Morrissey & anor v Health Service Executive & ors [2020] IESC 6 (19 March 2020)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

Judgment of Mr. Justice Clarke, Chief Justice, delivered the 19th of

1.1 The tragic circumstances underlying this appeal must be acknowledged. The

first named plaintiff/respondent (―Ms. Morrissey‖) is terminally ill from cancer. She

had undergone screening in accordance with the National Cervical Screening

Programme (―CervicalCheck‖) in August 2009 and again in August 2012. In both

instances, her smear test was reported as negative for abnormalities and she was

provided with a clear result. However, in May 2014, Ms. Morrissey attended at her

G.P. following symptomatic bleeding and was referred for further testing. A biopsy

and an MRI scan subsequently disclosed the existence of cervical cancer.

1.2 Following this diagnosis, the 2009 and 2012 smears provided by Ms.

Morrissey were audited and it was reported that the original results provided in

respect of both tests were incorrect. By 2015, the results of the audits had been

communicated to CervicalCheck. However, their results were not disclosed to Ms.

Morrissey until mid-2018, when Ms. Morrissey herself made inquiries as to whether

1.3 Ms. Morrissey and the second named plaintiff/respondent, her husband, (―Mr.

Morrissey‖, and collectively, ―the Morrisseys‖) commenced these proceedings

thereafter. The first named defendant/appellant (―the HSE‖) is, of course, the body

charged with the provision of health services in Ireland and, in the particular context

of these proceedings, with the promotion of the CervicalCheck scheme. It will be

necessary to go into the precise role of the HSE in due course, for its legal obligations

in respect of CervicalCheck are one of the issues which will need to be addressed.

1.4 From 2008, the HSE contracted out the testing of samples to various

multinational firms. The second named defendant/respondent, Quest Diagnostic

Incorporated (―Quest‖), carried out an examination of the sample taken from Ms.

Morrissey in August 2009, in one of its laboratories located in Grand Rapids,

Michigan in the United States. The third named defendant/respondent, Medlab

Pathology Limited (―Medlab‖), tested the sample provided in August 2012 in one of

1.5 The High Court (Cross J.), in a judgment delivered on 3 May 2019 (Morrissey

& anor v. Health Service Executive & ors [2019] IEHC 268), gave judgment in favour

of the Morrisseys in the sum of €2,152,508 against all three defendants and an

additional sum of €10,000 in nominal damages as against the HSE in respect of its

failure to notify Ms. Morrissey of the results of the audits of her earlier smear tests.

1.6 Each of the defendants/appellants sought leave to appeal directly to this

Court. An oral hearing ensued. For the reasons set out in a judgment of this Court

(Morrissey & anor v. Health Service Executive & ors [2019] IESC 60), leave to

appeal was granted although the issues became somewhat more confined in

1.7 It is, therefore, appropriate to set out the basis on which leave to appeal was

granted, for that defines the parameters of this appeal. Before this Court, it was

outlined that there were a number of other cases pending before the High Court which

have arisen in a similar factual context to that of these proceedings. It was also

stressed that the work of the CervicalCheck Tribunal, as established by law in July

2019, will involve the assessment of legal liability and will therefore require the

application of the same legal test as that determined in the courts. Thus, it was clear

that the Tribunal would be significantly constrained in progressing its work if there

remained doubt about the proper test to be applied in determining legal liability.

While it is usually appropriate that an appeal is heard in the Court of Appeal, where

narrowing and clarification of the issues of importance can take place, here this Court

held that the question of the standard of care to be applied in screening cases met the

constitutional threshold for leave to appeal generally and, having regard to the

urgency which attends its clarification, also met the additional criteria necessary to

1.8 In respect of whether the other grounds of appeal urged by the defendants

justified a grant of leapfrog leave, it was held that the interests of justice required

granting that wider leave. It should be noted that both Quest and Medlab had agreed

at the oral hearing to drop certain grounds of appeal as advanced in their applications

―Considering the standard of care grounds in the abstract, while issues

concerning some of the findings of the High Court remained alive before the

Court of Appeal, would be highly unsatisfactory. I am also satisfied that it

would be unfair to the defendants/appellants to require them to abandon all of

the other grounds as the price for obtaining leapfrog leave. Most of those

grounds are sufficiently closely connected with the standard of care grounds in

any event such that this Court is likely to need to at least get into the facts

relevant to those grounds to some extent. Insofar as the causation and, in the

case of Medlab, quantum grounds are raised, it seems to me that these issues

2.1 Before setting out the broad categories of issue which arise on this appeal, it

may be useful to make one general observation. At a superficial level, it might be

thought that a case such as this ought to be relatively straightforward. If a person gets

the all-clear as a result of a screening process and subsequently develops the disease

which the screening process was intended to detect, then it might be felt that it must

be fairly obvious that someone was at fault. However, that does not necessarily

2.2 I would emphasise that not all of the matters which I am about to note will

necessarily arise in each case and all of them do not arise in this case. However, there

are a range of reasons why it might, at least in theory, be the case that a person who

obtained a clear result on a screening process might subsequently be diagnosed with

the relevant disease in the absence of any negligence by any party.

2.3 First, there will almost always be some period of time between the test which

led to a clear result and the diagnosis of the disease in question. Depending on the

length of that time and the likely progress of the relevant disease, there may, in some

cases, be a possibility that whatever the relevant indicators for the disease in question

are, they may not have been present at the time of the screening process but have only

developed at a later stage. Next, in the context of a screening process such as that

with which this Court is involved on this appeal, it is always possible that a sample

taken may not contain any of the type of suspicious material which ought to have

given rise to a result suggesting that the test was not clear. In other words, suspicious

material might not actually have been captured during the taking of the sample, or at

least captured in a sufficient quantity that it would show up during the screening

process. Third, it should be noted, for reasons which will be addressed in this

judgment, that even screening processes which operate at the very highest standards

can give rise to different results by competent screeners. In addition, a retrospective

review of the screening process after someone has been diagnosed may well give

different results, possibly influenced by the difficulties encountered with hindsight. It

is thus possible, as the trial judge in this case found, that a competent screener

exercising ordinary care might give a clear result, even in circumstances where it

might transpire with the benefit of hindsight that there could have been suspicious

material on the slide. For these and doubtless other reasons, it is not simply a case of

inferring from the fact that someone obtained a clear result but subsequently was

diagnosed with the relevant disease that there was necessarily negligence on

someone‘s part. It follows that the assessment of whether there was negligence in any

particular case will involve findings of fact by the trial judge based on evidence. That

process may well be both difficult and complex and each case will, necessarily,

2.4 Against that background, I should start by noting that, at the oral hearing,

counsel on all sides were asked to agree that five sets of issues arose between the

parties. I did not understand there to be any disagreement about that proposition,

although there are undoubtedly sub-issues potentially arising in some, if not all, of the

2.5 The first issue concerned the proper approach to negligence in a case

involving medical screening such as this. As will become apparent, the trial judge,

relying on the case law of the United Kingdom, had applied what he said was an

―absolute confidence‖ test, although it is important to understand that the trial judge

was quite specific about the precise aspects of the case to which that test was to be

applied. The potential question which arose was as to whether the trial judge could be

said to have applied a test differing from the established case law of the courts in this

jurisdiction as regards the legal standard of care in medical negligence cases, which

finds its clearest example in Dunne (an infant) v. National Maternity Hospital

[1989] I.R. 91. However, in the course of the oral hearing, it became clear that there was no

significant dispute between the parties on this question. However, I am mindful of the

fact that a significant amount of public comment, some of it unfortunately

misinformed, followed from the judgment of the High Court and I do, therefore,

propose setting out the position in the hope of bringing some clarity to a potentially

2.6 Second, Quest raised certain points concerning the findings of the trial judge

which led to a conclusion that it was negligent in the way in which it carried out its

functions. In fairness to counsel for Quest, it was not argued that there was no

evidence from which the trial judge might have reached the conclusions which he

ultimately did in this case. However, it was said that there was a significant failure of

the trial judge to properly engage with certain aspects of the defence case, such that

2.7 Third, there were similar issues raised by Medlab concerning the finding of

negligence against it. In that context, it is of some relevance to note at this stage that

the finding against Quest related to the reading of the slides concerned. It was said

that had such reading been properly carried out, a negative result should not have

ensued. A like claim was made against Medlab, although it is of some importance to

note that the trial judge rejected that aspect of the claim against that party. However,

it is clear that there may be circumstances in which it is inappropriate to give a result

in respect of a particular screening where the cell count of the relevant sample is

inadequate to allow for its proper assessment. The finding of negligence against

Medlab related to what was said to be negligence in assessing the smear in question as

being adequate for such assessment. However, points are put forward on behalf of

Medlab to suggest that, as in the finding of negligence against Quest, there was a

failure to properly engage with certain aspects of the case which it put forward in that

2.8 Fourth, there were issues raised by the HSE concerning the finding of the trial

judge that the HSE was liable in negligence arising from the manner in which both

Quest and Medlab carried out their tests. The trial judge found that the HSE was

primarily liable in respect of the way in which the tests were carried out but also held

that the HSE was vicariously liable for the negligence found against the two

2.9 Fifth, and finally, Medlab raised some questions concerning particular

elements of the damages awarded by the trial judge. Neither of the other defendants

2.10 However, in order to properly understand both the approach of the trial judge

and, more specifically, many of the issues which arise before this Court, it is

appropriate to set out in a little detail the history of CervicalCheck and, in particular,

3.1 In 2008, the National Cancer Screening Service (―the NCSS‖) was responsible

for the establishment of the CervicalCheck programme, which provides free smear

tests for women between the ages of 25 and 60 in order to regularly screen for early

signs of the development of cervical cancer. In 2010, the Board of the NCSS was

dissolved and its functions were assumed by the HSE, as part of its National Cancer

3.2 From 2008, the testing of the samples was contracted out by CervicalCheck to

various multinational firms, including Quest and Medlab, which provided a faster

service than that of Irish labs. These contracts provide that the tests are to be carried

out in accordance with the NCSS Guidelines for Quality Assurance in Cervical

Screening, which require the use of the Bethesda System for Reporting Cervical

3.3 As part of the programme, audits are conducted in respect of the screening

histories of those patients who receive a diagnosis of cervical cancer. Laboratories

conduct a review of previous smears and their results where cancer is subsequently

detected. The purpose of these audits is educational, for the screeners and the

laboratory to look into the quality of the testing and to make any improvements

3.4 Smear tests are not diagnostic tests, but rather are screening tests to determine

whether the cells examined are healthy or whether they may have been contaminated

by the HPV virus. Where high-risk HPV is carried in the body for some two to three

years, a pre-cancer will develop. This pre-cancer may be high grade or low grade. If a

significant number of high grade lesions are not detected and treated, this will

progress and develop into an invasive cancer. The invasive cancers are typically either

squamous cell tumours or glandular cell tumours. It was accepted in evidence by the

High Court that such pre-cancer will usually progress to invasive cancer over an eight

3.5 Samples of cells are taken from the cervix by a patient‘s GP, using a system

known as Liquid Based Cytology, which are then reviewed by a laboratory. In the lab,

a slide is made from the sample using the ThinPrep Imaging System. The slide is

then examined by a qualified cytoscreener, who must first assess the sample for the

adequacy of its cell count. For a sample to be considered adequate, as required by the

Bethesda System, there must be a minimum number of 5,000 well-visualised

squamous cells. Most samples will clearly be seen to be adequate on a quick

overview, but in the event of any doubt, the slide must be assessed for adequacy using

3.6 The screener then examines the cells for abnormalities at various

magnifications. If any abnormal cells are detected, irrespective of the adequacy of the

sample, the slide must be classified as abnormal. The terms of the contracts agreed

between CervicalCheck and both Quest and Medlab provided that laboratories were to

screen samples using one of the following processes; a full manual primary and

secondary screening; a full manual primary screening and rapid review rescreening; a

full manual primary and rapid preview screening; or one of the above options with

3.7 The apparently uncontested evidence before the High Court suggested that the

practice of the laboratories operated by Quest at the relevant time in respect of those

slides which were reviewed under its contract with CervicalCheck was that two full

manual screenings of each slide took place. All parties agreed that the reports of the

Scoping Inquiry into the CervicalCheck Screening Programme, as conducted by Dr.

Gabriel Scally (―the Scally Reports‖), could also be considered by the Court. The

Scoping Inquiry was established by the Government in 2018 in order to examine and

establish the facts surrounding the issues which had recently come to light in respect

of the operation of CervicalCheck. I do note that a somewhat different description of

these procedures appears in the final Scally Report, published in September 2018. I

should emphasise that a court can only consider facts based on the evidence presented

to it. It may well be the case that, for the purposes of these proceedings, nothing turns

on the difference between this aspect of the evidence presented to the High Court and

3.8 However, it is again important to emphasise that, where a difference on the

facts is potentially relevant to the outcome of the case, it is important that sworn

evidence is presented in respect of the competing positions and that any relevant party

is given the opportunity to test the competing evidence before the judge. It is also of

particular importance to emphasise that the function of a trial judge is not the same as

the function of a person preparing a report. The trial judge is only required to make

findings of fact in relation to questions which are relevant to determining legal

liability. Trial judges sometimes specify additional facts by way of background or to

provide context, but the precise accuracy of such facts is only important where the

facts in question have at least the potential to influence the outcome of the case. In

that context, I would draw attention to the observations of this Court on the question

of whether it is necessary to correct factual errors which are not material (see, in that

regard, the judgment of this Court in Walsh v. Minister for Justice and Equality

3.9 Be that as it may, it was found by the trial judge that it is the practice of

Medlab that a primary manual screen is first conducted by a cytoscreener. If this first

screen highlighted any areas of suspicion, then a full manual rescreening is conducted.

However, if no areas of suspicion are detected by the first screener then there is no

manual rescreening, but rather an automated analysis of the sample is conducted by a

machine using the ThinPrep Imaging System. If there are any areas of suspicion,

these are highlighted on the slide using a marker, and the slide is sent to a pathologist

3.10 The results of the tests can be graded by a screener using either the Bethesda

System of classification or the British Society for Clinical Cytology CIN terminology,

both of which result in the same consequences for the patient. The system used by the

laboratories contracted to CervicalCheck is the Bethesda System. The grades to

which a test should be marked, according to the Bethesda terminology are;

unsatisfactory/inadequate sample, negative/NAD (no abnormality detected), ASC-US

(atypical squamous cells of undetermined significance), LSIL (low grade squamous

intraepithelial lesion), HSIL (high grade squamous intraepithelial lesion), squamous

cell carcinoma, AGC/AGUS (atypical glandular cells/of undetermined significance),

glandular neoplasia or broken/damaged/expired slides. The categories of ASC-US or

AGC/AGUS denote borderline abnormalities in squamous and glandular cells,

respectively. These do not of themselves denote either cancer or pre-cancer but do

represent a non-negative finding and require at least that a patient is rescreened within

3.11 The laboratories are obliged, under the terms of their contracts with the HSE,

to record the appropriate treatment alongside the results of the screening. If the test

finds that the cells are healthy, then the patient is referred for routine repeat

examination, which is usually every three years. Where the laboratory finds the cells

to be abnormal, then depending upon the result, the patient is either sent for early

repeat screening or else directly to the patient's doctor for colposcopy and, if

necessary, other treatment. In the High Court, it was held that the evidence of all

relevant experts was that in the event of any ambiguity, the lab ought to report the

3.12 It is clear that the screening is not infallible. For example, actual abnormal

cells in the patient‘s body may not have been located in the sample. In addition, as

noted by the High Court, there is also ―room in the analysis for genuine and non-

negligent divergence as to whether particular cells are negative or potentially

alarming‖. One study accepted in evidence before the High Court, which reviewed

the audits of over 6,000 English patients of the NHS Cervical Screening Programme

who had been diagnosed with cervical cancer, indicated that at least 44% (or 55%,

depending on whether samples labelled as ‗inadequate‘ are counted) of their previous

samples were incorrectly marked as normal at first screening. It bears noting that this

statistic does not indicate that there is anything close to a 50% chance of the results of

a screening test being wrong. The lives of many women have been saved as a

consequence of the detection of abnormalities in the course of such screening

programmes. It should be emphasised that this study refers to an audit of the prior

tests of the small proportion of patients who had subsequently received a diagnosis of

cervical cancer. In order to ascertain whether there has been any negligence or breach

of duty in any case, each screening test would have to be assessed individually.

3.13 Against that background, it is appropriate to turn to the history of the

4.1 Mr. and Ms. Morrissey are married with one young daughter. In August 2009,

Ms. Morrissey underwent a smear test which was processed by Quest, as part of its

contract with the HSE, in its laboratory in Grand Rapids, Michigan. The sample was

reported to be adequate and the result was negative for any abnormality. Ms.

