recall. It’s not like a mammogram report in a hospital where we would report that this is the density, this is how we’ve assessed it, etc. etc., we don’t make any report like that in a screening system.”

108. The lack of a report was not negligent but rather the normal procedure in BreastCheck and elsewhere. The fact that Professor Whitehouse was not aware of this left me with the impression that he was not familiar with the nature and purpose of breast screening, either in BreastCheck or more generally, and lacked an understanding of the process.

Moreover, he was willing to opine that two radiologists were negligent for not producing a report without checking whether reports were in fact normally produced in a screening context.

109. Third, having regard to the evidence of Dr Given-Wilson, Dr Coffey and Dr Pender, I find that all available mammograms should be viewed by the reader for a proper comparison.

In this respect, I was struck by the evidence of Dr Coffey that worrying features do not persist over four years unchanged. Yet Professor Whitehouse’s core conclusion i.e. that there were suspicious changes on the 2015 mammogram, appears to have been based exclusively on a comparison with 2013. In giving his evidence to the court, he provided a set of images that were exclusively from 2013 and 2015. He did not include any 2011 images, despite his own evidence that in cases with obvious change or suspicious factors it was necessary to go two series back. Here, he concluded there were features “highly suspicious” of cancer and yet he did not compare the 2015 mammogram with 2011.

Professor Whitehouse did not explain this approach. It is certainly true, as pointed out by counsel for the plaintiff, that he gave evidence on the 2011 images, but this was because he was specifically asked to do so in cross examination. When asked whether, if the abnormalities present in 2015 were also present in 2011, that would exclude suspicion about the 2015 abnormalities, on seven occasions he answered on the basis that 2011 looked different to 2015 but did not answer the question being put to him. Either Professor Whitehouse did not understand or was anxious to avoid the question. In either case, that exchange undermined my confidence in his evidence.

110. Fourth, unlike Dr Given-Wilson, he did not identify the impact of the variation on images from different mammograms (given that the compression will be subtly different each time) and did not factor that into his analysis of the images when analysing differences between the 2013 and 2015 mammogram.

111. Fifth, when referring to increased density on the MLO view, he did not distinguish between the upper outer quadrant of the breast and the remainder of the breast. Dr Given Wilson indicated that there were “islands of density” throughout the entirety of the breast and that she saw no distinction between those in the area where cancer had ultimately developed and the remainder of the breast. Professor Whitehouse did not distinguish between areas of increased density in the upper area from those in the other areas of the breast or explain why the density he saw in the right upper outer quadrant was different to the areas of density elsewhere in the right breast.

112. On the other hand, Dr Given-Wilson impressed me as a witness. First, she had run a screening service for ten years in St. George’s covering the whole of southwest London.

She has extensive screening experience. Second, she continues to screen on a regular basis and as such meets the guidelines set by the Royal College of Radiology. Third, she has published extensively in the field of breast screening. She was able to give context to many of the issues that arose, such as variations from screening to screening due to compression. The evidence she gave was consistent and comprehensible. She spent some time describing what she saw on the images and explaining why she did not consider there were any suspicious features on the mammogram. She looked at the series of mammograms and compared the 2011 mammogram with 2013 and 2015. That is in fact the approach recommended (but not adopted) by Professor Whitehouse i.e. where there is a suspicious feature one should go two series back.

113. Turning now to the correct approach to be adopted to the reading of the slides, in Penney Palmer & Cannon v. East Kent Health Authority [2000] Lloyds Rep Med 41, Lord Woolf M.R. established a three-part test in considering whether a screening had been negligent.

The first part required the court to identify what was to be seen on the slides. He observed:

“…the judge had the unenviable task of deciding as a matter of fact which of the experts were correct as to what the slides showed. This was a task which required expert evidence. However, the evidence having been given, the judge had to make his own finding on the balance of probabilities on this issue of fact in order to proceed to the next step in answering the question of negligence or no negligence”.

114. Considering the relevance of expert evidence to factual conflicts, Clarke C.J. in Morrissey observed that where the facts involve difficult questions on which expertise may bear, it will be often useful for the court to have the benefit of such expert evidence in assessing the facts. He went on to observe that the practical approach adopted in Penney Palmer may provide a useful guide to the way in which a court should adopt its task but that ultimately the question is whether a competent screener could have given a clear result.

115. Applying that approach to the present case, I must decide whether, on the images as presented, and having had the benefit of the evidence of both expert witnesses, Dr Pender and Dr Coffey could have given a clear result. In doing so, I will address each of the features that Professor Whitehouse identified as being suspicious.

Spiculation

116. Professor Whitehouse identified spiculation as being present on the CC view, although during his oral evidence he also identified spiculation on the MLO image. I was shown a diagram of a classic case of spiculation, as well as a mammographic image of spiculation.

A similar description of spiculation was given by both experts, being a star shape arrangement of lines radiating from a central density or mass into the surrounding breast tissue.

117. Professor Whitehouse accepted this was not a classic case of spiculation, that there was a vague star shape, that it was difficult to make out and that the lines were not radiating outwards in a classic star shape. It was not possible to see a set of lines in the mammograms for June 2015, nor indeed for any other year, emanating from any central point, either in a star shape or any other shape. I accept the evidence of Dr Given-Wilson that the multiple lines present throughout the entirety of the breast tissue were the normal ligaments of the breast tissue, not radiating from a central point or pull into one place but simply overlapping as part of the normal effects of compressing a breast from mammography. That pattern was similar throughout the whole of the breast tissue. There was no difference in pattern between the area identified by Professor Whitehouse, where the cancer was later found, and any other area in the right breast and he did not identify any such distinction. I find that it was therefore appropriate to reach the conclusion, as Dr Given-Wilson did, that the lines appearing in the upper right-hand quadrant reflected the normal structure of the right breast and were not spiculation. Her conclusion in this respect was reinforced by the fact that no spiculation was identified on the mammogram in December 2015 when the cancer was diagnosed.

Architectural Distortion

118. Professor Whitehouse explained that architectural distortion occurs where the cancer stimulates a dermoplastic response which has the effect of distorting the normal architecture of the breast. I found Professor Whitehouse’s evidence in respect of distortion inconsistent. On the CC view he said there was some distortion of the architecture but then referred to “normal architecture”, “probably some architectural distortion” and said that it was “sort of going with (the spiculation) really. It’s less specific but it’s a spiculation which is dominantly here”. On being asked to compare the 2011 and 2015 mammograms, he said there was “subtly, spiculation, but certainly architectural distortion” and then said there was a combination of architectural distortion and spiculation. It was not clear which was dominant: the spiculation or the architectural distortion. Professor Whitehouse did not describe the nature of the particular distortion being identified or its precise location. Dr Given-Wilson said she did not see any architectural distortion on the mammograms. Because of the inconsistency in Professor Whitehouse’s evidence in this regard and his lack of precision in respect of the asserted architectural distortion, and because of the clarity and certainty of Dr. Given-Wilson’s evidence, I find there was no evidence of architectural distortion on the mammograms that ought to have been identified by Dr Coffey or Dr Pender.

Density/Mass

119. Professor Whitehouse identified density and a mass in the MLO view but not on the CC view. Dr Given-Wilson gave evidence that to call something a mass, it is necessary to see it on both views (CC and MLO) and at the margins, one would expect to see spiculation coming in a star-like pattern to the centre of it. She said if it is only seen on one view, it is called a density or asymmetric density. Professor Whitehouse disagreed that for a density to be described as a mass, it must be seen from two views. His evidence on the definition of a mass was lacking in precision and confusing. Given that Dr Given-Wilson remains in active practice and meets the requirements of the Royal College of Radiologists for practice as a breast radiologist, while Professor Whitehouse does not meet those criteria having been out of practice for nine years, I prefer her evidence in relation to terminology. On that basis there is no evidence of a mass.

120. Professor Whitehouse identified a patchy ill-defined area of density within the upper half of the breast, with the diameter of what he described as a mass being approximately 2cm with slightly spiculated margins and indicated that a similar mass could not be seen for the equivalent image in 2013 (no. 1 on his book). He did not make a comparison with 2011. Dr Given-Wilson gave evidence that the plaintiff has very dense islands of glandular tissue throughout her right breast dense islands, with linear bits around the edge because of the ligaments that come into them. I accept the evidence of Dr Given-Wilson that there is a slightly subtly varying density that goes up and down from examination to examination as the breast is compressed and arranged in position slightly differently, with 2013 being somewhat less dense in appearance to 2015 but 2011 being more somewhat more dense, her conclusion being that there is no progressive change to a developing mass. Rather she noted the general pattern has stayed relatively similar between 2011 and 2015.

121. On the mammographic images, there was a great deal of white/grey material, being the breast parenchyma or tissue, and some black material, being fat. The greyish/white material seemed in general to be present in the same type of shape in 2011 and 2015, with some variation for 2013. I could not observe greater density in any given year although there were patches of brighter white as opposed to greyish white that varied from year to year. In particular, when one looks at slide 6 of the MLO images from 2007, 2011, 2013 and 2015, I could see no change over the series of films. Nor, as per the evidence of Dr Given-Wilson, could I see any difference in density between the area in the right upper quadrant (where the cancer ultimately manifested itself in December 2015)

and the remaining area of the breast. None was identified by Professor Whitehouse.

122. Therefore, accepting the evidence of Dr Given-Wilson of (a) the lack of a change in density from 2011 to 2015 and (b) the pattern of density not being any different in the place where cancer was diagnosed in December 2015 and the remainder of the right breast, I do not find that the images disclosed signs of a mass or density on the MLO such that a clear result could not have been given.

Asymmetry

123. Professor Whitehouse identified that the plaintiff’s breasts were asymmetric. It can clearly be seen on the series of mammograms for 2011, 2013 and 2015 (and indeed for 2007)

that the plaintiff had asymmetry between her right and left breast density over four years and Professor Whitehouse agreed with this, while stating that it had taken on an increased heterogenous density in 2015. My findings on density above therefore dispose of Professor Whitehouse’s concerns on the basis of asymmetry.

124. For the reasons set out above, I conclude that, as there were no suspicious features to be seen, Dr Pender and Dr Coffey acted appropriately in grading the plaintiff’s mammogram as R1. However, for the sake of completeness, I consider whether, if I am wrong in my conclusions and there were concerning features on the mammogram, a failure to identify same would have constituted negligence.

