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 £5, 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!

The decision of the Court of Appeal in Rowbottom v Royal Masonic Hospital has again shown up the problems encountered by the courts in applying the ‘knowledge’ provisions of the Limitation Act 1980, s 14 in relation to claims for personal injuries. The Court reached a majority decision. It will be argued that this neatly encapsulates the two approaches to the law which have made limitation decisions so unpredictable over recent years. The issue is made all the more important because the Law Commission has recommended that provisions analogous to s 14 should be at the heart of its proposed ‘core regime’ for limitation, which would apply widely to most claims, including virtually all founded in contract and tort (Law Com No 270). The government has announced its intention to legislate to enact the Law Commission’s proposals (Lord Chancellor’s Department, 2002) which brings the prospect that most English limitation law will be based upon the existing personal injury model, notwithstanding the difficulties of interpretation and application that have dogged those provisions.

The claimant sought damages for personal injuries from the defendants. His left leg was amputated as a result of an infection which developed following a hip replacement operation in February 1991. He alleged the infection to be the result of the defendant’s failure to administer prescribed antibiotics. Proceedings were initially issued in December 1993 but subsequently discontinued, and the writ which was the subject of this appeal was issued on 12 July 1996. The defendants pleaded, inter alia, a defence under the Limitation Act 1980. The Master, who tried the limitation point as a preliminary issue, considered that three documents were relevant to the claimant’s date of knowledge: an initial report from his instructed expert, Mr Millner, dated February 1993 (the first report), an opinion of counsel dated May 1993 and a second report from Mr Millner dated 22 July 1993 (the second report).

The claimant’s evidence, which was accepted by the Master, was that until receipt of the second report he did not know that his problems stemmed from a failure to administer antibiotics. Indeed, until then he said that he believed that he had received antibiotics. He had thought all along that his problems had been caused by something having gone wrong but had believed that a failure in the drain inserted in the wound was the most likely culprit. The first report largely exonerated the drain and implicated a possible failure to administer antibiotics. Mr Millner noted that the medical records did not show any antibiotics having been given but failed to go on to draw the conclusion that they were not. Counsel, in his opinion, suggested that Mr Millner be asked to ponder on that conclusion and also sought answers to some additional questions about the causative relevance of any failure to administer antibiotics. In his second report Mr Millner confirmed that it was reasonable to conclude that if the records failed to mention antibiotics being administered then they were not, and provided largely supportive answers to the causation questions.

Top | Contents | Bibliography

All three judges in the Court of Appeal regarded the relevant question as being when the claimant became aware that his injury was caused by the act or omission alleged to constitute negligence within s.14(1)(b), and that the relevant omission in this instance was the failure to administer antibiotics. The majority (consisting of Mantell LJ and Wall J) found that the claimant did not have knowledge until receipt of the second report. Prior to that he had only a suspicion that antibiotics may not have been administered and indeed believed that he had received them. Peter Gibson LJ, dissenting, considered that the Master had been right to hold that the claimant had had the requisite knowledge at the time of receiving counsel’s advice. Indeed he expressed the opinion that he would, had it been relevant, have found the claimant to have had knowledge earlier, when he received the first report. Peter Gibson LJ felt that the advice of counsel added nothing to the claimant’s state of knowledge which was not already available from the first report. However, as both dates left the claimant out of time, the issue did not need to be resolved.

Why then, was the Court of Appeal unable to come to a unanimous decision in relation to what it means to have ‘knowledge’ for the purposes of s.14? These were relatively straightforward and largely uncontested facts. It is a matter of concern, particularly to practitioners in the field, that the Court of Appeal was unable to agree how the law was to be applied to such facts.

