Wootton v J Docter Ltd & Anor

Despite taking an oral contraceptive dispensed to her by the respondent on 1 September 2003, the appellant conceived shortly thereafter and gave birth on 22 May 2004. The child is happy and well. But the appellant had a torrid time both in delivery and thereafter when she suffered post-natal depression. The pill, Logynon ED, dispensed to her by the respondent was not the pill she had been prescribed by her doctor. He had prescribed Microgynon ED. She claimed damages for care, expenses and loss of earnings flowing from the pregnancy and a period of post-natal psychosis. She sued the company, as first defendant, which had contracted to provide pharmaceutical services at the pharmacy where the respondent, the second defendant, had worked as a self-employed locum pharmacist.

His Honour Judge McKenna, at the Birmingham County Court, found that the second defendant, the respondent, was guilty of negligence in erroneously dispensing to the claimant Logynon ED rather than the Microgynon ED which she had been prescribed. The first defendant was acquitted of negligence. It plays no further part in these proceedings.

However, the judge concluded that the respondent’s negligent error had neither caused nor materially contributed to the contraceptive failure and the pregnancy of the appellant.

The appellant now appeals against the judge’s conclusions as to causation. She contends that his conclusion was not justified by the evidence.

The appellant also contended that because the state of scientific knowledge precluded her from establishing that but for the erroneous dispensing of Logynon ED she would not have become pregnant, she could invoke the alternative and exceptional principles of causation, identified in Fairchild v Glenhaven Funeral Services [2003] 1 AC 32, as explained by this court in subsequent jurisprudence.

The judge had to deal with far more detail than I believe to be necessary for the purposes of the appeal. It suffices to record that the appellant was first prescribed Microgynon ED oral contraceptive pills by her general practitioner on 1 November 2001. She continued to be prescribed and to take the oral contraceptive pill thereafter. On 27 August 2003 she obtained a repeat prescription for Microgynon ED. I should pause to note that there was a dispute as to the dates of the prescriptions on the basis of which the respondent sought to contend that the pregnancy was a result of the appellant failing to take the contraceptives which she had been prescribed. I do not deal with it since the judge refused to allow that line of argument to be pursued in the light of the respondent’s failure to cross-examine the claimant and of doubts as to the accuracy of the records on which he relied.

On 27 August 2003 the appellant obtained a repeat prescription for Microgynon ED. On 1 September 2003, at the first defendant’s pharmacy, the respondent dispensed to her Logynon ED rather than the Microgynon ED she had been prescribed. The respondent had disputed that he was responsible for this error; he denied that it was he who had dispensed the wrong box.

The appellant did not notice that she had been prescribed a different oral contraceptive, although both the packaging and the appearance of Logynon pills differ from those of Microgynon ED. The judge dismissed the allegations of contributory negligence which the respondent chose to make. The appellant took one Logynon on the night of 1 September 2003 and the following morning, 2 September 2003, after her husband had returned from work, they had intercourse. She only noticed later that day that the strip of pills she had been given differed from her usual prescription. She took a second Logynon on the evening of 2 September 2003. After reporting to the practice nurse at her GP’s surgery the dispensing error was corrected and she returned to taking Microgynon from 3 September onwards. She noted that she was pregnant on 25 September 2003. Subsequent history is not relevant for the purposes of the appeal.

The judge made no finding as to the date when the appellant became pregnant. But, in the light of the ultrasound scans and the fact that the appellant did not have sexual intercourse again until 10 September, the two expert gynaecologists called by the respondent accepted that the most likely date of conception was 2 September. I am prepared to take 2 September as being the date of conception. Accordingly, the contraception failed at a time when the appellant had taken one Logynon.

The appellant had been prescribed and believed she had taken Microgynon ED. Microgynon is a monophasic preparation. In other words, all the active pills within the formulation of that oral contraceptive contained the same amount and types of hormone. A prescription of one box of Microgynon ED (standing for “every day”) will contain twenty-one tablets containing 30mcg of ethinyloestradiol, a synthetic analogue of the female hormone oestrogen and 150mcg of levonorgestrol, which contains progesterone. After taking twenty-one tablets the recipient takes seven dummy tablets before starting a fresh course.

