Re: [Aeo-devel] Notes from the AEO/AERO call

[Ryan B. <rbr...@bc...>]

On 2011-09-04, at 4:51 PM, He, Yongqun wrote:

> Hi Melanie,
> Good notes! We had a very good meeting last Friday. I am glad that you agree the global term adverse event belongs to AEO and AERO will use the AEO term. 
> 
> As discussed, I have removed the may or may not sentence out of the AE definition. I have also added a note on the work to be done in the scope, notably regarding temporal association. 
> 
> I have updated the AEO definition of 'adverse event' to this:
> "a pathological bodily process that occurs after a medical intervention and is possibly induced by the medical intervention."
> One note: I still keep the second half of the definition (" is possibly induced by the medical intervention") for some more discussion. 

"possibly induced" is the same as "may or may not be induced" and for the same reason as above probably wouldn't this belong in the editor's note? 

> My reasons: 
> (1) I feel that our ultimate goal (or the goal in clinics) of study adverse events is to assess if the adverse event outcome is due to the medical intervention. Of course, there is different level of association. I would argue that all adverse events reported in VAERS are possibly associated with the medical intervention. This is why the data are collected, and assessment of the association needs to be done.  
> (2) Also, by including the second half of definition, we do not have to include any time frame in the definition any more. If the time is too long (e.g., one year) for one flu vaccine for a symptom (e.g, rash), we would consider that it is impossible (or very very low chance) to be related to the vaccination. Since it is impossible, according to the definition, we will not need to study it. For some new medical invention, since we don't know how long the incubation time would be, we will then need to monitor longer time.
> (3) Broadly, the definition also covers those AEs that were eventually found to have not association with the intervention. It this case, the probability is 0. 
> 
> In this sense, I generated a term 'causal adverse event probability' under 'information content entity'. I think for each adverse event, it can have some superclass:
> has_output some 'causal adverse event probability'.
> 
> under the term  'causal adverse event probability', we have two subclasses already:
> Naranjo ADR probability score
> DIPS score
> these two scores have been frequently used in drug adverse event probability assessment.
> 
> Any suggestions and comments are welcome. 
> Thanks!
> Oliver  
> 
> ________________________________________
> From: Melanie Courtot [mco...@gm...]
> Sent: Friday, September 02, 2011 7:14 PM
> To: He, Yongqun; Ryan Brinkman; Alan Ruttenberg; si...@um...
> Subject: Notes from the AEO/AERO call
> 
> A few notes from our call:
> 
> - global term adverse event belongs to AEO. Current definition to be updated to "a pathological bodily process that occurs after a medical intervention" with in an editor note "An adverse event can be caused by the medical intervention, or may not be caused by the medical intervention." (having may or may not in the definition logically cancels itself) An additional note will be added to indicate that some work is needed on how to restrict the scope of this term, notably regarding temporal association
> 
> - AERO will include "recorded AE" (such as those from randomized clinical trials, could be anything eg pregnancy etc), defines as "an AE that has some information entity about it recorded in an EHR"
> 
> - AERO will include "reported AE" to represent those recorded AEs that are selected and evaluated against guidelines to determine if they should be reported higher-up (such as in VAERS)
> 
> - those selection and evaluation processes will be added to AERO (including assessment of "serious AE" etc) as those are done against specific guidelines
> 
> - AERO will provide the framework to represent those against the guidelines, and will provide documentation to address Oliver's needs
> 
> - Melanie will clarify with her clinicians if "recorded AE" should be a subclass of AE or instead should be "recorded event" and be sibling (i.e., are all recorded events adverse events?) and will update above definitions as needed
> 
> - Oliver will produce a list of the types of statements AEO aims to make
> 
> Cheers,
> Melanie
> **********************************************************
> Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues 
>