The Debunking of 'Table 14'
Insanity Check: Discrediting Bogus Irish Government HPV Vaccine "Expert Group" Report.

HIQA
The Health Information and Quality Authority (HIQA) is an independent authority (established in May 2007) reporting to the Irish Minister of Health, "to inform and assist decision-making", and provide a "stamp of assurance for the public and the taxpayers' that the highest possible standards.. are adopted" . It is the statutory organisation in Ireland with the remit to carry out national health technology assessments. Its CEO reportedly earns €200k a year (2008) . 
In 4th June 2008, the HIQA published the Health Technology Assessment on "The role of HPV in reducing the Risk of Cervical Cancer in Ireland".  The purpose of this assessment was to establish the cost-effectiveness of a combined national HPV vaccination and cervical cancer screening programme compared to a cervical cancer screening programme alone. The Authority had commissioned the National Centre for Pharmacoeconomics(NCPE) to conduct the health technology assessment on its behalf. To lead and oversee the process and advise the Authority, a multidisciplinary Expert Advisory Group was convened.  You can view the report here.

"Start, Clear Advice - 52 Deaths will be averted (per year)"
The controversial politician Dr James Reilly is most known for his outspoken advocacy in favor of mass HPV vaccination among Irish schoolgirls.  The basis for his repeated claims on the death-prevention power of the HPV vaccine is the HIQA report, as he describes here himself:
".. a document that I have quoted from numerous times myself, and that is, the Health Information and Quality Authorities Assessment of the Cost Benefit Analysis  of proceeding with this vaccine. It is worth repeating this is the Government's own Health Information and Quality Authority and they say very clearly that:  Out of 93 deaths per year - 52 Deaths will be averted;  How can you ignore this stark clear advice from the Health Information Quality Authority ..." 

What does the Report say?..
We decided to check out the validity of his claims. Sure enough we found the section in the HIQA Report that is the source of these figures which are then illustrated in table 14:
Results, Section 4.3.2, p.44.
Cost-effectiveness of annual vaccination of all 12 year old girls without a catch-up programme
In the base case model, an annual average of 111 cases (56%) of cervical cancer and 52 deaths (56%) related to cervical cancer were averted, as a result of routine HPV vaccination of all 12 year old girls (Table 14).   Approximately, 30% of CIN 1 and 40% of CIN 2 cases were prevented.
In the model we assumed that 74% of cervical cancer, 50% of CIN 2/3 and 35% of CIN 1 are caused by HPV types 16 and 18. Furthermore, we assumed 95% vaccine efficacy, 80% vaccination coverage and a herd immunity effect.

 
The now infamous "Table 14".

But for anyone familiar with the expected impact of a HPV vaccine when added to an existing screening program - these figures are clearly impossible.  The screening program should be preventing 75-80% of cancer incidence with the vaccine then left to target the remaining 'unscreened' (i.e 20% of the female population). Using the same set of statistics and assumptions that HIQA base their 'model' on, we were able to quickly calculate a decidedly different set of results.  (See here for a breakdown of our calculations).  So, how did the HIQA come up with these results? 

Sanity Check
A useful sanity check is to do a comparison with the equivalent report from another similar sized european country.  If we examine the HPV Vaccine Technology Assessment report from Austria we see results for the following assumptions (on which they based their model): 
- both screening and vaccination occuring
- a booster after 10 years
- model calculated under best case assumptions 
They report  "Compared to screening only, screening plus vaccination of 12-year-old girls (and boys) would result in a median reduction of 10% (15% if boys included) fewer new cancer cases and 13% (20% if boys included) fewer cervical cancer deaths under best case assumptions over 52 years in the overall female population".

So these are the ballpark figures we should be expecting. Yet, according to the HIQA "results", hpv vaccination in Ireland is almost 6 times more effective than in Austria at reducing cancer cases - and that's without using a booster! (The Austrians thought it necessary to include a booster every 10 years in their model to show any kind of vaccine effect).

How do they do that?
The results in the Irish Report are clearly misrepresented to exaggerate the effectiveness of the vaccination.
The first point to make regarding this deception is that the figures presented as 'results' are entirely meaningless without additional qualifying information. They are presented as "an annual average of.." without saying which year they apply to, or whether it is a median average over the time horizon of the study (70 years).              (Again, refer to Austrian report to see how they properly qualify their results).
It seems as though they are selecting an average for one particular year which shows the geatest vaccine effectiveness - probably the last year of the model i.e. 70 years after vaccine 'intervention' (around 2078).  Given the medical advances expected to occur between now and the time the world is a dozen years short of the 22nd century, they probably decided to omit identifying the year to avoid accusations of redundancy and irrelevance (for a report based on 2004 statistics).  