125. The evidence of Dr Given-Wilson on this point was as follows. First, she observed that reading mammography was not entirely black and white, it was to some extent an art and there would be some decisions to be made about subtle things and whether they are recalled or not. Second, she identified that in the UK there is what is known as “duty of candour” legislation which obliges practitioners to go back to patients where an error has been made and has caused harm and to tell them proactively about it. She notes that where, on review, there are mammograms that have subtle changes but they would not be called an error, they are classified as “satisfactory with learning points” and that if one accepted what Professor Whitehouse observed, they were “very very subtle” and that she would put his reading of the mammograms into the UK category of “satisfactory with learning points”.

126. Equally, Professor Whitehouse characterised the evidence of speculation on the right CC slide in 2015 as “fairly subtle but present” stating “it is difficult to make out, perhaps, a main mass from which, a main nidus from which you are seeing spiculations appearing”.

On being asked about the mass he described in his report as being 2cm, he says he was worried about that because “it may well be that it has spread more than 20mm only you can’t see it. It is so subtle”.

127. Commenting on Dr Given-William’s report in cross examination, he agreed that the signs overall in June 2015 were “subtle, the signs, but appreciable” and on said “Oh yes these are subtle changes as I have always said but they are strong enough to submit the patient to further examination at an assessment unit”. On being asked in cross examination whether other radiologists might legitimately read it differently, he said that was clearly the case, on being asked whether there was room for two views on the mammograms, he said that did occur in practice and on being asked whether people could legitimately take a different point of view to him, he said he couldn’t answer that because he was not putting himself in their place.

128. I must also consider the undisputed evidence that:

• The plaintiff has either dense breasts (Professor Whitehouse) or very dense breasts (Dr Given-Wilson);

• The density of breasts makes it more difficult to see evidence of cancer on a mammogram;

• ILC is a disease that is difficult to diagnose on a mammogram because of its clinical presentation.

129. In those circumstances, even accepting Professor Whitehouse’s evidence of what could be seen on the mammogram (which, as made clear above, I do not accept), competent radiologists reading the plaintiff’s images could have come to different views about them given the subtlety of the signs and the relevant context i.e. the density of the plaintiff’s breasts and the way in which ILC presents itself.

130. Accordingly, the evidence does not establish that no radiologist of equivalent skill would have failed to identify the suspicious features identified Professor Whitehouse on the images in June 2015.

Conclusions on Clinical Presentation

131. The essence of this aspect of the case against Dr Coffey and Dr Pender is that they were negligent in not referring the plaintiff for clinical recall having regard to the clinical presentation of the plaintiff. The case against Ms. Mathews is that she failed to accurately record what the plaintiff told her on 17 June 2015.

132. Following the mammogram, the plaintiff’s records were sent to the Merrion Centre, along with her images. The records included a sheet with the plaintiff’s name and address, GP name and address and date of visit to BreastCheck typed on the sheet. Some boxes were ticked yes or no. The sheet contained a handwritten note from the radiologist that read as follows:

“Hx of cysts. Lady feels a cyst in the R breast. Couldn’t tolerate compression on R”.

133. As required by the Clinical Recall Guidelines, the plaintiff’s images were annotated with the letter “N” to alert the radiologists to the presence of the Note. The part of the radiology datasheet that provides for clinical recall was not ticked, thus informing the radiologists that the radiographer did not consider the plaintiff met the signs for recall identified in the Clinical Recall Guidelines. As the Guidelines make clear, a radiographer can, if appropriate, either identify clinical recall on the basis of signs and symptoms, or, alternatively make a note. They are mutually exclusive options.

Evidence of Professors Bundred, Ellis, O’Reilly and Hill on clinical presentation

134. At an early stage in the case, I was asked by the defendant to exclude Professor Bundred from giving evidence on the clinical presentation aspect of the case. Professor Bundred had provided a short report on this aspect on 17 February 2020. Objection was taken to his evidence on the basis that, as a breast surgeon, he was not a professional of the appropriate standing or skill as Dr Pender and Dr Coffey. A ruling was given permitting his evidence to be admitted but directing that a renewed application could be made at the

conclusion of the case in respect of the treatment of that evidence. A similar approach was taken to the evidence of Professor Ellis on this aspect of the case (although his report did not address this issue). In fact, although both professors gave evidence on this topic, both recognised the limitations of their experience. Professor Bundred accepted he had never been in the position of deciding whether to recall a woman for screening and that he had absolutely no experience of the circumstances that lead one to decide whether or not to recall a patient. Professor Ellis accepted he had never read a mammogram as a radiologist or reader in a screening clinic or had never read a note written by a radiographer as a radiologist or reader in a screening clinic.

135. The application to exclude the evidence of Professors Bundred and Ellis was renewed in closing written submissions, with the defendant arguing that neither was an appropriate expert to comment on the issue, not being radiologists or radiographers, with no experience or expertise in the field and no experience in writing or reading a radiographer’s notes. An additional point was made to the effect that a ruling had already been given that the defence could only have one expert on tumour doubling time (therefore excluding Professor Arnold Hill, breast surgeon, from also giving evidence on behalf of the defendant on tumour doubling) and that therefore the plaintiff should be confined to having one expert on the clinical presentation issue.

136. The defence also identified that Professor Ellis’s evidence was put in very late, in the face of opposition from the defendant, and the terms upon which I acceded to the application to permit him give evidence were that his evidence be limited to the topics identified in his report and that no further report be furnished by him.

137. In response, it was argued that, although it was accepted that only a radiologist could give evidence about the reading of slides, Professors Bundred and Ellis could give evidence in respect of the clinical presentation aspect of the case, since as consultants, they interpreted clinical notes with the sort of information contained in the Note on a regular basis. It was further argued because both professors participated in multidisciplinary meetings in respect of patients clinically recalled, they were in a position to give evidence on this issue.

138. In Morrissey, the Chief Justice stated that in a professional negligence case, the starting point requires “the identification of the standard of approach which would have been applied by a professional of the appropriate standing or skill as the person against whom the allegation of negligence is made” (paragraph 6.6). When the standard of approach is being identified, expert evidence should be given as to how professionals of the type in question would generally go about their work and the way in which they would have dealt with the case in question (paragraph 6.13). The Chief Justice described the standard of approach as being imposed by the profession itself (paragraph 6.14). I interpret that to mean by the profession carrying out the same type of work as the professional accused of negligence; and not a professional in a linked but different area. What is required is the standard approach of an expert with like specialisation and skill i.e. an equivalently situated professional of equal standing. This ensures the behaviour of the professional is evaluated by an appropriate standard and the evaluation is not being skewed by the application of irrelevant factors. (Indeed, despite seeking to introduce the evidence of Professors Bundred and Ellis on clinical presentation, in closing submissions, counsel for the plaintiff submitted when dealing with radiologists, the court must look at other radiologists – “the court takes the standard of approach of the profession of radiologists”.)

139. Neither Professor Bundred nor Professor Ellis work in a breast screening service, where decisions are made by radiographers and radiologists on who to recall, based on the results of the mammogram and any clinical indicators, following the Guidelines for Clinical Recall. They are not radiologists. They do not have the appropriate standing or skill as Dr Coffey and Dr Pender. The mere fact that they both are used to reading clinical notes does not alter this. All doctors read clinical notes. Equally, the mere fact that they attend multi-disciplinary meetings with other professionals including radiologists does not mean that they are professionals of equal standing or qualification. It would be inappropriate, unfair, and liable to lead to an unreliable result were they permitted to judge the behaviour of a non-equivalently situated professional.

140. Even if Professors Bundred and Ellis were equivalently situated to Dr Coffey and Dr Pender, a further obstacle arises by reason of Order 39, R. 58(3). The plaintiff already has a witness, Professor Whitehouse, to give evidence on this “particular issue” i.e. whether it was negligent for the radiologists not to clinically recall the plaintiff given the contents of the Note. No reason has been advanced as to why evidence on precisely the same issue from Professor Bundred or Professor Ellis is unavoidable to do justice between the parties as per the requirement of Order 39, R. 58(3).

141. Finally, insofar as Professor Ellis is concerned, I specifically ruled that he was permitted to be introduced on terms i.e. that he did not expand upon the topics covered in his report.

The defendant conducted its case accordingly. To permit him now to give evidence on the clinical presentation allegation would conflict with that ruling, where no good reason has been advanced to justify such a departure.

142. For those reasons, I do not propose to treat the evidence of Professors Bundred and Ellis as being of any significant weight insofar as that evidence relates to the clinical presentation aspect of the case. Similar considerations apply to the evidence on clinical presentation given during the cross examination of Professor O’Reilly and Professor Hill, witnesses for the defendant, for the same reasons, in circumstances where neither of them are radiologists with any experience in reading a radiographer’s notes following breast screening for the purposes of deciding upon clinical recall.

143. I should add that Professor Ellis’s evidence in this respect was focused largely on whether it is possible for a woman or a doctor to decide whether a lump is or is not a cyst simply by palpating it. He concludes it is not possible. As identified below, I fully accept that evidence. Professor Bundred took the same view in respect of cysts and indicated women identifying a cyst should be recalled for that reason. I explain in some detail below why the defendant was correct in not recalling for existing cysts, such as that identified by the plaintiff. Equally I address the points made by Professor Bundred about the relevance of pain on compression, the prevalence of cysts and the absence of complaint by the plaintiff on previous mammograms.

Allegations of negligence against Ms. Mathews

144. As noted above, the allegation against Ms. Mathews is that she incorrectly recorded the conversation between herself and the plaintiff. There is no specific allegation that she ought to have identified the plaintiff as requiring clinical recall but presumably that is an implicit part of the alleged negligence. No evidence was given on behalf of the plaintiff as to the correct standard of approach of a radiographer in this respect. I will proceed on the basis that, as identified by Dr Coffey, the radiographer should follow the applicable Guidelines on Clinical Recall.

145. First, I must decide whether, contrary to what the Note records, the plaintiff identified (a)

what she was feeling was new, and/or (b) she was feeling a lump. If the plaintiff had indeed identified a new lump, then she ought to have been clinically recalled. On the other hand, if the Note correctly reflected what she said, then, when reading the Note, the radiologists were entitled to proceed on the basis that she did not refer to a new lump.