Logynon ED is a triphasic contraceptive, a formulation with three separate types of tablets containing different amounts of two types of hormone. The first six tablets in the course contain 30mcg of ethinyloestradiol (oestrogen) but only 50mcg of levonorgestrol (progesterone). The second phase consists of five tablets with 40mcg of ethinyloestradiol and 75mcg of levonorgestrol and the third phase 30mcg of oestrogen and 125mcg of levonorgestrol. After those twenty-one tablets, a similar quantity of seven dummy tablets must be taken.

At the outset it is important to observe that by virtue of the change from Microgynon to Logynon the appellant took exactly the same quantity of ethinyloestradiol (oestrogen) but took 100mcg less of levonorgestrol (progesterone). On the second night of 2 September 2003, again, she took 100mcg less of progesterone but exactly the same amount of oestrogen as she would have taken had she continued with the prescribed contraceptive Microgynon. Thereafter, she returned, as I have said, to taking Microgynon. Thus the essential factual issue, leaving aside questions as to the correct formulation of the test of causation, is whether the failure of the appellant’s contraceptive cover can be attributed to the drop in intake of progesterone from 150mcg (which she would have received from Microgynon) to 50mcg on one, or, at most two nights, in consequence of the dispensing error.

The appellant relied upon the evidence of Dr Jackson, Reader in Clinical Pharmacology and Honorary Consultant Physician. Two Consultant Gynaecologists, Mr Afnan and Professor Dalton, were called on behalf of the defendants. Dr Jackson concluded that the intake of two Logynon tablets, instead of Microgynon, was the most likely explanation for the contraceptive failure. He based his view not merely on the relationship between the intake of those tablets and the appellant’s conception but upon observational studies relating to the lower efficacy of triphasics as opposed to monophasics and his interpretation of studies of those switching from one oral contraceptive to another and of those seeking an abortion after a failure of oral contraception. Based on those studies he concluded that there was a tenfold increased risk of failure as a result of the combination of a change from Microgynon and Logynon and the lower efficacy of triphasics as opposed to monophasic contraceptives.

Mr Thomson’s essential argument for the appellant is that Dr Jackson’s evidence ought to have been accepted in preference to that of Mr Afnan and Professor Dalton. They both disputed the inferences Dr Jackson drew from the studies on which he relied; both asserted that there was no biological reason to suggest that the intake of two Logynon pills increased the risk of contraceptive failure. It is, therefore, necessary to analyse in greater detail the nature of Dr Jackson’s evidence and the sources on which he relied.

The starting point was the figure which Dr Jackson took to be the inherent failure rate of oral contraceptives which combine hormones. The failure rate for each cycle (there are 13 cycles in each year) he put at 1 in 13,000, or an annual failure rate of 1 in 1000. He based that figure upon a paper by Trussell on Contraceptive Efficacy. Dr Jackson then considered the evidence that triphasic preparations were less effective than monophasic. In a paper by Ketting, in the Netherlands, in April 1998 an observational study was undertaken in relation to those seeking abortion following an unwanted pregnancy which the patient claimed was due to failure of oral contraceptive. The study observed that double the number of those requesting abortion in all age groups took the triphasic pill rather than the monophasic. It excluded those who suffered with conditions which interfered with the reliability of such pills, such as the use of other drugs, vomiting and diarrhoea. Its analysis excluded what it described as usage errors, reporting errors and those who could not identify which pill they had used. It reported that the number of failures was only 2 per 10,000 years of use. Only 2 of the 3 consecutive years, adopted for analysis, were statistically significant but, in those 2 years, double the number of those taking a triphasic pill, rather than a monophasic, had sought an abortion. But it did note that the risk of pill failure occurring in spite of correct use remained what it described as “marginal” and suggested that there was what it described as “slightly diminished reliability with the triphasic type”.

