Spire Healthcare Ltd v Royal & Sun Alliance Insurance Plc

This is the trial of the application by the Claimant insured, Spire Healthcare Limited (“Spire”) for declaratory relief as to the effect of certain provisions of the combined liability insurance policy (“the Policy”) it held with the Defendant insurer, Royal & Sun Alliance Insurance Plc (“RSA”). In particular, Spire contends that there is no operative aggregation clause so far as any limits on cover is concerned. If correct, the maximum cover is £20m. If there was such an aggregation clause the maximum cover for claims which would be encompassed by it would be £10m, which is the limit for any single claim. Spire also contends that if it was wrong on its primary contention, then there was aggregation in respect of the excess or contribution payable by it in relation to each claim, so that the contribution for a single claim of £25,000 only would be payable in respect of a group of aggregated claims. RSA takes the contrary position i.e. there is aggregation in respect of limits of cover but none in respect of contribution.

The reason why all of this matters is because of a large number of clinical negligence and related claims which have been brought by former patients of Mr Ian Paterson a Consultant Breast Surgeon who carried on his private practice at a number of Spire Hospitals between 2004 and August 2011. The individual claims brought against those hospitals, which form part of the insured under the Policy, allege that Mr Paterson carried out unnecessary and/or inappropriate and/or negligent procedures on a large number of patients. To date, claims have been notified to Spire in respect of 708 complainants. In respect of those complainants, about 86 claims in the High Court Queens Bench Division have so far been issued and are being managed together. The issue, therefore, as to whether Spire has total cover of £10m or £20m is of great significance. If it is £20m, then it is content to accept that there would be no aggregation of contributions which would mean that here, it would be liable for a capped contribution amount of up to £750,000.

If I were to find that there is no aggregation of claims for the purpose of cover, then it will not be necessary to go any further in this action. However, if I find that there is aggregation, there will need to be a further trial with evidence to investigate which of the extant and likely claims will fall within the precise words of the aggregation clause.

The Policy consists of various parts including schedules but all are to be read together as one contract and defined expressions have the same meaning wherever they appear - see page 1.

There are 6 heads of liability cover, namely Section 1 – Employers’ Liability, Section 2 - Public/Products Liability, Section 3 - Legal Defence Costs, Section 4 - Medical Negligence (the section in issue here), Section 5 - Financial Loss and Section 6 - Products Liability Medical Products. Sections 1 and 2 provide cover by reference to “Events” occurring in the period of insurance and sections 4 - 6 provide cover against claims brought against the insured and here it is on a “claims made” basis.

The Schedule entitled Limits of Liability (“the Schedule”) then sets out one or more financial limits to the relevant cover. A further schedule sets out the deductibles or insured's contribution in respect of each head of cover. The individual sections of liability then follow.

I set out here and all in one place, the numerous relevant provisions of the Policy.

£10,000,000 Limit of Indemnity Any one event other than claims

£5,000,000 Limit of Indemnity Any one Event arising directly or

£10,000,000 Limit of Indemnity Any one Event

£10,000,000 Limit of Indemnity All Events happening during the

£10,000,000 Limit of Indemnity All incidents considered to have

In the Policy, Proviso 5 to Section 4 is in fact without any lettered sub-paragraphs which have been added in this judgment for ease of reference. The paragraphs are however printed separately.

Some of the Policy wording above is inelegant or cumbersome. This is no doubt because certain phrases may have come from different policies in the past or have been changed over time so that the present Policy is something of an amalgam rather than a custom-made contract of insurance. For example, some clauses might have had more work to do in this Policy if there had been more separate limits of indemnity set out in the schedule.

Three aspects of the law of contractual interpretation need to be rehearsed.

It is not necessary for me to refer to any case other than Arnold v Britton [2015] AC 1619 for present purposes. At paragraphs 15-17 of his well-known judgment, Lord Neuberger said this:

“15 When interpreting a written contract, the court is concerned to identify the intention of the parties by reference to "what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean", … And it does so by focussing on the meaning of the relevant words,… in their documentary, factual and commercial context. That meaning has to be assessed in the light of (i) the natural and ordinary meaning of the clause, (ii) any other relevant provisions of the lease, (iii) the overall purpose of the clause and the lease, (iv) the facts and circumstances known or assumed by the parties at the time that the document was executed, and (v) commercial common sense, but (vi) disregarding subjective evidence of any party's intentions…

16

For present purposes, I think it is important to emphasise seven factors.