146. I have set out above in some detail the evidence given by the plaintiff and Ms. Mathews as to what the plaintiff told Ms. Mathews while she was having her mammogram. Having heard Ms. Mathew’s evidence, I conclude that her practice was to report accurately what the patient told her and that it is very likely that the account of her interaction with the plaintiff was an accurate reflection of the contents of same. In submissions, her counsel argued that it was not a full account of everything that took place in their conversation and criticised the Note on that basis. It was not of course a transcript and therefore would inevitably have omitted certain aspects of the conversation between the plaintiff and Ms.

Mathews. But the core aspects of what was said, specifically the references by the plaintiff to a history of cysts, to feeling a cyst in her right breast, to the issues with compression, were recorded. Equally importantly, there was no recording of a new mass.

147. I am reinforced in this conclusion by the plaintiff’s own evidence. In her direct evidence, the plaintiff agreed that the contents of the Note reflected what she had said. Even if her agreement with the term “reflected” means it was not a verbatim account, I find that if she had referred to a lump and/or to a new lump or new cyst, and the radiographer had not reflected that in the Note, the plaintiff would not have agreed that the Note “reflected” what she said.

148. Further, the Note reflects the evidence given by the plaintiff. The plaintiff made it abundantly clear in her evidence that she identified an existing abnormality rather than a new one, albeit one that was causing her unusual pain on compression. She did not give evidence of identifying a new mass. She did identify feeling something in the right breast.

She gave unequivocal evidence about failing to tolerate compression on the right side. All that is faithfully recorded by Ms. Mathews.

149. Indeed, the only point of difference between the plaintiff’s version of events and that recorded in the Note is that at certain points in the evidence (although not consistently)

the plaintiff referred to identifying a lump. At other times, she referred to identifying a cyst. I find that the plaintiff referred to a cyst rather than a lump for the following reasons:

• Ms. Mathew’s practice was to record carefully what the clients told her;

• all BreastCheck radiographers, including Ms. Mathews, were trained to record a lump because of the significance of a lump in the Guidelines on clinical recall;

• the plaintiff agreed that the Note reflected what she had told Ms. Mathews.

150. There was a significant amount of discussion about whether the lump was palpable i.e.

was a lump that could be touched or felt by the plaintiff. The plaintiff gave no evidence at any time of her feeling a lump with her fingers in the right upper quadrant, either before or after the mammogram. Therefore, I interpret her reference to “feeling” a cyst to be the fact that she experienced a sensation in the right upper quadrant during the mammogram.

151. From the summary of the evidence above, I find on the balance of probabilities:

• the plaintiff identified that what she felt was an existing issue and not a new area of concern;

• the plaintiff referred to cysts and did not identify that she felt a lump;

• the plaintiff identified that she felt something in her right breast;

• the plaintiff did not identify that she felt with her fingers or had touched the area of concern in the right upper quadrant either before or during the mammogram.

152. In the circumstances, I conclude that Ms. Mathews accurately recorded the substance of what the plaintiff told her and correctly made a note of same rather than making a clinical recall since the plaintiff had not identified symptoms requiring recall under the Guidelines for Clinical recall. Accordingly, the allegation of negligence against Ms. Mathews is not made out.

Allegations of negligence against Dr Pender and Dr Coffey

153. Proceeding on the basis that the information conveyed to Ms. Mathews by the plaintiff was correctly identified as suitable for a Note but not for a clinical recall, I must consider if the treatment of the Note by the radiologists was negligent. Counsel for the plaintiff submitted that the question of the radiographer’s negligence is a separate question from that of the radiologists. I accept that is the position.

154. Turning first to the expert evidence on standard of approach, I have the following observations about Professor Whitehouse’s evidence in this respect. First, it is clear from the responses of counsel for the plaintiff that Professor Whitehouse was provided with the Note, along with other medical records, in June 2017, despite his evidence that he did not have it until much later. He was asked to consider the Note at a pre-trial consultation on

12 February 2020 and, subsequent to that meeting, he provided the report of 18 February 2020. Given that he now considers Drs Pender and Coffey to have acted negligently in failing to refer the plaintiff for clinical recall having regard to the Note, it is difficult to understand how he failed to identify the Note as a matter of concern when he first received the medical records. Moreover, he does not explain whether he took the approach that the radiologists had read the Note and made the wrong decision, or whether he assumed they did not read the Note just as he assumed they had not read the images.

155. Second, I am considerably influenced by the fact that Professor Whitehouse said he was not criticising the “SOP” i.e. the Guidelines on Clinical Recall, thus accepting the validity of limiting clinical recall by a radiographer to identified significant signs and symptoms, but on the other treating the radiologists as negligent for failing to recall for a description of a symptom not coming within that list. Professor Whitehouse did not explain why, in his view, the radiologists were obliged to take a different approach to that identified in the Guidelines on Clinical Recall i.e. only recall for a new lump. On cross examination, he accepted that absent a new lump, there was no cause for concern. Professor Whitehouse did not appear to be familiar with the Guidelines on Clinical Recall, and his answers in respect of the applicable standard of approach of the radiologists were, in my view, made without a proper consideration of the Guidelines. I note that because of the lateness of the plea on clinical presentation, the Guidelines on Clinical Recall were only provided to the plaintiff’s solicitor on 20 February 2020. Nonetheless, Professor Whitehouse did not give evidence until 25 February and therefore had time to consider the Guidelines, being three pages in total. Nor did he appear to be familiar with the system in BreastCheck, in that he equated a note to a red dot, although a red dot is akin to a designation of clinical recall as it requires a patient to be referred to the assessment clinic, whereas a note is used to draw a matter to the attention of the radiologists.

156. Third, he gave no evidence about the normal standard of approach in a screening service by a radiologist in the U.K. in an equivalent situation. Professor Whitehouse did not draw my attention to the equivalent UK guidelines, or to the practice in breast screening clinics in the UK. Nor did he put his evidence in context at all. He did not avert to the distinction between a screening service and a symptomatic review. He did not comment on the specific approach of the Guidelines on Clinical Recall where radiographers are instructed not to ask questions and should simply record what the patient has told them or what they observe. He did not avert to the fact that radiographers do not – and, for obvious reasons, must not – carry out any kind of clinical examination or take a medical history.

Nor did he refer to the relevance, if any, of the BreastCheck literature sent to the client, which makes it clear that a very small number of mammograms do not pick up abnormalities and that clients should seek medical advice in respect of matters that concern them. Indeed, at one point, in response to a question on cross examination about the Note, he referred to the Note as being the “view of a GP”. This may have been simply a slip of the tongue but nonetheless it indicates it was not to the forefront of his mind that the Note emanated from a radiographer carrying out a screening mammogram rather than being generated in a symptomatic context.

157. Fourth, I found Professor Whitehouse’s evidence inconsistent in relation to the question as to whether it was acceptable for the radiologist to be satisfied not to recall the plaintiff after seeing a normal mammogram. He agreed in cross-examination that, because of the Note, the radiologists ought to be concerned about the right breast but that if the mammogram was normal, there would be no cause for concern and no need to recall. In his written report and on re-examination he indicated that in fact there was an obligation to recall on the Note alone, assuming it was a new lump, even if the mammogram gave no cause for concern. But he did so without referring to his earlier evidence and explaining his varying approach.

158. Fifth, Professor Whitehouse did not identify the “no doubt” standard that the plaintiff submits is the appropriate standard of approach (discussed below). It was suggested that he had addressed this issue in his answer on cross examination where he said that, if one is looking at a mammogram, one must ask “can I exclude cancer in this” and if not, it should go to an assessment clinic. However, when one looks at that passage in context, it is clear that he is talking exclusively about the mammographic images and not the clinical presentation of a woman.

159. On the other hand, the evidence of Dr Given-Wilson in respect of the relevant standard of approach was compelling. Having run a screening service for ten years, she clearly had extensive experience of the issues arising in respect of clinical presentation and her evidence reflected this. She clearly explained the distinction between a screening service and a symptomatic service. She explained the role of recall on the basis of clinical symptoms rather than the results of the mammogram. She was familiar with the Guidelines on Clinical Recall under which BreastCheck operated and explained the purpose and rationale of same. She was able to comment knowledgeably on the relevance of breast pain, the frequency of women describing their breast history during screening and the distinction between a radiographer designating a client for clinical recall and writing a note recording other signs and symptoms.

160. She identified that because some cancers don’t show up on mammograms but are producing a clinical lump, if a woman comes for screening and identifies a very significant symptoms, such as a new palpable lump, then there is an option for that to be brought to the attention of the radiologist for a clinical recall to be initiated. She gave evidence that one would not normally recall for a palpable lump that was not new, as a number of women have lumps that are longstanding.

161. Dr Given-Wilson was asked about the necessity for a clinical recall arising out of the Note a number of times during the course of cross examination and, on each occasion, she was consistent in her answers – if the plaintiff had referred to a new lump or a palpable new lump, there should be a clinical recall. She explained that it was very frequent when women come for screening for them to make comments about things like pain and lumpiness or some history that has always been there and that it was important to be focused on the real trigger symptoms, such as the patient saying “lump”.

162. For the reasons set out above, I prefer the evidence of Dr Given-Wilson on the applicable standard of approach to that of Professor Whitehouse, which I found incomplete, inconsistent and lacking in clarity. Accordingly, in the circumstances of this case, the relevant standard of approach appears to be as follows:

1. Clinical recall may take place even where the mammogram is treated as R1 or R2 i.e. not disclosing any concerning abnormalities;

2. Radiographers should only mark a mammogram for clinical recall on the basis of identified significant signs and symptoms, being (a) new lump (b) skin dimpling/tethering/inflammation (c) recent nipple inversion;

3. If there is a doubt about whether those significant signs and symptoms are present, the radiographers and radiologists should err on the side of caution and identify the client as a clinical recall;

4. Where the radiographers write a note to alert the radiologists to relevant information not covered by the category clinical recall the radiologist must read the note and use personal clinical judgement in deciding what action if any should be taken;

5. When exercising personal clinical judgement, the radiologist should have regard to the policy identified in the Guidelines for Clinical Recall at paragraph 4;

6. When considering a note that makes reference to “feeling a cyst”, the radiologist is entitled to rely on the expertise and training of the radiographer and may therefore proceed on the basis that the woman did not identify a new lump.

