PROVING THINGS 177: WHEN YOU ARE RELYING ON A MEDICAL EXPERT ON CAUSATION WHO “SHOOTS FROM THE HIP” (IT DOESN’T END WELL)

I have written many times about issues arising from expert witnesses.  In clinical negligence cases the role of the expert witness is often paramount.  The claimant, in particular, is almost wholly reliant on expert evidence in relation to causation. It is worthwhile  for litigators and experts alike to read the judgment in Jones v Ministry of Defence [2020] EWHC 1603 (QB).

 

“It reflects in my opinion a ‘shoot from the hip’ approach to the evidence that I have not found helpful when seeking to understand the complexities of the Claimant’s case.

THE CASE

The defendant had admitted negligence in failing to diagnose that the claimant had HIV.   The issue was whether the late diagnosis had caused the claimant major problems.  The judge was considering the expert evidence adduced by the claimant, Dr Croft.

THE JUDGMENT ON THE CLAIMANT’S EXPERT

Richard Hermer QC, sitting as a High Court Judge, was not impressed by the evidence he heard on behalf of the claimant.

    1. Dr Croft was not however an impressive witness. I conclude that his analysis on all the core disputed questions is flawed, unreliable and cannot be preferred to that of Professor Ross, for the following reasons:
    2. Firstly, Dr Croft displayed a marked lack of familiarity with basic aspects of HIV and its treatment. This judgment highlights a number of occasions on which this became apparent, but the starting point was his very use of the WHO system of classification for the staging of HIV, rather than the criteria set by the Centre for Disease Control (CDC).
    3. The classification, or staging, of the CD4 level is important not least because it forms a key component of Dr Croft’s opinion that the late diagnosis has had a very material impact on the Claimant. He opined that the Claimant was permitted, by the negligent failure to diagnose HIV, to slide from ‘Stage 3’ in November 2012 to ‘Stage 4’ by September 2013, and this had profoundly detrimental consequences for his health and life expectancy.
    4. It is however difficult to understand why Dr Croft was using the WHO classifications. As Professor Ross explained, and I accept, it is not one used at all by HIV practitioners in the UK, or the United States nor indeed any high-income nations. The UK, in common with many other high-income countries, applies the categorisation used by the US Centre for Disease Control (CDC), which relies on CD4 readings to rate the stage of HIV from A to C (C being the most severe).
    5. The WHO system was developed to permit the classification of HIV in countries that do not have widespread access to testing of CD4 levels – it is one based not on objective results of blood tests but rather the categorisation of a range of signs and symptoms to assist doctors working in less developed healthcare systems without access to testing facilities. It makes sense that absent access to blood sampling, treating doctors and public health officials would want to rely upon a system that permitted classification by reference to relevant signs and symptoms on physical examination. As Professor Ross explained, there would be no reason at all why anyone would use the WHO system to assess the Claimant when we have his CD4 tests. All clinicians assessing the Claimant would use a classification system based on CD4 levels, namely that provided by the Centre for Disease Control.
    6. This is recognised in the WHO standards themselves. The relevant document is “WHO Case Definitions of HIV for Surveillance and Revised Clinical Staging and Immunological Classification of HIV-Related Disease in Adults and Children” (2007). It notes that AIDS case reporting in middle and low income countries has been incomplete and of variable accuracy and this problem has been exacerbated by weak health information systems and the lack of diagnostic capacity. It also notes that countries should apply their own national case definitions and develop their own testing algorithms for diagnostic and surveillance purposes. It states that “WHO provides a simplified HIV case definition designed for reporting and surveillance.”
    7. There is nothing in Dr Croft’s reports that explains the reason why he departed from the standard system of HIV classification, indeed an uninformed reader would be forgiven for believing that terms such as ‘Stage 3’ or ‘Stage 4’ were the accepted terminology. The issue of classification was addressed at the expert meeting held with Professor Ross in December 2019 and Dr Croft stated that “there are two commonly-used staging systems” – in so far as this was seeking to describe the classification in the healthcare system in the country in which the Claimant lives, this is plainly incorrect.