17

First, the reliance placed in some cases on commercial common sense and surrounding circumstances (eg in Chartbrook [2009] AC 1101, paras 16-26) should not be invoked to undervalue the importance of the language of the provision which is to be construed. The exercise of interpreting a provision involves identifying what the parties meant through the eyes of a reasonable reader, and, save perhaps in a very unusual case, that meaning is most obviously to be gleaned from the language of the provision. Unlike commercial common sense and the surrounding circumstances, the parties have control over the language they use in a contract. And, again save perhaps in a very unusual case, the parties must have been specifically focussing on the issue covered by the provision when agreeing the wording of that provision.”

It is not in dispute that these are commonly found in insurance policies and that their effect is to treat multiple insured events or claims as single ones, either for the purpose of financial limits on cover or in respect of contributions or deductibles to be made by the insured, or both. It also cannot be doubted that the effect of such clauses is to put a cap on the relevant amount recoverable for or payable in respect of all such linked claims which is less than would otherwise be the case if each claim within the linked group was to count as a separate claim.

In Countrywide Assured Group Plc v Marshall [2003] Lloyds Rep. IR 195, Morison J observed at p202E that normally one would expect to see the aggregation of a claim for the purpose of limiting cover also to have the same effect in respect of the payment of any excess, so that the aggregation would benefit both insurer and insured in different ways. But whether that was so in any given case depended on the wording actually used. So one should approach any question of construction in relation to an aggregation clause without any preconceptions as to the result and to undertake it in a balanced way. Similarly, Morison J observed in Mabey v Ecclesiastical Insurance [2001] Lloyd’s Rep. IR 369 at page 372 that the fact that aggregating clauses are common and are in the insurers’ interests is not sufficient to imply an aggregation clause when none is stated expressly.

In AIG v The Law Society [2016] Lloyd’s Rep. IR 289, Longmore LJ pointed out at paragraph

I have been reminded of the caution that is to be exercised before applying the so-called

“argument from redundancy” in the context of commercial contracts, particularly those containing standard terms. The fact that on one interpretation a particular clause or part thereof would become otiose or unnecessary may be as much a reflection of what Lord Hoffmann referred to in Tea Trade v CIN Properties [1990] EGLR 155 as “linguistic overkill” by the draftsman as a sign that an alternative interpretation should be preferred. See also the decision of the Court of Appeal in Tektrol v International Insurance [2005] 2 Lloyds Rep. 701 and the judgment of Lord Neuburger MR in Macquerie v Glencore [2010] EWCA 697.

That, of course, is not to say that the redundancy argument can have no role to play at all. It all depends upon the construction issue in question, the effect of the alternative interpretation and the contractual context as a whole.

The first point to note is that the language of this provision is unquestionably that of aggregation. It is in the same words as the first example given by Longmore LJ in AIG cited in paragraph 14 above. It is also a replication of a significant part of the policy definition of “Event”.

The second point is that this provision makes sense in its own terms because it prescribes a

Accordingly, and as Mr Shapiro accepted, had the “Section 4” part of the Schedule contained a discrete limit said to apply to “claims arising out of one cause" or some such, that, together with Proviso 5 (a) would indeed have operated as an aggregation clause. However, he contended that such words are not there; all that the Schedule provides is one limit for “any one claim” being £10m and then £20m for “all damages costs and expenses arising out of all claims…” Thus there is a missing “third limit” which is what RSA is trying to introduce. Moreover, even if Proviso 5 (a) could be said to be referring to one of the two existing limits in the Schedule it is not clear which it should be and, applying the contra proferentem rule the default choice must be £20m.

However, this argument ignores three points. First, it is not really a question of three limits; rather it is a question of there being three categories of claim: a single claim, a number of claims not all falling within Proviso 5 (a) (“non-linked claims”) and those which do (“linked claims”). Only two financial limits are necessary because the linked claims are to be treated as a single claim. It is true that Proviso 5 (a) does not state that latter consequence in express terms but that consequence is what aggregating clauses are all about. Further, there is no real difficulty in deciding which financial limit applies because it is obviously the lower one than would otherwise be the case; this must follow, since the purpose of aggregation is to reduce cover in the case of linked claims and here the lower amount is specifically referenced to “one claim”. Finally, in my judgment there is no real ambiguity about the effect of Proviso 5 (a) so as to invoke the contra proferentem rule in any way – see Tektrol at para. 8.