163. Having established the correct standard of approach, I must then turn to the question as to whether Dr Pender and Dr Coffey complied with the above standard of approach. In this case, that is a relatively simple task having regard to the evidence of Dr Given- Wilson. She says that faced with that Note, she would decide whether to recall or not based on the application of the SOP (Guidelines on Clinical Recall) i.e. if it came within the SOP she would recall, but not otherwise. I must therefore consider whether the radiologists were entitled to conclude that the Note did not warrant a clinical recall having regard to the application of the Guidelines on Clinical Recall (SOP).

164. When analysing the material available to the radiologists, it is vital to understand that the Note must be read in the context of the absence of a clinical recall by the radiographer.

The system established by the Guidelines on Clinical Recall is intended to ensure, inter alia, that clients will be recalled when the tell the radiographer about a new lump.

Accordingly, in deciding not to recall the plaintiff, the radiologists knew that what she felt was not a new lump. In this respect, the role of the radiographer and the guidance provided by the Guidelines on Clinical Recall are crucial. Read cold, the words “Hx of cyst.

Lady feels a cyst in the R breast. Couldn’t tolerate compression on R” might suggest a distinct ambiguity about whether the abnormality “felt” in the right breast was part of the history of cysts, i.e. one that the patient was aware of and had been previously addressed or was in fact a new and distinct abnormality. It is only when one reads the Note in the context of the Guidelines and the training given to radiographers, i.e. when one understands that by not ticking clinical recall, Ms. Mathews had excluded the possibility of the abnormality being new, that it becomes clear that the radiographers were not failing to address the possibility the abnormality in question was new. Because they knew that new abnormalities would be made the subject of clinical recall by the radiographer, they could exclude the possibility that the plaintiff was referring to a new abnormality.

165. Accordingly, on the test posited by Dr Given-Wilson, the requirements of the SOP for recall had not been met and there was no obligation on the radiologists to direct the plaintiff for clinical recall. In fact, both radiologists gave evidence that they believed the Note had been provided in the context of whether the level of compression was satisfactory and was intended to provide an explanation in the event that the compression was not satisfactory on the mammogram. That does not alter my conclusion since their core assumption in considering the Note, i.e. that the plaintiff was not referring to a new lump, was a permissible one for the reasons identified above. They were therefore acting within the scope of the Guidelines on Clinical Recall and were observing the relevant standard of approach.

166. Having regard to the matters a court must consider as per the roadmap laid out in Morrissey, that disposes of the plaintiff’s claim of negligence in respect of this limb of the case. However, I must address some additional arguments made on behalf of the plaintiff.

Additional Arguments of the Plaintiff No doubt test

167. First, it was said that the standard of approach in relation to screening cases had been established in the Morrissey decision and that I should simply apply that standard, being that “a screener should not give a clear result if they have any doubt about either adequacy or the absence of suspicious material” (paragraph 6.34) – what I will describe by way of shorthand as the “no doubt” test. Linked to that argument was the contention that Dr Pender had in evidence identified the “no doubt” test as the appropriate standard of approach where she said that radiographers and radiologists should err on the side of caution. The former argument essentially amounts to a submission that the findings of fact in Morrissey in respect of the standard of approach applicable in that case applies mutatis mutandis to this case and therefore I do not need to – and indeed should not – make a factual finding as to the standard of approach in respect of recall for reasons of clinical presentation in breast screening. On that approach, I would simply ignore the evidence of the experts set out above in relation to the standard of approach as the matter has already been decided.

168. I cannot agree with that approach for the following reasons. First, Morrissey concerned screening by cytologists for cervical cancer. This case concerns screening by consultant radiologists for breast cancer. Second, and more significantly, Morrissey was concerned with the standard of approach on the reading of a slide. In this case, although there is a challenge to the reading of images, there is no dispute between the experts as to the standard of approach when so doing. The “no doubt” standard of approach is deployed here in respect of quite a different matter i.e. whether the information on the Note, read against the background of the Guidelines on Clinical Recall, warranted a recall. Nothing like that was at issue in Morrissey. I cannot treat the answer given by the Supreme Court on standard of approach as being determinative of the entirely different question raised in this case. The focus of the Supreme Court on the actual evidence given in Morrissey to identify the relevant standard of approach may be seen at paragraphs 6.21-6.30. The Chief Justice notes at paragraph 6.30 that “imposing such a standard of approach does not derive from the Court but rather from an assessment of the evidence… It is those experts, not the Court, who identified the standard expected of a normally competent screener as being one which precludes giving a clear result in a case of doubt”.

169. Counsel for the plaintiff laid heavy emphasis on paragraphs 6.34, 6.45 and 6.40 of Morrissey. Paragraph 6.40 provides inter alia, … “On the evidence, it was undoubtedly accepted by all witnesses that a competent screener must not give such a result if they had any doubt as to the presence of suspicious material or as to the adequacy of the material on the slide”. However, that does not appear to lay down an immutable rule as to the standard of approach in all screening cases, irrespective of the nature of the act or omission that is challenged. Rather it summarises the standard of approach applicable on the facts in Morrissey having regard to the evidence presented in that case. There, the unanimous evidence of the expert witnesses was that a woman should be recalled if there was any doubt after reviewing the slide. However, the Guidelines on Clinical Recall make it clear that a similar approach is not taken in respect of clinical presentation. The radiographers are instructed not to ask questions about breast history, and a woman is only to be selected for recall by them where specific signs and symptoms are reported by her or observed by the radiographer. That means a decision has been made that not all signs and symptoms that might conceivably be suggestive of cancer trigger a clinical recall. This is because the primary purpose of BreastCheck is to screen for cancer by way of a mammogram rather than by clinical presentation. Recall for non-mammographic reasons is essentially treated as a safety net, whereby if the patient mentions a significant sign or symptom they get a clinical recall. Accordingly, clinical information is treated quite differently in BreastCheck from the way in which it is treated in a symptomatic service. Thus, the factual evidence here demonstrates that, far from a “no doubt” approach, the approach to recalling for clinical symptoms is one that implicitly accepts a certain element of doubt. The Guidelines for Clinical Recall seek to recall only woman with “trigger” symptoms, and explicitly seek to avoid recalling women with other symptoms that have a very low incident of cancer. Moreover, that system, as identified in the Guidelines on Clinical Recall, is not challenged by the plaintiff.

170. Neither Professor Whitehouse nor Dr Given-Wilson identified a “no doubt” standard of approach in respect of clinical presentation. The only evidential basis identified by the plaintiff for a “no doubt” approach was certain portions of the evidence of Dr Pender. In cross examination, Dr Pender explained that when the radiologists feed back to the radiographers in respect of clinical recalls, they err on the side of caution and if there is any doubt about the clinical recall, unless there is a clear reason why it shouldn’t be a clinical recall, the woman is brought back for assessment. There is an important exchange at pages 76 to page 80 of Day 9 between counsel for the plaintiff and Dr Pender on the existence of doubt. In my view, the import of the totality of Dr Pender’s evidence on this topic is that if the radiographers have any doubt about whether the woman has a significant sign or symptom as set out in the Guidelines on Clinical Recall, they should mark them down as a clinical recall. That is not the same as a no doubt standard of the type the plaintiff is advocating i.e. that if there is any reference in a radiographer’s note to a sign or symptom not listed in the Guidelines that might conceivably, however remotely, be an indicator of cancer, the woman must be clinically recalled. Indeed, Dr Pender gave clear evidence that this was not the approach of BreastCheck where she said that in breast screening they don’t want to recall cysts as they are not usually cancerous and there is a very low incident of cancer and that if the mammogram is normal, it is their practice not to recall such a patient.

171. To summarise, Morrissey identified a “no doubt” standard of approach based on the evidence of witnesses in that case. But in my view, the Supreme Court were not imposing that as a universally applicable standard in every screening case, including those concerned with clinical recall for reasons unrelated to screening. Rather, in every case, including screening cases, the court must analyse the applicable standard of approach having regard, inter alia, to expert evidence. Here, there was no evidence given to suggest that the “no-doubt” standard either was, or should be, the standard of approach to be adopted by a radiologist when exercising her clinical judgment in a breast screening service with guidelines such as the Guidelines on Clinical Recall.

Relevance of identification of a cyst

172. A second argument made on behalf of the plaintiff is that because the plaintiff identified she felt a cyst, she ought to have been clinically recalled. This is because a woman cannot know whether what she is identifying as a cyst might be a suspicious lump or a potentially cancerous cyst. Effectively, it was submitted that the plaintiff’s reference to feeling a cyst should be treated as if it was a reference to feeling a lump and that therefore she should have been recalled on this basis.

173. Cysts can be any size from microscopic in an area of normal tissue to big palpable cysts.

20% of women getting a mammogram in BreastCheck are recalled for cysts visible on the mammogram as a new finding. Not all cysts are visible on mammograms. Cysts may be simple or complex. Simple cysts do not become a cancer. A complex cyst can, unusually, be cancerous. Evidence was given from various witnesses that neither the woman herself nor any doctor can, simply from palpating a mass, decide whether it is a lump (whether suspicious or not) or a cyst. I fully accept this evidence. During cross-examination of the defendant’s expert witnesses, counsel for the plaintiff put material gleaned from various websites, generally intended to inform the public about various aspects of breast cancer, to those witnesses. Much of this material, though not all, was about how breast cysts ought to be approached. This material had not been identified in the reports of any of the plaintiff’s experts, nor identified by them in their evidence. For that reason, I found this material to be of limited utility. Nonetheless, as identified above, the core point identified by that material, i.e. that it is not possible to know whether an abnormality is a cyst or a lump by palpating it, is accepted.