Ketting did not identify the contents of the pills blamed for the unwanted pregnancy. Dr Jackson relied in support on a further paper from the Medical Journal of Australia dated 15 May 1989 by Kovacs and others which supported the Ketting paper on the basis of a study of 209 inadvertent pregnancies in oral contraceptive users. It suggested that triphasic pills “may” have a lower margin of safety and thus result in a higher rate of inadvertent pregnancies than monphasic pills.

The studies in Australia and the Netherlands were in abortion clinics where, as Ketting himself acknowledged, subjects of the study were likely to be younger, unattached women who prefer triphasic preparations and were more likely to request an abortion. Moreover, Kovacs took no account of the effect of diarrhoea and vomiting. Neither study purported to be a randomised controlled trial. It was, doubtless, those defects which led to the review on which all the experts relied. This was a review by van Vliet and others comparing the efficacy of triphasic against monophasic oral contraception. It recorded the earlier studies to which I have referred on the basis of which it noted:-

“…whether triphasic oral contraceptives have higher accidental pregnancy rates than monophasic oral contraceptives is unknown.” (my emphasis)

Van Vliet sought to assess, on the basis of 18 randomised controlled trials, the hypothesis, based on the observational studies of Ketting and Kovacs, that triphasic oral contraceptives were less effective in preventing pregnancy than monophasic oral contraceptives. The result of its assessment of the 18 studies was that no significant differences could be found between the various pills and contraceptive effectiveness (see page 7 of the review). In its discussion it noted that the sample sizes of individual trials were too small to detect differences in effectiveness and in its plain language summary recorded that the evidence was not strong enough to say whether the three-phase pill was better than the one-phase pill for pregnancy prevention.

It included a graph, at the end of the review, on which Dr Jackson appears to have placed considerable reliance. This showed that over a six month period 3 out of 328 taking monophasic pills became pregnant as opposed to 2 out of 350 taking triphasic. However, in 5 studies over a twelve month period, 3 out of 2,094 taking triphasic pills became pregnant as opposed to only 2 out of 2,051 taking monophasic. That 1 extra pregnancy, when taking a triphasic pill, was the basis of a graph showing an increased risk of pregnancy of a factor of 1.35, described as an “odds ratio”. It emerged that that had formed the foundation, in part, of Dr Jackson’s view as to the increase of risk of unwanted pregnancy when taking triphasic as opposed to monophasic pills.

But in fairness to Dr Jackson, that was not the sole basis of his assessment of increased risk. He relied upon the evidence of unwanted pregnancy in those who had “switched” from one oral contraceptive to another. It is important, for the purposes of this appeal, to stress that the studies on which Dr Jackson relied for the evidence as to the consequences of switching were studies of those who changed from one type of oral contraception to another, not merely for two nights, but for the whole cycle. Although during the course of the appeal there was much discussion about switching, the appellant never “switched” from one contraceptive to another in the manner discussed by the study to which Dr Jackson drew attention. All that happened was a reduction of one out of the two hormones by 100 or, at most, 200mcg.

Dr Jackson had reported that changes in the brand of contraceptive could be associated with an increased risk of failure. He referred to a report in the Netherlands by Oddens and others as to the incidence of switching in the Netherlands over the period of four years from 1990-1993. They found that 37% of a sample of just over 4,500 women switched in the first 12 months after starting to take contraceptive pills. Dr Jackson then referred to a study by Sparrow and others in The New Zealand Medical Journal of February 1987, in which it was reported that in 11 cases out of 163 failures that had occurred in the first month of use, 7 patients had been using another brand in the month before. 6 had the predisposing factors to which I have already referred. Accordingly, only 5 women, without a predisposing factor, fell pregnant within the first month of starting a new pill. The study does not reveal how many of those five were taking triphasic pills at the time of conception. It was on the basis of that report, coupled with the information derived from the incidence of switching reported by Oddens, that Dr Jackson gave his view that there was a 13 times greater risk of an unwanted pregnancy following switching than when continuing with the same preparation (see page 10 of his written report).

