Epilepsy- cost of illness study

[hereticmnk]

Hello,
I am a graduate student in a health econ program. I have always had an interest in epilepsy. For my first paper, I decided to do a simple cost of illness study for epilepsy. I am not going to write out my entire research proposal here because no one is going to read it but I will try and go over the highlights.

Direct cost of epilepsy using claims data
- Stratifying cost based on age
- Stratifying cost based on epilepsy type
- Stratifying cost based on inpatient, outpatient, ED visit, drug use etc.
(To my knowledge, no one has stratified cost like this and may highlight some policy implications for a payer)
I plan to play with it a bit more and try and map the progression of disease as well.

I know this might seem like a semi garbage study but I have to start somewhere. As a clinician, what additional kind of information would you be interested in? it does not have to be costs. I have a large dataset to play with. Any opinions would be greatly appreciated
thanks!

---

Members don't see this ad.

 

[hereticmnk]

No one?

 

[HarveyCushing]

Do you have outcome(s) data of some sort? Also do you have data regarding which specific AEDs the patients are currently on and what they might have been on in the past?

 

[hereticmnk]

Do you have outcome(s) data of some sort? Also do you have data regarding which specific AEDs the patients are currently on and what they might have been on in the past?

Got everything
Data from 2000-2009. Originally procured by research advisor from Medicaid (all patients in one state, all pt. info included if pt. had one dx code of epi or one drug, criteria for inclusion in the study is obviously going to be much more restrictive). It was to see if AED's had any correlation with depression/suicidal ideation.Its a claims dataset so it includes all outcomes (billable ones, therefore clinically relevant outcomes captured) , drugs, hospitalization, ED visits etc. Advisor is done with that study and the data is just lying around for me to play with.

 

[lgroomes]

I realize I haven't started med school yet so take this with a grain of salt, but:
What about stratification based on how controlled a person's epilepsy is (if you have those data)? Epilepsy wouldn't cost much more than medication and dr. visits when under control, but for a drug resistant patient that means lots of tests, no driving, more potential ER visits, potential surgery, etc.

That's an awesome idea though!

 

[hereticmnk]

I realize I haven't started med school yet so take this with a grain of salt, but:
What about stratification based on how controlled a person's epilepsy is (if you have those data)? Epilepsy wouldn't cost much more than medication and dr. visits when under control, but for a drug resistant patient that means lots of tests, no driving, more potential ER visits, potential surgery, etc.

That's an awesome idea though!

Thanks for the input. I think we already have a pretty good idea that uncontrolled epilepsy is going to cost much more than its controlled counterpart. . There are ways to easily identify pharmacoresistant epilepsy using diagnosis codes (ends with 1 instead of 0). It is mostly a descriptive study but I would like to have some policy implications, which is my main interest. Thanks again

 

[tranzformer]

Sounds like you need an epileptologist on your project/manuscript to guide you in asking the best questions possible with your data set. No shame in that. Better than asking SDN.

 

[hereticmnk]

Sounds like you need an epileptologist on your project/manuscript to guide you in asking the best questions possible with your data set. No shame in that. Better than asking SDN.

I already am collaborating with an epi person . I have a bunch of outcomes data covering 200,000 epileptics that I can play with to answer any outcomes question anyone could have. I am also working on an a peds outcome study once that I will focus on once I finish this cost study. I was hoping to answer a burning question any specialist may have regarding this with my study including specialist utilization amongst epileptic patients, outcome of patients seeing a specialist vs. PCP for pharmacoresistant epilepsy (controlling for confounders using propensity score matching) etc. etc. I already know what I am looking for cost wise in this study, I was just looking for the proverbial cherry on top analysis that a clinician would also like to see in the study

 

[hereticmnk]

I realize I haven't started med school yet so take this with a grain of salt, but:
What about stratification based on how controlled a person's epilepsy is (if you have those data)? Epilepsy wouldn't cost much more than medication and dr. visits when under control, but for a drug resistant patient that means lots of tests, no driving, more potential ER visits, potential surgery, etc.

That's an awesome idea though!

And Boom....
http://www.ncbi.nlm.nih.gov/pubmed/23077014
Someone beat me to it
It seems that your idea was good enough to get published in Neurology which aint all that bad.
Thankfully, I have more ideas to play around with
(I have some problems with their methodologies since it seems they did not validate their method and also their way to measure uncontrolled epilepsy is a bit suspect)