Anonymous Survey Monkey

[Mary M]

Would it be useful to know CAmKII #?

[DaleS]

Here are the other questions I've heard from parents :

 

1) Overlay with Lyme incidence reports

2) Education level of mom [hey, what about Dads :-) ] -- came from Dr. K's web site

3) Likely correlation with incidence of acute rheumatic fever -- can we get that from CDC?

4) Check for confirmation of GABHS infection at time of onset/exacerbation?

5) Check for confirmation of mycoplasma pneumonia at time of onset / exacerbation?

6) Record age of onset

7) Record current age

8) Record age when parents heard about PANDAS

9) How did parents hear about PANDAS

10) Overlay with GABHS emm-type frequency

 

Are there other things -- I might not run the survey, but perhaps we can get one of the researchers to.

 

At the moment, the graph does seem correlated with population density -- lots of reasons why that might be true.

 

Buster,

Do the answers you get from this survey allow you to run correlations between the different questions? For example, it might be useful to know the success rate of IVIG (from original question #3 – IVIG put into remission) and overlay with Lyme. Or, for example, IVIG success rate and age of onset (and age of IVIG vs. onset). There may be hundreds of combinations of these questions and it may take a statistician to analyze (are there any smart statistician mothers out there?). Just knowing the % that of confirmation via IVIG, ABX, etc. may not be as useful as knowing the % with, say for example, early treatment. So my original question, stated differently, is: are you able to match the answers of all the questions to each respondent so that different correlations can be made or do you only know the % of each answer to the individual questions?

 

FYI, I have been trained in statistical methods of problem solving (specifically six sigma, Taguchi methods of design of experiments) which are very useful in finding correlation and confounding causes with small size populations (experiments). To get the most out of this type of a study, it would be ideal to make the question/answers as specific as possible and limit the answers to 3 (not including don't know).

Edited July 6, 2010 by DaleS

[Fixit]

Here are the other questions I've heard from parents :

 

1) Overlay with Lyme incidence reports

2) Education level of mom [hey, what about Dads :-) ] -- came from Dr. K's web site

3) Likely correlation with incidence of acute rheumatic fever -- can we get that from CDC?

4) Check for confirmation of GABHS infection at time of onset/exacerbation?

5) Check for confirmation of mycoplasma pneumonia at time of onset / exacerbation?

6) Record age of onset

7) Record current age

8) Record age when parents heard about PANDAS

9) How did parents hear about PANDAS

10) Overlay with GABHS emm-type frequency

 

Are there other things -- I might not run the survey, but perhaps we can get one of the researchers to.

 

At the moment, the graph does seem correlated with population density -- lots of reasons why that might be true.

 

Buster,

Do the answers you get from this survey allow you to run correlations between the different questions? For example, it might be useful to know the success rate of IVIG (from original question #3 – IVIG put into remission) and overlay with Lyme. Or, for example, IVIG success rate and age of onset (and age of IVIG vs. onset). There may be hundreds of combinations of these questions and it may take a statistician to analyze (are there any smart statistician mothers out there?). Just knowing the % that of confirmation via IVIG, ABX, etc. may not be as useful as knowing the % with, say for example, early treatment. So my original question, stated differently, is: are you able to match the answers of all the questions to each respondent so that different correlations can be made or do you only know the % of each answer to the individual questions?

[Fixit]

Here are the other questions I've heard from parents :

 

1) Overlay with Lyme incidence reports

2) Education level of mom [hey, what about Dads :-) ] -- came from Dr. K's web site

3) Likely correlation with incidence of acute rheumatic fever -- can we get that from CDC?

4) Check for confirmation of GABHS infection at time of onset/exacerbation?

5) Check for confirmation of mycoplasma pneumonia at time of onset / exacerbation?

6) Record age of onset

7) Record current age

8) Record age when parents heard about PANDAS

9) How did parents hear about PANDAS

10) Overlay with GABHS emm-type frequency

 

Are there other things -- I might not run the survey, but perhaps we can get one of the researchers to.

 

At the moment, the graph does seem correlated with population density -- lots of reasons why that might be true.

 

Buster,

Do the answers you get from this survey allow you to run correlations between the different questions? For example, it might be useful to know the success rate of IVIG (from original question #3 – IVIG put into remission) and overlay with Lyme. Or, for example, IVIG success rate and age of onset (and age of IVIG vs. onset). There may be hundreds of combinations of these questions and it may take a statistician to analyze (are there any smart statistician mothers out there?). Just knowing the % that of confirmation via IVIG, ABX, etc. may not be as useful as knowing the % with, say for example, early treatment. So my original question, stated differently, is: are you able to match the answers of all the questions to each respondent so that different correlations can be made or do you only know the % of each answer to the individual questions?

 

 

I was actually did quite well with statistics in college...but we won't talk about how long ago that was...i do enjoy numbers....

don't know if you'd want my help...i would have to refresh myself......i'll throw it out there...

[JAG10]

Suspected year of onset by area/region in addition to age.

[norcalmom]

Buster -

for what it is worth, we go back east to Rhode Island for 2-4 weeks per year. I am a dot in Northern Cal.

[Buster]

I'd have to rerun the survey - which is why I'm asking here what questions might be helpful. It also might be better to run the survey under a more secured manner - Too much correlation sort of crosses the line of privacy so I stuck with the weakest data (i.e. zip) at the moment.

 

 

Here are the other questions I've heard from parents :

 

1) Overlay with Lyme incidence reports

2) Education level of mom [hey, what about Dads :-) ] -- came from Dr. K's web site

3) Likely correlation with incidence of acute rheumatic fever -- can we get that from CDC?