174. If a “no doubt” standard of the type identified by the plaintiff was indeed the relevant standard of approach, then references to cysts by women while a mammogram was being carried out might require clinical recall since it cannot be excluded that a woman who identifies she is feeling a cyst might in fact be feeling a cancerous lump (although it seems likely that in the majority of cases a woman who told a radiographer she had a cyst would do so because it had been previously diagnosed as such). However, as Morrissey reminds us, in a case where there is no challenge to an accepted practice (such as the present one) a court has no role in imposing a standard of approach for a professional. Rather it is the standards of the profession itself that impose the standard required. Here, as identified above, the standard of approach, including for radiologists, is set by the Guidelines on Clinical Recall which do not require recall for the identification of a cyst or new cyst but rather for the identification of a new lump. Nonetheless, the evidence showed that it is the practice of the radiologists in BreastCheck to take into account that a woman cannot know with certainty whether an abnormality of concern is a lump or a cyst. Dr Coffey gave evidence that if the radiographer understood from the patient’s description that a patient felt a new abnormality in one or both breasts, then irrespective of whether the patient described it as a lump or cyst or in any other way, it would trigger a clinical recall.

175. Evidence was given by Professors Bundred and Whitehouse to the effect that cysts were very rare in post-menopausal women, to support the argument that reference to a cyst ought to have warranted a recall. Professor Whitehouse indicated that from his experience, he would not expect the plaintiff to have new cysts after the menopause.

That evidence was controverted by the witnesses for the defendant, with Dr Given-Wilson citing the ACRIN study to support the proposition that over one third of women having routine ultrasound screening will have cysts, and both Dr Coffey and Dr Pender saying that cysts were common in post-menopausal women. I prefer the evidence of the defendant’s witnesses in this respect as it is supported by scientific literature as well as by the extensive and ongoing experience of all three of them. I therefore do not find there was an obligation to recall due to the novelty of a cyst in a woman of the plaintiff’s age. A further argument was made on behalf of the plaintiff that a further factor warranting clinical recall was that no cyst was identifiable on the June 2015 mammogram and so another explanation for the plaintiff’s symptoms was required. But that argument fails to take into account that cysts may not be visible on a mammogram, that breast pain generally is a very common symptom for women and that pain during mammograms is equally very common.

Interpretation by radiologists of Note

176. As noted above, both radiologists gave evidence that they would have read the Note as indicating a difficulty in carrying out the mammogram due to the inability to tolerate the compression and alerting them to the possibility that the mammogram would be technically unsatisfactory. The radiographer did not give similar evidence: rather she identified that such a note would simply have been reflecting what she had been told by the woman in question. The plaintiff asserted that the fact that the mammogram was technically acceptable ought to have made the radiologists reconsider their interpretation of the note and consider clinical recall on that basis. However, that does not follow as a matter of logic. The radiographer may have been concerned about whether the mammogram would be technically acceptable and flagged that to them: the fact that it was in fact technically acceptable did not mean Drs Pender and Coffey had to reconsider their interpretation of the mammogram. A radiographer could quite legitimately have a doubt about the quality of the mammogram that turned out to be unfounded. Nothing in the evidence given as to how both radiologists interpreted the Note gave rise to an obligation to recall the plaintiff given the standard of approach to clinical recall for reasons unrelated to the mammographic images.

Compression intolerable

177. A further argument was made to the effect that the incapacity of the plaintiff to tolerate the mammogram ought to have triggered a clinical recall. The points already identified above apply here: pain is not a significant sign or symptom listed in the Guidelines for Clinical Recall justifying a recall. The case was made for the plaintiff that inability to tolerate compression was not akin to pain or discomfort but was something quite exceptional that warranted a different response from the radiologists. There was no evidence from any expert to support that approach. Moreover, the plaintiff did in fact tolerate the compression, resulting in a technically satisfactory mammogram. I am therefore treating the plaintiff’s inability to tolerate compression as equivalent to intense pain.

178. In relation to the relevance of the intensity of pain, again no evidence was presented to differentiate intense pain from any other level of pain during a mammogram. The vast preponderance of the evidence given on pain, whether intense or at a lesser level, or discomfort or tenderness, did not suggest it was a significant sign or symptom suggestive of possible breast cancer sufficient to trigger a recall. Professor Bundred accepted in cross examination that pain by itself was not a reason for clinical recall, even where a woman has not experienced pain previously. Professor Whitehouse said pain was a secondary symptom, that if it was just pain in the breasts alone, this was something for GP assessment and referral thereon and that the mere fact of pain in the breast did not give rise to the necessity for clinical recall. (Both Professors Bundred and Whitehouse were of the view that where pain was present along with a lump, the position was different). Dr Coffey gave evidence that she did not regard a woman complaining of pain, such that she couldn’t tolerate compression, as a significant sign or symptom. Dr Given-Wilson also gave evidence on the relevance of pain during a mammogram, saying that there were a whole number of different reasons for pain, that it was not associated with cancer and that cancers were much more commonly painless lumps. Professor O’Reilly, consultant oncologist for the defendant, did state that with hindsight the inability to tolerate compression was probably a significant symptom. It was not explored with him whether that meant that pain, when not considered with the benefit of hindsight, was considered to be a significant sign or symptom, although the evidence reviewed above suggests this is not the case. However, even if this is the case, given that pain was not identified as a significant sign or symptom in the Guidelines, the radiologists were entitled not to consider this a basis for recall.

179. A further argument was made by Professor Bundred that pain was relevant where it had not previously been experienced and that, because the plaintiff had experienced intolerance on this mammogram but not on previous mammograms, the radiologists ought to have been alerted to a potential abnormality, thus necessitating clinical recall.

But where pain is not itself a significant sign or symptom sufficient to warrant a recall, the presence or absence of pain on previous mammograms does not alter this. Moreover, evidence was given by Dr Coffey that it is not unusual for women to experience pain on some mammograms but not on others and I accept this evidence.

Alleged Inferences on the part of Dr Coffey and Dr Pender

180. Finally, at the very end of the case, a submission was made on behalf of the plaintiff that the reference to “feeling a cyst” was such an overwhelmingly obvious cause of concern that it was negligent not to clinically recall the plaintiff. In this context counsel for the plaintiff identified ten inferences that it was said the radiologists impermissibly drew when reading the Note. The first was that the pain felt by the plaintiff was from a cyst. On the plaintiff’s argument, no assumption could have been made that what the plaintiff was feeling was necessarily a cyst. As noted above, I accept that proposition. But as also noted above, the radiologists knew that what was being identified was not new, and therefore they were entitled to decide not to recall the plaintiff, having regard to the Clinical Guidelines on Recall. Second, it was said the radiologists inferred that the plaintiff knew what she was feeling and they ought not to have made that assumption. That is the same point in substance as the first.

181. Third, it was said an impermissible inference was drawn that the cyst was benign. No such inference was made. Rather, as per the Guidelines on Clinical Recall, a decision had been made in BreastCheck that women should not be clinically recalled for cysts because of the very low chance of them being cancerous. In not recalling despite an (existing) cyst being mentioned, the radiologists were following the appropriate approach.

182. Fourth, it is said the radiologists assumed that “feels a cyst” does not necessarily mean she palpates it and that this assumption was incorrect. This so-called assumption was drawn from the evidence of Dr Coffey, who gave evidence that a reference to “feeling” a cyst does not necessarily mean a woman can touch it with her fingers. I think rather than an assumption, this was an observation by Dr Coffey. That observation was an appropriate one in the circumstances of this case, where the plaintiff in fact never gave evidence that she had palpated the abnormality either before or after the mammogram.

The entirety of her evidence on this point suggested that the reference to “feels a cyst”

was that the action of the compression caused her to feel the presence of a cyst. This is supported by her evidence that she pointed to the relevant area during her meeting with Ms. Mathews. It is quite clear from the evidence of the time she spent with Ms. Mathews that at no point did she touch or feel with her fingers any abnormality.

183. Fifth, it was said an inference was drawn that the plaintiff couldn’t tolerate compression because of the entity called a cyst and that was an incorrect inference since the intolerance might have been caused by cancer. I agree that was the inference drawn by both radiologists as they both gave evidence of that, but I do not agree it was impermissible given the unanimous evidence of all the experts, referred to above, that pain on a mammogram is not considered to be a sign or symptom of cancer.

184. Sixth, it is said the radiologists inferred that not being able to tolerate compression is the same as pain and tenderness but that they ought to have inferred that not being able to tolerate compression is something that warrants another look. As addressed above, no evidence whatsoever was presented supporting the contention that inability to tolerate compression is different to pain, that it is suggestive of cancer or that it necessitates clinical recall.

185. Seventh, it is said the radiologists inferred that the underlining of the part of the Note by Ms. Mathews had no significance. Given that there is no suggestion that the radiologists failed to read the Note, or take into account the underlined words, it is difficult to see where this point goes. Their evidence did not show that they accorded no significance to the words “lady feels a cyst in her right breast”. Both in fact attributed a meaning to the entry on the Note, which meaning included an acceptance that she felt a cyst.

186. Eighth, it is said that the radiologists inferred that the placement of the Note beside the drawing of two breasts as opposed to in the radiographer’s comments section was of no significance and that meant no significance was placed on the fact that it was a clinical note. However, no evidence was given by any witness as to the relevance, if any, of the physical location of the handwritten note and in the absence of same, I cannot draw any conclusions on this issue. In any case, the radiologists clearly treated the Note as conveying clinical information and so again this point does not appear to lead anywhere.

187. Ninth, it is said the radiologists inferred that the Note related to a potential technical defect as opposed to a clinical recall. That is correct as per their evidence but for the reasons set out above, that interpretation of the Note was open to them and their

decision not to recall was consistent with the applicable standard of approach.

188. Finally, it is said that the radiologists infer that what the plaintiff describes as a cyst is not new, whereas in fact this could be an entirely new cyst. However, for the reasons set out above, the radiologists had a sound basis for proceeding on the basis that the cyst was not new: I have found as a matter of fact that the plaintiff did not alert the radiographer to a new cyst or abnormality. The radiographer correctly did not therefore consider the symptom to come within the Guidelines and so did not clinically recall the plaintiff.

Therefore, the radiologists were entitled to proceed on the assumption (or “infer” on the plaintiff’s version) that the cyst felt by the plaintiff was not new and therefore was not a basis for recall.