The current law of limitation in personal injury claims is found in ss.11, 14 and 33 of the 1980 Act. Section 33 deals with the court’s discretion to extend limitation periods. Its only relevance to this article is to note that this discretion was not available on the facts of Rowbottom because of the earlier House of Lords authority in Walkley v Precision Forgings Ltd [1979] 2 All ER 548. This case had decided that discretion to extend the limitation period was not available where the claimant had issued an initial set of proceedings within time but had not pursued them, and sought to issue fresh proceedings out of time. Section 11 provides that in relation to actions for personal injury arising out of negligence, nuisance or breach of duty the time limit shall be three years from the later of the date of accrual of the cause or action or the claimant’s date of knowledge. The definition of date of knowledge in s.14 is:

and knowledge that any acts or omissions did or did not, as a matter of law, involve negligence, nuisance or breach of duty is irrelevant.

The scheme of these provisions is to ask whether a claimant has actual or constructive knowledge of the s.14 facts.

Top | Contents | Bibliography

The very essence of limitation law is that it should bring a high degree of certainty. This is one of the main justifications traditionally offered for limitation statutes (see such diverse statements of policy as James, 1998; the Queensland Law Reform Commission 1998, p 24; Law Reform Committee 1974, paras 22-23). While certainty is often stated as being a protection for defendants, it is of equal interest to claimants. Practitioners advising claimants must be able to predict with a reasonable degree of confidence, based upon their knowledge and experience, the likely outcome of any limitation issue if they are to advise a client to invest the significant emotional and, perhaps, financial resources into pursuing a claim (and indeed in these days where conditional fee agreements have replaced legal aid in relation to most personal injury claims such practitioners may well be putting their own financial resources on the line as well). In the light of this need for a measure of certainty, inconsistency and confusion are, to say the least, undesirable.

However, the history of the courts’ attempts to wrestle both with what is meant by the concept of knowledge and how to define the facts a claimant needs to know, have not been entirely happy. In particular, the courts have had difficulties in pinning down how certain a person’s knowledge must be in order to satisfy the requirements of the Act. The continuum from not knowing at all to knowing with absolute certainty is a long one, and the selection of the point along its line at which a claimant has sufficient knowledge to trigger the start of the limitation period is, to a degree, arbitrary. This is particularly so when some of the matters a claimant will be considered to know for the purpose of setting his limitation period running are matters which may well be in dispute in the claim and which he cannot truly and finally know until the decision of the judge at the conclusion of the trial.

In Spargo v North Essex District Health Authority Brooke LJ noted (at p 242) that this area of law is ‘grossly over-loaded with reported cases’ and any analysis of the authorities should bear that wise caveat in mind. However, consideration of some of the main cases on the developing idea of what knowledge means will be helpful.

The early cases required a high threshold of knowledge in order for time to start running under s.11. In Davis v Ministry of Defence, The Times, 27 April 1985 the Court of Appeal noted that a claimant who believed his condition to be associated with his work did not have knowledge, because belief was not knowledge. It can be assumed, a fortiori, on the basis of that decision, that suspicion, which is less than belief, is also not knowledge. This was confirmed by the Court of Appeal in Wilkinson v Ancliff (BLT) Ltd [1986] 3 All ER 427 where it was said, in terms, that suspicion was insufficient to constitute knowledge. Those two cases are also authority for the proposition that attributable in s.14(1)(b) means ‘capable of being attributed to’, not ‘caused by’, and that the facts which a person needed to know were those constituting the broad essence of the case, not the detail sufficient to plead it.

This line of authority as to the meaning of knowledge first began to come in for serious challenge in Halford v Brookes and another [1991] 3 All ER 559. Lord Donaldson MR, in the Court of Appeal, (at 573) indicated that knowledge did not mean ‘know for certain and beyond possibility of contradiction’. The requisite knowledge was such as would justify a claimant in embarking on ‘the preliminaries to the issue of a writ’. Vague, unsubstantiated suspicion was not sufficient, but reasonable belief would be. In Nash and others v Eli Lilly & Co Ltd and another; Berger and others v same [1993] 1 WLR 782, the Court of Appeal went further when Purchas LJ (at 796) declared that the antithesis between knowledge and belief which had seemed to be present in earlier cases, was false. Knowledge, said Purchas LJ, is a state of mind and should be the subject of a factual enquiry by the courts.