Morrissey was advised to return for a routine repeat examination in three years‘ time.

4.2 Ms. Morrissey‘s next smear test was taken in August 2012, and processed by

Medlab, in its laboratory in Co. Dublin. It was reported that the sample was adequate

for assessment and that there was no evidence of the presence of new abnormal

growth tissues. Ms. Morrissey was again recommended for routine recall in three

4.3 Following an investigation as a result of symptomatic bleeding in May 2014,

Ms. Morrissey received a diagnosis of invasive squamous carcinoma of the cervix. A

further smear test was carried out in 2014 following the diagnosis, which also

returned a negative result. There is no suggestion that this interpretation was

negligent. Surgery was required and the cancer appeared to have been treated

successfully. In 2018, there was a serious recurrence of the cancer and, tragically,

4.4 As a result of Ms. Morrissey‘s 2014 diagnosis, audits were conducted of the

2009 and 2012 smears. In September 2014, the 2009 slide was reviewed by the senior

staff pathologist of Quest, who reported that the original test was incorrect and found

the slide to contain borderline nuclear abnormalities, marking it AGC/AGUS. Under

the heading in the report of ‗Factors likely to lead to false negative results‘, nothing

4.5 The 2012 slide was reviewed internally by the medical director of Medlab in

October 2014, and this review also found that the original result was incorrect. Under

the heading of ―Factors likely to lead to false negative results‖, the laboratory listed

the fact that the sample was ―scanty‖, a term used to indicate an inadequate number of

cells on the slide. The 2012 slide was also externally audited by an independent

reviewer in 2015, who reached the same conclusion that the slide had been

4.6 The results of the internal audit of the 2009 slide and the internal and external

audits of the 2012 slide were communicated to CervicalCheck by 2015.

CervicalCheck advised Ms. Morrissey‘s treating doctor of the audit results in June

2016. These were not discussed with Ms. Morrissey. When national publicity in

relation to another patient arose in mid-2018, Ms. Morrissey herself made inquiries as

to whether there had been errors in her case. At this juncture, she was advised of the

results of the audit by her doctor, who apologised to the Morrisseys and indicated that

4.7 It is Ms. Morrissey‘s case, as put before the High Court and again before this

Court, that had the 2009 and 2012 smears been accurately reported at first, then, as a

matter of probability, subsequent medical investigations would have disclosed a

precancerous condition, which would have been treated successfully, and Ms.

Morrissey would not be facing the same terminal prognosis which she has now

received. Each of the defendants deny liability and loss. Before the High Court, the

HSE admitted breach of duty limited to its failure to advise Ms. Morrissey of the

4.8 It is appropriate next to turn to the judgment of the High Court.

5.1 As a preliminary point, it is appropriate to draw attention to the timescale

within which these proceedings operated. Clearly, having regard both to Ms.

Morrissey‘s medical situation and the fact that there are a large number of other

similar cases pending before the courts, it was considered highly desirable that

5.2 Addressing the proper approach to be applied in a case involving medical

screening such as this one, the trial judge considered that the legal standard of care

and the factual standards and criteria to which the screener must adhere are different,

although interrelated, issues. The trial judge held that the classic statement of the

applicable legal standard of care in the context of medical negligence cases is set out

by Finlay C.J. in Dunne v. National Maternity Hospital, at p. 109:-

―1. The true test for establishing negligence in diagnosis or treatment on the

part of a medical practitioner is whether he has been proved to be guilty

of such failure as no medical practitioner of equal specialist or general

status and skill would be guilty of if acting with ordinary care.

2. If the allegation of negligence against a medical practitioner is based on

proof that he deviated from a general and approved practice, that will not

establish negligence unless it is also proved that the course he did take

was one which no medical practitioner of like specialisation and skill

would have followed had he been taking the ordinary care required from

3. If a medical practitioner charged with negligence defends his conduct by

establishing that he followed a practice which was general, and which

was approved of by his colleagues of similar specialisation and skill, he

cannot escape liability if in reply the plaintiff establishes that such

practice has inherent defects which ought to be obvious to any person

4. An honest difference of opinion between doctors as to which is the better

of two ways of treating a patient does not provide any ground for leaving

a question to the jury as to whether a person who has followed one

5. It is not for a jury (or for a judge) to decide which of two alternative

courses of treatment is in their (or his) opinion preferable, but their (or

his) function is merely to decide whether the course of treatment

followed, on the evidence, complied with the careful conduct of a

medical practitioner of like specialisation and skill to that professed by

5.3 The trial judge placed particular reliance on the decision of the English Court

of Appeal in Penney, Palmer & Canon v. East Kent Health Authority [2000] Lloyds

Rep Med 41 (―Penney Palmer‖), a case which also considered the allegedly negligent

misreading of smear tests by cytology screeners. It is apparent that the trial judge

considered the approach of Lord Woolf M.R. in those proceedings to be entirely

correct, and thus it is appropriate to briefly set out the relevant aspects of that

5.4 In that case, the Court of Appeal expressly approved the legal test for the

standard of care in screening cases, as had been stated in Bolam v. Friern Hospital

Management Committee [1957] 1 W.L.R. 582. This, together with Bolitho v. City and

Hackney Health Authority [1988] A.C. 232, was considered by Cross J. to be the

English equivalent of the Dunne test. Lord Woolf M.R. then stated the following in

relation to the questions to be addressed in the context of determining the negligence

―…the Bolam test has no application where what the judge is required to do is

make findings of fact. This is so even where those findings of fact are the

subject of conflicting expert evidence. Thus, in this case there were three

(ii) At the relevant time could a screener exercising reasonable care have

(iii) Could a reasonably competent screener aware of what a screener

exercising reasonable care will observe on the slide treat the slide as

5.5 Further, Lord Woolf M.R. agreed with the finding of the trial judge in those

proceedings, to the effect that the experts in that case endorsed a requirement of

―absolute confidence‖, being that, if there was any doubt in the mind of a screener as

to whether the slide was normal, the screener should not classify it as negative. Thus,

it was held by the Court of Appeal that a slide should not be classified as negative

unless the screeners had ―absolute confidence‖ that it was so.

5.6 The trial judge in these proceedings adopted a modified version of the three-

stage approach set out by Lord Woolf M.R. and held that, before applying the Dunne

principles, the court must consider what is on the slides, a question of fact which must

be determined on the balance of probabilities. Following such a determination, he

held that the court must then consider the Dunne principles, in reference to the

screener‘s practical obligation of ―absolute confidence‖, in the analysis of their slides.

This was elaborated on further at paras. 71 and 72 of the judgment of Cross J., as

―71. I hold that ―absolute confidence‖ is the screeners practical duty in

relation to their analysis of what is on the slide and indeed the adequacy

of the sample, and the legal issue is whether or not they have carried out

that duty in accordance with the Dunne principles. These extra tests set

out in Penney Palmer are combinations of factual and legal matters, but

I accept that a screening programme especially one such as in Ireland

which does not have annual retesting, is inherently deficient if screeners

ascribe as normal, results in which they are in any doubt. Accordingly,

to ascribe as normal, a slide which the screener has any doubt of that fact

even if he legitimately believes it to be normal on the balance of

probabilities, is to fall below the Dunne standards required of that

screener. Whether the screeners were right not to have any doubt is a

matter to be assessed at law in accordance with the Dunne principles.

72. In other words, if there is any room for doubt that the slide was normal

and the screener ascribes a normal result to the slide then the screener is

in breach of the Dunne principles as he has been guilty of such failure

that no professional scanner of equally specialist or general status and

skill would have been guilty of if acting with ordinary care. A screening

programme cannot operate safely if screeners are left to judge the slides

and whether they are safe merely on the balance of probabilities…‖

5.7 The overall approach which the trial judge considered was appropriate in the

assessment of negligence in the context of medical screening was then set out at para.

―[T]he legal standard to be applied on the issue of the liability of the

defendants is the Dunne test. Questions of fact, however, are for my decision

on the balance of probabilities. The questions of fact include what was to be

seen on the individual slides. Accordingly, as in Penney Palmer, the correct

(ii) whether a reasonably competent screener at the relevant time could have

(iii) whether a reasonably competent screener in the light of what he or she

Questions (ii) and (iii) above and any issues as to adequacy are to be decided

in the light of the ‗absolute confidence‘ test and thereafter, the test for

5.8 Before turning to consider the allegedly negligent analysis of the slides, the

trial judge noted the potential issue of hindsight bias in respect of any retrospective

review of the smear tests. It was held that a court should be wary of the potential for

retrospective bias in the course of an audit or an expert‘s review and thus treat their

results with caution. While noting that there are certain merits to conducting a blind

review in order to reproduce the original screening conditions, the trial judge held that

such a review is not compulsory and that the court should assess the professional

opinion of any expert, whether conducting a blind review or otherwise, and come to a

5.9 Employing the three-part test as set out above, the trial judge then turned to

consider the negligence of Quest in relation to the 2009 smear test. Answering

question (i), it was held that certain abnormal features of cells were visible on the

slide, and thus the trial judge had ―no doubt‖ that the cells on the slide were

AGC/AGUS. In this regard, he accepted the evidence of the Morrisseys‘ expert

witnesses, Ms. Tan and Dr. McKenna, and the conclusions of those who conducted

5.10 In respect of question (ii), the trial judge held that a reasonably competent

screener at the relevant time should not have failed to see what was on the slide,

where there were clear indicators of cell abnormality. In this regard, the evidence of

Dr. McKenna was also accepted. The trial judge held that the evidence challenging

this conclusion, which was given by American expert witnesses appearing on behalf

of Quest, should be treated with caution, in light of the influence of the Guidelines for

the Review of GYN Cytology Samples in the Context of Litigation or Potential

Litigation, which were issued by the American Society of Cytopathology (hereafter,

―the Guidelines‖). The trial judge found that the Guidelines were prepared in an

attempt to limit litigation and to provide a robust defence for cytoscreeners accused of

5.11 In particular, the trial judge disagreed with the provisions of the Guidelines

which seek to impose an obligation in all cases to carry out a blind review and which

suggest that, in the absence of a blind review, it is wrong to impute any screener with

negligence. Cross J. also disagreed with those provisions of the Guidelines which

deal with findings of ASC-US and AGUS cells. In respect of such findings, the

Guidelines state that such cases ―do not represent consistently identifiable

abnormalities and a reasonable basis for allegations of practice below a reasonable

prudent practitioner standard of care‖. Insofar as this provision proposes that it is not

negligent to report a finding of ASC-US or AGUS cells as a normal or clear result, it

was roundly rejected and criticised by Cross J. It is apparent that the trial judge

considered that this attitude towards the categorisation of ASC-US and AGUS cells

was prevalent amongst Quest‘s American expert witnesses. One of Quest‘s expert

witnesses, Prof. Austin, was involved in the drafting of the Guidelines, while another

expert witness, Mr. Feit, referred to the high degree of inter-observer variability in

respect of ASC-US and AGUS categories and stated that the categories are not used at

all in the U.S. for proficiency audits or examinations of cytoscreeners. Mr. Feit‘s

evidence that, in his view, the slide was normal, was dismissed as having been arrived

at on the balance of probabilities, utilising his professional skill and judgement. On

the basis that the Guidelines had affected, even subconsciously, the evidence of the

American expert witnesses, which could give rise to the belief that ASC-US or AGUS

cells need not be examined with the same scrutiny nor be subject to the requirement

of ―absolute confidence‖, the trial judge held that such evidence should be treated

5.12 Quest chose not to call the two screeners who reviewed the slide in 2009 or

the individual who conducted the 2014 audit of the slide to give evidence.

Considering question (iii), it was held that in the absence of any such evidence, the

trial judge had no knowledge as to what they did or did not see or how they did or did

not appraise the slide. Cross J. held that it was his belief that the Quest screeners

were utilising their professional skill and judgment and recording what they believed,

as a matter of probability, was the case. However, applying the requirement of

―absolute confidence‖, and on the basis of the abnormalities as identified on the slide

by Dr. McKenna, it was concluded that the slide ought not to have been categorised as

negative, and that Quest was negligent and in breach of duty in relation to the reading

5.13 As previously mentioned, Medlab was held not to be negligent in its reading

of the slide which had been sent to it. While concluding that abnormalities were

visible on the slide which suggested a categorisation of AGUS/AGC cells, the trial

judge held that these cells were not easily distinguishable from other normal cells and

thus that a screener exercising reasonable care could have failed to identify the same.

The High Court therefore held that the categorisation of the slide as negative for

abnormalities was not negligent nor a breach of duty. This finding is not in issue on

5.14 The trial judge then considered the issue of the adequacy of the 2012 slide. As

previously mentioned, under the Bethesda System a minimum of 5,000 well-

visualised cells must be counted on a slide for it to be considered adequate. The

adequacy of most slides can be determined by a brief visual examination. If this

visual inspection is insufficient, the Bethesda System prescribes that the ThinPrep

method is used to formally test for adequacy. This requires viewing the slide under a

microscope on a vertical or horizontal axis and magnifying random points on that axis

in order to count the number of cells present. From this, the number of cells are

multiplied appropriately in order to ascertain whether the minimum number of

required cells are present on the slide. If the screener finds less than that amount, then

the patient should be referred for a repeat smear test in one to three months.

5.15 Both the internal and external audits of the 2012 slide which were carried out

considered that the reason for the false negative result provided was that the slide was

―very scanty‖. The evidence of the Morrisseys‘ expert witnesses, Ms. Tan and Dr.

McKenna, and Medlab‘s expert witness, Prof. Pitman, also indicated that the slide

contained an inadequate number of cells. Dr. McKenna and Prof. Pitman both

reached this conclusion having carried out a formal test for adequacy as prescribed

under the Bethesda System. Ms. Stowe, another expert witness who appeared on

behalf of Medlab, gave evidence that she found a sufficient number of cells on the

5.16 The trial judge dismissed the evidence provided regarding a study carried out

in St. Vincent‘s Hospital, Dublin, and a second study carried out by Prof. Pitman,

both of which purposively looked for cells, and counted and multiplied the number of

cells located in order to achieve a purported total. It was held that this was not a valid

approach, given that neither study sampled on a random basis in order to calculate an

5.17 The trial judge also dismissed the evidence of Dr. Madrigal, a pathologist who

conducted a computer analysis of the slide on behalf of Prof. Pitman and determined

that there were, in fact, over 35,500 cells on the slide. This alternative system of

counting cells was held to be irrelevant, in circumstances where there has been no

study conducted, subject to peer review or otherwise, to demonstrate that such a

computer analysis is suitable or indeed to determine what number of cells must be

found on the slide by the computer in order to establish adequacy. Further, the trial

judge placed reliance on the fact that Medlab was required, under the terms of its

agreement with CervicalCheck, to utilise the Thin Prep method, as prescribed under

5.18 Turning to consider the 2012 slide, the trial judge held that, on a visual

examination, it was significantly different from all the other slides he had witnessed

and that it appeared to have significantly fewer cells on it. It was held that the

―absolute confidence‖ test is also applicable in the context of the slide‘s adequacy,

meaning that in the event of any doubt on the part of a competent screener, the slide

ought to have been formally tested for adequacy. In light of the evidence put before

the High Court, and the trial judge‘s observations of the 2012 slide, it was held that

the failure to test the slide for adequacy in accordance with the ThinPrep method was

negligent and in breach of duty. In accordance with the Dunne principles, it was held

that no screener of equal specialist or general status would have, if acting with

ordinary care, failed to subject the slide to an appropriate test as to adequacy.

5.19 While acknowledging that the nature of the formal adequacy test carried out

by Ms. Stowe was unclear and was not referred to in any report, the trial judge

accepted that she conducted her test in accordance with the Bethesda System.

However, he held that it was unlikely that any such random review would have found

the slide to be adequate. As Medlab chose not to call the two individuals that initially

reviewed the 2012 slide, there was no evidence provided as to whether an adequacy

test was carried out. The trial judge held that it must be presumed no such test was

conducted. On the balance of probabilities, Cross J. held that, had a test been carried

out by the cytoscreeners, in accordance with the ThinPrep method and the Bethesda

System, it would have resulted in a finding that the slide was inadequate. Thus,

Medlab was found to be negligent and in breach of duty in its failure to properly

5.20 In light of the foregoing findings of negligence on the part of Quest and

Medlab, the trial judge turned to consider the consequences of the negligent

misreading of the slides. On his analysis of the evidence as to causation and having

regard to the fact that Ms. Morrissey was diagnosed with cancer in 2014, the trial

judge held, on the balance of probabilities, that the HPV infection was present for at

5.21 The trial judge went on to hold that had the 2009 slide been reported by Quest

as abnormal, that is, with a categorisation of AGUS/AGC cells, Ms. Morrissey would

have been referred for a colposcopy or, at least, been advised to attend for a repeat

smear in six months‘ time. It was accepted by the High Court that, on the balance of

probabilities, such a follow-up screening or colposcopy would have returned

abnormal results. It was further accepted that the detection of such pre-cancer would

have resulted in the performance of a large loop excision of the Transformation Zone

(LLETZ) procedure. This would have led to a complete excision of the cancerous

cells, with only a 5% chance of the recurrence of precancerous cells, and less than a

1% chance of invasive cancer developing. On the balance of probabilities, the trial

judge therefore determined that the serious prognosis which Ms. Morrissey now faces

5.22 In relation to the 2012 smear, the trial judge reached the same conclusion. He

held that, had the sample been tested in accordance with the ThinPrep method, as

prescribed by the Bethesda System, it would have been deemed as inadequate and Ms.