    8. Had it been the case that Dr Croft considered that the application of the WHO categories would give some particular insight into the Claimant’s plight that could not be derived from the standard CDC classification then one would have expected some explanation of this in his reports, both identifying that he was adopting a different classification to the norm and why it provided greater or additional insights. No such qualifications or explanations were provided. I consider that Dr Croft’s almost exclusive reliance on the WHO classifications and his belief that they enjoy parity in the UK with those of the CDC reflects a lack of familiarity with the subject matter that does little to instil faith in the quality of the rest of his analysis.
    9. Secondly, and more importantly, Dr Croft has in any event misapplied the WHO classification system in a manner that fundamentally undermines core aspects of his wider analysis.
    10. This is because in applying the WHO categories, he failed to have regard to the Claimant’s CD4 counts, which as Professor Ross explained need (where available) to be factored into the WHO criteria themselves. The WHO recognises that late stage HIV (AIDS) should (where available) be classified by reference to CD4 levels and not exclusively symptoms/signs. The guidance states that:
“The immune status of a child or adult living with HIV can be assessed by measuring the absolute number or percentage of CD4+ cells. And this is regarded as the standard way to assess and characterize the severity of HIV related immunodeficiency”.
A Table in the WHO document defines the criteria for diagnosis of ‘severe HIV’ as being a CD4 count of less than 200 in an adult. Severe HIV is classified as ‘Stage 4’ under the WHO criteria.
    1. Although no CD4 measurement was taken from the Claimant in November 2012 both experts (now) accept that it can readily be estimated. Dr Croft belatedly accepted that it would have been around 40, even lower than the 54 estimated by Professor Ross. This means that under the WHO criteria, the Claimant would have been properly classified as having already been well within Stage 4 rather than Stage 3 by November 2012 and probably for some time before that.
    2. Dr Croft only accepted that the Claimant’s likely CD4 count was around 40 in November 2012, when he had his first joint meeting with Professor Ross. Having done so it might have been thought that Dr Croft would have wished to review his reliance on the WHO staging before coming to give evidence. He did not. In fact, in his answers recorded in the memoranda of the meeting with Professor Ross, Dr Croft went further. He not only maintained that the Claimant was only in ‘Stage 3’ when his diagnosis was missed but that, applying the CDC criteria, he would have been ‘Category B’ on the CDC scale in November 2012 not ‘Category C’.
    3. Professor Ross made plain, by reference to the CDC published criteria, that any individual with a CD4 count below 200 would be classified as ‘Category C’. When it was put to Dr Croft in cross examination that his classification was unsustainable in light of the CDC’s published criteria he answered that he would only ‘potentially‘ agree. I found the reason he gave for qualifying his answer troubling. He stated that the qualification as to his agreement was because he had not in fact considered the CDC published categories notwithstanding that Professor Ross had referenced them in his report and he himself had used them to classify the Claimant in the joint meeting. It thus appears that although Dr Croft stated that in November 2012 that the Claimant was at CDC Category B, he did so without actually checking the CDC criteria, and was still unfamiliar with them when he came to give his evidence. This show not only a lack of rigour in the formulation of Dr Croft’s opinion but (again) a lack of familiarity with the subject matter.
    4. In cross examination, Dr Croft bowed to the inevitable. He accepted that, contrary to that which he had expressed in writing, the Claimant was probably in both CDC Category C and WHO ‘Stage 4’ long before November 2012. Although he did not accept the Defendant’s contention that the threshold (200) was probably crossed in 2010 he did accept it would have been reached by March 2011, about 20 months before the failure to diagnose in November 2012. Had Dr Croft properly considered the impact of his own assessment of a CD4 count of 40 in November 2012 then these are views he should have appreciated long before stepping into the witness box.
    5. In circumstances in which so much of Dr Croft’s analysis was pinned to his view that real damage was done because the delay in diagnosis deprived the Claimant of an opportunity to be treated whilst still only Stage 3 or Category B, his ultimate concession, and his failure to properly analyse this point earlier does little to instil faith in the solidity of the remainder of his analysis.
