Science can prove a negative--- 20 March 2005
Clifford G Miller,
Lawyer, graduate physicist, former examining Univerisity lecturer in law

Send response to journal:
Re: Science can prove a negative
Dear Sir,


What Mr Jacobs says (1) could mislead my fellow lawyer Ian Barnett (2). Regrettably, and with my respects to him, but Mr Jacobs has got it wrong yet again. He claims 'in medical research, you can never prove a negative'.

If scientific method is applied to medical research, science can prove a negative in more than one way.

We can also establish that MMR vaccine causes autism, and that has been done (3). Therein lies the problem for the pro-MMR camp and why they fall back on regrettably inaccurate epidemiology which proves nothing, as Mr Jacobs so readily agrees (1):-
'There is indeed no evidence that proves there is no link between MMR and autism.'

What has been produced so far is so inaccurate that it is shameful and that is before taking into consideration all of the detailed criticisms of the papers produced, which criticisms can be read elsewhere.

Standard procedures for the assessment of adverse drug reactions are well established and applied worldwide (except by our government when it comes to the MMR Children). Government, pharmaceutical companies and the medical professional hierarchy will not accept that evidence, despite it being obtained by the application of their own well-established standard procedures and science which they also refuse to apply. 'Dechallenge' and 'Rechallenge' are significant parts of those procedures and science.

Mr Barnett will easily understand what powerful evidence of causation a 'dechallenge' case series is if he considers by analogy the famous case of R v Smith, better known to the public as 'The Brides in the Bath' case (4). R v Smith is one of the leading cases on similar fact evidence and is applicable by very close analogy to a 'dechallenge' case series. What is particularly striking is that proof by dechallenge requires only three well documented cases of positive dechallenge. What is also striking is the high standard of proof it affords. Similar fact evidence is proof beyond reasonable doubt, so this will demonstrate to the non-scientist why dechallenge is so powerful.

'Dechallenge' sounds fancy but its essence is simple if we strip away the embellishments. The drug is administered (challenge). The subject suffers a reaction. The drug is no longer administered - the subject recovers. With only three well documented cases of positive dechallenge, causation can be established. A simple search through numerous medical journals will show this is well-known in pharmacological research, but it is sometimes not so well-known to clinicians.

More interestingly, whilst dechallenge is powerful in its own right, an even more powerful proof of causation is positive well documented rechallenge. Only one case of positive rechallenge establishes causation. Rechallenge is where the drug is readministered and the subject suffers the reaction again. Rechallenge was not a feature of R v Smith. It is not a simple matter nor was it necessary, after having succeeded, to murder the same person twice.

In the ‘brides in the bath’ case, the defendant was accused of one murder but evidence was offered of two more. The admission of this evidence followed similarly from the improbability that three different women with whom he had gone through a form of marriage, and who had made financial arrangements from which he would benefit, had all drowned in the bath by accident shortly afterwards.

There was no direct evidence the accused committed the murder charged, but similar fact evidence proved it causally to the high criminal standard of proof. Similar fact evidence is evidence of such similar and unique prior behaviour of the accused, that it is proof of guilt of the accused beyond reasonable doubt. This is exactly the same with MMR where no one has seen the 'murder' being committed by MMR but whilst everyone is trying to find the murder weapon in MMR's 'hands', dechallenge and rechallenge is the scientific pharmacological equivalent of 'similar fact evidence' as proof of causation and notably, the criminal standard of proof 'beyond reasonable doubt' is analogous to the very high scientific standard. Notably, in the brides in the bath case, not unsurprisingly, there was no rechallenge (ie. 'brides in the bath' was dechallenge only) and it is not strictly dechallenge because it is impossible to 'undrown' the brides in the cases in question. However, we do know that anyone can recover from drowning if rescued and resuccitated in time, so it is not essential in this particular example - we know full recovery was possible. Further, there were three examples just like a dechallenge case series which requires only three well documented case histories to prove an adverse event was an adverse drug reaction.

'similar fact evidence' is an exception to the rule that such evidence cannot normally be adduced in a criminal trial. It is not evidence that is directed to proving the specific offence charged. It relates to other offences beyond the scope of the trial. Hence it is normally excluded as it would have a prejudicial effect that outweighs it probative value for the specific offence charged. In Mr Jacobs' example, it would not normally be relevant to prove he committed this particular burglary that he had committed 20 prior burglaries but it would most likely be prejudical to a fair trial and outcome.

So, what did happen when a rechallenge case series was presented to the authorities?

Andrew Wakefield presented a case series in 2001 to the IoM Immunisation Safety Review Committee. He is quoted (5) relating:-

'... they became anxious and agitated. They then asked for the data to be provided to their closed session the following day. .... Yet they didn’t mention the relevant parts of any of my testimony or data in the final report. .......

...... I raised this issue at the second congressional hearing on autism and vaccines in front of Marie McCormack, Chair of the Institute of Medicine’s Immunization Safety Review Committee. Representative Dan Burton told Dr. McCormack that they needed the transcript from that closed session. She explained that it was not their policy to release transcripts. He said in his own idiosyncratic style that they would then be subpoenaed, though he put it more forcefully than that. So, they were subpoenaed and the tapes were sent.

They just happened to be blank. These were copies of the original tapes, so Representative Burton said, “Okay, then in that case we will have the original tapes.” So they requested the original tapes—and they were blank as well."

As for Mr Jacobs further points, these are argument without substantiation and substance. I would suggest that others might wish to post weblinks to the commentaries which pull to pieces the epidemiology so that Mr Barnett and any other interested parties can read them and see how flawed the government position is.

Mr Jacobs had a go himself at defending the Madsen paper, when even Madsen did not bother (no doubt Madsen realising it is a completely lost cause, but not before grabbing the headlines first). Mr Jacobs' points, the sole defence of Madsen appearing the Journal of American Physicians and Surgeons late last year, did not survive criticism, sad to say. They ended up the same way as did the paper he was attempting to defend. That's statistics for you - its all a game of chance.

Email: bmj050319"replace 'at' sign"
to email replace "replace 'at' sign" with '@'


1) Adam's Reply Re: Simply a concerned parent 19 March 2005

2) Simply a concerned parent 18 March 2005

3) See text to footnote 5)

4) R v Smith, 1915, (11 Cr App R, 229)

5) P.7 ‘Autism and the MMR Vaccine - An interview with Andrew Wakefield, MD’ - Fall 2001 -


Competing interests: Close relative with life threatening food allergy