Morrissey would have been required to undergo a repeat smear within one to three

months. On the balance of probabilities, it was held that the repeat smear would have

resulted in a referral for a colposcopy and that the cancer, which in 2012 was certainly

developing, would have been identified and Ms. Morrissey would have undergone the

same procedure which could have taken place in 2009, with the same benign results.

Accordingly, the negligent misreading of both the 2009 and 2012 slides were found to

have caused the entirety of the Morrisseys‘ claims for injuries and loss.

5.23 The trial judge dismissed the defendants‘ submission that cervical cancer

develops more rapidly in younger women than in older women. On the basis of such

a suggestion, it had been argued that Ms. Morrissey‘s cancer was not present in pre-

cancerous form in 2009 or 2012. The trial judge held that there was no evidence to

support the proposition that Ms. Morrissey‘s cancer had developed more rapidly than

average. The High Court judge further rejected the argument of the defendants that,

as Ms. Morrissey developed squamous cancer and the abnormalities identified on the

reviewed slides were glandular in nature (being categorised as AGUS), the

development of the two were unrelated. In this regard, the trial judge accepted the

evidence of Prof. Wells to the effect that it is highly likely that where one sees

glandular pre-cancer there is an associated squamous pre-cancer, as the same

5.24 The trial judge held that, along with Quest and Medlab, the HSE was liable to

the Morrisseys, in light of its organisational role in the CervicalCheck programme.

Given that it determines the standards to be applied in relation to screening and to

which laboratories must conform and provides for the manner in which screening is

conducted and how it is reported, the HSE was held to have responsibility for ―all

aspects‖ of the screening programme. Further, the trial judge emphasised that the

HSE chose to contract out the screening of smear tests to Quest and Medlab and,

citing Byrne v. Ryan [2007] IEHC 207, [2009] 4 IR 542, he held that a party cannot

evade its liability merely by engaging competent professional persons to perform

5.25 Thus, the High Court held that the HSE has a primary liability towards the

Morrisseys. The trial judge dismissed the HSE‘s submission that primary liability is

based on agency and stated that he accepted the principles of Woodland v. Essex

County Council [2013] UKSC 66, [2014] AC 537 and the five-part test there

outlined by Lord Sumption. The trial judge held that the HSE was also vicariously

liable for the activities of Quest and Medlab, by virtue of the control which it holds

over the laboratories and their performance, as a result of their contractual

5.26 The High Court held that it was impossible to differentiate between the losses

resulting from the failure to properly assess the 2009 and 2012 slides and therefore,

the three defendants were deemed to be ‗concurrent wrongdoers‘ under the Civil

Liability Act 1961, each being responsible for the same damage or for damage that

cannot be distinguished. However, as a result of the terms of the contracts between

the HSE and the other defendants, the HSE was found to be entitled to an indemnity

against the laboratories for all matters other than its liability in relation to the non-

5.27 Assessing the quantum of damages to be awarded, the trial judge held that Ms.

Morrissey was entitled to both general damages and special damages. Noting that the

cap on general damages was most recently fixed at €450,000 by Quirke J. in Yun v.

Motor Insurers Bureau of Ireland [2009] IEHC 318, during the time of an economic

recession, the trial judge considered it appropriate to raise the cap on general damages

to €500,000. Ms. Morrissey was awarded special damages in respect of, amongst

other things, the cost of her future care and the loss of her earnings for the remainder

of her life. The High Court held that the admitted breach of the HSE in its failure to

advise Ms. Morrissey of the results of the audits did not result in any further personal

injury being suffered by Ms. Morrissey and therefore that only a finding of nominal

5.28 The trial judge held that the Ms. Morrissey was entitled, as against all

defendants, to: general damages of €500,000; miscellaneous special damages (agreed)

of €12,508; cost of home adaption (agreed) of €70,000; occupational therapy costs

(agreed) of €55,000; loss of earnings in the sum of €50,000; care costs of €60,000;

giving rise to a total to Ms. Morrissey of €747,508; together with €10,000 in nominal

damages against the HSE in respect of its negligence in relation to the audit. The total

amount of damages awarded to Ms. Morrissey was in the sum of €757,508.

5.29 The trial judge held that Mr. Morrissey was entitled to general damages for

loss of consortium and to special damages arising from that loss, specifically in

respect of the future care that he and his daughter will require after Ms. Morrissey‘s

death and the loss of Ms. Morrissey‘s future earnings throughout her career. The

defendants submitted that, on the basis of comments made by Geoghegan J. in

Coppinger v. Waterford County Council [1998] 4 I.R. 243, a spouse is only entitled to

damages for loss of consortium in respect of his losses while Ms. Morrissey is alive.

This was dismissed by the trial judge, who held that a spouse is entitled to damages

for future pecuniary losses under the heading of loss of consortium and that it was

irrelevant that such damages are, in certain circumstances, available in an action taken

under Part IV of the Civil Liability Act 1961, in relation to fatal injuries.

5.30 The trial judge then referred to the decision of Lavan J. in Mahon v. Burke &

anor [1991] 2 I.R. 495, which held that, where a deceased person has brought an

action in negligence and was awarded damages, a statutory dependent was not entitled

to maintain an action under Part IV of the 1961 Act, as to hold otherwise would be to

subject a defendant to more than one case arising from the same cause of action. It

was then pointed out that this authority can operate to exclude the dependants of an

injured party from making claims for future financial losses because the injured party

had vindicated her rights during her lifetime in a personal action claiming damages

for personal injury. A situation where an injured party was confronted with a choice

to either vindicate her own rights in a personal action claiming damages for personal

injury, therefore depriving her dependants of any separate claim in the future, or

alternatively forego her personal claim so that her dependants could achieve

compensation for their losses, was described by the trial judge as ―grossly unfair‖.

5.31 To remedy this situation, the trial judge suggested that the principle of ―lost

years‖ could be employed to compensate a plaintiff whose working life expectancy

has been cut short for the loss of her future earnings and the loss of her ability to care

for her children. If Mr. Morrissey is not, in fact, entitled to damages for future losses

under the heading of loss of consortium, then the trial judge held that he would have

found that the same damages would be recoverable by Ms. Morrissey under the

heading of ―lost years‖. The trial judge also highlighted the potential application of

the ―lost years‖ principle in circumstances where a claimant does not have a spouse

that can recover such damages for loss of consortium but does have child dependants

5.32 The trial judge held that the Mr. Morrissey was entitled, as against all

defendants, to: general damages for loss of consortium in the sum of €60,000;

damages for loss of Ms. Morrissey‘s income of €600,000; loss of pensions, company

car and share options in the sum of €150,000; costs of care for their child of €500,000;

costs of domestic assistance of €75,000; retrospective care costs of €13,432 and

bereavement counselling costs of €6,532; thus giving a total to Mr. Morrissey of

€1,405,000. The overall sum of damages awarded to Mr. and Ms. Morrissey was,

therefore, €2,162,508. The trial judge assessed that sum as being fair and reasonable

5.33 As will have been seen, most of the five issues identified earlier as arising on

this appeal involve quite distinct questions. The only exception are the two issues

raised by, respectively, Quest and Medlab, which suggest that the trial judge failed to

adequately or appropriately engage with their case on the question of negligence.

While the precise manner in which those defendants analysed the purported failure of

the trial judge in that regard involves different considerations, there is clearly a great

5.34 I will, therefore, deal with those two issues together. However, it seems to me

to be appropriate to deal separately with each of the other issues and to set out the

arguments of the parties in respect of those different issues as part of the discussion

under each heading. I propose, therefore, to turn first to the question of the legal

standard of care in clinical negligence cases and its application in the context of

6.1 As noted earlier, developments at the oral hearing suggested that there was

ultimately little difference between the parties as to the proper approach to be adopted

by a screener in a case such as this. However, as this matter has generated significant

controversy and was, indeed, one of the reasons why this Court considered it

appropriate to allow a leapfrog appeal directly to the Court, it is important that the

6.2 I have already cited the so-called Dunne principles. All parties ultimately

agreed that those principles continue to represent the law in this jurisdiction. It is also

clear that there are significant similarities between the Dunne principles and the legal

test for the standard of care identified in the United Kingdom in Bolam and Bolitho.

6.3 Before addressing the precise principles to be applied, it is appropriate to

briefly make a number of preliminary observations. The first is that care needs to be

taken in the use of the term ―standard of care‖. As the trial judge pointed out, that

term has a precise legal meaning and represents, at the level of principle, the legal

duty which applies in any particular circumstance. However, given that, in clinical

negligence cases, a court is dealing with ―care‖ in a medical sense, the phrase

―standard of care‖ might at least colloquially be used to define the appropriate

standard by reference to which the approach of a relevant professional to a particular

problem should be assessed. To avoid any possible confusion between the term

―standard of care‖ in its precise legal meaning and what might colloquially be called a

standard of care in a clinic setting, I will use the term ―standard of approach‖ to mean

the standard which, in practise, has been shown to be required of a particular

6.4 One further observation may also be useful at this stage. There are, of course,

similarities between the Dunne principles as applied by the courts in this jurisdiction

and the Bolam and Bolitho principles applied in the United Kingdom. I have already

cited a passage from the judgment of Lord Woolf M.R. in Penney Palmer in which he

indicated that ―the Bolam test has no application where what the judge is required to

do is make findings of fact‖. It seems clear that, in the context in which Lord Woolf

M.R. was commenting, the Bolam test is in reality a reference to the fourth point

identified in the judgment of Finlay C.J. in Dunne (as already cited). Thus where, in a

number of judgments, the U.K. jurisprudence suggests that the Bolam test has no

application in particular circumstances, it does not seem to me that it follows that the

Dunne principles may not apply, for the specific point decided in Bolam forms but

6.5 In any event, it is important to start by indicating that, while the principles

identified by Finlay C.J. in Dunne are expressly related to negligence against medical

practitioners, they mirror the test applied across the board in cases involving

allegations of professional negligence. For example, a similar approach to legal

advice allegedly negligently given was adopted by this Court in Roche v. Peilow

6.6 Thus, the starting point in any professional negligence case 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. Accepted practice is highly relevant, although

departure from normal practice may be found not to give rise to negligence where an

equivalent professional might reasonably have followed such a course of action while

exercising reasonable skill. Likewise, following normal practice may not absolve the

professional from a claim in negligence if it can be shown that there was an inherent

defect in the practice which should have been obvious to professionals of the type in

question. Finally, it is emphasised that the question will not come down to one of a

judge deciding which of two or more possible courses of action might have been

considered, often with the benefit of hindsight, to have been preferable, but rather

whether the course of action actually adopted was consistent with the exercise of the

ordinary care which could reasonably be expected of a professional of the type under

6.7 In my view, the various aspects of the Dunne test identified by Finlay C.J. can

be reduced to one overarching principle with a number of subsidiary considerations

which impact on the application of that overarching principle in particular

circumstances. The overarching principle is to be found in point (1); the standard of

approach of a medical professional is to apply a standard appropriate to a person of

equal specialist or general status acting with ordinary care. A failure to act in that

6.8 Each of the other points made by Finlay C.J. derive from that overall

obligation. Thus, even if the medical practitioner concerned deviated from what

might have been established to be a general and approved practice, the practitioner

concerned will not be found to be negligent unless it can also be shown that no

practitioner of equal status exercising ordinary skill might not also have deviated in

the same way. This represents point (2) of the matters identified by Finlay C.J. and

can be seen to represent the application of the overarching principle to circumstances

where the relevant practitioner has, in fact, deviated from an established practice.

6.9 A similar comment can be made in respect of point (3). Here, Finlay C.J.

makes clear that following an accepted practice may not absolve a practitioner from

negligence if it can be demonstrated that there were inherent defects in the practice

which ought to have been obvious to a professional giving the matter due

consideration. This, again, can be seen to derive from the overarching principle, for a

professional cannot be said to have applied the required standard of approach, being

to exercise the ordinary skill which would be appropriate to a person of the status in

question, if it can also be shown that any such person ought to have realised that the

6.10 In addition, a similar observation can be applied to points (4) and (5) as

identified by Finlay C.J. in Dunne. Where there are two or more schools of thought

as to what course of action should be adopted in particular circumstances, it is clearly

the case that a professional exercising reasonable care may adopt one of them even

though there may be others who would take a different course. These principles, in

effect, represent what might be called the Bolam element of the Dunne test and can

6.11 Finally, it is important to note that Finlay C.J. went on, at p. 109 of his

judgment in Dunne, to note that the relevant principles ―apply in identical fashion to

questions of diagnosis‖ as well as to questions of treatment. There was no argument

put forward on this appeal that different principles would apply in the case of

screening. There are some passages in judgments of the courts of the United

Kingdom which do seek to make a distinction between treatment cases and diagnostic

cases. It is, of course, the case that the application of general principles to particular

circumstances may give rise to somewhat different approaches, which stem from the

very fact that circumstances may be different. It may very well be the case that

questions concerning deviation from accepted practice, following inherently defective

practices or choice between two respectable professional opinions may be much more

likely to arise in the context of decisions on treatment rather than decisions on

diagnosis (or, indeed, it should be said, screening). But it is not, for example,

impossible to envisage that there might be circumstances where there were two

different approaches to diagnosis in a particular set of circumstances, both of which

had their reasonable adherents. I cannot see that those aspects of the Dunne principles

which deal with such questions would not have equal applicability to a case of alleged

misdiagnosis in such circumstances. There seems to me, therefore, to be strong

grounds for favouring the view adopted by Finlay C.J. in Dunne which does not seek

to distinguish, at the level of principle, between different types of clinical tasks.

However, it does need to be acknowledged that how those general principles apply

may well be different in particular circumstances. If there are not two schools of

thought about how to go about a particular diagnosis, then the question of the

applicability of those aspects of the Dunne principles which address such questions

(and are at the heart of what are described as the Bolam principles in the United

Kingdom) simply has no application. That is so not because there is a difference in

principle between the approach to treatment cases and other cases, but rather it simply

derives from the fact that, in practice, some of the Dunne principles may have no

application in a particular type of case because they do not arise on the facts.

6.12 No issue arose in these proceedings concerning the adoption of a common

practice which had inherent defects which ought to have been obvious. Likewise,

there was no issue concerning there being two schools of respectable thought or of an

acceptable deviation from common and established practice. The issue of the

standard of approach in this case came down, therefore, to the same issue which

almost inevitably arises in any professional negligence case. 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? The test to be applied in this

case is, therefore, no different to the test which would properly be applied in any

professional negligence case. Ultimately, in the particular circumstances of this case,

the question came down to establishing what the approach of relevant professionals to

screening should be and whether it had been demonstrated that the actions of any of

the screeners in this case had fallen below the reasonable standard which

professionals of that type could be expected to have applied, having regard to that

6.13 Next, it must be emphasised that the question of the standard of approach

which should be applied by an ordinarily competent professional is ultimately a

matter of fact. It requires expert evidence 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. It follows that, at least in many cases, the court has no

role in determining the standard to be applied other than to assess the evidence given

by professionals as to the standard to which they themselves regard as being

appropriate to someone of the standing and skill of the defendant. There may be some

scope for the court reaching a further assessment in those limited cases where it is

said that a professional did comply with an accepted practice but where there were

also said to be inherent defects in the practice concerned. As already noted, no such

6.14 It follows that, in a case such as this, a court has no role in imposing a standard

of approach on a professional. Rather, it is the standards of the profession itself, as

demonstrated by the evidence, which impose the standard required.

6.15 I have felt it important to emphasise these factors to ensure that there is no

misunderstanding about the court‘s role in cases such as this. Some public

commentary on the decision of the High Court in this case seems to have ignored the

6.16 It seems to me next to be appropriate to address the way in which evidence

was given in the High Court for, as I have pointed out, it is that evidence, given by

professionals themselves, which determines the standard to be applied by the Court to

6.17 In that context, it is of some importance to emphasise that the analysis of the

Court of Appeal for England and Wales in Penney Palmer was based on the evidence

given by relevant professionals to that court in that case. Further, that analysis was

also expressly put to many of the experts who gave evidence in this case, such that its

terms actually form part of the evidence given, in the sense that witnesses gave their

testimony, at least in part, by reference to the conclusions reached in Penney Palmer.