    6. Thirdly, having belatedly accepted that had the Claimant been diagnosed earlier in November 2012, his CD4 count would still have been very low, I found Dr Croft’s continuing justification for why late diagnosis nevertheless made a significant difference to anything other than short-term suffering, unconvincing.
    7. Dr Croft, having accepted in cross examination that his earlier classification of the stage of the HIV in November 2012 was probably misplaced, nevertheless went on to state:
“… but we’re still left with the situation that there was a delay in a diagnosis and the CD4 count, okay was perhaps lower than I’d originally made allowance for and it became still lower and then still lower and then still lower again. That’s really the whole point and during that 11 months of continuing worsening, irrespective of what is happening with the symptoms, with every further missed opportunity the ultimate outcome was going to be worse and the life expectancy was being shortened. So I think we’re perhaps getting a little caught up on whether it’s the symptoms that are the important factor.
The important factor is just the delay and the immeasurable and dramatic worsening that was occurring in the CD4 count, which was, if you like, the pivot of Mr Jones’ ability to function with an immune system that would protect him in this life.”
    1. Professor Ross, building on his expertise as a researcher and clinician in the field applied a markedly different analysis. He opined that in a patient responding well to ART, with increasing CD4 levels and a non-detectable viral load, one would not expect to see a patient suffering directly from fatigue after approximately 12 months and that such a patient will return to the same track as they would have done had treatment started earlier. This applies irrespective of the CD4 count on commencement of treatment – once the patient with a very low count is out of the danger zone, they follow the same track as other HIV patients.
    2. Dr Croft’s analysis was premised on a repeated assertion that the later the start of treatment (and thus the lower the CD4 level) the worse the outcome for the patient, which he considered applied to fatigue. Had the Claimant been treated earlier then his prognosis would have been better and the fatigue accordingly either resolved or less significant. This was obviously not an opinion (in contrast to Professor Ross) based on clinical experience of observing differing outcomes in patients with different CD4 levels, and it is therefore important to look carefully at what Dr Croft relied upon to justify his continuing belief that late diagnosis continues to explain the Claimant’s poor health.
    3. Dr Croft relied on two published documents to support his conclusion. The first is what he described as a ‘short paper‘ by ‘Stevens et al‘ published in the Journal of the American Medical Association (JAMA), which he described as “the number two medical journal in the world. It’s very prestigious.” The second document is an academic paper co-authored by Dr Dufty which drew on her own treatment of the Claimant.
    4. JAMA is indeed one of the world’s leading medical research journals and I took it from his repeated references to its high standing that Dr Croft was seeking to elevate the importance of what he described as the “short paper”. The document relied upon is not however an academic paper let alone the type of peer reviewed study that JAMA reputation is founded upon. Rather, it is a one-page document on the “JAMA patient page” entitled “HIV Infection: The Basics.” It is designed to help members of the public understand the basics about HIV. It is written in the plain language of the layperson as one would expect of a public health pamphlet, for example explaining how HIV is transmitted and how it can be prevented in highly practicable terms such as “If you inject drugs, seek treatment and do not ever share needles with others“. In medical publication terms it is the antithesis of a peer reviewed study.
    5. In any event the document does not support the proposition that a delay in diagnosis leads to worse long-term outcomes in terms of fatigue.
    6. The particular lines relied on by Dr Croft in the patient page are a comment that:
“Symptoms tend to increase in severity and number the longer the virus is in the body if the individual remains untreated. Symptoms may include…. Persistent tiredness….”
    1. To rely upon this passage as providing authoritative support for the proposition that a delay in diagnosis results in a worse outcome, once a patient has responded to treatment, including for tiredness, misconstrues what it is saying (which is no more than the longer you leave a symptom the worse it gets) and wrongly elevates a public health pamphlet into the status of a peer reviewed article. To describe it as Dr Croft did as a “paper by Stevens et al”, emphasising the prestigious standing of JAMA is to seek to give an impression of its status neither its single author (Stevens), nor the association, could possibly have intended.
