Clusterheadaches.com Message Board


    
      
        Clusterheadaches.com Message Board
        (http://www.clusterheadaches.com/cgi-bin/yabb/YaBB.cgi)
      

        New Message Board Archives >> 2005 Cluster Headache Specific Posts >> Clusters - How Common ? (FR)
        
(Message started by: floridian on Feb 10^th, 2005, 10:39am)

Title: Clusters - How Common ? (FR)
Post by floridian on Feb 10^th, 2005, 10:39am

The latest numbers from Italy suggest an overall rate of 2.8 people per thousand - more than previously believed (many put the estimate at 1 or 1.5 per thousand). The gender breakdown is 40% female vs 60% male ... again, different from what lots of people think.

Quote:

Neurology. 2005 Feb 8;64(3):469-74.

Cluster headache prevalence in the Italian general population.

Torelli P, Beghi E, Manzoni GC. Headache Center, Section of Neurology, c/o Ospedale Maggiore, Padiglione Barbieri, Via Gramsci, 14, 43100 Parma, Italy. paolatorelli@libero.it.

BACKGROUND: Prevalence of cluster headache (CH) is estimated at 56 to 69 per 100,000. OBJECTIVE: To calculate the CH lifetime prevalence in a sample representative of the Italian general population over age 14 years. METHODS: Possible CH cases according to the diagnostic criteria of the 1988 International Headache Society classification were screened from a sample of 10,071 patients (5,311 women and 4,760 men; mean age 50.4 years, SD 19.7 years) registered in the lists of seven Parma-based general practitioners (GPs), using a previously validated, specially designed, self-administered questionnaire. RESULTS: Seven thousand five hundred twenty-two subjects (74.7%; 3,971 women and 3,551 men; mean age 50.8 years, SD 19.0 years) responded to the questionnaire in their GP's office (n = 3,338; 1,885 women and 1,453 men) or at home by mail (n = 1,914; 1,030 women and 884 men) or by phone (n = 2,270; 1,056 women and 1,214 men). Of the 111 suspected cases (76 women and 35 men), 105 were seen by a neurologist and 6 were contacted on the phone. The diagnosis of CH was confirmed in 21 (9 women and 12 men), including 7 already followed at the authors' center for CH. Seventeen patients had episodic CH, and four (all men) had chronic CH. The estimated prevalence rate was 279 per 100,000 (95% CI 173 to 427), 227 per 100,000 (95% CI 104 to 431) in women, and 338 per 100,000 (95% CI 175 to 592) in men. CONCLUSION: These results point to a higher cluster headache lifetime prevalence than previous reports.

Title: Re: Clusters - How Common ? (FR)
Post by Frank_W on Feb 10^th, 2005, 10:44am

I have long suspected this myself, Floridian; that this condition is a lot more common than previously thought, but that people spend so much time being misdiagnosed. Again, this is why education needs to remain at the forefront of OUCH's mission statement. Education of patients, of doctors, of neurologists, of ENT specialists, dentists, and on down the line.

Title: Re: Clusters - How Common ? (FR)
Post by Tiannia on Feb 10^th, 2005, 11:07am

Thank F. This might really help when dealing with a new doc now for me. It is a crap shoot if i can find a doc that has ever death with HA's at all. Let alone one that has dealth with clusters.

Title: Re: Clusters - How Common ? (FR)
Post by sandie99 on Feb 10^th, 2005, 1:29pm

Thanks for sharing this article. It's always fascinating to read about new research concerning CH. One neuro said to me that CH is one of the neurological conditions that the medicine will be concentrading in future. Let's hope that's true.

Title: Re: Clusters - How Common ? (FR)
Post by BikerBob on Feb 10^th, 2005, 3:31pm

More questions about prevalence and misdiagnosis...

The 2.8 per 1000 in Italy is based on CH diagnosis being confirmed in only 21 people. How accurate can the extrapolation be?

If the estimate in the U.S. was 1-1.5 per 1000 and 2.8 per 1000 in Italy, could other factors explain the difference? i.e. double the prevalence of cigarette smokers in Italy?

Let's split the difference and assume 2 per 1000. I've read that the prevalence of migraine is 14% of the population, or 140 per 1000. This would mean that doctors who treat headache patients see 70 meegrainers for every CHer? No wonder there's so much misdiagnosis.

Bob

Title: Re: Clusters - How Common ? (FR)
Post by floridian on Feb 10^th, 2005, 4:29pm

on 02/10/05 at 15:31:45, BikerBob wrote:

More questions about prevalence and misdiagnosis...

The 2.8 per 1000 in Italy is based on CH diagnosis being confirmed in only 21 people. How accurate can the extrapolation be?

...

Let's split the difference and assume 2 per 1000. I've read that the prevalence of migraine is 14% of the population, or 140 per 1000. This would mean that doctors who treat headache patients see 70 meegrainers for every CHer? No wonder there's so much misdiagnosis.

