[2024] UKUT 170 (AAC)

It has to be said at the outset that the basis on which this third ground of appeal has been advanced by counsel on behalf of the Appellant has shifted somewhat over time.

First, the Appellant’s original grounds of appeal accepted that the approach to causation as set out in JM was correct, but sought to argue that the FTT had misdirected itself in the application of that test to the facts of the case. For example, it was argued that there was an inherent contradiction between the FTT’s finding that the IMEG report did not try to define a latency period (paragraph 18.7) and its conclusion a few sentences later that “sun exposure post April 2005 might have led to some form of melanoma, but not for many more years” (paragraph 19).

Secondly, however, the Appellant’s skeleton argument then raised the possibility (at §16) that JM had no application in the case of a ‘long-tail’ condition (“It is difficult to conceive of how the application of a predominancy test would apply to an ‘all or nothing’ (and hence indivisible) condition”). On this basis, the Appellant’s “primary case” was then put rather differently – namely “if JM is found not to be applicable”. By analogy with Holmes v Poeton Holdings Ltd [2023] EWCA Civ 1377; [2024] 2 WLR 1029, it was argued that the AFCS causation test should be whether service has made a “material contribution in fact” to the outcome. Failing that, and as it was put, “even if” JM remained good law, the Appellant’s case in the alternative was that his appeal should have succeeded by reference to factors such as genetic susceptibility and current medical knowledge.

Thirdly, and finally, Mr Rawlinson’s oral submissions at the Upper Tribunal hearing involved a further change of tack. He reverted to the position that the approach to causation as set out in JM was correct and accordingly eschewed any suggestion that the common law test of a “material contribution in fact” was sufficient to meet the predominancy test. In concluding, he argued that the application of the four-fold staged process in paragraph 118 of JM only admitted of two possible outcomes. The first was to accept that there was nothing to do but throw one’s hands in the air as a long-tail cancer case would never satisfy the predominancy test. The second was to acknowledge that the predominancy test was satisfied in this case by reliance on the majority of periods of intense sun exposure having taken place in the post-2005 period. Mr Rawlinson averred that the latter was both the correct and the just approach.

As I understood them, Mr Rawlinson’s submissions necessarily proceeded on the following basis. As to the first element of the four-stage test in paragraph 118(i), both pre-2005 UVR exposure in service and post-2005 UVR exposure in service were potential process causes. As to paragraph 118(ii), there were no other possible process causes in play as intense exposure to the sun during service (at whatever date) was the only operative process cause (childhood exposure having by now been discounted). As to paragraph 118(iii), this likewise posed no problem as there was no relevant non-service cause. As to paragraph 118(iv), Mr Rawlinson submitted that here the FTT had misapplied the predominancy test. Thus, as it had been framed in the grounds of appeal, it was argued that “the evidence demonstrated that the overwhelming balance of the exposure post-dated 2005” (at §69). This comes perilously close to a perversity challenge.

In support of the proposition that the overwhelming balance of the Appellant’s UVR exposure post-dated 2005, Mr Rawlinson provided (at the Upper Tribunal oral hearing) a document entitled ‘Tables of Exposure’. This sought to tabulate the respective periods of sun exposure during the Appellant’s first and second periods of service. It purported to show 14 months of UVR exposure in the 1988-1997 period and 32 months of sun exposure in the second period of service (namely after 2004). Quite properly Mr Rawlinson acknowledged that this was not a document that had been agreed by Mr Hays, but, in any event, he argued that it was simply collating the evidence as it was before the FTT and presenting it in a compendious manner. For present purposes I leave to one side the facts that (i) no formal application was made to admit this document in the Upper Tribunal; (ii) the Respondent has not agreed it (and so Mr Hays has not made submissions on it); and (iii) self-evidently it was not put before the FTT, at least in the present form.