It is in that context that the term ―absolute confidence‖ came into play. It is of the

highest importance to understand the way in which that term was used by the courts

of England and Wales, in the evidence in this case and, it follows, by the trial judge.

6.18 First, it should be said that it is clear that the term ―absolute confidence‖

derived from what the trial judge in the Penney Palmer proceedings (see Penney,

Palmer and Cannon v. East Kent Health Authority [1999] Lloyd‘s Rep Med 123)

considered to be the agreed evidence of the experts before him. Reviewing the

―All five [experts] agreed that if the screener was in any doubt about

what he saw on the slide he should not classify the smear as negative. In

their evidence before me each expressed the point differently but the

6.19 The requirement of ―absolute confidence that the smear is within the normal

range‖ before a screener ascribes to it a clear result was said to be the appropriate

practice of a screener, as advanced by one such expert, a standard with which the trial

judge stated that all the other experts were in agreement. Likewise, the adoption by

the Court of Appeal for England and Wales of the same term followed on from the

finding of the trial judge, based, as it was, on the evidence. At para. 40 of his

―The judge's reliance on the absolute confidence test is understandable. The

phrase 'absolute confidence' was no more than shorthand for the approach

which on examination of the transcripts it seems to us all the experts

6.20 It follows that the adoption of the ―absolute confidence‖ test in England and

Wales stemmed from an assessment of evidence given by relevant professionals on

both sides of the Penney Palmer case. It was not itself a court imposed obligation but

one which, on the evidence, the trial court assessed as being the agreed position of

6.21 Turning then to the evidence before the High Court in this case, it is important

to note that the question of there being potentially a difference between the

appropriate standard applied in the United Kingdom and that applied in the United

States did form some part of the debate on the evidence. It was in that context that

evidence was given of the fact that an ―absolute confidence‖ test was applied in the

United Kingdom (including Northern Ireland) and was also the subject of discussion

with some of the United States witnesses called on behalf of the defendants.

6.22 Dr. McKenna, a consultant cellular pathologist based in Northern Ireland who

gave evidence on behalf of the Morrisseys, was referred to the Penney Palmer

authority during his examination in chief before the High Court on 30 January 2019:-

―Q. Can I just ask you about something… it‘s English decision called

Penney, Palmer and Cannon v. East Kent Health Authority and it‘s a

decision of the English Court of Appeal of 16th December 1999. Are you

Q. There‘s just one particular thing said in that judgment I just want to draw

attention to at this stage and it was that the English Court of Appeal said

that all the experts endorsed the ‗absolute confidence‘ test, under which

if there is any doubt in the mind of the cytoscreeners as to whether the

slide was normal, he or she should not classify it as negative. Now, what

exactly is the English Court of Appeal saying there? Like, what is the

A. I think what it‘s saying is that unless you are absolutely – and that is

100% -- convinced that the slide is negative, do not call it negative.

Q. So, in other words, what I am getting at here is that when a cytoscreeners

has a slide in front of him or her, is there any acceptable level of doubt –

A. No, you can only – you should only report something as negative when

6.23 When the ―absolute confidence test‖ was put to Dr. McKenna once again the

following day, under which ―if there is any doubt in the mind of a cytoscreeners as to

whether the slide was normal he or she should not classify it as negative‖, he stated

―A. And I fully agree with that statement, and in fact those are the very

words that I always use to my staff; if you have any doubt, then pass it

6.24 Ms. Tan, a cytotechnologist based in New York who also gave evidence on

behalf of the Morrisseys, stated that there is no tolerance for any doubt in the

assessment of the normality of the cells. The evidence tendered by the experts

appearing on behalf of Medlab also indicated their agreement with the ―absolute

confidence‖ standard. Prof. Pitman, the lead consultant cellular pathologist for

Medlab, agreed that a cytotechnology must have absolute confidence that a smear is

normal before reporting it as normal. Ms. Stowe, a cytotechnologist based in

Wisconsin, also had the following exchange during cross examination:-

―Q. Now let‘s be clear about this. Do you agree with me that the test is that a

cytotechnologist or a cytoscreeners must have absolute confidence that

6.25 The experts who appeared on behalf of Quest, both based in the United States,

expressed some reservation with the ―absolute‖ nature of the standard, however. Prof.

Austin, a consultant cellular pathologist based in Pennsylvania, expressed the

cytoscreeners must have absolute certainty to pass cells as normal?

A. The cytotechnologist should be confident if a slide is negative or if it is

6.26 While Prof. Austin was under cross-examination, the following exchange in

―Q. …You see, that decision in the UK, the Penney, Palmer and Cannon v.

East Kent Authority, did you know that experts from The United States

Q. All right. Did you know that they all agree, all the experts actually

agreed in that case that the test is "the absolute confidence" test? Did you

Q. So what I just want to be clear about is this, because we are dealing with

a slide which is in a laboratory in Grand Rapids in Wyoming in August

2009; are you saying that the cytotechnologist who looked at that slide

because I wasn't there, you know, instructing the cytotechnologist in the

Q. MR. JUSTICE CROSS: When you said that a competent screener when

looking at this slide in 2009 might or could have reported it as normal...

MR. JUSTICE CROSS: ...were you saying that that could have been done by a

competent screener assessing his or her professional view and coming to

a decision on the balance of probabilities or were you saying that a

competent screener could say with absolute confidence that it was

A. I can certainly say with confidence. To be honest, this idea of absolute

A. It is not something we routinely deal with. So I am not quite sure how to

say the difference between 'confidence' and 'absolute confidence'.‖

6.27 After the trial judge‘s intervention, the cross-examination continued and Prof.

Austin appeared to ultimately express agreement with the standard of ―absolute

―Q. Now what I just want to be very clear about here, I want to be clear that

the standard that applied in Grand Rapids in Wyoming in August 2009,

all right, that‘s what we are dealing with here in this case, all right. Now

are you suggesting that a cytoscreener, when he or she is looking at a

slide, should be anything other than absolutely confident if they are

6.28 Mr. Feit, a cytotechnologist based in Wisconsin, also expressed misgivings in

relation to the ―absolute‖ nature of the confidence required:-

―A. Well, this is something that -- so absolute confidence for me, as a

scientist, absolute confidence is very hard to have because, I mean,

generally in science, you generally say the evidence supports this or

Q. MR. JUSTICE CROSS: And there can be levels of support, can't there?

A. Yes, yes, but I don't think -- I mean, you probably wouldn't even have --

A. Pretty close, probably, pretty close. But I mean, in science, it's just, you

confidence is something I just, it's very difficult for me. I know it

6.29 From that review of the evidence, it is first clear that there was evidence that

the actual standard applied by relevant professionals in the United Kingdom was one

of ―absolute confidence‖. Given some of the ―doomsday‖ predictions which followed

on from the judgment of the High Court in this case, I do feel it necessary to say that

at least some of the more extreme comments are very hard to reconcile with the fact

that there was clear evidence that such a standard is actually applied in the United

Kingdom generally and, to the particular knowledge of an important expert witness, in

Northern Ireland specifically, without the screening systems in those countries

6.30 It is also absolutely clear that, while there may have been some quibbling on

the part of witnesses called on behalf of the defendants with whether the term

―absolute‖ was appropriate, all of the witnesses agreed that any doubt about whether

what is seen on a slide is normal must not allow the case to be reported as a clear

result. Again, it must be emphasised that imposing such a standard of approach does

not derive from the Court but rather from an assessment of the evidence given on all

sides concerning the standard actually applied by professionals in the area in question.

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

6.31 Next, it is perhaps of particular importance to note again that the trial judge in

this case clearly accepted that it was possible for competent screeners, applying an

appropriate standard of approach, to come to different views about whether there was

even a doubt as to whether a clear result could be given. It would, indeed, have been

difficult for the trial judge to have come to any different view having regard to the

evidence given in respect of the United Kingdom‘s screening programme (which is

accepted as being one of the very highest standard), where, notwithstanding those

high standards, reviews suggest of a different view to that of the original screener

being taken in a significant proportion of those cases where a diagnosis of cancer

6.32 But if any proof were needed of the practical acceptance by the trial judge of

the fact that there could be genuine differences of opinion between competent

screeners, even applying a no doubt test, this can be found in the very finding of the

trial judge in respect of the claim for negligence against Medlab relating to the

reading of the 2012 slide. It will be recalled that, while the trial judge found Medlab

negligent in respect of the adequacy issue, the trial judge was not satisfied that it had

been demonstrated that Medlab was negligent in respect of the manner in which the

screener did not identify any suspicious material on the slide in question. The trial

judge came to that view notwithstanding the fact that there was expert evidence,

tendered on behalf of the Morrisseys, which suggested that a competent screener

would have identified the slide as problematic. It follows that this case itself provides

a very clear example of how, even applying what the trial judge considered to be an

―absolute confidence‖ test, there was nonetheless room for a finding, contrary to at

least some of the evidence, that this was the sort of case where competent screeners

could come to different views and where there was, therefore, no negligence. The

suggestion, therefore, that the trial judge was applying an ―absolute confidence‖ test

in a way which effectively meant that every error (even with the benefit of hindsight)

must result in a finding of negligence is demonstrably wrong and would clearly be

seen to be such to anyone who had taken the trouble to read the judgment. If the trial

judge were applying the test in that way then it is impossible to see how he could have

found against the Morrisseys in respect of the allegation that Medlab had negligently

read the slide, as opposed to the finding which he actually made, which was to the

effect that Medlab was only negligent in treating it as adequate.

6.33 Finally, it is necessary to touch on a question on the standard of approach issue

in respect of which there was some debate at the hearing. That debate concerned the

use of the term ―test‖ in relation to ―absolute confidence‖. Counsel for the Morrisseys

accepted that the Dunne test was the appropriate test and was anxious not to be heard

to suggest that Penney Palmer provided a different legal test. In that view, I consider

counsel to be correct. Penney Palmer does not establish any different legal test.

Rather the term ―absolute confidence‖, which derives from Penney Palmer, can be

seen as no more than a synopsis of the evidence given in that case and not as

representing any different or separate test. However, it is worth noting that, in the

written submissions filed in the High Court, counsel on behalf of the Morrisseys did

describe it as ―the absolute confidence test‖. That may well explain why the trial

judge also described it as a ―test‖ in his judgment. However, it is clear that it does not

represent a separate or different legal test. The only test for the standard of approach,

at least so far as relevant in this case, is based on evidence as to the standard actually

applied by competent professionals in the area. In describing that standard, based on

the evidence, it is important not to get too hung up on the words used, for to do so can

create an erroneous impression that the language used has a status amounting to a

6.34 Indeed, it may well be that the introduction of the term ―absolute confidence‖

into the debate has created more confusion than clarity. Ultimately, the overall

question which a court has to address in a case such as this is as to whether a

reasonably competent screener could have given a clear result, both having regard to

adequacy and the absence of suspicious material. On the evidence in this case, it

seems clear that the standard of approach to be adopted by a screener in such

circumstances is that a screener should not give a clear result if they have any doubt

6.35 That is the only test to be applied. Therefore, the question for the court‘s

assessment is whether, on the evidence in this case, it has been shown on the balance

of probabilities that a competent screener could not have concluded that there was no

6.36 It does not seem to me that there is any basis for suggesting that the trial judge

6.37 It may be useful, at this stage, to also comment on the role of expert evidence.

For the reasons already addressed, it is clear that, in any case of dispute, it will be

necessary to tender expert evidence as to what the appropriate standard of approach

should be for a professional of the standing and skill of the one against whom an

allegation of negligence is made. But expert evidence may play another role. It may

assist the court in deciding on the facts of the individual case whether the relevant

professional has lived up to whatever standard of approach has been established.

Expert evidence of that type may play a role in many types of negligence actions.

6.38 For example, in the case of an industrial accident, expert evidence may be

tendered to establish the appropriate standards which should be applied by employers

in the industry concerned. But expert evidence might also be tendered to assist the

court in deciding whether, as a fact, the accident occurred in a way which could be

said to be attributable to a failure to apply those standards. For example, an engineer

might be called to give evidence about whether it was likely that an accident could

have happened in the way in which it was described by a witness. It might be said

that the workings of whatever industrial process was involved would not make it

possible for an accident to have occurred in a particular way. Similarly, medical

evidence might be tendered as to whether injuries actually suffered were or were not

consistent with an accident having occurred in a way described by a witness. In those

latter cases, the experts are not giving evidence as to the standards to be applied but

rather are giving evidence which may assist the court in conducting its role of

6.39 It is worth emphasising that, where expert evidence is tendered in that latter

respect, there is no place for the application of the Bolam test or, in the Irish context,

the third item of the Dunne test. The experts may express different views about the

likelihood or otherwise of something having happened in a particular way or the

likelihood or otherwise of a competent professional having acted in a particular way.

But when such views are expressed, they only represent evidence to assist the court in

deciding on the facts. The court must choose between any such competing evidence

and reach an overall assessment on the facts on the basis of all the evidence tendered.

In such a case, the court is not involved in identifying that there are two respectable

schools of thought about the proper course of action to be adopted, such that a

professional will be absolved from liability if following either. Rather, in such a case

the expert is simply tendering evidence which may be useful, to a greater or lesser

extent, in assisting the judge to reach a conclusion on the facts. Such evidence may

not even be absolutely necessary in every case. To take but a simple example, it

might be established in the circumstances of a particular negligence claim that it was

accepted practice that a doctor should have taken a particular course of action if a

patient had a temperature above a certain level. Whether the patient actually had a

temperature above that level when measured by the relevant doctor is a question of

fact. There might be competing evidence. A doctor might give evidence of one

reading, whereas a nurse who was present might suggest that the doctor‘s evidence

was incorrect. In such a case, the court would have to make a finding of fact based on

that evidence in circumstances where expert evidence would be of no assistance.

However, where the facts involve difficult questions on which expertise may bear,

then it will, of course, often be useful for the court to have the benefit of such expert

6.40 That being said, I would also accept that the practical approach adopted in

Penney Palmer may well provide a useful guide to the way in which a court should

approach its task in circumstances of this case. Ultimately, however, the question is

as to whether a competent screener could have given a negative, or clear, result. The

first question for this Court is, therefore, as to whether there was evidence to support

the trial judge‘s conclusion that a screener must not give a clear result when in any

doubt. 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. It

follows that the second question, in respect of the slide tested by Quest, was as to

whether there was evidence and analysis to support the finding that a competent

screener could not have formed the view that there was no doubt but that the slide in

question was clear. Likewise, the second question in respect of Medlab was as to

whether there was evidence and analysis to support the finding that a competent

screener could not have formed the view that the sample was adequate.

6.41 Those issues obviously involve considering the slides themselves, potentially

with the benefit of expert evidence, and considering whether any element of those

slides was such that it should not have left a competent screener without a doubt.

Ultimately it does not seem to me that the approach of the trial judge in that regard

was materially different to the approach which I have just outlined. On that basis, it

does not seem to me that the standard of care issue itself could provide any basis for

allowing an appeal from the decision of the trial judge in this case. In those

circumstances, it is next necessary to turn to the two connected issues, which concern

whether the trial judge failed, as both Quest and Medlab argue, to properly engage

7.1 While the specific issues raised by, respectively, Quest and Medlab were

different and depended very much on the factual issues arising in the separate cases as

against them, there is, nonetheless, a similar approach adopted by both of these

7.2 While it will be necessary, therefore, to turn to the specific issues which arise

in the different case made against both of these defendants, it is appropriate to start

with some general observations which are applicable to both sets of issues.

7.3 The legal position is clear. The classic statement of the correct approach to be

taken by an appellate court in respect of findings of fact made at first instance is as set

out by McCarthy J. in Hay v. O’Grady [1992] 1 I.R. 210 at pp. 217-218. In particular,

it was held that an appellate court is bound by the findings of fact made by the trial

judge when they are supported by credible evidence. Further, McCarthy J.

emphasised the importance of a clear statement by the trial judge of his findings of

primary fact, the inferences to be drawn and the conclusion that follows.

7.4 This latter obligation on the trial judge underpins the case law which

subsequently developed on the question of the appropriate engagement on the part of

a trial judge with the competing arguments of the parties to litigation. Any party is

entitled to a judgment which states why the party concerned won or lost. In Doyle v.

Banville [2012] IESC 25, [2018] 1 I.R. 505, this Court held that, to this end, it was

important that the judgment of the trial court engages with the key elements of the

case made by both sides and provides a reasoned conclusion as to why the case on the

facts made by one or other side is preferred. In my judgment in that case, a

distinction was drawn between circumstances in which there may have been a

significant and material error in the way in which the trial judge reached a conclusion

as to the facts, in respect of which an appellate court can and should intervene, and a

case where the trial judge was simply called on to prefer one piece of evidence to

another and does so for a stated and credible reason. In the latter case, it is not the

function of the appellate court to revisit the trial judge‘s findings.