Accordingly, I do not accept Spire’s point that Proviso 5 (a) is incomplete as an aggregation clause because, while it contains a description of the claims to which it applied (i.e. those arising from a single cause) it failed to contain a “doing" provision by which the consequences of claims falling into that category were set out. The consequence was that such claims could not exceed the limit of indemnity stated in the Schedule which, for the reasons given above, is £10m. So there is nothing in this point.

Thus I conclude that the language of Proviso 5 (a) is certain and clear when taken with the Schedule so that it acts to aggregate linked claims.

Spire’s interpretation of Proviso 5 (a) is that it has the same effect as Proviso 5 (b): “all claims" in either category are subject to the £20m limit. But if so, both provisions are entirely unnecessary because Proviso 1 has already “enacted” such limits.

In this particular context, that seems an odd result because it is apparent that these two elements of Proviso 5 appear separately in other sections but have been brought together here. And in both elements, there is a change from “the” Period of Insurance (as is found in Proviso 3 to Section 1 and Proviso 5 to Section 6) to "any". This is clearly done deliberately, in order to reflect the fact that cover in Section 4 is on a “claims made” basis.

Having said that, I have to accept that there is still degree of duplication because even on RSA’s interpretation, Proviso 5 (b) would itself be redundant having regard to Provisos 1 and 4. All that can be said is at least both elements are not redundant and also that Proviso 5 (b) might serve to highlight the rule in Proviso 5 (a) about linked claims, by distinguishing them from non-linked claims. As for Proviso 4 itself this may not be redundant (having regard to Proviso 1) because it may be making clear for the avoidance of doubt that just because more than one person is making the claim for indemnity, this does not increase the total limit beyond £20m.

However it is also submitted that there is duplication elsewhere in the Policy. In Section 1 cover is based on Events already defined so as to include a number of events which arise from the same cause. On that footing, Mr Shapiro argues that on any view, Proviso 3 (the equivalent to Proviso 5 (a)) is unnecessary here and duplicative in that context. I see that, because both financial limits in the Schedule are concerned with “any one Event” and because Proviso 1 already says that the total amount payable under Section 1 shall not exceed the limit of indemnity. However, it may be that Proviso 3 adds something by making it plain for the avoidance of doubt that the aggregation rule applies not just to linked Events but also to linked claims arising out of one Event. It may also be that the end of Proviso 3 (which is the same as Proviso 5 (c) to Section 4) may be of limited value where something turns on the number of parties otherwise than for the purpose of any limits of cover.

Equally, it is submitted that Proviso 5 to Section 6 (the equivalent of Proviso 5 (b)) is duplicative because Proviso 1 thereto, which follows three possible types of claim themselves set out in the preceding sub-paragraphs (a)-(c)), has already established that for all claims the limit is as set out in the Schedule. And the Schedule provides only one limit of indemnity i.e. that for all claims. I see that also.

Again, Proviso 1 to Section 2 is otiose here because all three applicable limits in the Schedule are the same.

However, the fact that a particular clause may be duplicative in other parts of the Policy does not necessarily mean that it is also merely duplicative in Section 4. Indeed, if one treats Proviso 5 (a) as an intended aggregating clause, then, unlike Proviso 3 to Section 1, it is not duplicative of an earlier aggregating clause since there is none – that is because there is no definition of “Claim” like there is of “Event”. Proviso 5 (a) only becomes duplicative if one accepts Spire’s interpretation of it as such.

The truth is that such provisions (including Proviso 5 (a) and Proviso 5 (b)) may or may not be duplicative or unnecessary depending on the context and in particular the Schedule. Thus, for example, in Section 6, sub-paragraphs (a)-(c) might be important if there were separate limits of indemnity for such items in the Schedule (as there are for Section 2 items, albeit in the same amounts). It all depends upon what logically precedes the relevant provisos, as it were. Otherwise one would have to conclude that certain provisions in the Policy with the wording as contained in Proviso 5 (a) and Proviso 5 (b) are in fact always otiose, wherever and however they appear – but that is an unsustainable conclusion in my view.

So it does not follow that the provisos in Section 4 are necessarily duplicative. And it is, after all, that section, and not the others, which must be the primary focus of attention.

