Select Committee on Health Minutes of Evidence

Memorandum by Mr David Thrower - continued

Yellow Card Reports

  227.  Under the yellow card scheme:

    ". . . doctors, dentists, hospital pharmacists and coroners are asked to provide details of suspected—NOT NECESSARILY PROVEN—ADRs they encounter during their regular work. For established drugs, report only serious reactions". (source: Committee on Safety of Medicines factsheet, 1998, their emphasis).

  228.  There is a Sub-Committee on Pharmacovigilance (SCOP), which meets every two months and which advises the main committee on the safety and risk/benefit considerations relating to vaccines and other medicines. The Committee has eight members with appropriate specialist medical backgrounds, but there is concern by parents that they may lack sufficient specialist expertise in statistical methodology in their original in-the-field data-collection, and may have taken insufficient measures to search-out cases of suspected degeneration, because degeneration into autism was not recognised as potentially connected with vaccination.

  229.  There seems to be a contradictory attitude within the MCA towards the yellow card system:

    "The yellow card system . . . has a proven record in quickly identifying new safety concerns" (letter of MCA of 18 June 1998).

    "The yellow card system is well established and has been shown to be effective in identifying previously unrecognised adverse drug reactions" (letter of MCA of 21 August 1998).

    "The level of under-reporting varies with the condition being reported . . . For very serious events which are caused by a medicine, then our evidence suggests that under-reporting will be much less" (my emphasis. Letter of MCA of 21 August 1998).

  230.  But then, apparently in direct contradiction:

    "In most situations there is considerable under-reporting of reactions. It has been estimated from various surveys that only 10-15 per cent of serious ADRs are reported" (my emphasis. Source: Adverse Drug Reaction Information Service; Guidance On Interpretation Of Yellow Card Data, published by the Committee On Safety Of Medicines and the Medicines Control Agency, 1997).

    "Spontaneous reporting systems such as the UK yellow card scheme provide a major data source . . . (but) . . . it is accepted that spontaneous reporting schemes have limitations, including the level of under-reporting)". (Letter of MCA of 29 March 1999.)

  and perhaps even more significantly:

    "In our whole database . . . autism has been very rarely reported as an adverse drug reaction . . . These figures are unsurprising since autism is not a recognised ADR to any particular medicinal substance" (letter of MCA of 29 March 1999, my emphasis).


    "Evidence from the yellow card scheme is unlikely to resolve the issue as to whether or not autism could be causally associated with MMR vaccine" (letter of MCA of 29 March 1999).

  231.  An indication that the level of under-reporting of adverse drug reactions could be extremely high, even higher than is being acknowledged by the Medicines Control Agency and the Committee on Safety of Medicines, is provided by the West Midlands Centre for Adverse Drug Reactions Reporting. The West Midlands Centre has reported:

    "A recent survey from a major teaching hospital in Oxford gives a clearer idea of the proportion of adverse reactions seen on the general medical wards and reported to the CSM . . . 1,420 adverse reactions were collected from 20,695 consecutive admissions, a rate of 60 per 1,000 admissions. Of these, 477 would have fulfilled the CSM criteria for reporting, but only 30 yellow cards seem to have been sent, accounting for only 6.3 per cent of the identified `reportable' reactions"

  232.  The West Midlands Centre offers three possible explanations for this extremely poor rate:

    —  the reporter has not proved the relationship between drug and reaction;

    —  the reaction is not severe enough; and

    —  the reaction is already well known, even when severe; and is therefore not "news".

  233.  It is suggested here that the first-mentioned could be particularly important when applied to degeneration into autism, because:

    —  the Department is actively attempting to persuade doctors that there is not a relationship (though this paradoxically could alert them to the possibility of one);

    —  many parents would not suspect a relationship, though this is now changing;

    —  some doctors will be anxious not to be "taking on" their Department and Chief Medical Officer over the degeneration issue; and

    —  degeneration could be spread over successive weeks/months, and thus be missed altogether.

  234.  The extremely low level of reporting is also backed up by a study based upon the whole UK yellow card database (Belton et al, Journal of Clinical Pharmacology 1995, 39, 223-6). This found that in any one year, 11-13 per cent of all UK doctors sent in at least one yellow card. The study estimated that about 60 per cent of all doctors would report at some stage in their working lives.