Neither Mr Afnan nor Professor Dalton regarded those sources as support for Dr Jackson’s conclusion. On the contrary, they could identify no biological reason for any increased risk. The oestrogen component remained the same. Oestrogen, Mr Afnan reported, was the main hormone preventing ovulation. Combined with oestrogen, progesterone increases the barrier effect of cervical mucus.

The appellant’s riposte to the suggestion that there is no biological reason for a reduction in efficacy was to rely upon a paper by Spona dated 1987 on the efficacy of low-dose oral contraceptives containing, amongst other things, Levonorgestrol, which sought to determine the borderline dose for ovulation when taking Levonorgestrol. That showed that at 50mcg, whilst ovulation was inhibited in 3 out of 3 cases, in one of those 3 cases follicular maturation, the precursor of ovulation, did take place. Moreover, it was accepted by both Professor Dalton and Mr Afnan, in evidence, that there was an increased likelihood of ovulation at the start of a woman’s cycle, in other words at the time the appellant took her reduced dose of progesterone.

The final element of evidence on which the appellant relied related to the warnings given to the manufacturers of oral contraceptives and in other medical advice in relation to occasions when pills were missed or when there was a change from one type of pill to another. The appellant placed considerable reliance upon warnings given when changing from one type of contraceptive to another. The Microgynon advice, when changing from a 21-day combined contraceptive pill to Microgynon, is to finish that pack and then start taking Microgynon the next day. The advice is not to leave an interval between packs; it asserts:-

“By starting in this way you will have contraceptive protection at once.”

The manufacturer’s advice when changing from a 28-day pill is to finish by taking all the active pills but to throw away the inactive, or dummy pills, and then start the Microgynon pack. Again, it is asserted that by starting that way the recipient will have contraceptive protection at once. When changing from another type of oral contraceptive to Logynon the same advice is given, in short not to leave a gap between active pills. The Monthly Index of Medical Specialities of February 2002 recommends additional precautions if what are described as the placebos, the dummy pills, have been taken before the switch. Similar advice is given in the APBI Data Sheet Compendium of 2001-2002 and the British National Formulary of March 2003.

The advice given by Professor Guillebaud, an acknowledged leader in this field, in his work “Contraception: Your Questions Answered” indicates that some women will ovulate if pills are missed for more than seven days, in other words, after the period when dummy pills are taken. In his experience, follicular maturation will start in a small percentage of women at the beginning of the cycle.

However, I must also record, as the judge himself did, that in advice dated 1990 Killick and others noted that whilst increasing a pill-free interval from seven to nine to eleven days allowed more follicular development, ovulation did not in fact occur. They also reported that no differences could be detected between the different types of combined oral contraceptive. Moreover, the manufacturers of Logynon advise that contraceptive protection should not be affected if one active pill is missed for twelve hours or less, provided the late pill is taken at once, even if it means taking two pills in one day. Similar advice is given by the Royal College of Obstetricians and Gynaecologists in its April 2005 document which asserts that a woman does not require additional contraceptive protection even if a pill is missed provided she remembers to take the most recent tablets thereafter. As the judge recorded, even if the appellant had missed two of her Microgynon pills, according to the Royal College’s advice she would still have had satisfactory contraceptive cover.

The judge held that Dr Jackson was:-

“…simply wrong to conclude that the efficacy of triphasics is lower than monophasics.”

He concluded that the literature on which Dr Jackson relied does not help and he accepted the evidence of the defendant’s experts.

He accepted that the manufacturers’ advice demonstrated that the manufacturers believed that a change from one combined oral contraceptive to another increased the risk of a woman becoming pregnant. He pointed out that the advice would have a safety margin built in, as an application of the precautionary principle. But he concluded that the advice did not, in reality, cover the appellant’s situation. He pointed out that the increased risk was unquantified and was contradicted by the Killick paper.