    2. That this is one of the key ‘papers’ relied upon by Dr Croft does nothing strengthen the robustness of his conclusions. His attempts to elevate its importance do little to instil confidence in his analysis of the epidemiology generally.
    3. Dr Croft also misinterpreted Dr Dufty’s article which can certainly be described as a ‘paper’. He sought to rely both in his reports and in evidence on the paper’s findings that a delay in diagnosis results in worse outcomes. The paper does indeed stress the importance of early diagnosis and notes in respect of the Claimant that the delay “… put this soldier at risk of potentially irreversible complications of advanced HIV infection.” At the time the paper was written (about a year post diagnosis) the Claimant was still immunocompromised, as it notes:
“His CD4 count remains below 200 cells/mm3 and may not ever fully recover. It is likely he will be unable to remain within the Army because he remains immunocompromised and he struggles with symptoms as a direct consequence of the late diagnosis.”
    1. When Dr Dufty came to give her evidence, she was cross examined on the contents of the paper and it was suggested to her (consistently with Dr Croft’s views) that the paper supported the proposition that the late diagnosis has led to long-term problems for the Claimant, including fatigue. Dr Dufty disagreed. She contended that the late diagnosis put the Claimant at grave risk of death and exposed him to risks of opportunistic infections but only whilst he remains severely immunocompromised. Since she wrote the paper he has responded very well to treatment and can be deemed to be immuno-reconstituted. Whereas she was previously concerned that he would not recover his CD4 levels sufficiently, or even worse that he might die, that is no longer the case. The expressions about worse outcomes for those diagnosed late reflect the fact that some of these patients will either die or will have worse outcomes because they do not respond well to ART. Once a patient has responded well to ART, with a consistent rise in CD4 levels and drop in viral load, the outcome for morbidity and mortality is essentially the same as if they had been diagnosed earlier. This explanation is entirely consistent with that of Professor Ross, namely that the late diagnosis exposed the Claimant to a short-term period of significant risk but because he has responded so well to treatment, from an HIV perspective he can be considered healthy.
    2. Dr Croft indicated that he had read Dr Dufty’s evidence but maintained that in so far as she considered the Claimant had now immuno-reconstituted, she was wrong and maintained that her paper supported his views. For the reasons given, I disagree.
    3. Fourthly, I found Dr Croft’s ’11th hour conversion’ to a CD4 of 500 as marking the threshold for a ‘normal range’, unconvincing and his explanations for it revealing a lack of independence of thinking.
    4. As referred to earlier, in the first joint memoranda of December 2019, Dr Croft had stated that the normal CD4 range in a healthy body was between 400 to 1400. The Claimant reached that level, and has remained close to it, since July 2018[5].
    5. Dr Croft first indicated his change in view on the CD4 threshold in a healthy population during his cross examination. He explained that he had taken the opportunity to review the trial bundles and had read the WHO paper relied upon by the Defendant, which referred to a threshold in the healthy population as being CD4 500-1500 rather than the 400 he had previously identified. It might be thought surprising that Dr Croft had not considered this WHO paper beforehand as it is the source for the staging criteria (i.e. Stages 1-4) that he had based much of his written analysis upon, but in any event he explained at trial that he now considered the WHO criteria of 500-1500 to represent the ‘gold standard’.
    6. In his oral evidence, Dr Croft placed very considerable significance on the WHO figure of CD4 500. A count of CD4 500 was not simply the beginning of the range in healthy individuals but it importantly, he said, marked the point at which an individual has achieved full immuno-reconstitution. He stated that had the Claimant attained this figure (which he believes he would have done with earlier intervention) then he would have reached the ‘sunny uplands‘ and could have been considered to be reconstituted. He stated, for example, that “if his CD4 reading rose above 500 he would not have immunodeficiency and so therefore it’s not likely to have tiredness to the same degree.” He also stated, in respect of the change from his previous opinion that the threshold was 400:
“…. since then I think I’ve considered the gold standard figure, the slightly authoritative figure is different to that, and its 500 to 1500, and I think that explains a lot about this case, that explains why Mr Jones is still unwell, he still unwell, he’s still fatigued, because he’s got HIV associated immunosuppression.”