  235.  But turning these statistics around, it seems remarkable that in any one year 87-89 per cent of doctors do not put in a single yellow card report, and, even more remarkably, that it is estimated that 40 per cent of doctors would not expect to submit any yellow cards throughout their entire working lives. This could theoretically imply that almost all medicines are totally safe, but alternatively, it is also possible to conclude, that many parents do, that the yellow card system is extraordinarily ineffective—a viewpoint clearly endorsed by the West Midlands figures quoted above.

  236.  Also it is interesting to note that, in the United States, the view is:

    "Accumulations of events reported to a passive surveillance system do not allow incidence rate calculations, due to the generally unknown extent of under-reporting . . . On the basis of VAERS data alone, we don't have proof that vaccines are not contributing to these problems and we certainly don't have proof that they are contributing" (source: Caveats to the Interpretation of VAERS Data, Center for Biologics Evaluation and Research, Vaccine Adverse Event Reporting System (VAERS), 1998, my emphasis).

Other Warning Mechanisms That Miss Autism

  237.  Correspondence has revealed that the MCA appears to be placing faith in other auxiliary mechanisms that do not seem likely to identify autism as a potential serious adverse effect of vaccination for a minority of individual children. For example:

    "It is accepted that spontaneous reporting schemes have limitations, including the level of under-reporting. For this reason the MCA employs a variety of other data sources . . . For example, the MCA also monitors the safety of medicines through . . . a number of world-wide data bases. There are two large computerised GP databases in the UK which are used by the MCA routinely to evaluate drug safety concerns." (source: letter of MCA of 29 March 1999).

  238.  However, the above does not address the obvious criticism, that GP databases are highly unlikely to contain details of slow degeneration into autism following vaccination, and that once again, the MCA may not be able to grasp the numerous cases of degeneration that may have occurred, because its systems are pointing at (in the context of autism) the wrong database targets.

  239.  The MCA has also pointed to improvements to the yellow card adverse reaction reporting system. However, the improvements that it cited in its letter of 29 March 1999 were

    —  the extension of the yellow card scheme to hospital pharmacists;

    —  collaboration with GP prescribing system suppliers;

    —  pilot projects to extend the scheme to community pharmacists and nurses.

  240.  None of the above would have substantive relevance to the detection of any increased incidence of autism.

  241.  The limitations of the pharmacist scheme were again pinpointed by the West Midlands Centre for Adverse Drug Reaction Reporting, in September 1997:

    "Reporting By Pharmacists—A Slow Start . . . all registered hospital pharmacists (but not pre-registration pharmacists) and community pharmacists in this Region can report adverse reactions to CSM West Midlands providing they are directly involved in the care of the patient" (their emphasis).

  242.  As hospital or indeed any other pharmacists are unlikely to have any involvement in the care of autistic children, it is not obvious how this measure relates to detecting degeneration into autism in the weeks and months following vaccination, nor to associated gut and food hypersensitivity problems. Once again, it appears to be missing the autism-degeneration or autistic-enterocolitis target, though no doubt it is a sensible measure in relation to other health conditions.

  243.  If the above are the most quotable improvements on offer, it implies that the yellow card scheme is still missing autism, and that measures to amend it to pick up autism as an ADR to vaccination have yet to be taken. If post-vacine degeneration into autism continues to not be recognised as connected with vaccination, then minor improvements to the yellow card scheme will have no impact whatever.

Actual Yellow Card Figures

  244.  The yellow card system has produced figures for measles/MMR vaccine adverse reaction reports that, at first glance, seem to suggest that adverse reports are low, and that very few children are affected.

  245.  However, if the numbers of reports are a reflection of the level of under-reporting quoted (all types of report) by the MCA and the CSM in their literature, then the figures are rather more disturbing.

  246.  In 1988-97, some 3,029 reports were received by the MCA for MMR (each report relates to one child, but can include more than one specific reaction, and the receipt of a report does not conclusively provide evidence that it was linked to vaccination). Over the nine and a quarter years 1988-97, 3,029 reports relates to eight million children approximately, giving a report receipt rate of 1 in 2,645 approximately.

  247.  However, if under-reporting is running at 10 per cent to 15 per cent of total suspected instances, as indicated by the MCA/CSM (admittedly an overall, all-reaction, non-MMR-specific quote), then the 1 in 2,645 adverse report rate could potentially be up to 10 times higher, or up to 1 in 265.