The judge concluded by saying that :-

“the evidence of Mr Afnan and Professor Dalton as to the absence of any biological reason that the taking of Logynon ED pills would increase the risk of contraceptive failure is very persuasive. Mr Afnan summed up his view in the course of his evidence in this way, when he explained that the background risk, that is, of intrinsic failure of the pill, was far greater and far outweighed any possible additional risk due to the taking of two wrong pills. That evidence is in my judgement compelling. Whilst it is superficially attractive to infer from the fact that the claimant had not got pregnant whilst taking any of the preceding Microgynon pill packs, but did fall pregnant in a very close relationship with a dispensing error which was itself at an important and vulnerable time at the beginning of her cycle after seven days of inactive pills, there is simply insufficient evidence to conclude that the drug error caused or materially contributed to the pregnancy.”

He repeated his conclusion in the following paragraph (104):-

“As it seems to me Dr Jackson’s estimate of the increased risk by a factor of 13 is simply not supported by the material on which he relies and the methodology is invalid. To the extent that there is any possible additional risk resulting from the taking of two wrong pills that risk was far outweighed by the intrinsic contraceptive failure risk and in the circumstances it cannot be shown that the dispensing error caused or materially contributed to the claimant’s pregnancy and I would therefore dismiss the claim.”

In my view there was ample evidence upon which the judge could reject Dr Jackson’s estimate of increased risk and the inferences which the appellant suggested should be drawn from the manufacturer’s and other advice. The criticism advanced by Mr Afnan and Professor Dalton of the sources on which Dr Jackson relied emerge from the very material deployed by Dr Jackson. The conclusions in the van Vliet analysis go far to destroy any basis for concluding that there is an increased risk. Dr Jackson himself accepted that previous non-randomised studies were too small.

Moreover, those studies did not relate to the circumstances of the appellant’s case. She had not switched from one oral contraceptive to another, still less from a monophasic to a triphasic contraceptive, over the space of one or more cycles. She had merely reduced the level of one of the two hormones supplied, progesterone, by 100mcg or, at most, 200mcg. None of the studies, however small, cover that situation. Added to that is the absence of any biological explanation as to why the reduction in the level of progesterone to a level, which even Spona found to be adequate to prevent ovulation, should have caused pregnancy in this case.

The judge was entitled, on analysis of the underlying material and on the basis of the evidence of both Mr Afnan and Professor Dalton, to reject the views of Dr Jackson. Mr Thomson was unable to advance any reason as to why this court should substitute a different conclusion to that of the judge either as to the efficacy of Microgynon over Logynon or as to the consequences of taking what he described as “two wrong pills”.

Nor was Mr Thomson able to satisfy me as to why the warning given on change by the manufacturers demonstrated an increase in risk. The manufacturers do not advise any extra precautions if pills are missed. Their advice deals with a different situation, namely a change lasting throughout a cycle from one type of oral contraceptive to another. More fundamentally, the basis upon which their advice is given has never been discovered or disclosed. Dr Jackson did not make enquiries of the manufacturers. There was debate at trial as to whether the advice applied where the underlying type of progesterone remained the same. There was evidence that the warning was given where the type of progesterone changed from one oral contraceptive to another but it was never clear as to whether it should apply, other than in the application of the precautionary principle, when the type of progesterone remained the same, as it did in the instant case.

There is, in my view, no basis for challenging the judge’s findings or conclusions of fact. There was ample foundation, both in the analysis of the written sources on which Dr Jackson relied and in the evidence for his preferring the experts advanced by the defendants.

That seems to me sufficient to dispose of this appeal but I should mention one further argument advanced by Mr Thomson.