The key contextual difference as far as Section 4 is concerned is that the Schedule contains two different financial limits, one of which is for single claims. And for the reasons given above, it

It is only if one instead aligns the higher limit of £20m to Proviso 5 (a) as well as Proviso 5 (b) that one reaches the conclusion that both elements are merely duplicative or confirmatory. Nor do I accept that because both of these provisions each refer to “all claims” they must both be referring to the £20m limit in the Schedule - because this ignores the aggregating language of Proviso 5 (a). 34. The treatment of Proviso 5 (a) (to show that it is simply confirmatory) in paragraph 121 of Spire’s written submissions but substituting the two actual limits in the Schedule for the words “the Limit of Indemnity” does not work because one would not put both limits in but only one. On Spire’s (but not RSA’s) case, there is only one and it is £20m.

It is important to note that Spire is not contending (at least primarily) that Proviso 5 (a) should be disregarded because in effect it is meaningless or nonsensical, as opposed to being harmlessly duplicative. But as noted above, the latter conclusion does not square with the aggregating impetus of its language towards the lower limit.

It is also important to note that the facts of this case are very different from those in the cases relied upon by Spire as examples of the courts applying an interpretation which gave rise to an element of redundancy, of which it is necessary to refer to only three.

First, in Tektrol itself the Judge at first instance had held that the physical theft of a computer claimed for on the insurance policy was within an exclusion clause which applied to “erasure, loss, distortion or corruption of information on a computer system”. This was on the basis that unless “loss” could refer to circumstances such as the theft of the computer itself, the term was redundant since any “loss” of data would be covered by the separate term “erasure”. It is easy to see how that reasoning made far too much use of a redundancy argument as the majority in the Court of Appeal held especially where, on that interpretation, the exclusion clause was now made to apply to something completely different from all the other events contemplated i.e. physical loss of the computer itself as opposed to elimination of data.

Equally, in Arbuthnot v Fagan [1994] 3 ReLR 168, the question was whether a “pay now claim later” clause in a Names’ agents contract meant that the Names were prohibited from making any kind of negligence claim against the agent until they had paid a recent cash-call made by the agent (they had not) or whether non-payment only precluded an action brought in relation to the propriety of the payment requirement itself. The Court of Appeal held that the latter was obviously the correct interpretation and so the Names were not barred from commencing proceedings against the agents for negligent misstatement. The Court dismissed one particular argument which was that if this was the correct interpretation then certain phrases within the relevant clause would either be redundant or would overlap with provisions elsewhere. In this context, and where all the other factors pointed to the narrow interpretation it was therefore “no justification for construing the language so as to apply it to a situation which on a fair reading of the general purpose of the clause was not within the target area” i.e. using the redundancy argument to improperly extend the ambit of the exclusion clause. But again, that reasoning has no bearing on this case. There is no serious dispute here as to the ambit or scope of Proviso 5 (a) when read with the Schedule. It plainly consists of aggregating language. The dispute here is between interpreting it so as to make it work as an aggregating clause or interpreting it so that it does no work at all.

Finally, in the case of Bindra v Chopra [2008] EWHC 1715, a property and trust case, Etherton

I therefore reject the notion that in this context, the fact that Spire’s interpretation would render Proviso 5 (a) redundant, is insignificant.

I accept that it would have been much neater and elegant if linked claims were defined to constitute one single claim by an appropriate definition of “Claim” as was done with “Event”. But in frequently used, modified and revised policies of insurance, neatness and elegance are often lost. And one returns to the point that the actual aggregating language of Proviso 5 (a) is the same as that employed in the definition of “Event". This case is very different from the facts of Mabey where the question was not the proper interpretation of what appeared to be an aggregating clause, but rather whether to imply one where no such clause was expressed at all.

Equally, the fact that there is a standard notification clause for linked claims found in Proviso 7 to Section 4 (there is a similar proviso in Section 6) does not mean that because the aggregating clause with all its consequences is not set out in a similar way within Section 4, Proviso 5 (a) should not be interpreted as such. Again, that is to assume too much for stylistic or structural consistency in respect of a policy such as this.

The same applies to the additional limits contained in the separate provision dealing with Associated Companies in Section 1.

It is clear that, since the Schedule in respect of Section 4 is dealing with the limits of cover for claims in respect of “all damages costs and expenses” Proviso 5 (b) needs to be interpreted with the words “costs and expenses” added after the word "damages". Mr Shapiro contends that while this is correct, if RSA contends for that conclusion so that Proviso 5 (a) and Proviso 5 (b) are consistent with each other in that regard, then it follows that they must be regarded as consistent with each other for all other purposes, i.e. that they are both directed to the same financial limit and that they are both equally redundant. I reject that argument as a non-sequitur.