7.5 This obligation on the trial judge to engage with and adequately address the

competing arguments of the parties on the facts was restated by this Court in Wright v.

AIB Finance & Leasing and ors [2013] IESC 55 and Ulster Bank v. Healy

[2015] IESC 106. Importantly, in Leopardstown Club Ltd. v. Templeville Developments Ltd

[2017] IESC 50, [2017] 3 I.R. 707, MacMenamin J. set out in clear terms the

approach to be taken by an appellate court when reviewing the engagement of the trial

judge with the arguments of the parties to litigation, at paras. 109-111:-

―109. Save where there is a clear non-engagement with essential parts of the

evidence, therefore, an appeal court may not reverse the decision of a

trial judge, by adverting to other evidence capable of being portrayed as

110. ―Non-engagement‖ with evidence must mean that there was something

truly glaring, which the trial judge simply did not deal with or advert to,

and where what was omitted went to the very core, or the essential

validity, of his findings. There is, therefore, a high threshold. In effect,

an appeal court must conclude that the judge's conclusion is so flawed,

to the extent that it is not properly ―reasoned‖ at all. This would arise

only in circumstances where findings of primary fact could not ―in all

reason‖ be held to be supported by the evidence (see Henchy J. in V.C. v.

J.M. and G.M. [1987] I.R. 510, at p. 523, quoting his earlier judgment in

engagement‖ will not, therefore, be established by a process of

identifying other parts of the evidence which might support a conclusion

other than that of the trial judge, when there are primary facts, such as

here. Each of the principles in Hay v. O'Grady [1992] 1 I.R. 210 is to be

111. The task faced by the judges of our appeal courts is already too onerous.

reduced to a piece-by-piece analysis of the evidence, in an effort to

show, on appeal, that the trial judge might have laid more emphasis on,

or attached more weight to, the evidence of one witness, or a number of

witnesses, or one document, or a number of documents, rather than

others on which he or she relied.‖ (emphasis included in original)

7.6 In the context of that case law, it is, however, important to emphasise a

number of features which are of some relevance to the issues which arise on this

7.7 First, it is clear that what is spoken of as a lack of engagement in those

authorities relates not so much to the way in which a trial judge conducted the

proceedings but rather to the way in which the trial judge determined the issues in the

judgment. A failure to engage, in the context in which it is used in the relevant case

law, clearly refers to the failure on the part of a trial judge to set out the reasons why

central or important aspects of the case of one or other party on the facts were not

7.8 I did not understand counsel on any side of this case to be in the least way

critical of the way in which the trial judge conducted these proceedings. Even a

general reading of the transcript makes it clear that the trial judge was fully involved

with the evidence and the issues and, in that sense, was clearly engaged with the case.

However, that is not the sort of engagement to which the case law referred to earlier is

directed. Rather, the criticism suggested both by Quest and by Medlab is directed to

the question of whether the trial judge gave adequate reasons in his judgment for

rejecting what are said to be central aspects of their respective cases on the facts.

7.9 Second, it is worth adding that it is clear from that case law that it is far from

sufficient for a party seeking to appeal a decision of a trial court to search through the

undergrowth of the pleadings and evidence so as to find some tangential or minor

aspect of the case which is not expressly referred to in the judgment. The test,

therefore, involves, as MacMenamin J. pointed out in Leopardstown Club, a high

threshold which requires the court to address the question of whether, taking that

party‘s case as a whole, can it be fairly said that the trial judge has significantly failed

to adequately address the reasons for rejecting the appellants‘ case on the facts?

7.10 Finally, before going on to consider the specific issues raised by the respective

defendants, it is also of importance to emphasise that it was accepted by both Quest

and by Medlab that there was evidence on which the trial judge could have concluded

that the facts were as alleged by the Morrisseys. In other words, there was evidence

that a competent screener would, in the case of Quest, not have given a clear reading

and, in the case of Medlab, not considered the slide adequate. This is not the sort of

case, therefore, where the primary issue identified in Hay v. O’Grady arises. In most

cases in which factual decisions are sought to be overturned, it is asserted that there

was no evidence on which the decider of fact could have reached a particular adverse

conclusion. This is not such a case. Rather, this case is solely concerned with

whether there was sufficient engagement in the judgment with the case on the facts

made by the two laboratories so as to give a reasonable and broad explanation as to

why what were said to be important elements of their respective defences on the

7.11 Against those general observations, it is necessary to turn separately to the

8.1 It was argued on behalf of Quest that there was a failure on the part of the trial

judge to engage with its case in respect of the interpretation of the 2009 slide, which

was to the effect that a reasonably competent cytoscreener could have reported the

2009 slide as clear or negative. Quest submits that the trial judge reached two

erroneous conclusions in relation to its U.S. expert witnesses, resulting in a failure on

the part of the High Court to engage with the cumulative effect of their evidence.

8.2 First, as previously referred to, the trial judge considered that the evidence

tendered by American expert witnesses called by Quest should be treated with caution

and as being ―subconsciously affected‖ by the Guidelines issued by the American

Society of Cytopathology and its view on the reading of ASC-US and AGUS cells.

Quest submitted that this conclusion was not supported by evidence and that at no

point in these proceedings was it submitted that the Guidelines should set the legal

standard for negligence in cytoscreening. While Quest‘s expert cytotechnologist, Mr.

Feit, and its expert pathologist, Prof. Austin, were members of the American Society

of Cytopathology, it was submitted that their conclusions as to the presence of

abnormalities on the 2009 slide were derived from honestly held opinions rather than

bias towards the categories denoting borderline abnormalities arising from the content

8.3 Second, Quest argued that the trial judge also erred in concluding that Mr. Feit

and the Quest cytoscreeners reached their conclusions as to the 2009 slide on the

balance of probabilities, in circumstances where, it was submitted, there was no

8.4 As a result of these errors, Quest submitted that the conclusions of the trial

judge in respect of the interpretation of the 2009 slide contained no real engagement

with the evidence of Prof. Austin and Mr. Feit and that there was no proper

explanation as to why the evidence of the Morrisseys‘ expert witnesses was preferred

8.5 Furthermore, it was argued that there was a failure to engage with the evidence

adduced by Quest of the blind review which Mr. Feit coordinated of the 2009 slide.

This review was submitted to have reproduced the original screening conditions

insofar as possible in order to assess how the reasonably competent cytoscreeners

would interpret a slide. The 2009 slide was reviewed by eight cytoscreeners,

alongside nine other slides, and the screeners did not know which slide was of

interest, nor by whom they were instructed. Six out of eight screeners deemed the

8.6 Quest submitted that the trial judge‘s approach to the blind review was flawed.

While it was accepted that Cross J. had engaged at a high level with some of the

―hazards‖ said to be associated with blind reviews in general, the trial judge was said

not to have engaged at all with the methodology or results of the Quest blind review

and not to have explained how its results can be consistent with the finding of

negligence which was arrived at. This occurred, Quest submitted, in circumstances

where the expert witnesses on behalf of the Morrisseys also accepted the merits of

blind review. In addition, it was submitted that the trial judge expressly refused to

consider the evidence of Prof. Roese, thus failing to appreciate the need to mitigate

8.7 Finally, Quest submitted that there was also a failure on the part of the trial

judge to engage with their case in respect of causation. The cumulative effect of the

evidence adduced by Quest in relation to causation was said to point to the probability

that neither squamous abnormalities nor squamous pre-cancer was present in 2009.

8.8 In this regard, it was submitted first that the trial judge erred in his conclusion

that a rescreening or colposcopy in 2009 would have revealed squamous pre-cancer

cells which ultimately developed into the cancer with which Ms. Morrissey had been

diagnosed in 2014. This conclusion was based on what Quest argued was an

inaccurate and unsubstantiated assertion made by Prof. Wells that both glandular and

squamous pre-cancer existed in 2009, whereas the evidence was said to indicate that

only glandular abnormalities may be identified from the 2009 slide.

8.9 In addition to this error, counsel for Quest argued that the trial judge failed to

engage with the aggregate evidence in relation to the 2009 and 2012 slides and the

2014 smear, all of which returned negative results, and the 2014 trachelectomy

specimen. This specimen, acquired during the radical trachelectomy procedure

performed on Ms. Morrissey in 2014, was said by Prof. Austin to have indicated

limited evidence of CIN3 cells, and its result was not referred to in the judgment of

the High Court. Prof. Austin gave evidence on behalf of Quest that the features of

Ms. Morrissey‘s case suggested that Ms. Morrissey‘s cancer had progressed more

rapidly than usual. This was accompanied with evidence to the effect that cancer may

develop more rapidly amongst some younger women. It is submitted that the trial

judge failed to engage with the totality of this evidence when reaching his conclusion

8.10 In response to the submissions of Quest in relation to the evidence of its U.S.

expert witnesses, counsel for the Morrisseys submitted that, on the basis of the

evidence which had been adduced in respect of the influence of the Guidelines, the

trial judge was entitled to form his assessment of the appropriate weight to be given to

the evidence of Prof. Austin and Mr. Feit In particular, counsel referred to the

affirmation to comply with the Guidelines which other U.S. expert witnesses, who

were members of the American Society of Cytopathology, confirmed in evidence that

they had given. Further, counsel relied on the evidence of Dr. McKenna to the effect

that borderline categories of abnormal cells cannot be underestimated and can be

8.11 Further, it was argued on behalf of the Morrisseys that the trial judge was

entitled to assess the appropriate weight to be given to the expert evidence in relation

to blind reviews and hindsight bias. Evidence was adduced on behalf of the

Morrisseys in relation to the inadequacies of blind reviews generally and Dr.

McKenna further stated in evidence that the blind review of the 2009 slide had been

compromised as the slide had markings on it and therefore was not a true reproduction

of what was before the original screener. Prof. Roese, it was submitted, had no

experience in cytology screening and did not engage with the facts of this case in

giving evidence, and counsel argued that the trial judge was entitled to determine that

he did not require the assistance of Prof. Roese in respect of the impact of hindsight

8.12 Finally, in respect of the trial judge‘s analysis as to causation, counsel on

behalf of the Morrisseys submitted that there was no error in his finding that

squamous pre-cancer was likely to have been in existence by 2009 and that this would

have been identified following a rescreening and would, as a matter of probability,

have led to a referral for a colposcopy had the 2009 slide been correctly interpreted.

This finding was based on the evidence of Prof. Shepherd that the lesion was likely to

have developed over an eight to twelve-year period and his evidence to the effect that,

where glandular abnormalities are found it is highly likely that squamous

abnormalities will co-exist, as the same oncogenic stimulus is affecting the stem cells

of the cervix, which differentiate along either squamous or glandular lines.

8.13 Further, counsel for the Morrisseys denied that there was a failure to engage

with the totality of the evidence adduced by Quest. That borderline abnormalities

were identified on the 2009 and 2012 slides, it was submitted, did not lead to the

conclusion that there were no high grade abnormalities or precancerous lesions in

existence at that time. Prof. Wells gave evidence that there was not always a close

correlation between the cytology and the subsequent histopathology so that cytology

screening might show minor abnormalities whereas histopathology may often show

more serious disease. The negative result of the 2014 smear, which was taken

following Ms. Morrissey‘s cancer diagnosis, was said by Prof. Shepherd to be readily

explicable as it is said not to be uncommon that smears taken from a cancerous cervix

are obscured by the presence of inflammatory cells, blood and pus. In addition, it was

submitted that there was no inconsistency between the findings of the 2014

trachelectomy specimen and the abnormalities identified on the 2009 slide. Prof.

Wells gave evidence that any glandular abnormality could have been overgrown by

8.14 Finally, it was argued on behalf of the Morrisseys that the trial judge was

entitled to dismiss Quest‘s contention that Ms. Morrissey‘s cancer was one which

rapidly developed, particularly in light of the evidence of both Prof. Wells and Prof.

Shepherd to the effect that the trachelectomy specimen indicated that Ms. Morrissey‘s

cancer was a moderately differentiated one and did not bear the characteristics of an

aggressive or rapidly growing cancer. Both experts also rejected the proposition that

8.15 Before leaving the question of causation, I should make one point which is

applicable to the case against both Quest and against Medlab. It was accepted by

counsel on all sides that this case proceeded before the High Court on the basis that

the task of the trial judge was to determine causation as a matter of probability. In

other words, was it more probable than not that, had any relevant negligence not

occurred, a benign or improved outcome would have ensued. Counsel agreed that no

case had been made which suggested that, following the jurisprudence identified in

Philip v. Ryan [2004] IESC 105, [2004] 4 IR 241, it might, arguably, be appropriate

to assess causation on the basis of the likelihood or otherwise of a benign or improved

outcome and to measure damages accordingly. There certainly are some types of

cases where the proper approach is not to decide what consequences would have

resulted from the absence of negligence by determining that question on the balance

of probabilities, such that full damages would be awarded if there was a slightly

greater than 50% chance of a benign or improved outcome but no damages at all

would be awarded if there was a slightly less than 50% chance of such a result.

Rather, the approach identified in Philip v. Ryan is that, in the types of cases to which

it applies, it is appropriate to award damages which are broadly proportionate to the

likelihood of a benign or improved outcome, so that the relevant damages would

approximate to full damages if there is a very significant likelihood of a benign or

improved outcome in the absence of negligence but that limited but proportionate

damages would be awarded even if it there was a less than 50% probability of a

benign or improved outcome, but some realistic possibility of such a consequence.

8.16 I would decline, therefore, to express any view as to whether cases of this type

might fit in the jurisprudence identified in Philip v. Ryan. I feel it appropriate to make

that comment lest, by not dealing with the issue, it might be later asserted that I had,

by implication, agreed that the approach adopted in this case was necessarily the

correct one. That is an issue which I would leave to any case in which the issue was

9.1 The first issue raised by Medlab is in respect of the trial judge‘s determination

of the factual question of what was to be seen on the 2012 slide. This, the trial judge

held, in accordance with the guiding principles set out in Penney Palmer, was to be

assessed prior to any consideration of the standard of care required under the Dunne

test. On the evidence given by Dr. Madrigal, his computer-generated analysis of the

2012 slide identified over 35,500 squamous cells. It was Medlab‘s case that on this

evidence there were more than 5,000 cells on the slide, as required under the Bethesda

criteria, and therefore the slide was, as a matter of fact, adequate. Thus, in

circumstances where Medlab‘s cytoscreeners arrived in 2012 at what is submitted to

be a correct conclusion as to the adequacy of the slide, it was argued that there could

not have been a finding of negligence by reference to the particular method used to

9.2 Medlab contended that, in his judgment, the trial judge had accepted as a fact

that there were over 35,500 cells on the 2012 slide. While not strictly falling under

the ―failure to engage‖ jurisprudence, it was submitted that he erred in subsequently

dismissing Dr. Madrigal‘s evidence as to the alternative method of counting cells as

―irrelevant‖ and in failing to answer the question as to how many cells were on the

slide. Medlab further argued that the trial judge erred in concluding that, in the event

of any doubt as to adequacy, the slide must be examined using the ThinPrep method,

because Medlab was held to be obliged under the terms of its contract with Cervical

Check to do so. The terms of this contract, Medlab submitted, cannot be relied on in a

negligence claim. It was also argued that the contract, in any event, obliged Medlab to

test in accordance to the Bethesda System, which method was said not strictly to

require a formal adequacy test to be carried out in any particular manner.

9.3 Medlab further submitted that there was a lack of engagement on the part of

the trial judge with its case in respect of the proper approach to be applied by a

reasonably competent cytoscreener when assessing the adequacy of a slide. In the

first instance, it was submitted that the evidence adduced by Medlab demonstrated

that the Bethesda System does not strictly require a formal adequacy test to be carried

out and that applying the Dunne principles, the reasonably competent screener could

have found the 2012 slide to be adequate without conducting a formal assessment.

Two expert cytoscreeners, Ms. Frasch and Ms. Drew, both gave evidence on behalf of

Medlab that they had conducted a blind review of the 2012 slide and had assessed it

as adequate without considering it necessary to carry out a formal adequacy review.