The next deception is the omission of the existence of a screening program in the presented results.  The 'deaths-prevented' figure they are presenting can only be as a result of both screening plus vaccination but they imply it is the result of vaccination only: "..as a result of routine HPV vaccination of all 12 year old girls"  and in the sentence that follows this, they also deliberately omit to include screening in the set of assumptions  "..we assumed 95% vaccine efficacy, 80% vaccination coverage and a herd immunity effect". 

How am I so sure that these figures are for screening plus vaccination when they are clearly presented as the effect of vaccine only?
In reality, there will never be a future situation where cervical cancer screening of the general female population does not occur.  Therefore noone would seriously consider modelling for a scenario with vaccination only.  The only possible scenarios of relevence are 'screening alone' or 'screening plus vaccination'.  So I am pretty sure that the 'results' (including 'Table 14') which are represented as 'vaccination only' are actually the figures which result from the combined effect of 'screening plus vaccination'.  
In fact, this is confirmed by the "Foreword" on the first page of the report:  
"The purpose of this assessment was to establish the cost-effectiveness of a combined national HPV vaccination and cervical cancer screening programme compared to a cervical cancer screening programme alone.."
And this is indeed the scenario they use in the base case model: 
Section (4.2.3): 'Data Inputs' :
Vaccine coverage of 80% was included in the base case analysis...In all cases it was assumed that the vaccine would be combined with a screening programme, the aim of which is to cover 80% of the population aged 25 to 60 years.
with accompanying Table 12 showing:
Summary of key parameters included in the base case:
80% screened every 3 yrs, 25-44 yrs
80% screened every 5 yrs, 45-60 yrs.

But if you don't read all these earlier sections and just skip through to the 'Results' section (as seemingly every doctor, journalist and politician who have seen this report have done), then you will be taken in by the conology.
Thus can the misguided Dr James Reilly write that "The Health Information and Quality Authority has stated unequivocally that this vaccine would save 52 lives and prevent 111 cancers annually".    
 
As a further confirmation, I found this Irish independent report  which included the following line:
"The Health Information and Quality Authority estimated that if the vaccine was introduced, along with the national screening programme, 52 deaths would be prevented".
So we have the Irish Independent quoting the HIQA saying that the 52-deaths-prevented figure relates to a vaccine plus screening scenario, so we know that at least someone in HIQA is aware that Table 14 is misleading.

More Bias?
Apart from the crude attempts at obfuscating the results (detailed above), there are a few other examples of bias in attempting to get the numbers to come out 'right'. Note the excessive time horizon of 70 years chosen for the model (compared to 52 years for the Austrian report).  Again, this was taken in order to try and get some 'respectable' numbers on the vaccine impact. To illustrate this point, if the HIQA had taken a 43 year horizon for their model, and with the average age of mortality of 56 for girls getting vaccinated at 12, then no girls would statistically be old enough to die of cervical cancer yet (on average).  So the impact of the vaccine would be nil.   
Other examples of bias in favor of getting the 'right' result is the selective use of statistics. For example, even though the report was published in 2008, they chose to use cervical cancer mortality statistics from 2004 (there was a spike in the figures that year at 93 deaths).  Presumably the more up-to-date 2005 mortality figure of 73 deaths  was not deemed high enough. 

Lifelong Protection?
The third point that undermines the report is the assumption of lifelong protection from the vaccine even though they themselves admit in the report that there is no data to justify this assumption.
Section 3.6 Duration of protection of vaccines
Clinical trials of both vaccines have evaluated long-term efficacy against HPV infection to a maximum of five years. Long-term immunity beyond this is unknown and therefore, it is not yet clear whether
booster doses of the vaccine will be required...In this economic evaluation, lifelong protection from the three-dose course of the HPV vaccine is assumed in the base case analysis.
There is evidence that protection cannot be expected to last beyond 10 years. Dr Diane M. Harper, professor at Dartmouth Medical School who led clinical trials of the Gardasil HPV vaccine writes in this 2007 article that:
"We know that Gardasil is an alum-based vaccine; we assume its efficacy will last for about 10 years".
The model in the Austrian report was designed so that a  booster vaccination was implemented after 10 years in the base case. The HIQA choose a highly speculative hypothesis of lifelong protection in order to better justify mass vaccination.  This assumption more than any other negates the usefulness of the report. 
If the protection from the vaccine disappears after 10 years, mass vaccination of 12 year olds would obviously be futile - protection would have worn off decades before the girls can be expected to be at risk from cancer (average age of diagnosis is 44).