Mr Shapiro also contends that if one did not so construe Proviso 5 (b) so that on the face of it both provisions were different in that respect, then they would cease to be redundant because while Proviso 5 (a) allows for costs and expenses to be part of the limit in linked claims, Proviso 5 (b) does not in the case of non-linked claims. This does not follow either, because in fact one must construe Proviso 5 (b) so as to include costs and expenses – otherwise it conflicts with both the Schedule and the introductory paragraphs to Section 4. Furthermore in my view there would still be a need to give effect to Proviso 5 (a)’s aggregating language.

Mr Shapiro also contended that there was no pleaded commercial purpose to RSA’s construction of Proviso 5 (a). That may be, but the claimed commercial purpose was obvious from its submissions and is plain in any event - that is, in order to give effect to a purported aggregating provision which would limit RSA’s liability in such cases. Moreover, it would be consistent with the same aggregating intention which flows from the definition of an “Event".

It is also said that in contrast to Section 4, aggregation is necessary for Section 1 because otherwise there is no overall limit. It is true that by having aggregation (through the definition of “Event”) there is one kind of overall limit, but if there is a number of unlinked claims, then there still remains no total limit. Similarly, aggregation will produce one kind of total limit in Section 2. But either a total limit overall is already there under the rubric of (b) in the Schedule or if rubric (b) does not catch every claim, for example one in relation to public liability rather than product liability, then there remains no overall total for unlinked claims of that kind. So I do not consider that a case can be made that aggregation is somehow necessary for Sections 1 and 2 but not for Section 4.

Equally, the fact that Sections 3-6 happened to impose total amounts for all claims, whether linked or not, does not mean that no contractual intention to have aggregation as well must be assumed for Section 4. The fact remains that Section 4 does have a separate limit for individual claims as well but unlike Section 2, the limits are different.

Finally, I consider that RSA’s interpretation best fits the indicia set out by Lord Neuberger in Arnold: (i) the natural and ordinary meaning of the clause is that of aggregation (ii) that is consistent with other parts of the Policy which give effect to aggregation for other liability cover (iii) it is consistent with the overall purpose of Proviso 5 (a) i.e. aggregation and the applicable Schedule which here has different limits for different kinds of claim (iv) no factual matrix evidence has been adduced but aggregation here, as well as in other sections of cover in the Policy might well be expected, objectively, given the many different hospitals insured and the fact that it is hardly inconceivable that numerous claims could arise from one essential cause - but even if one ignores this, there is certainly no countervailing factual matrix point put forward and (v) it is certainly consistent with common sense.

Accordingly, in my judgment Proviso 5 (a) operates as an aggregation clause so that claims falling within it are subject to the lower limit of £10m.

The Schedule in respect of Insured's Contribution is set out in paragraph 7 above.

It is common ground that the operative “Aggregate Insured's Contribution" here is £750,000. Since “claim" is not defined as the word “Event" is, it is not possible to say that the use of the word “claim" here connotes and includes a group of linked claims. What that means is that, up to the limit of £750,000, Spire will have to make a contribution of £25,000 in respect of each separate claim liability even if linked. There is no other form of express wording in this Schedule to reduce the contribution to merely one payment of £25,000 where there are linked claims.

Notwithstanding this, Mr Shapiro submits that if (as I have found) Proviso 5 (a) operates as an aggregation clause in respect of cover limits, it would be illogical if there was no equivalent aggregation for Section 4 claims in terms of contribution. I see the force of that point. I can also see that if it were otherwise, it is somewhat unfortunate given that the defined expression “Event") will achieve parity of aggregation, as it were. However, to achieve that result for claims would simply amount to rewriting this part of the Policy in a way which does too much violence to the language. The observations, once more, of Morison J in paragraph 13 of his judgment Countryside are pertinent:

“.. Normally, I accept, the policy will be worded so that the aggregation of the claim will involve an aggregation of the excess in respect of claims so that claims are aggregated both for excess and limit. But that will depend upon the wording of the policy, and the circumstances which surround it being underwritten. I am not persuaded that the plain and ordinary meaning of the words in this policy should be corrupted, as I think they would have to be, to make this policy work “normally”.

I take the same view here. Accordingly, RSA succeeds on this issue, too.

There must therefore be judgment in favour of RSA and against Spire. I am most grateful to both Counsel for their very helpful oral and written submissions and I will hear them on the precise form of declaratory order together with any other consequential matters, upon the handing down of this judgment.