Ms. Tan, the Morrisseys‘ expert cytoscreener, also gave evidence that the 2012 slide

could have been passed for adequacy without a formal assessment. Medlab argues

that this evidence was not properly addressed by the trial judge when, at paras. 123

and 124 of his judgment, he concluded that no screener of equal status or skill acting

with ordinary care would have failed to subject the slide to an appropriate test as to

adequacy and made the assumption that the two screeners who originally reviewed the

9.4 Further, and in the alternative, Medlab submitted that the trial judge‘s finding

that, as a matter of probability, a formal test would have resulted in a finding of

inadequacy, made at para. 125 of his judgment, cannot be reconciled with the

evidence of Medlab‘s expert cytoscreener, Ms. Stowe, who conducted a formal

adequacy review in accordance with the Bethesda system and found the slide to

9.5 Medlab also argued that the trial judge failed to engage with the evidence on

causation which was tendered by expert witnesses on behalf of both Quest and

Medlab. The trial judge concluded that, had Ms. Morrissey been referred for a repeat

smear following a finding of inadequacy in respect of the 2012 slide, it would have

resulted in colposcopy which would have revealed the existence of squamous pre-

cancer cells. This finding, it was submitted, was reached in the absence of any

evidence as to the existence of glandular pre-cancer cells in the 2012 slide or in the

2014 trachelectomy specimen, which would have indicated the likely coexistence of

squamous pre-cancer. Further, this conclusion was said to have been reached without

addressing the evidence adduced by Quest which indicated that Ms. Morrissey‘s

might have been a particularly rapidly-developing form of cancer.

9.6 In response to Medlab‘s arguments, counsel on behalf of the Morrisseys

submitted that the findings of the trial judge in relation to the adequacy of the 2012

slide were supported by credible evidence. First, it was said that the factual question

as to adequacy to be determined by the trial judge related to whether there was a

sufficient number of well-visualised cells on the slide. In accordance to the method

prescribed by the Bethesda system, the slide must contain 5,000 ―well-visualised‖

cells and be ―satisfactory for evaluation‖. These cells must be well-visualised for the

cytoscreener operating in laboratory conditions, it was contended, rather than for a

computer conducting an automated analysis. Both the terms of Medlab‘s contract

with Cervical Check and the ThinPrep manufacturer‘s specifications require

compliance with the Bethesda System. Further, the Morrisseys rely on the evidence

of Dr. McKenna to the effect that an alternative method of testing adequacy, such as

that employed by Dr. Madrigal, has yet to be clinically shown to be of the highest

standard and proven to be as safe as the prescribed method that is being departed

9.7 In addition, counsel on behalf of the Morrisseys referred to the evidence given

by a number of expert witnesses to the effect that the slide was not adequate for

cellularity, including that of Prof. Pitman, who appeared on behalf of Medlab.

Counsel for the Morrisseys also pointed out that Ms. Tan had given evidence that she

reported the sample as adequate but only because she had identified abnormalities on

9.8 The submissions of the Morrisseys in respect of the trial judge‘s findings in

relation to causation are as stated in full above. It is submitted that the trial judge‘s

finding that a colposcopy carried out in 2012 would have been likely to identify

squamous pre-cancer was based on credible evidence. In particular, counsel for the

Morrisseys referred to the evidence of Prof. Shepherd to the effect that the cancerous

cells were likely to have been developing over an eight to twelve-year period and that,

on the balance of probabilities, they progressed to an invasive cancer in early-mid

2013. Prior to this, Prof. Shepherd stated, a high-grade lesion would have been in

existence. Further, counsel for the Morrisseys relied on the evidence of Prof.

Shepherd to the effect that where glandular abnormalities are found, it is highly likely

that squamous abnormalities will co-exist, as the same oncogenic stimulus is affecting

9.9 Having set out in some detail the precise contentions put forward both by

Quest and by Medlab concerning what was said to be a failure of the trial judge to

engage adequately in his judgment with certain aspects of their cases on the facts, it is

necessary to address whether any of the points made are sufficient to warrant allowing

10.1 The starting point has to be to reemphasise the fact that the threshold for

successfully establishing that there has been a sufficient lack of engagement by a trial

judge with the case made by a party who appeals against the first instance decision is

a high one. As MacMenamin J. pointed out in Leopardstown Club, as referred to

earlier in this judgment, an appellate court may only reverse the decision of a trial

judge ―where there is a clear non-engagement with essential parts of the evidence‖.

As MacMenamin J. also pointed out, the ultimate obligation on a trial judge is to give

a reasoned judgment. If the analysis of the evidence and the basis on which certain

evidence is preferred to a conflicting account fails to meet the requirement that the

judgment is ―truly reasoned‖ then an appellate court must set aside the judgment

concerned. But it is clearly insufficient to that end to simply identify some piece of

relevant evidence that is not mentioned in the judgment or to suggest that the basis for

a trial judge preferring certain evidence over some other piece of evidence is not as

fully set out as it might be. The threshold is much higher than that.

10.2 On the other hand, it is clear that an attack on the judgment of a trial judge on

the basis of a ―failure to engage‖ is different from a suggestion that there was no

evidence on which the trial judge could have come to the conclusion which he/she

reached, under what might be described as a ―pure‖ Hay v. O’Grady point. Much of

the argument put forward on behalf of the Morrisseys centred on identifying evidence

which, it was said, the trial judge was entitled to accept and which, if accepted, led to

the conclusions of fact set out in the judgment. However, I did not ultimately

understand either counsel for Quest or counsel for Medlab to suggest that there was

no evidence from which the trial judge could have reached the conclusions of fact

which are to be found in the judgment. Therefore, pointing out that there was such

10.3 It seems to me that there really are three questions which need to be addressed

under this heading. First, can it be said that the trial judge did fail to address any

aspects of the respective cases made on the facts by Quest and Medlab? Second, and

to the extent that it may be possible to say that there were aspects of those respective

cases which were not fully addressed, were the issues in respect of which any such

failure might be established sufficiently central to the case as a whole to warrant

holding that it was an ―essential part‖ of the case on the evidence in the sense in

which MacMenamin J. used that term in Leopardstown Club? Third, was the extent

of any non-engagement sufficiently serious that it can properly be said that the trial

judge has not really given reasons as to why one side succeeded and the other failed

10.4 In addressing the issues which arise under this heading, it is important to start

by emphasising the particular circumstances in which the trial of these proceedings

was conducted in the High Court, insofar as those circumstances might be said to

have had a reasonable impact on the way in which the judgment in this case was

10.5 First, it must be said that, commendably, all concerned were anxious that

judgment be given as quickly as possible. There were obvious reasons why it was

important that the Morrisseys should know the result of their case at the earliest

possible time. There is also the fact that at least some of the issues which arise in

these proceedings have the potential to also arise in other cases brought before the

High Court claiming negligence arising out of CervicalCheck or where similar cases

10.6 In that context, the unusual step was taken of dispensing with oral closing

submissions after the parties had filed written submissions subsequent to the evidence

being completed. It should be emphasised that this is an unusual procedure. In

straightforward cases, oral submissions will frequently take place immediately after

the evidence completes. In more complex cases, it may be considered advantageous

to give the parties an opportunity to file written submissions after the evidence has

finished. While it is true that, in many complex cases, written submissions may be

filed in advance, such submissions are necessarily based on the evidence the parties

consider is likely to be given. There will almost inevitably be some difference

between the evidence as it has emerged at the trial (including questions over which

evidence should be preferred in cases of conflict) compared with how either party

might have assumed, in advance, that the evidence was likely to turn out. For that

reason, written submissions which reflect the evidence as actually given, as opposed

to the evidence which might have been anticipated, can be useful.

10.7 But in such cases, it almost inevitably follows that there is some oral debate

after closing written submissions have been filed. The purpose of that debate is to

tease out issues raised and allow the court to clarify, often not least in its own mind,

the precise issues which need to be addressed in the judgment. However, all parties

went along, for understandable reasons, with the truncated procedure adopted in this

10.8 While it is entirely commendable that all concerned sought to bring the

proceedings to a close as quickly as possible, it must be acknowledged that truncating

the process does carry with it the risk that not everything will be done quite as

thoroughly as might otherwise have been the case. While proceedings should always

be concluded as quickly as is reasonably possible, it must be emphasised that

truncating proceedings may well have consequences. Where the parties go along with

such a truncated process, then it equally follows that at least some of the

consequences cannot be taken to give rise to any legitimate complaint. In saying that,

it must, of course, be acknowledged that there is an irreducible minimum below

which it is not permissible to go. In the context of the issues which arise under this

heading, a judgment might so lack in reasoning or demonstrable engagement with the

issues that an appellate court could not allow it to stand notwithstanding the process

adopted even with the agreement of the parties. But it does seem to me that it is

appropriate for this Court to take into account the truncated process which was

adopted in these proceedings in assessing whether the reasoning of the trial judge was

10.9 As already noted, the first point made under this heading by Quest concerned

the approach of the trial judge to the evidence given by certain American expert

witnesses. It may be that it would have been preferable if the trial judge had gone into

greater detail concerning the actual evidence given by those experts and had given

more detailed reasons, connected with their evidence and the competing evidence

given by the corresponding experts called on behalf of the Morrisseys, for preferring

the case made on behalf of the Morrisseys in that regard. But the trial judge did give

some reasons for preferring the evidence tendered by the Morrisseys‘ experts. In that

context, it does not seem to me that it can be said that there was a ―clear non-

engagement‖ by the trial judge such that the threshold identified by MacMenamin J.

in Leopardstown Club has been met in respect of this aspect of the case.

10.10 The next aspect of Quest‘s case in respect of non-engagement concerned the trial

judge‘s approach to blind reviews. There is no doubt that Cross J. did offer some

general comments on the topic of blind reviews and did put forward reasons for his

conclusion that a blind review was not necessary. It would undoubtedly have been

preferable, however, if the trial judge had explained in more detail why he did not

consider that the blind review evidence tendered in this case was preferable to the

views expressed by the experts called on behalf of the Morrisseys. If the process

followed in this case had followed the normal course of events, then there might have

been a basis for suggesting that the judgment in that regard was insufficiently

reasoned. However, it does not seem to me that the judgment falls below the

irreducible minimum of reasoning which would require this Court to set aside the

judgment, whether or not the parties have gone along with a truncated process.

10.11 The third area concentrated on by counsel for Quest related to the trial judge‘s

finding in respect of causation. I am of the view that it would have been appropriate

for the trial judge to have set out in more detail the reasons for preferring the evidence

of the Morrisseys‘ experts on the causation issue. However, certain reasons were

given for the trial judge‘s conclusions on causation and they do not seem to me to fall

below the irreducible minimum of reasoning which, for the reasons already set out, I

consider to be the appropriate threshold having regard to the process adopted in the

10.12 Turning then to the issues relied on by Medlab, it should be recalled that the

first such issue concerned the actual number of cells present on the relevant slide.

Using a computer-based method, Medlab‘s expert witness concluded that there were

35,500 cells on the slide, which would be well above the threshold for adequacy of

5,000. It seems to me that the debate on this issue became overly-centred on the first

question identified in Penney Palmer, which suggests that a court should initially

decide what the factual situation actually was. It was certainly open to the trial judge

to take the view that the question which must be addressed by a screener in

determining whether a slide is adequate is as to whether there is a sufficient number

of well visualised cells, in accordance with established practice. Whether or not the

method adopted by Medlab‘s expert could be said to establish adequacy on that basis

may be debatable. It would, again, have been preferable if the trial judge had given

more detailed reasons for not considering that evidence to be relevant. However, yet

again, some reasons are given and they do not seem to me to fall below the irreducible

10.13 The next issue raised on behalf of Medlab complains that there was what was

said to be an inadequate engagement by the trial judge with the evidence given by two

expert screeners called on its behalf, who gave evidence that they had subsequently

looked at the slide in question and had passed it for adequacy. That evidence arose in

the context of the question as to the proper process to be followed in reaching an

adequacy assessment. All of the experts appeared to agree that it was open to a

screener to reach a positive conclusion in respect of adequacy without doing a formal

analysis. This was said to be the proper approach where adequacy was clear. A

formal approach to the assessment of the number of cells was only required to be

adopted where the initial assessment could not be said to have shown the sample to be

clearly adequate. The two relevant witnesses had indicated that they did not consider,

when reviewing the slide in question, that a formal analysis would have been

required. However, as counsel for the Morrisseys pointed out, one of Medlab‘s own

witnesses had considered that the slide was not adequate and another witness had only

considered it to be adequate because of finding abnormalities. While, again, it might

be said that more detailed reasons could have been given for preferring the evidence

of some of the witnesses over others, the trial judge did give some reasons and, in my

10.14 Finally, Medlab also made complaint about the trial judge‘s engagement with

the evidence on causation as set out earlier in this judgment. Here, again, there was

competing evidence as to what would have happened had Medlab assessed the

relevant slide as being inadequate, thus leading to a further sample being required.

There was undoubtedly a significant difference of medical opinion as to what might

have occurred in those circumstances. The trial judge did address those issues in his

judgment but did so in a way which might well have fallen short of the level of

reasoning which might have been required had this case followed a normal process.

However, again, it does not seem to me that the level of reasoning of the trial judge

10.15 It will be seen, therefore, that I would consider that, in respect of some of the

complaints made, it would ordinarily be justified to regard relevant aspects of the

judgment as being insufficiently reasoned, even allowing for the high threshold which

applies in reaching such a conclusion. However, I do not consider that any lack of

reasoning falls below the irreducible minimum which would lead the Court to

overturn the decision of the High Court, irrespective of the manner in which the trial

was conducted. If the judgment of the trial judge had been delivered in circumstances

of ordinary expedition and with the benefit of final oral submissions, it might well

have been impossible to regard the judgment as sufficiently reasoned to withstand

challenge. Of course, the very fact that there might have been oral submissions

would, in those circumstances, cast some light on that very question. The focus of the

debate during those oral submissions might give some guidance to the court as to the

extent to which any particular issues truly remained as significant and material at the

final close of the proceedings. Such an assessment is more difficult to reach in the

absence of closing oral submissions directed towards teasing out the true issues which

remain for debate between the parties and which, therefore, impact on the questions

10.16 That being said, I consider that, in the particular circumstances of this case,

and having regard specifically to the somewhat expedited and truncated procedure

adopted, the reasoning of the trial judge is adequate to withstand challenge.

11.1 As noted earlier, the HSE has appealed the findings of the trial judge in

respect of its liability for the alleged negligence of Quest and Medlab. As mentioned,

in this regard the trial judge held that the HSE was primarily liable in respect of the

way in which the tests were carried out but also held that the HSE was vicariously

liable for the negligence found against the two laboratories. The statutory functions

of the National Cancer Screening Service Board, which was responsible for the

establishment of the CervicalCheck programme in 2008, were set out in the National

Cancer Screening Service Board (Establishment) Order 2006 (S.I. 632/2006) (―the

2006 Order‖). Article 6(b) thereof provides that the Board was required to:-

―[C]arry out or arrange to carry out a Programme for the early diagnosis of

cervical cancer, and arrange for the primary treatment of cervical cancer in

such classes of women as may be determined by the Minister from time to

11.2 On the dissolution of the National Cancer Screening Service Board in 2010

(under s. 28 of the Health (Miscellaneous Provisions) Act 2009), the HSE assumed

the functions of the Board. In respect of the trial judge‘s finding that the HSE had a

primary liability in respect of the acts of the laboratories, the HSE submitted that the

extent of its duty under the statutory provision quoted above is to carry out, or to

arrange to carry out, a competent cervical screening programme and that this duty had

been complied with. The HSE, it was argued, had not sought to assume for itself a

duty to perform the relevant functions of the laboratories, being the screening and

reporting of smear samples, rather it had only arranged for their performance. In

support of its contention that its duty was solely an organisational one, the HSE

submitted that the operation of the CervicalCheck programme was dependent on

contracting out to third party service providers which could provide services with a

turnaround time which the HSE was unable to match. It was further submitted that

while the HSE had sought to impose obligations of quality assurance, this did not in

11.3 The HSE disputed the finding that it held a non-delegable duty to take

reasonable care in the interpretation or reporting of smears by individual cytologists

employed by the contracted laboratories. It was accepted that a hospital could be said

to owe a non-delegable duty to provide its patients with skilful treatment and

therefore be liable in negligence for the acts of all of those who administer treatment

in that hospital, whether they are employed by the hospital or engaged as independent

contractors (see, to this end, the judgments of Lord Greene M.R. in Gold v. Essex

County Council [1942] 2 K.B. 293 and Denning L.J. in Cassidy v. Ministry of Health

[1951] 2 K.B. 343). However, the HSE argued that the service provided by the

CervicalCheck programme is not akin to a hospital service and that the persons who

take part in a screening programme are not comparable to patients in the care of a

hospital. To this effect, the HSE relied on the decision of the Court of Appeal for

England and Wales in Farraj v. King’s Healthcare NHS Trust, [2009] EWCA Civ

348, [2010] 1 WLR 2139, where it was held that the defendant hospital was not

liable for the negligent acts of a third party laboratory which had been contracted to

perform tasks relating to genetic screening. This finding was made on the basis that

the claimants in the case had not been admitted to the hospital for treatment and what

was described as the ―special duty‖ that exists between a patient and a hospital, where

the hospital undertakes the care, supervision and control of persons those who are in

11.4 The HSE also sought to distinguish these proceedings from the facts in Byrne

v. Ryan, as was cited by the trial judge in support of his conclusion, which concerned

the imposition of a primary liability on the defendant hospital for the negligence of a

consultant in the course of his treatment of the plaintiff. This case, the HSE

submitted, was in fact decided by Kelly J. on the basis of the existence of an

employment relationship between the hospital and the consultant, who was a

practitioner in the full time service of the hospital in question.