Untutored by the evidence and the judge’s findings of fact, this case might have appeared to be another staging post in the Court of Appeal’s quest to identify the principles of causation. The judge founded his approach on the President of the Queen’s Bench Division’s reliance in Clough v First Choice Holidays and Flights Limited [2006] EWCA Civ 15 on Bonnington Castings v Wardlaw [1956] AC 613:-

“What was required was for the plaintiff to make it appear at least ‘that on a balance of probabilities, a breach of duty caused, or materially contributed to, his injury’.” (§ 44)

In a supplemental skeleton argument Mr Thomson now contends that because of the current state of scientific knowledge it is “inherently impossible” (his words) for the appellant to prove exactly how her injury was caused. The negligent dispensing of the Logynon was capable of causing a failure of contraception and materially increased that risk. Her pregnancy was the result of the kind of risk created by the respondent’s wrongdoing. Moreover, her pregnancy was caused by what he describes as “the same agency, namely ovulation, which the reduction in levonorgestrol permitted to occur. These submissions were formulated to align the case for the appellant with principles identified by Lord Roger in Fairchild v Glenhaven Funeral Services [2003] 1 AC 32 and Barker v Corus UK Limited [2006] AC 572 as explained by Lady Justice Smith in Sanderson v Hull [2008] EWCA Civ 1211 at § 53.

Inevitably, these arguments were met with the riposte that the appellant had to prove, on the balance of probability, that had the respondent not negligently dispensed Logynon, she would not have become pregnant, in short, an application of the simple “but for” test. This was not, contends the respondent, a case in which it was impossible to show why contraception failed because of the current state of scientific knowledge. The biology was understood; statistical analysis failed to establish an increased risk due to a reduction in the intake of progesterone by 100mcg or at most 200mcg. Neither observational nor random tests are likely to be able to replicate circumstances where so limited a change takes place over so short a period.

In my view, this is not a case in which to add to the teaching of the Court of Appeal on causation. The reason is obvious. The appellant has failed to establish that the erroneous intake of two Logynon pills materially increased the risk of contraceptive failure. She cannot invoke any principle to be derived from Fairchild or Barker to overcome the judge’s finding, which I believe to be unassailable, that the reduction of 100mcg or 200mcg or progesterone did not increase, let alone materially increase, the risk of contraceptive failure. Whether this case is to be regarded as a case in which the appellant must satisfy the “but for” test for causation or whether the principle in Fairchild and Barker can be extended and applied to a case such as this is immaterial. The appellant has not proved that the reduction in level of Logynon in that short period increased the risk of contraceptive failure.

Mr Thomson urged this court to conclude that the judge had found as a fact that the intake of two Logynon pills gave rise to an increased risk. He founded this argument upon the reference, which I have already cited, in §§ 103 and 104, to the judge’s conclusion that:-

“The background risk…of intrinsic failure of the pill was far greater and far outweighed any possible additional risk due to the taking of two wrong pills.”

The judgment must, of course, be read as a whole. The judge did not find any additional risk. In the passage I have cited in his ante-penultimate paragraph the judge was doing no more than to refer to the evidence of Mr Afnan, the gynaecologist called by the first defendant, to which he had referred in the previous paragraph. Mr Afnan had distinguished between two types of risk, namely an inherent risk of contraceptive failure and a distinct risk caused by the changed and reduced intake of progesterone. Mr Afnan was cross-examined about Dr Jackson’s reference to an increased risk by a factor of 13. Mr Afnan, whilst not accepting Dr Jackson’s figures, suggested that the risk due to the change was a lower risk than the inherent risk and was asked, on behalf of the appellant:

“So you’re saying that according to this, that the additional risk is less than the background risk (the inherent risk)?”

Answer: “That’s right”.

The judge then intervened:-

“Exactly. And therefore on the balance of probabilities it is more likely the background risk was the risk which was effective. I think that summarises your position?”

Answer: “That’s right.”

I have every suspicion that this part of the cross-examination was based upon Mr Afnan’s interpretation of Dr Jackson rather than the propositions which Dr Jackson sought to advance. Dr Jackson was referring to an increased risk of an inherent failure by which he meant failure due to an unknown cause. It therefore made no sense to describe the additional risk as less than the background risk. The additional risk increased the background risk.

But I do not think it matters. Whether the risk due to the reduced intake of progesterone is to be regarded as a risk distinct from the inherent risk or merely a risk which increases the inherent risk, the appellant failed because she was quite unable to establish any risk, attributable to so limited a reduction of progesterone, at all.

It is for those reasons that I would dismiss this appeal.

I agree.

I also agree.