Causation

189. The question of causation only arises on the hypothesis that the defendant was negligent in not clinically recalling the plaintiff, either because there were irregularities on the mammogram that ought to have been identified or because the information in the Note ought to have prompted recall. Because of my conclusion that there is no negligence on the part of the defendant in either respect, the question of causation does not arise.

However, I propose to make findings on it for the sake of completeness.

The witnesses

190. First, I should identify the evidence I considered in this respect. The plaintiff called the following witnesses to give evidence on the question of whether the tumour had been there in June 2015 and whether it would have been discovered had the plaintiff been clinically recalled: Professor Bundred, breast surgeon, Professor Ellis, pathologist and Professor Whitehouse. The defendant called Professor O’Reilly, medical oncologist.

Professor Arnold Hill, breast surgeon and Dr O’Keane, pathologist. Dr Given-Wilson was cross examined on the issues.

191. The question of tumour doubling time is relevant to the resolution of the questions on causation i.e. how quickly a cancerous tumour grows, since when one knows the size of the tumour on diagnosis, one can seek to extrapolate from the statistics on tumour doubling time its size at an earlier point in time. The science of tumour doubling is a medico-legal construct and, as described by Professor Ellis and other witnesses, is not used in the practice of medicine. Two scientific papers were referred to in these proceedings, one by Peer and one by Michaelson , both of which measured a selection of tumours at a particular point in time. The tumours had been allowed to develop (either because of non-detection or other reasons) and were then measured again. This allowed an estimate (varying depending on age) as to how fast an average tumour would grow over an identified period of time. All witnesses accepted that these rates were averages only and that individual tumours might differ from the norm. Professor Ellis proposed a confidence interval of 10% plus or minus in the result arrived from applying tumour doubling times.

192. Because tumour doubling time is not used in the practice of medicine, none of the experts who gave evidence on it are strictly speaking experts in this field. Rather their expertise in their own area permits them to give evidence on the likely existence and size of the tumour in June 2015. For that reason, I do not treat Order 39, R (58)(3) as requiring me to identify one expert from each party since each expert has their own, different, area of expertise. Nonetheless, where two experts from the same side, albeit from different areas of speciality, give effectively the same evidence on a point, I propose to focus on the evidence of one of them rather than repeating the evidence.

193. Professor Ellis, for the plaintiff, is a professor of cancer pathology at the University of Nottingham and practices at the Division of Cancer and Stem Cells, Department of Histopathology at Nottingham City Hospital as an honourary consultant pathologist. He spends 50% of his time in research and 50% in clinical practice. His first post as a consultant pathologist was funded by the NHS breast screening programme as a histopathologist working in the screening programme in 1987. He was appointed to set up a national training programme in breast pathology to support multidisciplinary working in breast cancer screening in the UK. His unit was the first one to roll out breast screening in Britain in 1989. He was the Chair of the National Breast Screening Pathology Committee funded by the national screening programme for over 20 years until recently.

He has run the Quality Assurance Programme for breast pathology in the UK since the 1990’s and is the lead for Quality Assurance in breast pathology in the UK. In respect of his research work, his focus is the classification of breast cancer and along with his colleague, Chris Elston, he developed the grading system for breast cancer now the internationally adopted grading system across the world. This is the grading system discussed in more detail below. His was one of the first groups to publish on the impact of HER2 protein alteration in breast cancer. 90% of his publications are in breast cancer. His day to day work clinical work involves making diagnoses of breast disease, and participating in multi-disciplinary meetings.

194. Professor Bundred, for the plaintiff, was described as a surgical oncologist, meaning a surgeon who specialises in the treatment of cancer. He qualified as a doctor in 1980 and started his training in cancer surgery in 1982, worked in a number of hospitals around the UK and then became Professor of surgical oncology in Manchester. He is an international Guest Fellow of the American College of Surgeons. He worked in both symptomatic breast clinics and also in recall clinics following breast screening. He is an emeritus professor of the faculty of the National Institute of Health Research, having been attached to that Institute as Senior Investigator between 2008 and 2018. He is the Clinical Research and Development lead in the Greater Manchester and Cheshire Cancer Research Network, that manages clinical trials on cancer, including breast cancer, in Manchester. He is also the Director of the Northwest Surgical Trials Centre.

195. Professor O’Reilly, for the defendant, is a medical oncologist in Cork University, Mercy University and South Infirmary Victoria hospitals. He trained in medical oncology in John Hopkins University in Baltimore, Maryland and was on the faculty there for six years. He then returned to Ireland and set up the medical oncology service in Waterford and then set up the medical oncology service in Cork in 2001 and has been there since. He is one of two medical oncologists attached to the Breast Centre in CUH (Cork University Hospital), which is one of eight national cancer centres. About half his practice involves management of patients with breast cancer, being supervision of patients after breast cancer surgery and patients with metastatic disease. They have a clinical trials unit which has 22 full time staff and they conduct breast cancer trials. Professor O’Reilly is involved in establishing and enrolling patients in trials. It has the largest number of patients entering clinical trials and studies in Ireland. He is the co-national speciality director in medical Oncology for the Royal College of Physicians. He explained that a medical oncologist is involved in the medical management of patients with cancer, in particular the use of drugs and other treatments in breast cancer therapy. He participates in weekly multi-disciplinary meetings.

196. Professor O’Keane, for the defendant, is a consultant histopathologist at the Mater Hospital since 1993. He qualified in medicine in 1981, and trained in Boston in anatomic pathology, practising there for three years and returned to the Mater Hospital in 1993. He reports across a range of specialities including reporting on breast cancers for both BreastCheck and the Symptomatic Breast Cancer. Since 2019, because of the evolution towards subspeciality reporting in breast pathology, he no longer does breast reporting.

But his diagnostic skills in breast pathology were externally validated by the UK EQA breast screening programme on a satisfactory basis. He is the pathology lead at general oncology multidisciplinary meetings. He is the cofounder and current member of the steering group of the Irish National Speciality Quality Improvement Programme in Histopathology. He explained that histopathology is the pathology of tissue samples, as opposed to cytology, which is a sub branch of histopathology and concerns the examination of individual cells.

197. Professor Arnold Hill is a consultant surgeon with a special interest in breast cancer. He was the lead surgeon of the Merrion Unit of the National Breast Screening Programme in St. Vincent’s University Hospital for seven years until 2005. He was then appointed as Professor of Surgery in the Royal College of Surgeons and is currently consultant Surgeon at Beaumont Hospital, and for the last seven years Head of the Medical School in RCSI.

For the last ten years Professor Hill has been the National Surgical Advisor to the National Cancer Control Programme and is the President of the Society of Academic and Research Surgeons in the UK and Ireland, a major academic research group. Professor Hill has published extensively in his field.

198. There was an objection to the evidence of both Profession O’Reilly and Professor O’Keane inter alia on the basis that they were both employees of the defendant, that Professor O’Keane was a member of the BreastCheck quality assurance multidisciplinary consultants group and engaged in the public relations efforts of the Royal College of Physicians in Ireland following the decision in Morrissey, and that Professor O’Reilly identified himself as a defence expert. The first objection cannot be upheld. The defendant, being the HSE, employs a very significant proportion of medics in Ireland. A blanket exclusion of them all from every giving expert evidence would be disproportionate to the object to be achieved i.e. ensuring that witnesses giving evidence are independent and objective. I was satisfied having heard the evidence of both Professor O’Reilly and Professor O’Keane that their employment by the defendant did not affect the objectivity of their evidence.

199. In relation to Professor O’Keane, the plaintiff raised his membership of the BreastCheck Quality Assurance Multidisciplinary Consultants Group. Professor O’Keane explained that he participated in same for the purposes of continuous professional education in circumstances where pathologists already appointed in the hospitals involved in BreastCheck deliver the pathology service and therefore BreastCheck ran an annual study day to look at the data from BreastCheck to see how they were performing. He emphasised that he had never been employed by BreastCheck. The plaintiff also raised the fact that Professor O’Keane made a national representation in his capacity as a public advocate qua director of quality and clinical care in the College of Physicians in Ireland in respect of possible implications of the decision of the High Court in Morrissey. When cross examined about this, Professor O’Keane indicated that the report he gave was focused on specific issues with respect to pathology and that he was very conscious of making no reference to screening whatever. The evidence Professor O’Keane gave was exclusively in relation to causation and in particular in respect of the response of the tumour to chemotherapy and the relevance of that response to the likely size of the tumour in June 2015. Having regard to the confined nature of his evidence and his demeanour as a witness, I am satisfied that his previous advocacy for the Royal College of Physicians and his participation in the quality assurance group run by consultants by BreastCheck did not affect the independence of his evidence.

200. In relation to Professor O’Reilly, to suggest he identified himself as a defence expert is a serious misreading of the evidence. The relevant transcript extract makes it quite clear that he was simply identifying that he was acting as a witness for the defendant in this particular case. Indeed, he had already given evidence that in the previous two cases where he gave medico-legal evidence, he had acted for the plaintiff in one and identified liability of the HSE in another.

201. The defendant also called Professor Arnold Hill, consultant breast surgeon at Beaumont Hospital and Professor of Surgery in the Royal College of Surgeons as an expert witness.

There was an extensive objection to his evidence on behalf of the plaintiff as summarised in the written submissions. Partly because his evidence on tumour doubling time was excluded to avoid duplication of evidence during the trial, it was ultimately not necessary for his evidence to be relied upon in resolving the issues in these proceedings.

Accordingly, there is no necessity to rule upon the matters raised by the plaintiff.

Existence of tumour in June 2015

202. The first question to be addressed is whether there was any cancer in existence in June 2015. The evidence of Professor O’Reilly for the defendant, who was called to address the causation issues, is clear in this respect: he accepted that there had to have been a tumour present in June 2015 as to go from nothing to something between June and December was not plausible. He said “biologically there had to be” and identified a size of 1.75cm. He said that taking the doubling rates from the Peer and Michaelson papers, it would suggest the tumour that was there in June was half the size it was when diagnosed and that because the tumour was not behaving regularly: “A. So I suspect the tumour figure of 3.5 cm based on 50% of 7cm has been said, but I suspect it may have been smaller than that. But I think also, because...(interjection)

Q. To what extent, Professor, doing the best you can?

A. Probably half of that”

203. The belief that the tumour was present in June 2015 was also expressed by Professor Bundred and Professor Ellis based on the size of the tumour when diagnosed and the nature of the cancer. Given that experts for both the plaintiff and the defendant agree on this question, I conclude that the plaintiff had a cancerous tumour present in June 2015 in the location where it was diagnosed in December 2015.