    1. Professor Ross was having none of this. In answer to a question from me he stated:
“… it’s a very artificial distinction, as I mentioned, I think the cut-off for what normal CD4 count is is not very relevant with regards to your risk of future illness and when you are in therapy the viral load is less than 40. However, normal will change in different laboratories. Laboratories use different kits to measure CD4 counts, they have different ranges, the average is around 400 to 500, at my lab it’s 350 to 450. It varies. WHO has given a figure not to be used in this way, however, it’s merely a guide to what we think is normal.”
    1. Dr Croft appeared to be suggesting in his oral evidence that the WHO was not simply the ‘gold standard’ but thereby the only standard that regard should be had to. When it was suggested to him that there is a broad range of opinion (an example given was the evidence from Dr Dufty that her laboratory uses a figure of 300,) Dr Croft robustly disagreed:
“No, there isn’t a broad range of opinion. There’s an authoritative ruling from the WHO as to what constitutes a normal range. Now, it may be that in individual practice clinicians will be inclined to blur the margins a bit and they may do that for their own encouragement or to encourage their patients. But I would have thought, Mr Fortt, that that is the gold standard definition, the WHO definition for the normal range in adults and adolescents for a CD4 count and textbooks may say something else, but that’s because they perhaps have a different focus. So I think it would be helpful to stick with what’s on page 794 of the bundle and 795” [the WHO document]
    1. As Dr Croft’s was so clear in his oral evidence that (i) 500 was in fact the applicable standard and that importantly (ii) it marked the point at which someone could be understood to be effectively healthy, it is reasonable to carefully examine this final iteration of his opinion.
    2. It is clear to me that the WHO was not seeking to set any particular standard in its paper, let alone could it be described, as it was by Dr Croft, as some form of ‘pronouncement’ or “an authoritative ruling”. It is simply a single sentence in the 46 page document produced in 2007 which states, as part of general description of immune status in adults that “The normal absolute CD4 count in adolescents and adults ranges from 500 to 1500 cells per mm3 of blood“. The paragraph in which it is contained, then goes on to deal with decreasing CD4 counts in HIV individuals and makes no further reference to the CD4 range in the healthy population. None of the sentence, paragraph or document can be read as setting down (expressly or by implication), a universal definition, ruling, or pronouncement as to what the range of CD4 counts is in the non-HIV infected population. It equally does not suggest that if an HIV patient reaches an CD4 count of 500 then that is the point at which they are clinically deemed to have recovered.
    3. Dr Croft stated in cross examination that his previous opinion, that the threshold was 400-1400, was based on the fact that this was the standard set out in medical textbooks. Asked to explain the difference between the textbooks and the WHO his answer was:
“… I can see now why it is that textbooks are really aimed at clinicians will say that – they will say normal range is 400, and that’s partly because clinicians in the field want to be optimists, they want to motivate their patients to take the drugs and so they will introduce just a bit of a fudge factor by saying 400 is kind of getting on for normal, so let’s call it 400 to 1400 as a normal range and that also incorporates this consideration of the fact that a given reading might fluctuate from one area to the next.”
    1. I found this to be a bewildering answer. It suggested that medical textbooks deliberately provided doctors with false information about the range of CD4 in the healthy population in order to encourage them to give inaccurate advice to HIV patients that they were making progress. In order to make good such a dramatic assertion one would have expected to see some form of substantiation in evidence, for example reference from a credible source that there was an acknowledged pattern in medical textbooks of giving inaccurate data to clinicians in order to nudge them into providing encouraging advice to patients. Unsurprisingly there was none.