  248.  Some of the above could be reports relating to very minor reactions. However, even the above could overall be an underestimate. Degeneration into autism may have been missed altogether, because it may have taken place gradually over weeks or months, and because symptoms such as language loss and acquired hyperactivity almost certainly would not have been connected by parents as potentially linked to vaccination. In my son's case, it was almost a decade before the connection was made and an adverse reaction report filed.

  249.  The MCA has also confirmed (letter of 18 June 1998) that it monitors medicines, including vaccination, through:

    —  evaluation of a variety of epidemiological studies;

    —  published and unpublished literature;

    —  a number of worldwide databases; and

    —  two large computerised GP databases in the UK.

  250.  However, it would appear that few or any of the above will have picked up degeneration into autism after vaccination, or alerted the MCA to it, with the possible exception of certain literature.

  251.  Experience indicates that the MCA is not always in possession of all the relevant research literature. As for studies, the problem there is that they have not been done. It seems unwise to depend upon studies when none have been commissioned.

  252.  The MCA has pointed to its "new method", developed by Dr Paddy Farrington (a statistician at the Open University) and others, as a way of obtaining a good estimate of whether a particular adverse event occurs more frequently immediately after a vaccination than it does prior to a vaccination. The method requires known vaccination histories for all those children who are believed to be affected by a particular adverse event, whether associated with a vaccine or not. The "new method" uses separate databases, for hospital admissions for serious events and for vaccination records, to do this.

  253.  However, a serious potential flaw may be what is defined as "immediately after". If the Farrington method defines "immediately after" as up to six months, then it may detect a few cases of degeneration into autism. However, even that would depend on the child appearing on a database during that period, which may (given the time taken for paediatricians to assess and diagnose children) be extremely unlikely. Many children are not formally diagnosed for several years, or even at all.

Failure to Contact Families

  254.  It is a source of great surprise to parents such as myself that the MCA does not closely investigate the health background of families of children who have degenerated into autism, to see if there is a pattern of background circumstances that could help to provide a focus for future investigation, and future research.

  255.  In the case of Oliver Thrower, a "yellow card" report was filled out by the GP, and then no further contact made. No direct contact at all was made with members of the extended family of Oliver Thrower, although his GP was contacted and his own individual case was added to ADR figures for the year he was immunised.

  256.  Despite the above case, and other cases, the MCA clearly has adopted the viewpoint of the Department of Health, under which it is controlled for overall management purposes, that there is no link between vaccination and autism, based upon the seminar of March 1998. However, as it is believed that no actual evidence was presented to the seminar, the seminar's conclusion that there was no evidence was unsurprising, and the MCA's reliance on this conclusion may prove in time to have been misplaced.

  257.  What is interesting, to parents seeking investigation as to how their child degenerated into autism following vaccination, is that the MCA's various safety-nets do not appear, even in combination, to be focused upon the vaccine/autism issue. This, and the past failure to recognise degeneration into autism as an ADR, and the shortness of timespan for ADRs to apparently be required to fall in, may explain how the entire syndrome may have come to have been missed.

Failure to Actively Seek Out Potential Cases

  258.  It is also disturbing to parents seeking an investigation of their child's degeneration that the MCA is not actively seeking out details of every single potential case of vaccine damage, as is being done in the reasonably comparable instance of suspected Gulf War Syndrome.

  259.  The MCA has explained its policy of not following up all the reported cases on a proactive basis as follows:

    "Suspected ADRs reported in association with vaccines may not necessarily mean `vaccine damage' because causation remains unproven. Reports range in seriousness from minor symptoms such as nausea or fainting to serious neurological reactions such as encephalitis. Following receipt of every yellow card, a letter is sent to the reporter reminding them to forward any further relevant information when available. Each yellow card is evaluated by experienced assessors within the MCA with a focus on following up serious cases, those for which more information is needed to allow for a full assessment, and those which are not recognised ADRs". (full text of response to this question: source: letter of MCA of 29 March 1999.)

  260.  The above, which has not properly addressed the original question in the correspondence, begs several questions:

    —  how does this approach, of only reacting to yellow cards, deal with parents' anecdotal cases?

    —  how does it deal with cases of autism following vaccination where the parents have not connected the two events, but where the one followed on from the other?