Whether tumour could have been diagnosed in June 2015 if plaintiff had been recalled

204. The next question is whether that tumour would have been diagnosable if the plaintiff had been clinically recalled in June 2015 to the BreastCheck triple assessment clinic. The answer to that question depends to a certain extent on the size of the tumour. However, even if one proceeds on the basis that the tumour was 1.75cm, as estimated by Professor O’Reilly (which is hotly disputed by the plaintiff), I conclude that on the balance of probabilities it would have been diagnosed had it been identified on the mammogram. My

conclusion in that respect is based on the evidence of Dr Coffey and on the fact that, as per the Quality Assurance Guidelines, the aim of BreastCheck is to make an early diagnosis of breast cancer i.e. detections of lesions less than 1.5 cm in diameter.

205. Dr Coffey’s evidence sets out the methods used to identify same once a person is clinically recalled. First, imaging with dedicated mammography using digital breast tomosynthesis is carried out on the woman recalled. Following that, a focused ultrasound is done on the abnormality seen on the initial mammogram. If the abnormality is not seen there, a biopsy is done using a modified mammography machine with an attachment that computerises and identifies exactly the depth and position of the abnormality. The woman is also examined by a surgeon who would identify a lump if palpable. Counsel for the defendant, when faced with that evidence, said it was not specific to the plaintiff’s case, that that evidence doesn’t focus upon the size of the lump and that evidence was not directed to whether the plaintiff’s lump would have been found. But the fact that the evidence was not focused on the plaintiff’s situation seems to make no difference to the position since the plaintiff would have been referred to the triple assessment clinic run by BreastCheck and Dr Coffey was giving evidence as to the normal practice of that clinic.

Given the stated aim of BreastCheck to diagnose small tumours on the balance of probabilities and the methodology employed to do so, it seems likely it would have been diagnosed.

206. If there was no irregularity seen on the mammogram, and the plaintiff had been referred on the information in the Note, then if the tumour was only 1.75cm, the position is less clear as to whether it would have been detected. It is certainly possible the tumour would have been diagnosed following triple assessment. However, because the evidence is inconclusive on this point, I will go on to consider what was the likely size of the tumour in June 2015, as that will assist in determining whether it could have been diagnosed had it not been apparent on the mammographic images.

Tumour doubling time

207. The plaintiff’s experts carried out an exercise on the likely size of the tumour in June 2015 based on tumour doubling times. Professor Ellis noted this was a methodology used in medico-legal practice to determine the likely size of a tumour at an earlier point. The patient’s age can influence the tumour doubling time as can the characteristics of the tumour. He explained that he had taken the size of the tumour at the time of starting the adjuvant chemotherapy, being 64mm from the MRI on 12 January 2016. On the basis of a tumour doubling time of 157 days, being the appropriate time period for a woman of her age, the cancer seven months earlier in June 2015 would have been 47mm in size. But there is a confidence interval in terms of the accuracy of the times, so he gave a band of between 43mm and 50mm. He also used tumour doubling times for woman under the age of 50 i.e. 75 days tumour doubling time. That would have meant the tumour would have been 34mm in June with a confidence interval of between 21mm to 46mm. He considered that because of the patient’s age, the true size would be between 34mm and 47mm in June 2015 based on tumour doubling times. Professor Ellis considered that having regard to the size of the tumour on diagnosis and the rate of growth of ILC, it was more likely closer to 47mm than 34mm. Professor Bundred, again using tumour doubling times, estimated the tumour size in June 2015 at between 37 and 41mm.

208. The defendant’s experts took a different approach to the likely size of the tumour in June 2015. First, they noted that tumour doubling is not used clinically. Professor O’Reilly highlighted various factors demonstrating the inaccuracy of applying estimated tumour doubling times to an individual case, including that the studies and resultant estimates of tumour doubling are based on mammogram measurement which is imprecise. Professor O’Keane cautioned that they are generic statements based on moderate numbers of patients and they do not calculate a specific response in individual patients. Therefore, their application to an individual case is imprecise and unreliable. However, it is also the case that Professor O’Reilly said he would have accepted that the tumour was between 37mm to 40mm in June 2015 on the basis of tumour doubling time, were it not for the way in which the tumour behaved during chemotherapy which lead him to believe there were fast growing aspects of it. I am therefore going to permit reliance upon tumour doubling times, while conscious of the imprecision of this method of estimation.

Response of tumour to chemotherapy

209. Returning to Professor O’Reilly’s evidence, he took the approach that because of the tumour’s surprisingly good response to chemotherapy, this indicated that it was not in fact a slow growing tumour with a low mitotic rate but rather a fast-growing tumour (or at least parts of it were) since such tumours respond much better to chemotherapy because of their rapidly production of cells. That in turn suggested to him that it was significantly smaller than estimated by the plaintiff’s experts, somewhere in the region of 17.5mm. The four reasons identified by the witnesses for the defendant for the asserted excellent response of the plaintiff’s tumour were (a) the reduction in the grading of the tumour post chemotherapy from grade 2 to grade 1; (b) the reduction in size and cellularity of the tumour (c) the response of the lymph nodes to treatment and (d) the description of the tumour’s response to chemotherapy in various correspondence from the treating physicians to other physicians.

210. Professor O’Reilly conceded there was a slower growing component to the tumour that remained after chemotherapy. He said significant numbers of tumours are heterogenous where they have different components to them. However, in support of his contention there was a fast growing aspect to the tumour, he emphasised the reduction of the marker node in the lymph nodes. He noted on the MRI performed on 25 April 2016 the mass is less avid post neoadjuvant and there had been a marked improvement from the previous examination in January 2016. He was struck by the fact that the lymph node went from negative to positive in a short period of time which was concordant with a more rapidly developing tumour than a standard lobular cancer. He considered the 10% cellularity indicates a reduction in cancer and was an excellent response. He pointed to the original grade being 2 and it being downgraded to 1. He was of the opinion that part of the Plaintiff’s cancer was fast growing and part of it was slow growing.

211. Professor O’Keane thought the response to chemotherapy was unusually effective and was at the extreme end of partial response. He said the fact that the residual tumour foci were all grade 1 supported the opinion that the higher grade elements previously recorded as grade 2 had responded completely to chemotherapy. His view was that the tumour components least sensitive to the treatment i.e. the grade 1 component, is what was apparent at the time of the mastectomy. This response was evidenced by the complete response in terms of the lymph nodes and internal mammary nodes. It was further evidenced by the considerable response to the tumour demonstrated by the reduction in invasive tumour cellularity to 10%.

212. Dr Given-Wilson was asked whether, if the plaintiff had been called back in June 2015, assuming the existence of a tumour, would it have been picked up. She said she did not know since it was all quite subtle in December, apart from the lymph node, and the indrawing of the nipple. She thought the answer would depend on whether the lymph nodes would have looked abnormal at that point and she could not speculate on that.

213. The plaintiff’s experts disagreed with the approach of the defendant’s experts, with Professor Bundred focusing on the lack of a complete pathological response to demonstrate that the plaintiff had not had a surprisingly good response to chemotherapy.

Professor Ellis focused on the description of the plaintiff’s tumour after the chemotherapy, her residual cancer burden score and the nature of the tumour to justify his conclusion that the plaintiff did not have a particularly good response to the chemotherapy and indicated that her response did not affect his conclusions on the likely size of the tumour in June 2015. Complete pathological response (“PCR”) means that chemotherapy has not been wholly successful in eliminating the tumour. It was agreed by all that the plaintiff had not had a PCR. However, that does not exclude the possibility that the tumour had a much better than expected response to chemotherapy, even if not a PCR.

Core characteristics of the plaintiff’s tumour

214. To test the defendant’s approach, it is necessary to start by identifying the core characteristics of this tumour. First, this tumour was a grade 2 ILC ER positive, HER2 negative. Classic ILC, as this tumour was, is a slow growing tumour. As noted by Professor Ellis, the WHO Classification of tumours 2019 describes the response of ILC as follows: “Following neoadjuvant chemotherapy lower rates of pathologically complete responses (pCR) for ILC compared to invasive carcinomas NST have been found, possibly favouring neoadjuvant hormonal therapy. However this relative resistance to cytoxic therapy may be more caused by its molecular characteristics and especially its lower proliferation (growth) rate than the histological subtype per se”. Its cells do not divide quickly.

215. Second, pre-treatment, tumours are given a grading to assist in determining treatment, in particular whether the tumour will respond to neoadjuvant chemotherapy. This system of grading was devised by Professor Ellis along with a colleague. However, the scoring systems for looking at the effect of chemotherapy after it is given do not tend to use grading and Professor Ellis gave evidence that it is not relevant to the effect of neoadjuvant chemotherapy. The grading is on a scale of 1 to 3. 1 is the lowest grade, then 2, and 3 is the most rapidly dividing grade. Grade is important because grade shows how aggressive a tumour looks under the microscope, how distant it is from normal. The grade is arrived at by adding up scores in three different areas: the tubules, the nuclear pleomorphism and the mitoses. In the plaintiff’s case the score was T3, N2 and M1 giving her a total of 6. A total of 5 or less will be assigned a score of 1. The mitotic score of 1 confirmed that the tumour was a slow dividing one and therefore slow growing. Mitoses are used as a measure of how fast a tumour is turning over and producing new cells. A low score for mitoses indicates a slow growing tumour. The score allocated to the plaintiff’s tumour did not alter following chemotherapy. So biologically, this tumour was slow growing.

216. Those two factors i.e. the nature of the cancer and the low mitotic rate, suggest that the plaintiff’s tumour was slow growing, thus supporting a view that the tumour was more likely to be in the range suggested by the plaintiff’s witnesses than that suggested by the defendant’s witnesses.

217. Third, this tumour was ER positive and HER2 negative, meaning that it was responsive to hormone treatment but not to Herceptin. As attested to by Professor Ellis, ER positive, HER2 negative invasive lobular cancers very often have only a moderate response to neoadjuvant chemotherapy and the plaintiff’s response was what he would have expected for a tumour of this kind.