    2. One might also have expected a bit more caution or humility from Dr Croft before dismissing the notion that the different ranges referred to in textbooks, and the practice of experts such as Dr Dufty, should be so summarily disregarded. Such caution might have been thought warranted in circumstances in which it appeared he himself was blissfully unaware of the ‘gold standard’ until shortly before he stepped into the witness box. I am forced to the conclusion that Dr Croft’s stance, in particular his explanation for the ranges found in textbooks, was no more than an unedifying attempt to find any argument, however ill-informed, to justify his position. It was certainly not the result of informed, considered or well researched analysis.
    3. It was not the only example of Dr Croft’s paucity of reasoning when seeking to explain away evidence that on its face appeared inconsistent with his opinion. It was put to him that, contrary to his own view that the Claimant responded poorly to ART, his treating doctors, including Dr Dufty, contemporaneously recorded the opposite. He was taken to a letter Dr Dufty wrote to the Claimant’s GP in February 2014 where she noted that CD4 had started to go up, his viral load had gone down to undetectable levels, that he was fully adherent to treatment and that these “were all good prognostic indicators“.
    4. When asked whether or not Dr Dufty’s entries were evidence that the Claimant was responding well to treatment, Dr Croft told the Court that she had sent the letter to the GP knowing that the Claimant would read it and:
“… what the clinician wants to do is to get the patient enthusiastic about their course of treatment and about the drugs so one can see there’s a certain amount of Pollyannaishness about this letter, that presenting a rosier picture, I suspect than Colonel Dufty really had in her mind, and it’s with good clinical reason.”
  1. Dr Croft was suggesting that the treating Consultant was providing inaccurate information to the GP in order not to discourage the patient. None of this was put to Dr Dufty. One would have thought it reasonable to assume that a treating clinician would consider it important to communicate an accurate picture to a GP and that the latter would need to know whether their patient was, or was not, responding well to treatment. In my assessment, Dr Croft in these answers and others was clutching at straws in attempts to seek to justify his position. That is confirmed in my view by the fact that Dr Dufty’s assessment that the Claimant was responding well to treatment, is reflected by an objective reading of her notes and findings including the fact that he continued to see his CD4 levels rise and he continued to maintain his non-detectable viral load. It was also the view of Professor Ross who has long experience of observing responses to treatment in HIV patients.
    1. I far prefer the analysis of Professor Ross. The assessment of whether HIV is a likely cause of ill-health is not undertaken by asking whether or not s/he has reached a CD4 of 500 nor indeed by primary reference to the CD4 range in the general population. It is by assessing whether a patient is responding well to treatment, has a good trend in CD4 count that is above 200 and is maintaining an undetectable viral load.
    2. Fifthly, I found Dr Croft’s explanation for his highly pessimistic views on life expectancy to be very unconvincing indeed.
    3. The analytical framework that he employed shared some characteristics with that of Professor Ross in that both started with a figure for general life expectancy for HIV males drawn from the literature and then modified that number to take account of the Claimant’s individual characteristics.
    4. In evidence, Dr Croft stated that his initial very pessimistic estimate of as little as 6 years until death, which has now moved up to about 10 years, relied on two papers (i) Narrowing the Gap Expectancy between HIV-Infected and HIV Unaffected Individuals With Access to Health Care by Marcus et al[6] and (ii) Impact of late diagnosis and treatment on life expectancy in people with HIV-I: UK Collaborative HIV Cohort (UK CHIC) Study authored by May et al[7]. However, on even cursory examination, these papers did not justify his low estimate of life expectancy and his analysis of the May paper in particular, betrayed a basic misreading of the study.
    5. The Marcus paper notes a dramatic increase in lifespans for those diagnosed with HIV since the introduction of ART. The reasons why life expectancy has not reached ‘normal levels’ are noted to include the heightened risk associated with the social demographics of many suffering from HIV and the high prevalence of other risk factors such as drug use, hepatitis etc that lead to co-morbidities and affect survival – indeed one study cited in this paper noted that now HIV-infected men lose more life years through smoking than HIV infection itself. The study did not consider directly the impact of late diagnosis but rather concluded that the gap in life expectancy for HIV infected individuals is now about 8 years even for those who start treatment with high CD4 levels of above 500. It noted disparities on the basis of race and lifestyle most particularly drug abuse and smoking.