Bob

Your analysis of the clusters vs migraine is right, and it explains alot of why misdiagnosis is so common.

Although the study identified and verified only 21 clusterheads, that is from 5252 questionaires (1914+3338). There will always be sampling error, but 5000+ data values should give a good estimate, even from a population of 58 million.

Along with the estimate of 279 / 100,000 was this in parentheses: 95% CI 173 to 427. That is the 95% confidence interval .... ie, 279 is their best estimate and they are 95% sure that the true value is somewhere between 173 and 427.

Here's a good approach to understanding how small sample sizes can be useful for very large populations:

Quote:

The best way to figure this one out is to think about it backwards. Let's say you picked a specific number of people in the United States at random. What then is the chance that the people you picked do not accurately represent the U.S. population as a whole? For example, what is the chance that the percentage of those people you picked who said their favorite color was blue does not match the percentage of people in the entire U.S. who like blue best?

The important number is not the number of people in the United States (or in Italy). It is the the number of the sample size, the number of people that you ask. If you only ask 10 people, the odds are low that you can 'know' (within reasonable approximation) what percent of the population likes blue best. If you ask 100 people, you are likely to get closer, and 1,000 will put generally put you really close to the true value, even in a nation of many millions.

Quote:

The margin of error in a sample = 1 divided by the square root of the number of people in the sample

Both quotes on statistics from: http://www.robertniles.com/stats/sample.shtml

As the number of people in a sample increases, the margin of error drops quickly at first, then very slowly.

Rare events do complicate things. If you wanted to estimate the number of people whose favorite color was 'salmon' or 'avocado' you might get zero if you ask only 10 or 100 people or even 1000 people. That estimate of zero would be close to the truth, but maybe not accurate enough for your purposes - in a nation of many millions, 0.01% would translate to more people than my neighborhood could hold. But I think with 5000+ questionaires and 21 confirmed occurences, they got a decent estimate.

Title: Re: Clusters - How Common ? (FR)
Post by debs on Feb 10^th, 2005, 8:58pm

Thanks floridian, for keeping us up on the new literature. I really appreciate seeing this stuff on the MB, at this point in my cycle, I can't search it out myself.... so thanks again for taking the time for us! :D

floridian wrote:

Quote:

Another aspect of this is representative sampling --as the authors of the paper are trying to obtain... If one asks enough people, the chance of getting close to the right answer is good as floridian says. But, this also depends on the question being asked and of whom.

For instance, if 1 million Texans were asked who they planned to vote for president last fall, the result would be much different than the results of 1 million people from Massachusetts, and again different than the results of 1 million from both states combined.

So, I have been curious how this applies to CH (like floridian says, a rare event). Is the sample population biased towards those who go to the doc for headaches? Towards people who who have headaches who bother to fill out surveys? Do most people with CH go to the doctor? (probably...) Do people in Parma have more headaches cause they eat lots of Parmesan cheese?

I don't know, but I agree with you all, I have a feeling the prevelance of CH is underestimated in general, but how to get to a better estimate? It's a tough problem. And yes, I too think the methods in this study seem to be getting towards a decent estimate.

And as Frank_W says --education and outreach is key. There are probably lots of people out there suffering, that don't need to be. Something that has been in the back of my mind...

Just a few random thoughts... happy Thursday and PF nite to ya'll,

Debbie

Title: Re: Clusters - How Common ? (FR)
Post by Pinkfloyd on Feb 11^th, 2005, 2:35am

Looks more like 2.08 per thousand to me. 21 out of the 10,071 they started with.

I'd think the 3000 people that didn't fill out the cluster questionaire, would have, had they been offered the chance, if they had clusters.
If they couldn't find them to ask, they aren't seeing their doctors and probably don't have clusters.

Add to that, the percentage of people that don't see GPs and the percentage goes down further.

OTOH, I'm sure some cluster sufferers in Italy see doctors other than GPs for their treatments and wouldn't show up on the lists. That would skew it higher.

Just don't think working off the current patient lists of GPs necessarily gives them a "general population."

just my opinion.
Bobw

Title: Re: Clusters - How Common ? (FR)
Post by floridian on Feb 11^th, 2005, 8:05am

Debs and PinkFloyd are right - this wasn't a pure representative (random) sampling. The method that used to gather people probably has some effect on the results. I don't know if the people who declined to participate were asked to be part of a health study or a headache study - even the way that initial question is worded could bias the results one way or the other.

Title: Re: Clusters - How Common ? (FR)
Post by Mr. Happy on Feb 11^th, 2005, 1:01pm

If I remember my statistics class back in `71, a group had to include a minimum of 25 random samples to be considered valid. I got a "D."

Whether _completely_ accurate or not, the male:female thing sticks out. Shot this extract to my Wingate virgin, since she started out with "In my research I found that the majority of cluster headache sufferers are males."

Thanks Floridian,
RJ