Brooke LJ, in Spargo, attempted to draw together these and other authorities in a four-point summary of the principles (at 242):

Establishing the required certainty of a claimant’s knowledge is, of itself, of limited value without a consideration of what a claimant needs to know. The relevant facts are set out in s.14(1). It is the issue of attributability in s.14(1)(b) which has presented the main problems. The proviso to s.14(1) expressly provides that the claimant need not know that the act or omission is negligent, but he must know that there is some causal relationship between what was done or not done and the negligence he ultimately pleads. This raises two issues. First, as indicated by Brooke LJ in the fourth Spargo principle, a claimant does not have sufficient knowledge if he is ‘barking up the wrong tree’. Hence, Mr Rowbottom’s firmly held belief that his problems stemmed from a failure of the drain would never have been sufficient to give him knowledge, because he was aiming at the wrong target. Secondly, and perhaps more importantly, is the problem which arises where the claimant knows something about the act or omission but does not know everything about it. The courts have had difficulties in establishing the degree of specificity which is required. The Court of Appeal’s initial view on the matter seemed to be that a high degree of specificity was needed. In Nash v Eli Lilly the Court of Appeal approved Hidden J’s first instance formulation of the relevant act or omission. He rejected the idea that knowledge in a vague and generalised sense was sufficient, suggesting instead that ‘...the plaintiff ...must know that his injury was capable of being attributed...to the very act or omission which his lawyers say in his eventual action constitutes negligence...’ ([1991] 2 Med LR 169 at 177).

The Nash line of reasoning was, however, to be rejected by the Court of Appeal in subsequent cases. In Dobbie v Medway Health Authority [1994] 1 WLR 1234 the claimant had a lump removed from her breast. The surgeon believed that it appeared cancerous and, without waiting for microscopic examination, proceeded to remove the whole breast. The lump was subsequently found to be benign. Years later the claimant discovered that her breast did not need to be removed until after microscopic examination. The Court of Appeal found her case to be time-barred. Sir Thomas Bingham MR criticised the high specificity Nash test on the grounds that it introduced ideas of fault, which were expressly excluded by the Act. All that Mrs Dobbie needed to know was that the removal of the breast was attributable to something which the defendant had done, which she knew virtually as soon as she came round from the operation.

The Court of Appeal has, however, found itself distinguishing Dobbie. In Smith v West Lancashire Health Authority [1995] PIQR P514 the negligence was alleged to lie in the prescribing of conservative treatment rather than surgery for an injury. Although Dobbie might suggest that the claimant had sufficient knowledge at the time the treatment was carried out, the Court of Appeal considered that the relevant omission was the failure to operate, and that a claimant cannot know of an omission until he knows what it is that has been omitted. Until he became aware that there might have been an operation, therefore, he did not have sufficient knowledge.

Top | Contents | Bibliography

The attempt by Brooke LJ to distil the state of the various authorities into a workable test has, regrettably, caused confusion rather than clarity. The divergence between the majority in the Court of Appeal in Rowbottom, and the dissenting judgment of Peter Gibson LJ demonstrates the contradictions and uncertainties inherent in the test of knowledge propounded by the courts in the line of cases detailed above. Mantell LJ and Wall J found the claimant not to have knowledge until receipt of the second report because until that time he did not believe that his condition was caused by a failure to administer antibiotics. If anything, he believed it to have been caused by the failure of the drain inserted into the wound. Wall J (paras 40-41) categorises the claimant as ‘barking up the wrong tree’ until receipt of that second report, and therefore within the fourth Spargo principle. Peter Gibson LJ, however, also placed reliance on Spargo, in his case on principles one and two. He found that the claimant had broad knowledge of the essence of the causally relevant omission to which the injury was attributable in that he knew from the first report that the injury, in the form of the infection, was capable of being attributed, as a real possibility, in whole or in part, to the omission to administer antibiotics (para 56). Put another way, from the first report he knew that it was a real, rather than a fanciful, possibility, that a failure to administer antibiotics was the cause. Mr Rowbottom did not know for certain, but then he didn’t need to.