The reality, however, is that on closer scrutiny the document does not prove what it seeks to assert. In particular, the second period of the Appellant’s service is not congruent with post-2005 service, as the first entry for the second period (6 months in Afghanistan, October 2004-April 2005) properly counts as pre-AFCS service. Furthermore, the last entry for the second period of service (10 months in Afghanistan, September 2017-June 2018) cannot realistically be included as relevant causative sun exposure, given both (a) the FTT’s factual finding about this period (“This winter deployment was primarily office based indoors”: paragraph 14.7) and (b) what is known about latency (an issue to which I also return below). It follows that providing an aggregate duration of sun exposure of 14 months (for the first period of service) and 32 months (covering the second period of service) is of no relevance and is potentially misleading. The self-same data can be more accurately recategorized respectively as 20 months’ exposure (being the original 14 months + 6 months from 2004/05) for pre-April 2005 service and 16 months’ exposure (32 – 16 (6 + 10)) for post-April 2005 AFCS service. Looked at in this more nuanced way, the evidential case for post-April 2005 exposure as being predominant in comparison to the pre-April 2005 exposure is not at all obvious.

Be all that as it may, it is not the Upper Tribunal’s role to review and redecide the facts where the right of appeal is confined to an error of law jurisdiction. As rehearsed above, the FTT here made detailed findings of fact as to the Appellant’s UVR exposure both before and after April 2005 in the section of its decision headed ‘The evidence relating to the Appellant’s sun exposure’ (which actually contains a combination of a narrative of some of the key evidence together with the FTT’s factual findings) – see paragraph 24 above. Those findings amply supported the FTT’s overarching and determinative findings of fact as neatly summarised at paragraphs 18.1 and 18.2 of its decision:

All this notwithstanding, the outcome of the application of the causation test cannot be determined by a somewhat arid arithmetical exercise of counting the respective numbers of years and months of UVR exposure both pre- and post-April 2005 (and even putting to one side the obvious point that the intensity of exposure at any given time and at any given location will vary according to both geographic and climatological factors). Instead, the assessment of causation is a complex multi-factorial exercise, taking into account ‘What we do know’ (paragraph 16.6) and bearing in mind ‘What we don’t know’ (paragraph 16.5). The question of latency is part of this wider process of multi-factorial assessment. In this context two principal submissions were made on behalf of the Appellant as regards the FTT’s approach to latency.

The first, as outlined above, was the submission that there was a contradiction between the FTT’s finding that the medical evidence did not “define a latency period” (paragraph 18.7) and its conclusion just several sentences later that sun exposure in the Appellant’s AFCS service “might have led to some form of melanoma, but not for many more years” (paragraph 19). This argument goes nowhere. In stating that the medical evidence did not “define a latency period”, the FTT was simply observing that the medical evidence did not specify a set number of years (or a range of years) for latency. Rather, the UVR Synopsis stated in more guarded terms that “‘childhood sunburns … have been linked to melanoma in later life’ implying a long latency period” (paragraph 16.5, 4^th bullet point). This understanding was reinforced by the acknowledgement that exposure as a child or young adult “is a potent risk factor” (‘What we do know’, paragraph 16.6, 1^st bullet point). In short, the FTT was recognising that a trigger event may occur in youth or early adulthood but not manifest itself until much later in life and indeed it noted that “Mr C’s age profile is consistent with a ‘peak incidence in the fifth decade’” (‘What we do know’, paragraph 16.6, 3^rd bullet point). It follows that there is no inconsistency between the FTT finding there is no defined (in the sense of definitively specified) latency period and its conclusion that e.g. the mole first identified in 2013 was unlikely to be attributable to (relatively recent) exposure to UVR after April 2005.

The second main submission on latency was that the FTT had failed to define what it meant by the “basic injurious process” (paragraph 16) and failed to distinguish between what Mr Rawlinson referred to as latency type 1 and latency type 2. By latency type 1, he meant the time lag after the exposure to the harmful agent and the period during which no changes to the body took place. By latency type 2 he referred to the later period that elapsed after the affected cells begin to mutate and the cancer spreads. There are at least two reasons why this submission is unpersuasive. The first is that this was not how the case was run before the FTT, so the tribunal can hardly be properly criticised for failing to make this distinction. The second is that on any fair reading of the FTT’s decision as a whole it is perfectly clear what it meant by the “basic injurious process”. It was referring to the entirety of the period from the time of the relevant causative exposure to the time that the melanoma manifested itself, namely, as the FTT put it, “the timeline between exposure and onset” (paragraph 16.5, 3^rd bullet point), in other words the latency period as that term is commonly understood.

I therefore conclude that the FTT did not err in law in its approach either to the issue of latency nor to the application of the JM predominancy test.