A new study used the World health Organization’s global database of reports of adverse events associated with the use of a medicine (VigiBase), including vaccines, to look at reporting patterns after HPV vaccine.
The cases in VigiBase are collected for 125 member states, including NZ, who participate in the WHOs International Drug Monitoring Programme. The data from all the member states are pooled together. The pooled database is extensive (it is really bigly 14.468 million reports as of yesterday, although the paper mentions that in 2010 it was 10.3 million.
The ‘cases’ in the database are reports of events that have occurred either in association with, or following, the use or administration of a medicine. These cases may or may not have been caused by the medicine.
What did this study do?
The authors wanted to test a novel computer algorithm to see if they could group cases reporting similar symptoms such as dizziness or skin rash. They used HPV vaccine reports from around the world to do this (VigiBase). This is potentially a very cool tool.
It is important to note that the cases the publication based its research on are derived from data extracted from reports from member states. These reports are identified using a dictionary of possible events terms from case summaries submitted. The level of clinical detail available for each report is not necessarily always comprehensive. In other words there is a certain degree of garbage in garbage out because the data going in is highly variable (imagine all the different doctors, nurses and members of the public from all the different countries filling in the report forms, in different languages!). Making a diagnosis based on this data is near impossible. The authors note this themselves
The greatest limitation of this study is the lack of information on many of the reports in VigiBase. Missing data means that it may be difficult to make a clinical judgment regarding case relevancy.
What did the study find?
The study found that the algorithm could indeed place symptoms into clusters. Unfortunately it is not clear from the paper how successful the methodology was at separating different conditions into different clusters.
Did the study find that HPV vaccine causes CRPS or POTS?
No. There are three reasons the study did not determine this. One is because that was not the research question being asked in the first place; two, the study was not designed to answer that question; and three, it is an accepted fact that databases such as VigiBase cannot be used in this way (causality assessment).
The authors themselves state in their conclusion that they cannot use this analysis to draw any conclusions regarding a causal relationship between the HPV vaccine and reported adverse events.
A causal association with the HPV vaccine remains uncertain
It is common to identify safety signals with vaccines, most of them are dismissed after further investigation. This signal has already been investigated further in NZ and EU and no evidence has been found to show that HPV vaccines cause this syndromes [click here for the NZ Medicines Adverse Reaction Committee minutes].
But the study found reports of CRPS and POTS.
Yes, the study identified these conditions. However, it is also essential to note:
- These conditions occur every year in every country, particularly in adolescent girls, so some cases will inevitably occur after vaccination by chance and find their way to this reporting system. This is especially relevant where there is high uptake of the vaccine in this population group.
- Interestingly, Japan and Denmark have contributed a significant number of reports. These countries have also experienced extensive media coverage and lobbying on the topic of these specific conditions. Even more interesting is that when cases from Japan and Denmark are removed from the analysis for serious headache+dizziness+fatigue or syncope the significance disappears (this result is found in the supplementary material).
- Anti-vaccination groups have been actively collecting cases of these conditions. I am making an assumption these will be reported to the adverse event reporting system.
- There is no description of the timing of onset of these cases from the point of vaccination. If there is a biologically plausible causal association between vaccine and onset of symptoms there should be a consistent time between vaccination and development of symptoms. This was not addressed in this study which is another reason it cannot be used to infer causality.
This study describes a novel methodology in the earliest stages of development that has great potential benefit. However, as yet its true ability to identify meaningful clusters to inform further research has not been established. At best the authors have identified clusters of symptom constellations observed in association with HPV vaccination. However, as with other publications of case reports of related symptoms, further and more in-depth scientific investigation of these associations (signals?) is required to establish whether these associations are causal and if so, the mechanisms by which they occur.
I think these are the main messages about this study. It did not find that HPV vaccine causes anything. Also, in my opinion, the authors have extended their claims about the study and made statements that are unsupported by their methods and results. I think they should have left it as a methods paper and considered a title that better reflects the study.
Additional links on the safety of Gardasil from NZs Medicines Safety Authority, MedSafe:
- See here for MedSafe info for consumers QA’s on Gardasil9
- See here for Prescriber Update on Gardasil9