    —  how does this approach, of only reacting to yellow cards, compare with the approach with GWS suspected victims, where it is understood that every GP has been approached on a cold-canvas basis in the search for potential cases?

    —  as the MCA itself has already stated the view that autism is not a recognised ADR, then the following up of yellow cards presumably treats these children as having experienced an unconnected event.

  261.  The above could be expected to severely depress the MCA's figures. The MCA quoted figures on autism/ADR as at 29 March 1999 as:

    —  37 cases of autism;

    —  12 drug substances cited;

    —  29 of the 37 associated with MMR.

  262.  At present, there appears to be some suggestion of a culture of "no news is good news", with the results that statistics on adverse reactions are minimised. This may lead to a "circular argument of self-resassurance":

    —  statistics are low;

    —  this implies that there are few problems;

    —  because there are "few" problems, the Agency does not need to proactively seek out cases;

    —  because the Agency does not seek out cases, statistics are low.

  263.  The lack of proactive searching-out of autism cases therefore appears to place the MCA in a chicken-and-egg trap, whereby it does not identify a problem because of the lack of reports, and it does not seek the reports because it has not identified that there is a problem. This is extremely disturbing.

  264.  The MCA does acknowledge that:

    "Despite pre-marketing trials to establish the safety and efficacy of a medicine, some drug safety hazards will only emerge after widespread use of the medicine in clinical practice. . .This is particularly true for rare adverse drug reactions" (source: letter of MCA of 29 March 1999).

  265.  This implies that the MCA requires a very large scale detailed database on autism and other forms of sudden degeneration to be able to identify whether autism and vaccination are causally as well as temporally linked.

Large Linked Databases

  266.  The need for large linked databases (LLDBs) was recently acknowledged by the US Institute of Medicine in the publication Vaccine Safety Forum: Summaries of Two Workshops, National Academy Press, Washington DC, 1997:

    To improve detection of adverse events associated with vaccination. . .CDC established LLDB in 1990. This database is a CDC-coordinated linkage of large databases of four large health maintenance organisations. Comprising automated data from more than 500,000 children aged 0 to 6 years in Oregon, Washington and California, LLDB enables researchers to link vaccine exposures to medical outcomes and thereby to estimate the rates of occurrence of adverse events following vaccination. . An elevated rate following vaccination, in comparison with a rate at other times, would suggest a possible causal relationship to vaccination."

  267.  The absence of such a database in the UK to date must leave the Department of Health's present publicly-stated conclusion, that there is no link, open to question in the minds of parents seeking the investigation of their child's condition.

  268.  It should also be pointed out that even in the case of the US LLDBs cited above, parents are concerned that the criteria for qualifying as a possible adverse event are so tightly drawn as to exclude many potential adverse reactions. This may be particularly important in the instance of autism, where degeneration may be gradual over a period of several weeks.

  269.  There are many areas where the MC could become more proactive, and actually seek out improved data, and cases of possible damage.

  270.  Concern by parents on these issues is not limited to the UK. At the Vaccine Safety Forum hosted by the US Institute of Medicine in Washington DC in 1997, the published report of the proceedings noted:

    "A parent noted that no government representative had ever contacted him personally seeking more information about his son's death following vaccination with hepatitis B" (the death had also been misclassified in the US adverse event reporting system, as Sudden Infant Death Syndrome).

  271.  The parents themselves could be brought into the system, again as noted at the US 1997 Vaccine Safety Forum:

    "Noting that most paediatricians lack the time to fill out long forms, a physician suggested that the number of reports . . . could be increased by placing the burden of such reporting on the parents of vaccinated children".

  272.  A criticism of the US vaccination adverse event reporting system (VAERS) has been that its data is not easily accessible either to the medical and research establishment or to the public, with parents having to use the US Freedom of Information Act to gain access to it. In the UK, there seems to be availability of information from the MCA on demand, but this could be readily improved by its being posted on the Internet, so that parents can interrogate it from their homes. It is suggested that this be done.

  273.  Publicly releasable information should also be put out in a more user-friendly and self-explanatory form, including graphs and indications of estimated true levels, allowing for known under-reporting. The MCA could work with parents' groups to develop an acceptable format. This would then help parents to make an informed choice.

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries

© Parliamentary copyright 1999
Prepared 27 July 1999