Factors identified by defendant to support excellent response of tumour

218. Turning now to the factors identified by the defendant to support a surprisingly good response, insofar as grading is concerned, the score given to the plaintiff’s tumour was assigned downwards to 1. This was because although the tubules score and the mitosis score remained the same, the nuclear pleomorphism was revised from a 2 to a 1. Nuclear pleomorphism refers to the appearance of the cells in the tumour. The pathologist must examine the normal breast nuclei and see whether the tumour cells resemble and are the same size as the normal nuclei whether they are bigger or vary in size or shape. Professor Ellis gave evidence that the reduction in grade following chemotherapy is normally caused by a change in the score for tubules or for mitoses. He said it was very rare and somewhat exceptional for the score of nuclear pleomorphism to be altered and that of the three variables in grading, it is the most subjective. In the teachings and guidance they give, they advise pathologists to think very seriously when allocating a score of 1 as it is very rare in breast cancer to obtain a score of 1 as cancer cells are generally larger than normal cells. He added that there is a tendency for non-specialist pathologists to score nuclei lower than specialist pathologists. That evidence was not contradicted by Professor O’Keane, the pathologist called on behalf of the defendant. Professor Ellis’s evidence in respect of downgrading of the score given to the plaintiff’s tumour led me to conclude that the downgrading of the tumour from a grade 2 to a grade 1 was not a particularly significant factor to be considered when analysing the response of the tumour to chemotherapy and whether that response indicated that it was in fact a fast growing tumour.

219. In relation to the reduction in size and cellularity, the conclusion of a mammogram and ultrasound reported on 25 April 2016 was that there had been improvement in the conventional imaging findings since the diagnostic imaging of 11/12/15 but no appreciable change since the study 29/2/2016. That report noted that there was architectural distortion and microcalcification of the right breast which extended over 5cm. Professor Ellis gave evidence that having regard to a residual mass of 5cm he did not agree with Professor O’Reilly’s view that the plaintiff had a dramatic clinical, radiological and pathological response to chemotherapy. He took the view that her response is what he would have expected for an invasive lobular cancer, ER-positive, HER2-negative. He also placed emphasis upon the pathology report that identified a residual invasive carcinoma of 70mm (28 May 2016 (page 81 of Book of Medical Records). He noted there was no tumour cellularity measurement on the needle core biopsy but that in a lobular cancer it would be in the region of 40-50% so that there has been a reduction in the percentage of tumour cellularity. He concluded that this was a partial response and that using the residual cancer burden score, being a measurement system to assess the degree of cancer left behind, she would have a score of 2. Category

0 is a complete pathological response and Category 3 is evidence of no response, with 1 and 2 being levels of response. He concluded that the plaintiff’s response, indicating a less responsive tumour, would fit exactly with the type of cancer she has. He indicated this was not a complete pathological response (PCR) where the breast goes back to normal or there may be a small fibrous scar remaining. Professor O’Reilly accepted that the plaintiff had a residual cancer burden of Category 2 but considered that was evidence of an objective response to therapy.

220. I am persuaded by Professor Ellis’s evidence in respect of the responsiveness of the tumour, both because of his extremely impressive background, which included an early discovery of the relevance of Herceptin and the formulation of the grading system for cancer used to classify the plaintiff’s tumour, and because of the precision and comprehensiveness of his evidence. In particular, his conclusion that a residual 70mm invasive carcinoma could not be described as an excellent result was persuasive in circumstances where the size of the tumour prior to chemotherapy was 64mm, even taking into account Professor O’Reilly’s evidence that the population of cancer cells would be depleted in the residual carcinoma after chemotherapy. Moreover, both Professor Ellis and Professor O’Reilly appeared to agree that there was a residual cancer burden of Category 2, thus evidencing at best only a medium level of response.

221. In relation to the responsiveness of the lymph nodes, that is taken into account when calculating the residual cancer burden and thus the excellent response in that respect has already been factored in. However, even with that, the plaintiff is in Category 2, thus indicating that other parts of the cancer were less responsive to treatment. Professor O’Reilly accepts that some of the plaintiff’s tumour must have been slow growing.

222. Finally, insofar as the descriptions of the effect of chemotherapy in the various correspondence by the plaintiff’s doctors are concerned, in circumstances where objective evidence of the effects is available, the best evidence rule suggests that I should avail of the primary sources rather than relying on an interpretation of those sources. I therefore place little weight on that material.

223. In conclusion I conclude that the plaintiff’s tumour (or at least part of it) responded to the chemotherapy in the same way as expected from 52% of similar tumours as identified by Professor Ellis. There was a very good response on the axillary nodes and on the internal mammary node. There was a moderate response on the tumour. Even if part of the tumour was a fast-growing tumour, and responded very well to the chemotherapy, there was clearly a slower growing part to it (as accepted by Professor O’Reilly) that did not respond so well. The evidence on tumour doubling time made it clear that the methodology is based on estimates and that it necessarily lacks precision when sought to be applied to an individual case. Nonetheless, it permits an estimate for estimating the growth rate of a tumour. The defendant asks me to assume, on the basis of Professor O’Reilly’s evidence, that half of the tumour was fast growing and therefore at most the slow growing part of it would have been 1.75cm. No evidence was put forward by Professor O’Reilly to substantiate the figure of 50% of the tumour being fast growing.

Conclusion

224. Having regard to all the evidence indicating that the plaintiff’s tumour was slow growing, including the nature of the tumour, the specific mitotic rate of the tumour before and after chemotherapy, the evidence of Professor Ellis that the tumour did not have a particularly good response to chemotherapy and the lack of hard evidence in respect of the portion of the tumour said to be fast growing by Professor O’Reilly, I find that on the balance of probabilities the tumour was larger than 1.75cm in June 2015. Rather, I find that even on the most conservative tumour doubling time, the tumour would have been at least 34mm and therefore very likely to have been palpable. On that basis, having regard to the evidence of Dr Coffey on the triple assessment process, I consider the tumour would likely to have been diagnosed on triple assessment, even if it had not been present on the mammogram. Accordingly, had I found the defendant negligent, then that negligence would have been causative of harm since the tumour was likely present at a size in excess of 34mm and would have been detected in June 2015 had the plaintiff been sent for clinical recall, whether there was an indication on the mammogram or whether the recall was on the basis of the note alone.

Minute of January 2016

225. I have averted above to the minute of the meeting of 6 January 2016 by Professor O’Doherty (the “minute”) and identified that it was not of importance in respect of the issue of breast density, given that all experts, and the plaintiff herself, agreed that she had either very dense, or dense, breasts. By way of background, the minute was provided to Professor O’Reilly when he was briefed as an expert along with other material. On 21 January 2020, the plaintiff wrote seeking voluntary discovery in respect of certain categories from the defendant. There was a continuing exchange of correspondence on discovery but no application was made for discovery or Order made. The plaintiff’s solicitor sought the minutes of the 2016 meeting on 6th February 2020. That had not been provided by the start of the trial on 28 February 2020. The plaintiff was crossexamined about the meeting of 6 January 2016 and complaint was made about the failure to provide the minute. Ultimately, the minute was provided to the plaintiff’s solicitors at the end of Day 2 of the trial. Complaint was renewed by counsel for the plaintiff on the last day of the trial in respect of the failure to provide the minute earlier, either by way of a data access request in 2017 or discovery. Subsequently, the defendant’s solicitors wrote explaining the minute was not considered to form part of the plaintiff’s records and therefore did not come within the data access request and nor was it considered to form part of the review of 4 January 2016.

226. In closing oral submissions, counsel for the plaintiff raised the failure to provide the minute at an earlier stage, submitting that (a) the approach of the defendant in this respect indicated that it was excessively defensive in its approach overall to the case; (b)

the minute was of evidential value in respect of the question of the density of the plaintiff’s breasts and ought to have been provided to the plaintiff’s witnesses considerably earlier (c) it was inappropriate to give it to Professor O’Reilly as it was not being treated as a medical report and it therefore affected his independence and (d) the court should intervene to protect its own process and the defendant’s failure merited policing and investigation by the court. The defendant rejected those points, submitting that it was not provided initially because the defendant did not consider it as part of the plaintiff’s records, that it was of limited evidential value, that there was no order for discovery and no obligation to provide it and that the plaintiff had the document during most of the hearing.

227. To the extent that there is a complaint about an alleged failure to comply with data protection obligations, that is not a matter for this court. Equally, there was no discovery order made so there is no question of a breach of same. Rather, there is a complaint that there was a request for a document that remained unanswered and the document was only provided after the case started. Both those points are correct insofar as they go but the defendant did not have an obligation to provide the document and therefore its late delivery is not a cause for sanction. Given that it was provided to Professor O’Reilly, it would have been desirable for the defendant to disclose it even in the absence of an obligation to do so. However, the failure to do so has not in my view caused any prejudice to the plaintiff since all her experts had an opportunity to see it before giving evidence and more importantly, there was nothing significant in the minute. For the reasons identified earlier, the two (conflicting) references to the density of the plaintiff’s breasts in the minute were not material to my conclusions on density. In the circumstances, no action is necessary in respect of the furnishing of the minute.

Conclusion

228. The plaintiff was undoubtedly very breast aware, going for her mammograms at two yearly intervals and checking her breasts regularly. She had a mammogram in June 2015 that was reported as clear. She was then diagnosed with breast cancer in December 2015 after she identified symptoms and went to her GP who referred her for assessment to St.

Vincent’s Hospital. This was undoubtedly a very difficult time for the plaintiff, as was her subsequent treatment over the following months. The plaintiff considers that the defendant was negligent in failing to clinically recall her following her mammogram in June 2015.

229. However, for the reasons set out in this judgment, I find that the plaintiff’s cancer was a true interval cancer, i.e. it showed normal or benign features in the screening mammogram of June 2015 and therefore reporting the mammogram as normal was not negligent.

230. Equally, the defendant acted appropriately in not identifying the plaintiff for clinical recall following her screening in June 2015, having regard to her clinical presentation and given the Guidelines on Clinical Recall within which Ms Mathews, Dr Pender and Dr Coffey were operating.

231. Accordingly, the plaintiff’s claim falls to be dismissed.