    6. The Marcus study does not, without more, support Dr Croft’s argument that the Claimant’s life expectancy will dramatically fall. Its conclusion is that (albeit in those who start treatment early) general life expectancy without other risk factors is about 8 years less than average. Dr Croft’s analysis is that as the Claimant was diagnosed late one must ‘knock off’ decades more.
    7. In order to justify a drop in life expectancy of approaching 30 years, Dr Croft relied upon a second study which did seek to differentiate outcomes according to CD4 levels at the time of commencement of treatment. That was the study by May et al.
    8. This study focused on life expectancy of those whose were diagnosed late, which was defined as having a CD4 count of less than 200 at the start of ART. The cohort studied was comprised of those who started treatment between 1 January 1996 and 31 December 2008 (the paper was published in 2011) and I consider that Professor Ross was right to note that some considerable caution needs to be exercised over the results because life expectancy has significantly increased over this period because of increased drug efficacy.
    9. In the course of his evidence, Dr Croft was taken to this study and asked to explain how he utilised it to extrapolate his low estimate of life expectancy. He took the Court to a passage in which the authors discussed their results and sought to explain how it demonstrated a greatly reduced life expectancy for those diagnosed with a CD4 below 200. Dr Croft explained how the authors showed that life expectancy reduced markedly by 10 years with every drop of 100 CD4 before the commencement of treatment supporting, in his view a reduction of about 38 years in the Claimant’s life expectancy. It became apparent however that Dr Croft had very materially misread the passage. In fact, as became clear when he was shown the authors illustration of this part of their discussion (‘Figure 2’) their results in fact suggest that a 30 year old man who started ART therapy between 2000-8 with a CD4 count of below 200 could expect to live a further 30 years. This was a drop in life expectancy of some twenty years, more than the later Marcus paper showed but still far off the figures that Dr Croft estimated and indeed had wrongly stated this paper supported. Dr Croft’s misreading of the paper, which he himself identified as one of two foundation stones of his analysis of life expectancy (and the only one looking at different outcomes dependent on the CD4 level at the start of treatment) undermines the credibility of his opinion.
    10. Dr Croft explained that the papers provided a form of launchpad from which he could base a more nuanced life expectancy based on individual factors relevant to the Claimant. Key to his more nuanced calculation was, in his opinion, the very low CD4 count on diagnosis. He said (correctly) that predicting life expectancy is not an exact science and that his initial estimate of 6-9 years included a certain amount of guesswork. What he could not adequately explain was how he could apply the figures in the two papers that he identified and articulate a reason why he thought that the Claimant’s likely life expectancy was so dramatically worse. He was repeatedly asked to provide a reasoned basis for his conclusion which self-evidently cannot be explained by the two studies alone, but in my view wholly failed to provide a cogent, reasoned basis. He pointed to the unhappy condition of the Claimant, what he considered (wrongly in my view) to be the poor response to treatment, the fact that his CD4 remained below the WHO ‘norm’ of 500 (see above) and placed some reliance (again wrongly in my view) on Dr Dufty’s view in her paper that the Claimant’s prognosis was uncertain.
    11. Dr Croft’s analysis does not stand comparison with that of Professor Ross on this issue. Professor Ross relies upon a correct reading of the two studies that he identified as relevant (see below) and applies a bespoke analysis to the Claimant based on the absence of presenting risk factors (he is compliant with his medication, has no relevant co-morbidities, and a non-smoker) and he brings to this issue a lifetime of research and clinical practice.
    12. Sixthly, I was deeply unimpressed by Dr Croft’s attempts to critique the studies relied upon by Professor Ross in his report.