11.5 Further, the HSE took issue with the acceptance by the trial judge of the

principles of Woodland v. Essex County Council, a decision of the U.K. Supreme

Court which approved of the doctrine of the non-delegable duty, as identified in the

―hospital cases‖, Gold v. Essex County Council and Cassidy v. Ministry of Health,

cited above. The ―essential element‖ giving rise to this duty, as described by Lord

Sumption at para. 24 of his judgment, is the defendant‘s control over the claimant for

the purpose of performing a function for which the defendant has assumed

responsibility. The criteria required to establish the existence of a non-delegable duty

were set out in the preceding paragraph of his judgment as the following:-

―(1) The claimant is a patient or a child, or for some other reason is

especially vulnerable or dependent on the protection of the defendant

against the risk of injury. Other examples are likely to be prisoners and

(2) There is an antecedent relationship between the claimant and the

defendant, independent of the negligent act or omission itself, (i) which

defendant, and (ii) from which it is possible to impute to the defendant

the assumption of a positive duty to protect the claimant from harm, and

not just a duty to refrain from conduct which will foreseeably damage

the claimant. It is characteristic of such relationships that they involve

an element of control over the claimant, which varies in intensity from

(3) The claimant has no control over how the defendant chooses how to

perform those obligations, i.e. whether personally or through employees

(4) The defendant has delegated to a third party some function which is an

integral part of the positive duty which he has assumed towards the

claimant; and the third party is exercising, for the purpose of the

function thus delegated to him, the defendant‘s custody or care of the

(5) The third party has been negligent not in some collateral respect but in

the performance of the very function assumed by the defendant and

11.6 The HSE argued first that none of the above criteria were fulfilled in respect of

the role it performs in the provision of the CervicalCheck Programme, other than

factor (3), as it was conceded that Ms. Morrissey did not have any control over the

screening of her samples or her treatment. Furthermore, the HSE contested the trial

judge‘s approval of the doctrine of non-delegable duty as a significant development in

Irish law which, it submitted, has not been expressly adopted by the courts in this

jurisdiction, and which represented a form of organisational or enterprise based

liability which would be more appropriately adopted by way of legislative

11.7 Finally, the HSE argued that any imposition of a non-delegable duty must be

justified by a strong policy rationale. To this end it was submitted that, in fact, a

finding of primary liability would have negative consequences for the operation of a

number of HSE programmes that involve the outsourcing of health services and that

the associated increased burden of risk on the HSE would effectively lead to a

11.8 Counsel for the Morrisseys, in response, submitted that the non-delegable duty

of care owed by the HSE, both under statute and at common law, arises from the

nature of its relationship with the women participating in its cervical screening

programme. It was argued that a statutory non-delegable duty arises from the

wording of Article 6(b) of the 2006 Order, in which it is said to be implicit that the

HSE has a duty both to take reasonable care in the provision of the programme and to

ensure that those carrying out services on its behalf exercise reasonable care. It was

submitted that this statutory duty cannot be evaded on the basis of the argument that

the HSE‘s obligations were to be discharged by an independent contractor.

11.9 At common law, the Morrisseys submitted, a non-delegable duty of care arises

under the criteria set out in Woodland, as quoted above, and further suggested that

such a duty has been recognised in this jurisdiction in Byrne v. Ryan. Counsel for the

Morrisseys disputed the HSE‘s interpretation of the decision in Byrne as one which

was determined solely on the basis of an employment relationship. Particular reliance

was placed on para. 126 of the judgment of Kelly J., where it was held that the

hospital owed a primary duty to the plaintiff given the fact she was a public patient

who was referred to the hospital and not to an individual consultant. In that regard,

Kelly J. cited the following comments of Denning L.J. in Cassidy v. Ministry of

―I take it to be clear law, as well as good sense, that, where a person is himself

under a duty to use care, he cannot get rid of his responsibility by delegating

the performance of it to someone else, no matter whether the delegation be to

a servant under a contract of service or to an independent contractor under a

…[T]he hospital authorities accepted the plaintiff as a patient for treatment,

and it was their duty to treat him with reasonable care. They selected,

employed and paid all the surgeons and nurses who looked after him. He had

no say in their selection at all. If those surgeons and nurses did not treat him

with proper care and skill, then the hospital authorities must answer for it, for

it means that they themselves did not perform their duty to him.‖

11.10 The Morrisseys argued that the five criteria as set out in Woodland, which are

required to establish the existence of a non-delegable duty, had in fact been fulfilled

in this case. First, it was submitted, Ms. Morrissey was one of a number of persons in

the care of the CervicalCheck programme for the purpose of their protection from

developing cervical cancer and she was in a position of vulnerability as a result,

because a misreading of her samples would give rise to a risk of significant harm.

Counsel suggested that the HSE had recognised this responsibility to protect Ms.

Morrissey against such a risk by providing detailed and extensive quality assurance

11.11 Second, it was submitted that an antecedent relationship existed between Ms.

Morrissey and the HSE, as the CervicalCheck programme was a public service which

involved the provision of medical care. The programme, it was said, involved both

the screening and, if required, treatment of the relevant population and independent

laboratories had been contracted to screen smear samples as well as to direct the

future treatment of those screened. Counsel sought to distinguish these facts from the

situation of the claimants in Farraj, as relied on by the HSE. In that case, the

claimants resided outside the U.K. and sought to have a sample of foetal tissue tested

for analysis. Their consultant in Jordan sent what was thought to be an adequate

sample to the NHS, which did not have the capacity to clean and culture the sample

and so this sample was sent on to an independent laboratory for that purpose. It was

held that the NHS did not have a non-delegable duty towards the claimants, and

counsel for the Morrisseys submitted that this was due to the absence of the central

requirement of an antecedent relationship between the parties to the case.

11.12 The fulfilment of the third criterion, that of an absence of control on the part of

Ms. Morrissey over the performance of the functions of CervicalCheck, was conceded

by counsel for the HSE. Fourth, it was submitted by counsel for the Morrisseys that

the HSE had delegated to the laboratories the task of screening and reporting of

smears, which was said to be an integral part of the HSE‘s contended duty to take

reasonable care in the provision of the programme. That the HSE had retained, in the

terms of its contracts with the laboratories, full supervisory jurisdiction over the

provision of these services, by means of systems of quality assurance, audits and site

inspections, was suggested by counsel to be indicative of an assumption of

responsibility on the part of the HSE over the performance of these functions.

Further, it was contended that the fact that the HSE had delegated one of its functions

to a third party because it did not have sufficient capacity to perform that function is

immaterial. Finally, it was submitted that, if the preceding criteria have been fulfilled,

it followed that any negligence established against Quest and Medlab arose in the

performance of the function assumed by the HSE and delegated to the laboratories,

that is, in the screening and interpretation of Ms. Morrissey‘s smear tests. I will turn

to a discussion of the issues arising in respect of the possible primary liability of the

HSE in due course. However, I will first set out the position of the parties in respect

11.13 As noted earlier, the HSE also contested the trial judge‘s finding that it is

vicariously liable for the activities of Quest and Medlab by reason of the control

which the HSE exercised over the laboratories under the terms of their contractual

arrangements. It was submitted that vicarious liability often arises in an employment

relationship, or one analogous to the same, and that the default position under tort law

is that a party is not generally liable for the acts of an independent contractor such as

11.14 On the facts of this case, the HSE submitted that it possessed no direct control

over the laboratories which would suggest that their relationship was analogous to that

of an employment relationship, as was said to be required to establish vicarious

liability in the judgment of Hardiman J. in O’Keeffe v Hickey [2008] IESC 72,

[2009] 2 IR 302. In particular, the HSE contended that it had no real and exercisable power

of control over the acts of the cytoscreeners. Further, the HSE disputed the contention

that the provision of quality assurance guidelines in its contracts with the laboratories

was indicative of vicarious liability. It was said that these provisions do not ―guide

the hand‖ of those carrying out the expert functions of the laboratory.

11.15 Moreover, it was argued that the concept of control is subsidiary to the

ultimate question as to whether a relationship akin to an employment relationship

exists. On that issue, the HSE submitted that, by the terms of their contracts, Quest

and Medlab were in business on their own account and provided a service that the

HSE could not provide in terms of turnaround speed. This was said to be indicative

of the engagement of an independent contractor for its skills and expertise rather than

of the existence of a relationship analogous to employment. Furthermore, the HSE

contended that the laboratories were neither ―part of the organisation‖ nor

―integrated‖ into the business of the HSE in a manner which could give rise to the

imposition of vicarious liability. The HSE pointed to a number of features of the

business relationship between itself and the laboratories which, it was said, indicated

that the true nature of the relationship involved the engagement of an independent

contractor. Amongst those relevant factors was said to be the fact that the laboratories

provided their own equipment and had responsibility of the management of their

affairs, together with the fact that the laboratories had assumed the financial risk of

the enterprise and that the laboratories had the opportunity for profit through the

11.16 Finally, it was submitted on behalf of the HSE that the policy considerations

which usually underpin a finding of vicarious liability, such as risk creation and

enterprise liability, as referred to by Hardiman J. in his judgment in O’Keefe, are

absent in the present case. The HSE, in its arrangement of the CervicalCheck

programme, is not operating for profit and is not ―creating‖ any risk, in the

11.17 On behalf of the Morrisseys, it was submitted that, by virtue of what was said

to be the extraordinary degree of its control over every aspect of the laboratories‘

screening of the slides, the HSE is vicariously liable for the negligence of Quest and

Medlab. It was contended that the essence of vicarious liability is control, as held by

Hardiman J. at para. 47 of his judgment in O’Keefe, and that the question of vicarious

liability is no longer resolved simply by the question of whether the tortfeasor is an

11.18 In dictating the methodology by which the slides would be reviewed by the

labs, it was submitted, the HSE exercised a degree of control that would not ordinarily

be exercised over a truly independent contractor. Counsel for the Morrisseys

contested the HSE‘s reliance on the decision of the Supreme Court in O’Keefe, where

it was held that the State was not vicariously liable for the sexual assault of the

claimant perpetrated by the claimant‘s national school teacher. That case, it was

submitted, was based on an entirely different factual matrix, being that the sexual

assault committed by the relevant teacher evidently fell outside the scope of his

employment. In addition, the rules which the Minister for Education had laid down

for national schools were general in nature and did not go to the governance of the

detailed activities of any individual teacher. On the facts of this case, it was

submitted, the acts of the cytoscreeners clearly fell within the scope of their

employment and the HSE exercised a high level of control over the screening

activities at the laboratory. Furthermore, counsel indicated that the application of the

―close connection‖ test, as adopted by Fennelly J. in O’Keefe and with whom the

majority of the Supreme Court agreed, which considers the closeness of the

connection between the work which the individual was engaged to carry out and the

tortious act, would also lead to a finding of vicarious liability in this case.

11.19 Finally, on the level of policy, it was argued on behalf of the Morrisseys that

the HSE ought not to be able to divest itself of any responsibility for failings in the

delivery of public health services through the outsourcing of its services

11.20 In the light of those submissions, I propose to turn firstly to the question of

whether the trial judge was correct to hold that the HSE was vicariously liable for the

12.1 Both sides sought to place reliance on the judgment of Hardiman J. in O’Keefe

v. Hickey. There was, however, something of a dispute between the parties as to

whether the key component in the relevant test centred around the legal relationship

between the parties or on the degree of control which one party exercised over the

12.2 It may be said that where difficult or novel issues arise concerning the

application of well-established concepts in common law to particular circumstances, it

may be useful to identify what the fundamental principle or purpose behind the law

concerned may be, for that may give valuable guidance as to how the practical rules

regarding the application of the relevant aspect of the common law might be applied

in such circumstances. However, there are areas where it may be difficult to discern

any overarching but consistent fundamental principle. Rather, in some areas, the case

law has grown up as part of an attempt to deal with a range of different circumstances.

In some such cases it may well be that the proper approach to the evolution of the

common law in such an area is to avoid over-radical developments, but to extend the

parameters of the established case law to novel or evolving areas in a manner which is

at least analogous to the way in which the existing case law has developed.

12.3 The former might be described as a ―back to first principles‖ model. The

latter might be described as an ―evolution by analogy‖ approach. In my view, the

former approach is preferable should it prove possible to discern, with any real degree

of clarity, what the underlying principles are. However, the latter approach may be

more appropriate where it is not really possible to identify a coherent underlying basis

12.4 It will be necessary to return in due course to the most recent judgment of this

Court on the issue of vicarious liability, being Hickey v. McGowan and anor

[2017] IESC 6, [2017] 2 I.R. 196. However, in this context, it is appropriate to note that

O'Donnell J. reached his conclusions on the proper principles to be applied in that

case by adopting what he considered ―to be the cautious and incremental approach

outlined by Fennelly J. in O’Keefe‖. O'Donnell J. indicated that he proposed adopting

that approach, having considered what seemed to be the different approaches adopted

12.5 In this same context, it may be appropriate to refer to Robinson v. Chief

Constable of West Yorkshire Police [2018] UKSC 4, [2018] 2 WLR 595. While the

Supreme Court of the United Kingdom in that case was concerned with determining

whether a duty of care existed in the particular circumstances of that case, it did make

comments which seem to me to be appropriate regarding the broader issue of how one

should approach the evolution of the common law. In Robinson, it was indicated that

―[a]n approach based, in the manner characteristic of the common law on

precedent, and on the development of the law incrementally and by analogy

I would only add the qualification that, where the existing case law can be seen to

derive from some underlying principle or principles, it may be more appropriate to

start by seeking to identify how those principles should direct the incremental

evolution of the common law. However, that being said, it seems to me that Robinson

provides cogent reasoning as to why the ―evolution by analogy‖ approach may well

be appropriate in many circumstances not confined to the narrow question of the

determination of the existence or otherwise of the duty of care which was at issue in

12.6 There is, in my view, a strong argument for the view that the law of vicarious

liability may be more properly dealt with under the ―evolution by analogy‖ approach.

That it may be appropriate to impose liability on one person for the negligence of

another is, of course, well-established as a legal principle. It would be surprising

indeed if, for example, a supermarket could avoid liability for an injury suffered as a

result of negligence by simply asserting that the fault was all that of a specified

employee. However, there is a strong basis for suggesting that the categories of case

in which vicarious liability has been established do not easily give rise to being

explicable by reference to some underlying principle or policy. In Imperial Chemical

Industries Ltd v. Shatwell [1965] AC 656 at p. 685, Lord Pearce said that ―the

doctrine of vicarious liability has not grown from any clear, logical or legal

principle‖. Likewise, Tipping J., in the New Zealand Court of Appeal case of S. v.

Attorney General [2003] NZCA 149, [2003] 3 N.Z.L.R. 450, commented that the

literature was ―replete with comments concerning the lack of any coherent or agreed

jurisprudential underpinning‖. I respectfully agree with those comments. It seems to

me to follow that any evolution of the law of vicarious liability or the application of

that law to new or evolving circumstances requires an approach much closer to the

―evolution by analogy‖ method to which I have already referred. It is in those

circumstances that it is appropriate to look at how the law on vicarious liability has

developed both in this jurisdiction and in other relevant common law jurisdictions in

12.7 It is probably correct to say that, in its original form, the concept of vicarious

liability stemmed principally from the legal relationship between the actual

wrongdoer and the party who was held to be legally responsible for the actions of the

wrongdoer even though not personally at fault. Thus, the relationships of employer

and employee, principal and agent or partner and the other members of a partnership

have been all held to potentially give rise to the possibility of liability being placed on

a person who was not themselves directly responsible for any relevant wrongdoing.

12.8 However, this Court, in Moynihan v. Moynihan [1975] I.R. 192, did extend the

concept of vicarious liability to a situation where there was sufficient control by one

family member over the actions of another family member, who committed the

negligent act in relation to the pouring of tea, which led to an injury. It is clear,

therefore, that, for at least the last 45 years, the Irish courts have recognised that

informal relationships involving some degree of delegated function and control can

12.9 Likewise, the courts in Canada (see John Doe v. Bennett [2004] 1 S.C.R. 436)

and England (see for example Various Claimants v. The Catholic Child Welfare

Society & Ors [2012] UKSC 56, [2013] 2 AC 1) have also accepted that vicarious

liability can attach to relationships that are considered to be ―akin to employment‖. It

would seem that many of the Canadian and English cases stem from a desire to

ensure, as Lord Phillips put it in Catholic Child Welfare Society, that, insofar as it was

fair, just and reasonable, liability for tortious wrong was borne by a defendant with

the means to compensate the victim. I have to say that I am not convinced that,

laudable as that aim may be, such analysis adds very much to the difficult task of

determining the boundaries of vicarious liability. In that context, I would very much

agree with the observation of O'Donnell J., which is set out at para. 43 of his

judgment in Hickey v. McGowan, regarding the function of vicarious liability.