The vaccination costing detailed in the report is as follows:  
The cost of a three-dose vaccine schedule (€100 per dose), including an administration fee of €30 per dose, would be approximately €9.73 million per year for a cohort of 12 year old girls with a vaccine
coverage of 80% (Table 15). These costs will recur every year.
So using the HIQA's own figures, funding the mass vaccination requires a total wealth transfer from the Irish taxpayer to some chosen pharmacutical company of €680million (9.73m*70) over the course of the model 'time horizon'.  However, without boosters, all this funding will probably be completely wasted expenditure in terms of preventing cervical cancer...unless one is ready to add tens of millions more to cover this additional cost of boosters. 

In any case, we at ComeLook.Org have produced our own independent calculation of the effect annual mass vaccination of 12 year olds has on cervical cancer deaths, where screening also takes place , and accepting (for illustration purposes) the assumption of lifelong immunity.  We can reveal that in the year 2057, the first death to be prevented by the effect of the vaccine alone, can be expected. Then every 5 years after that one death will be prevented up to the model time horizon of 70 years.  So instead of 52 lives being saved annually, we have 5 lives saved over the course of 70 years at a cost of € 680million.  The method of calculation can be reviewed here.

Vaccine Safety 
This same HIQA report also has a brief section on Vaccine Safety which, although trying to downplay the risks, still does not make reassuring reading (noteGuillian-Barre Syndrome is a form of paralysis, mostly temporary but may be permanent):
"Systemic adverse events such as headache, fatigue, gastrointestinal upset and rash occurred in 69% to 86% of recipients and were only partially causally related.   A possible association between Gardasil and Guillian-Barre Syndrome is being investigated by the US VAERS. In the US, three deaths were closely related in time to immunisation with Gardasil. No causal relationship was established between the deaths of the young women and the administration of Gardasil. On the 24th January, 2008, the European Medicines Agency (EMEA) issued a statement relating to the safety of Gardasil following reports of sudden, unexpected deaths in two women who had previously received GardasilTM 92. The cases in question occurred in Austria and Germany. In both cases, the cause of death could not be identified".
For a breakdown on the latest number of deaths and serious adverse reactions reported in the US in relation to Gardasil (the vaccine chosen for the Irish mass HPV vaccination campaign  see here).

Summary
There is evidence of bias running all through the report, whether it be selectively using statistics that promote the vaccination agenda (eg 2004 mortality rate with it's unusually high spike) or 'misinformation by omission' as in the presentation of the results.   Through speculation, misdirection, equivocation, obfuscation and contrived ambiguity, this HIQA report manages to demonstrate spectacular life-saving potential for a HPV mass vaccination program when in fact it is highly unlikely that the mass vaccination of 12 year olds (as presented in the reports base case model) will actually prevent a single cervical cancer mortality. 
It is also worth noting that curiously, noone from the National Centre for Pharmacoeconomics (NCPE) put their name to this report, even though this was the body that actually carried out the assessment (being commissioned on behalf of the HIQA to do so).  At least one of the NCPE chief researchers apparently believe their own propaganda though, and was seen to join a crusading Facebook Group with the title "Harney must reinstate cervical cancer vaccine" which organised candlelit vigils in Dublin and Cork in late 2008.

These bogus figures formed the basis for the subsequent charade that played out in the Dail (Irish parliament) debates on the mass vaccination campaign.   This report was the fuel that powered the politically motivated misinformation campaign that ultimately forced the Government to reinstate the mass vaccination program after it had been initially cancelled due to government cutbacks.  I am not aware of any members of the HIQA who spoke out to offer a correction during the 6 months of public discussion on the issue. 
 
Last month the HIQA hosted the 7th Annual Meeting of the Health Technology Assessment International (HTAi) conference in Dublin in partnership with 'Platinum sponsor' (€ 25k) GlaxoSmithKlein - one of only two companies currently marketing a HPV vaccine.

Contact
justin@comelook.org
.

---