    13. As noted above, Professor Ross relied in his assessment of life expectancy on findings in two studies, one by Gueler et al and the other by Lohse et al. Professor Ross noted that they showed that with successful ART patients were enjoying life expectancies not far off that of the general population. He identified and explained the contents and significance of these studies in his first report. He also explained why their data was more relevant to the Claimant than those relied upon by Dr Croft.
    14. In Dr Croft’s second report he addressed these studies. As they were important to Professor Ross’ analysis one might have expected a detailed response from Dr Croft and a thorough epidemiological critique. In fact, the studies were dismissed by Dr Croft, in the space of a sentence, on a rather dramatic basis:
“the two papers that Prof Jones [sic] encloses that might seem to support his thesis (his second and third enclosures) were both sponsored by drug companies, and hence too are prone to commercial bias”
    1. That was the total of his analysis. There was no attempt to explain how funding bias might have actually impacted on the integrity of either paper, let alone any wider critique of their methodology. That was bad enough but worse was the fact that the accusations were simply wrong.
    2. As Professor Ross noted in his second report, Gueler was not funded by a drug company but by the Swiss National Science Foundation which has clear policies to ensure scientific integrity. The study was peer reviewed and published in a high-ranking medical journal. Lohse et al was also not funded by a drug company but rather the Danish AIDS Foundation. It too was peer reviewed and published in a high-ranking medical journal. Professor Ross explained that some authors in both studies received funding from drug companies for other unrelated work but that this was common for high quality researchers and was declared appropriately and transparently. All this is apparent from the face of the articles themselves, which contain specific sections identifying the funders and also potential conflicts of interests. Had Dr Croft read the articles carefully this should have been obvious to him.
    3. Rather than backtracking, or seeking to substantiate his critique, Dr Croft dug in, moderating only to the extent required by the indisputable facts, to record in the expert memorandum that because some of the authors had openly declared in the papers that they financial links (his phrase) to drug companies it “seriously undermine the paper’s credibility.” Had Dr Croft’s critique been part of a more comprehensive analysis on the impact of industry funding on biases in academic papers, supported by evidence, then it would have been an issue that the Court would have had to grapple with. As it is, all that Dr Croft’s position on these highly relevant papers amounts to is an unevidenced, unexplained and inaccurate, assertion. It reflects in my opinion a ‘shoot from the hip’ approach to the evidence that I have not found helpful when seeking to understand the complexities of the Claimant’s case.
    4. It was not unfortunately the only example of a somewhat cavalier approach to important evidence. One more example suffices. In his second report, Dr Croft critiqued Professor Ross’ calculation of the likely CD4 level in November 2012. As noted above, Professor Ross had based his calculations on two studies one from the UK Health Protection Agency and the other by May et al showing an average decline of CD4 63 per annum in an HIV patient who was not being treated with ART. In his second report Dr Croft was dismissive of this whole approach:
“I do not consider that Prof Jones [sic] retrospective calculations as to Mr Jones’s likely CD4 count at various timepoints are helpful (or indeed valid, given the observed and very wide biological variation in individual patient responses to infection with HIV); the calculations are based on one study only (May 2009) which looked almost exclusively at non-white patients. Mr Jones if [sic] of course “Caucasian”.”
  1. Two things are noteworthy about this trenchant criticism of Professor Ross. Firstly, as Professor Ross pointed out he relied on two studies not just one. In the UK study, which was consistent with the other study by May et all, 80% of the cohort was comprised of Caucasians – entirely negating Dr Croft’s criticisms. Secondly, by the time of the expert memoranda, Dr Croft himself was quite prepared to estimate CD4 levels at November 2012 and indeed provided a figure below that given by Professor Ross.
  2. It will be evident from all that I have set out that I found Professor Ross to be authoritative and fair. His analysis was cogent, clear and corroborated by both the epidemiology and his very significant clinical expertise. It will also be apparent from the foregoing analysis that I took a very different view of the evidence of Dr Croft.

WEBINAR ON EXPERTS

On the 29th June I am presenting a webinar on the Perils and Pitfalls of expert evidence.  This case may well feature.  Details of how to book on the webinar are available here.