The categorisation of this case as an omission case seems to have led the majority to the conclusion that Mr Rowbottom required a high degree of specific knowledge. But, in general, the distinction between acts and omissions is problematic. In Smith v West Lancashire the negligence was characterised as being an omission, the failure to operate. But it is just as possible to regard it as an act, the decision to prescribe conservative treatment. And, conversely, in Dobbie, is the failure to carry out microscopic examination not just as relevant an omission as the decision to remove the breast is an act? The arbitrary distinction between acts and omissions, and the different outcomes which seem to result according to the classification used, has introduced a further element of uncertainty into an already confused area.

The confusion arises, it is submitted, because Spargo, and the line of authority it claims to summarise, contains two separate tests. These tests use different types of concept and, in factual situations such as arose in Rowbottom, they are contradictory. The first and second Spargo principles speak of knowledge (as does the Act). The third principle speaks of belief and the fourth principle, while not using the word believe, describes a situation antithetical to the third, and therefore, by implication, is also concerned with belief rather than knowledge. But knowledge and belief make uncomfortable bedfellows. A person can know that X is capable, as a real rather than a fanciful possibility, of being attributable to Y, without believing it to be so. Conversely, a person can believe X to be attributable to Y without any knowledge that it is so. If I have a bad chest and I smoke, it might well be said that I know, as a real rather than a fanciful possibility, that my chest problems may be capable of being attributed to my smoking, but I may not believe it to be so. I may firmly, but irrationally, believe my symptoms to be caused by some chemical to which I have been exposed at work, even though I cannot identify any specific substances in my workplace known to be harmful. I cannot, however, be said to have knowledge in any meaningful sense that my bad chest is caused by my work, not least because it is probably not. The essence of the problem is a confusion of language which has led the courts to conflate different mental processes into an uncertain blend of unspecific ideas. Knowledge is a fact-based concept, whereas belief is an opinion-based concept. The New Shorter Oxford English Dictionary (1993) defines knowledge, inter alia, as ‘acquaintance with a fact or facts’; ‘a state of being aware or informed’. Belief, however, is defined as ‘mental acceptance of a statement, fact, doctrine, thing etc as true or existing’. It is that element of mental acceptance which makes belief different from knowledge. The courts, however, have been inconsistent about whether mental acceptance is an element in the test of knowledge under the Act, and if it is whether it is a sufficient element or merely a necessary one. Spargo suggests that belief is enough, but it is only enough if it is a belief in something which turns out to be correct in the end. Consequently, firmly and reasonably believing something will not constitute knowledge if in fact it is a false belief, because such belief will be ‘barking up the wrong tree’. Firmly but unreasonably believing something will constitute knowledge, even though the unreasonableness of the belief will effectively prevent the claimant litigating earlier because of a lack of supportive evidence. Indeed, in Spargo, the claimant’s belief was entirely unreasonable but it was held to constitute knowledge. Because her claim fell at the limitation hurdle the court was never called upon to adjudicate whether or not her belief was true. Had her belief turned out to be untrue then the limitation decision is plainly wrong, because she would have been ‘barking up the wrong tree’. In those circumstances, however, her claim would have failed on causation – a classic ‘no-win situation’.

Suspicion is, equally, a fact-based concept rather than an opinion-based concept. Mr Rowbottom knew, or at least suspected, at the time of the first report, that failure to administer antibiotics might, as a real possibility, be the cause of his condition, because Mr Millner floated it as an idea. But he did not believe it to be so. Peter Gibson LJ found against him because of that knowledge, whereas the majority found for him because of that lack of belief. Both can find support for their conclusions in Spargo, but as a matter of interpretation of the Act they cannot both be right. It is of course, perfectly possible that both may be wrong, and I submit that they are.