12.10 Some persons are injured through the fault of others who are neither insured

nor have the means to compensate. Obviously the advisors of such persons will seek

to extend liability, if it be possible, to anyone who might be a so-called ―mark‖ for

damages. But the real question concerns how far it is possible to extend the concept

of vicarious liability. Stating that it will be done when it is ―fair, just and reasonable‖

does not really contribute very much to the analysis, for it simply begs the question as

to the kind of circumstances that can be regarded as coming within that criteria. The

real issue is to identify the type of situation which may legitimately give rise to

12.11 There were undoubtedly good reasons for departing from an overly-technical

view which confined liability to cases coming within specific legal relationships.

Persons may well not fit, strictly speaking, within the definition of, for example,

employee or agent but nonetheless may be doing something very similar to that which

would be done by an employee on behalf of an employer or by an agent on behalf of a

principal. There are sound reasons why vicarious liability should not depend on

whether, technically, a person might be regarded in law as an employee or as an

12.12 For example, it is interesting to note that a number of the Canadian and

English cases arose in the context of sexual abuse within religious organisations.

Clearly, the concepts of employee or agent do not always fit neatly within the

structures typically encountered within religious orders. Furthermore, such

organisations are not normally run as businesses and do not give rise to the kind of

relationships with a commercial enterprise that would exist in, for example, a private

company. But the substance of the relationship is similar. A member of a religious

order who teaches in a school operated by that order may or may not, strictly

speaking, have been considered to have been an employee, but their relationship with

the school and those who manage or own it would be very similar to the relationship

between a teacher employed in a privately owned school and the owners of that

12.13 Similarly confining the analysis to the question of control might also lead to

unsatisfactory results. Obviously, employers not only identify the work which

employees are to do but also exercise a significant degree of control over how it is to

be done. However, the extent to which actual control, as opposed to a theoretical

entitlement to control, may be exercised in practice may vary greatly from

employment to employment. This will particularly be the case where an employee

possesses a particular skill or expertise which may not be shared by those who

manage the affairs of the employer. The very point of employing the expert

professional may be to ensure that certain types of work will be carried out in a way

which that professional deems appropriate in circumstances where the employer

would not be able to have the work carried out with the requisite expertise in any

other way. The reality of the employer exercising any real degree of control over how

the work is to be carried out in such a case may be highly theoretical. Thus, just like

an overly-technical reliance on the legal basis for the relationship between the parties

may present an unduly narrow focus, so also a complete emphasis on control may not

12.14 In those circumstances, it is important to return to Hickey v. McGowan. In that

case, the issue was as to the potential liability of a religious order for acts of sexual

abuse committed by one of its members. In assessing the question of whether the

order concerned might be vicariously liable in such circumstances, O'Donnell J. said

“At a crude level the question of whether a relationship between [the first

defendant] and [the second defendant] in any particular case is sufficient to

give rise to vicarious liability can be addressed by asking how closely the

relationship approximates to the classic case of employer/employee. Some of

the cases have taken this approach. There is however in my view something

slightly absurd in seeking to draw comparisons between the case of religious

orders and businesses. Furthermore, the tests and language applicable when

considering the case of employment and analogous relationships, such as

‗enterprise‘ and ‗risk‘ are not easily applicable in the case of religious orders.

Indeed, to apply tests drawn from the relatively modern world of commerce

and industry to religious organisations which have existed for centuries is in

my view, to miss the sheer scale and impact of religious institutions on

peoples' daily lives, particularly in the Ireland of the first three-quarters of the

20th century. The relationship between members of an order and his or her

fellow members and indeed the order itself was much more intense, constant

and all pervasive than the relationship between an employer and an employee,

or in the old language of the late Victorian cases, a master and his servant.

Everything in the organisation of religious orders is directed towards

emphasising the collective. The vow of obedience involves subjugation of

individual will to that of the superior. The vow of poverty has the effect of

making the member dependent upon the order's collective resources. The vow

of celibacy emphasises the focus of the member on relationships with the

order and with God. The objective of teaching young people is not merely

incidental to the work of an order, it is indeed the manner in which the order

seeks to achieve its object. For a member of the order, teaching was not

merely a job it was a religious vocation. There can no doubt that Brother

Cosgrove was in the classroom in Sligo between 1969 and 1972 because he

was a member of the Marist Order. That was known, understood and accepted

by pupils and parents, and when such individuals looked at the various

brothers who staffed the school at any given time, they saw, and were intended

12.15 O'Donnell J. went on to conclude that, at least at the level of principle, a

religious order could be vicariously liable in such circumstances. The reasoning in

―Looked at in this way, it should be apparent that this is in fact not the most

difficult issue in the case. There is evidence that the second defendant was a

member of the Marist Order. It is accepted that the Order was a teaching order,

and supplied teachers to the school. That was how the second defendant came

to be in the school. Teaching was not simply an occupation, but an important

and central part of the mission of the Order. The Order was established, and its

members bound, not merely by rules, but solemn vows taken by them and

considered sacred and binding. Those vows included chastity, obedience,

poverty and celibacy. I think we are entitled to take cognisance of the fact that

members of religious orders at that time normally wore habits of standard

design, identifying them as members of orders, and indeed correspondingly

reducing their individuality, while emphasising their part in a collective.‖

12.16 On that basis, it seems to me that the overall test must be one which considers

whether the nature of the relationship, including the question of control, is such that it

is similar in substance, if not in form, to the types of relationships which have

traditionally been regarded as giving rise to vicarious liability.

12.17 The Morrisseys argued for greater attention to be paid to the degree of control

exercised by the HSE which, it was said, was significant. The HSE emphasised the

fact that the laboratories were independent contractors and that the screeners were

employees of the laboratories and not of the HSE. There is, in any event, often quite a

close correlation between the legal relationship between two parties which arises

where one carries out work on behalf of another and the degree of control exercised.

Indeed, one of the criteria which courts consider in determining whether a relevant

contractual relationship involves one of an independent contractor, sometimes

described as a contract for services, or of employment, often described in this context

as a contract of service, is the degree of control exercised. It was often said that a key

distinction in that context lies between a case where, on the one hand, the employer

determined what the other party was to do and, on the other hand, determined not only

what the other party was to do but also how they were to do it. There can obviously

be grey areas between circumstances where one party simply contracts that a task is to

be carried in return for payment and leaves it entirely up to the other party to choose

the method of delivery compared with a case where the contracted party is under

some form of direct supervision. Skilled employees may very well be left to a large

extent to their own devices. Many contracts which are very much at arm‘s length will

12.18 In my view, the ultimate question which the Court must address is as to

whether the level of engagement by one party with the way in which the other party is

to carry out a task entrusted to it is sufficient to conclude that there is a real extent to

which it can be said that the contracted party is closely integrated into the activities of

the employer, not just in respect of the ends to be achieved but as to the manner in

which those ends are to be pursued. It is, in my view, difficult to give any true red

line, for the type of circumstances which will arise will be many and varied. The

legal relationship will certainly be relevant. So, too, will the degree of control. But

neither are necessarily decisive particularly where the substance of the practical

12.19 In that context, it is clear that there were undoubtedly significant provisions in

the relevant contracts between the HSE and the laboratories which specified certain

standards to be applied. But it does not seem to me that those requirements can

properly be characterised as involving the HSE in the task of directing how the

laboratories were to go about their work. In any complex situation it would be

extremely surprising if parties entering into a contract of the type with which this case

is concerned did not impose measures by reference to which the quality of the work to

be done could be assessed. Such contracts, for example, frequently specify that work

is to be carried out in accordance with a particular recognised standard, even if that

standard is not legally binding. It can hardly be said that a clause of that type implies

that one contracting party is taking control over how the other is to do their work.

The manner in which the second party is to ensure compliance with the quality control

12.20 Taking a broad view of the relations between the HSE and the laboratories, it

seems to me that both the relationship between those parties (which is clearly that of

independent contractors) but also the level of control exercised by the HSE is such

that it cannot give rise to vicarious liability. I would, therefore, hold that the HSE was

not vicariously liable for any negligence established against the laboratories.

However, that leaves the question of whether it can be said that the HSE was

primarily liable for any such negligence. I, therefore, turn to that question.

13.1 There are a number of ways in which a party who has an established

relationship with someone such as a customer, client or a patient, but who ultimately

arranges for relevant work to be done by a third party, may become primarily liable in

negligence if something goes wrong. Perhaps the most common such circumstance

arises where it can be said that the party with the primary relationship with the

customer knew or ought to have known that there was a real risk that the work which

was passed on to a third party would not be carried out in an appropriate fashion. For

example, a failure to contract with persons of proven ability in the area might well

provide the basis for a claim in that regard. However, no such claim is advanced in

13.2 However, there is a second, and it might be said potentially developing, area

of primary liability which comes under the heading of a so-called ―non-delegable

duty‖. Under this heading, a party who has the primary responsibility with the

customer, client or patient may be held, in certain circumstances, to have accepted a

duty to ensure that any relevant arrangements will be carried out in a non-negligent

way and may be held to have done so in circumstances where that duty remains in

place irrespective of whether the contracting party chooses to carry out its side of the

13.3 It seems to me that three questions arise under this heading in the particular

circumstances of this case. The first is as to the extent to which the concept of ―non-

delegable duty‖ arises in Irish law. The second is as to the parameters of the

applicability of such a duty, should it be held to exist in principle. The third is as to

whether it can be said that the arrangements of the HSE in respect of CervicalCheck

can be said to come within the scope of any such principle so as, in turn, to lead to the

HSE being primarily liable for any negligence established in respect of laboratories

whom it contracts to carry out work in the context of that scheme.

13.4 The principle which gives rise to the existence of the non-delegable duty has

been accepted by the courts of the United Kingdom, most notably in the decision of

the U.K. Supreme Court in Woodland v. Essex County Council, as referred to above.

In that case, the duty of a local education authority towards pupils in its care was held

to be non-delegable in circumstances where a pupil suffered serious injury in the

course of a school swimming lesson conducted by swimming instructors who were

not employed by the education authority and whose services were provided to the

authority by an independent contractor. Whilst that decision represents the most

comprehensive analysis of the doctrine of non-delegable duties provided by the U.K.

Supreme Court to date, a number of U.K. authorities had previously indicated that, in

certain circumstances, the duty of a party can extend beyond taking reasonable care in

the performance of their tasks, to ensuring that reasonable care is taken by anyone to

13.5 As indicated by Lord Sumption in his judgment in Woodland, the

characterisation of non-delegable duties has its origins in the law of nuisance and in

the rule in Rylands v. Fletcher (1868) LR 3 HL 330, where strict liability may be

imposed irrespective of fault on the defendant‘s part and where a defendant can be

held responsible even when the acts complained of are committed by independent

contractors. The category of non-delegable duties was subsequently extended to

include situations involving a duty to take reasonable care, as in Hughes v. Percival

(1883) 8 App Cas 443, where Lord Blackburn held in the House of Lords that the

defendant had a duty to ensure that reasonable skill and care was used when carrying

out work which involved a risk to the party wall which divided his house with that of

the plaintiff‘s. It was held that this duty could not be discharged by delegating the

performance of the work to a third party and thus the defendant was found liable for

the negligence of the builder whom he employed to do the work which had caused

13.6 Subsequent to this, the principle of the non-delegable duty was further

developed in the case law with reference to the character of the relationship between

the plaintiff and the defendant and consideration of whether the defendant has

assumed responsibility for the exercise of due care by a third party to whom the

defendant may delegate the performance of his or her functions. This analysis first

arose in the employment context, where it was held that an employer had a personal

duty to provide a safe system of work and that, where he or she has appointed an

agent to perform this duty, the employer remains responsible for the agent‘s negligent

acts (see Wilsons & Clyde Coal Ltd. v. English [1938] AC 57). This principle was

affirmed by the House of Lords in McDermid v. Nash Dredging and Reclamation Co

Ltd. [1987] AC 906, where the defendant company, who had employed the plaintiff

as a deckhand, was held to be liable for the negligence of an independent contractor,

the captain of the ship on which the plaintiff worked, who had failed to operate a safe

13.7 In the ―hospital cases‖ previously mentioned, a series of decisions of the Court

of Appeal for England and Wales, it was suggested that a hospital holds a non-

delegable duty to take reasonable care in the provision of care to its patients. In Gold

v. Essex County Council, a local authority which operated a voluntary hospital was

held liable for the negligence of a radiographer which it employed. While the

majority of the Court of Appeal decided the case according to the principles of

vicarious liability, in his minority judgment Lord Greene M.R. held that the hospital

had assumed the obligation to treat the plaintiff patient, stating that this was the

―natural and reasonable inference‖ to be drawn from the authority‘s method of

conducting their affairs and the nature of the radiographer‘s engagement, and that the

authority was therefore liable if the persons employed by them to perform the

obligation on their behalf act without reasonable care. At pp. 301-302, Lord Greene

―… [T]he extent of the obligation which one person assumes towards another

is to be inferred from the circumstances of the case. This is true whether the

relationship be contractual (as in the case of a nursing home conducted for

profit) or non-contractual (as in the case of a hospital which gives free

treatment). In the former case there is, of course, a remedy in contract, in the

latter the only remedy is in tort; but in each case the first task is to discover the

extent of the obligation assumed by the person whom it is sought to make

liable. Once this is discovered, it follows of necessity that the person accused

of a breach of the obligation cannot escape liability because he has employed

another person, whether a servant or agent, to discharge it on his behalf; and

this is equally true whether or not the obligation involves the use of skill. It is

also true that, if the obligation is undertaken by a corporation, or a body of

trustees or governors, they cannot escape liability for its breach, any more than

an individual can; and it is no answer to say that the obligation is one which on

the face of it they could never perform themselves. Nor can it make any

difference that the obligation is assumed gratuitously by a person, body or

corporation which does not act for profit: Mersey Docks Trustees v Gibbs….‖

13.8 In another case involving the negligent acts of employed medical staff,

Cassidy v. Ministry of Health, again the majority of the Court of Appeal reached its

decision on the application of principles of vicarious liability. As quoted above at

para. 11.9, Denning L.J. in his minority judgment considered that the critical factor

was not the hospital's relationship with the doctor or surgeon, but its relationship with

the patient, arising from its acceptance of the patient for treatment. This, he

concluded, meant that the hospital authorities are liable for the negligence of a

hospital worker in the course of the treatment of a patient, regardless of whether the

worker is employed under a contract of service or a contract for services. Denning

L.J. restated this analysis in his minority judgment in Roe v. Minister of Health

[1954] 2 QB 66, at p. 82. While the principle of a hospital‘s non-delegable duty towards its

patients was referred to with approval in an obiter comment of Lord Browne-

Wilkinson in his X (Minors) v. Bedfordshire County Council [1995] 2 AC 633, at p.

740, the underlying principle identified by Lord Greene M.R. and Denning L.J. was

13.9 In Woodland, reflecting on the foregoing development of the law in this area,

Lord Sumption considered that there are two broad categories of case in which a non-

delegable duty has been held to arise. The first such category was described as a

―large, varied and anomalous class of cases‖ in which a defendant employs an

independent contractor to perform some function which is either inherently hazardous

or liable to become so in the course of his work. The second category of cases were

held to involve circumstances where a protective relationship exists between the

defendant and the plaintiff. Considering the latter category as that which was relevant

to the facts of Woodland, Lord Sumption referred to three critical characteristics in

identifying the existence of such a duty, at para. 7 of his judgment:-

―First, it arises not from the negligent character of the act itself but because of

an antecedent relationship between the defendant and the claimant. Second,

the duty is a positive or affirmative duty to protect a particular class of persons

against a particular class of risks, and not simply a duty to refrain from acting

in a way that foreseeably causes injury. Third, the duty is by virtue of that

relationship personal to the defendant. The work required to perform such a

duty may well be delegable, and usually is. But the duty itself remains the

defendant's. Its delegation makes no difference to his legal responsibility for

13.10 In an effort to articulate the exceptional circumstances in which a non-

delegable duty of care may arise, Lord Sumption clarified that the question cannot

depend simply on the degree of risk involved in the relevant activity. As previously

referred to, five ―defining features‖ of a non-delegable duty were set out at para. 24 of