On the wording of the statute it is difficult to support the majority. The Act speaks of knowledge, not of belief, but it is the belief passages in Spargo to which the majority look. But it is also difficult to support the dissenting judgment of Peter Gibson LJ because, again, the Act speaks of knowledge, not suspicion, and suspicion is the most that the claimant can be said to have had at the time that Peter Gibson LJ would have found him to have had knowledge. Indeed, had Mr Millner offered several alternative causes for the infection then, on the logic of Peter Gibson LJ, Mr Rowbottom would presumably have been found to have knowledge of all of them, because they were all raised as suspicions. It is an odd use of the word to say that someone can simultaneously have knowledge of several answers to the same question, even if they may be contradictory.

It is submitted that the Court of Appeal had the proper interpretation of s 14 in the early cases such as Davis v Ministry of Defence and Wilkinson v Ancliff. Knowledge is a straightforward English word. It is not belief and it is not suspicion. It imports a significant degree of certainty and a high level of specificity. The trend of the Court of Appeal to gloss the definition of the word with additional meanings has not led to clarity. The Spargo test has proved itself unhelpful in cases such as Rowbottom, and a test which is not reasonably easy to apply and predictive of outcome is not likely to satisfy the requirements of certainty which are central to the aims of limitation laws.

As a matter of justice as well as interpretation, the Spargo test is deficient. The substitution of suspicion for knowledge is clearly prejudicial to claimants because it will commence the running of time at an earlier point. But even though the majority in Rowbottom found for the claimant, the introduction of the concept of belief into the consideration of knowledge also contains substantial dangers for claimants. In Spargo itself the claimant firmly believed that she had been wrongfully confined to a psychiatric hospital as a result of a misdiagnosis. Despite the fact that she consulted several experts to advise her, and all told her that she was wrong in her belief, the Court of Appeal held that her belief was sufficient to trigger the running of time against her. This is despite the fact that she had acted entirely reasonably in seeking advice and cannot meaningfully be said to have known there to be a misdiagnosis when the weight of evidence was strongly against it.

It seems, therefore, that the majority in Rowbottom reached the right result for the wrong reason. Certainly justice in that case lies with the outcome favoured by the majority. It would appear from the report that this was a well-documented case (as is often the position in clinical negligence claims) where negligence would turn largely on expert argument over the interpretation of the medical records, and it seems right that the issue should be tried.

Departure from the earlier Court of Appeal authority as to what is meant by knowledge, and confusion of the definition with ideas of belief and suspicion are to be regretted. Belief and suspicion should have no part in establishing the trigger point for a limitation period. Limitation laws are, in many ways, punitive in nature. They deprive claimants of their rights irrespective of the substantive merits of the case or the culpability of the defendant. The trend of the Court of Appeal to lower the threshold level of requisite knowledge should be reversed in the interests of justice. At the same time the confusion in the application of the law which has followed from the introduction of ideas of belief and suspicion has made the task of advising litigants more difficult for practitioners on both sides of the fence.

Top | Contents

James R (1998) ‘The Limitation Period in Medical Negligence Claims’ 6 Med L Rev 62.

Law Commission Report No 270 (2001), Limitation of Actions ( London: TSO) HC 23 of 2001-02. http://www.lawcom.gov.uk/files/lc270.pdf

Law Reform Committee (1974) Interim Report of the Law Reform Committee on the Limitation of Actions in Personal Injury Claims (London: HMSO) Cmnd 5630.

Lord Chancellor’s Department (16^th July 2002) press release ‘Government Accepts
Law Commission Proposals On Time Limits For Bringing Cases’.
http://www.gnn.gov.uk/gnn/national.nsf/LC/2F35657328C89F2980256BF800554D57?opendocument

Queensland Law Reform Commission (1998), Review of the Limitation of Actions Act 1974, (Brisbane) Report No 53. http://www.qlrc.qld.gov.au/publications/r53.pdf