4) Check for confirmation of GABHS infection at time of onset/exacerbation?

5) Check for confirmation of mycoplasma pneumonia at time of onset / exacerbation?

6) Record age of onset

7) Record current age

8) Record age when parents heard about PANDAS

9) How did parents hear about PANDAS

10) Overlay with GABHS emm-type frequency

 

Are there other things -- I might not run the survey, but perhaps we can get one of the researchers to.

 

At the moment, the graph does seem correlated with population density -- lots of reasons why that might be true.

 

Buster,

Do the answers you get from this survey allow you to run correlations between the different questions? For example, it might be useful to know the success rate of IVIG (from original question #3 – IVIG put into remission) and overlay with Lyme. Or, for example, IVIG success rate and age of onset (and age of IVIG vs. onset). There may be hundreds of combinations of these questions and it may take a statistician to analyze (are there any smart statistician mothers out there?). Just knowing the % that of confirmation via IVIG, ABX, etc. may not be as useful as knowing the % with, say for example, early treatment. So my original question, stated differently, is: are you able to match the answers of all the questions to each respondent so that different correlations can be made or do you only know the % of each answer to the individual questions?

 

FYI, I have been trained in statistical methods of problem solving (specifically six sigma, Taguchi methods of design of experiments) which are very useful in finding correlation and confounding causes with small size populations (experiments). To get the most out of this type of a study, it would be ideal to make the question/answers as specific as possible and limit the answers to 3 (not including don't know).

[Mary M]

I know I'd be willing to forego my privacy (and my family's) for some of these answers.

 

I know the "institution" requires privacy but if you could see what I look like today, bruised from head to toe, you'd understand why privacy is of little solace to me. Answers are the key now.

[MomWithOCDSon]

I know I'd be willing to forego my privacy (and my family's) for some of these answers.

 

I know the "institution" requires privacy but if you could see what I look like today, bruised from head to toe, you'd understand why privacy is of little solace to me. Answers are the key now.

 

Buster -

 

I'm sure I could find a copy of the HIPA privacy language that appears with all our insurance claims paperwork; you could just put that in small print at the bottom of your next survey!

 

More seriously though, since we're all participating voluntarily, is privacy a huge concern?

 

And maybe it's just me, but I tend to suspect that when someone makes a law to "protect" you from something (as HIPA is supposed to protect you from having personal information released without your knowledge), chances are that "something" has already been violated, and is fully vulnerable to being violated again in the future, and somebody is just trying to cover themselves from being held accountable.

[Buster]

Hi, I updated the map with latest data.

 

We're now at 163 results. I'll probably close down the survey soon and try to construct a better one -- I was quite impressed that this many responses came in over a weekend.

 

Buster

[becjonz]

Is there a way to find info about incidence of Strep and/or MycoP per state over say even the last 5 years? I'm just curious if our results correlate in regards to states with higher infection rates or conversely those states where there aren't any pinpointed PANDAS cases. (yes, I know we are simply going by who is on this board etc but the cluster of states without is rather interesting to me.

[Mary M]

Hi, I updated the map with latest data.

 

We're now at 163 results. I'll probably close down the survey soon and try to construct a better one -- I was quite impressed that this many responses came in over a weekend.

 

Buster

 

 

Impressive total especially since it was a holiday week end...

 

oh, but alas perhaps PANDAS families don't get a holiday week end?!

[momto2pandas]

Actually, we are working on a big comprehensive survey - will let you all know when it's out there. It's already drafted, actually, and may be implemented using a sophisticated platform commonly in use in the pharma industry. I want to follow HIPAA in implementation and get IRB approval, so that we can publish the results, if called for. There are a bunch of people (professionals) involved. I will post more as it becomes more developed.

[momto2pandas]

Re. HIPAA, my understanding (and I'm not exactly an expert, so take it with a grain of salt) is that you can collect whatever you want with proper consent, but what you can do with it - in particular, sharing it - will depend on whether it was collected in a HIPAA-compliant manner. For example, my understanding is that the experts will not read Diana Pohlman's collection of cases because it would violate HIPAA rules for them to be shared (don't quote me on this). That is why we are trying to develop a sophisticated, HIPAA-compliant method for organizing and analyzing all of the self-reported information that I'm sure we can get. We've actually gotten pretty far!

 

In the mean-time, Buster, if you do want any help analyzing anything you're collecting, I've been analyzing survey data for 6 years so could probably help. I also have the statistical software to do anything - as long as I could get data in Excel.

[Buster]

I've gotten a 175 responses to this survey. I think I'll shut down the survey monkey in another week. The map is interesting. I've plotted relative to population density and it correlates (which is to be expected). That could be because:

• with enough samples, someone in the town is likely to have participated in the survey
  
• with enough people, strep gets to the 4-6% that has such a predisposition
  
• strep (and other triggers) tends to spread in dense areas of populations
  
• ...
  

 

Buster

 

 

Hi folks,

 

Please take a moment to answer 4 questions on this survey monkey. My goal is to try to plot a map.

 

http://www.surveymonkey.com/s/V96RVXT

 

This should take < 1 minute to complete.

 

Please use zip code where you were when your child first exhibited symptoms. If you aren't in the US, please provide city/country.

 

This isn't yet automated -- so refresh won't be immediate

 

I'll see if there is a way to maintain a shared map once this version is done.

 

Buster

 

Current results for 163 responses are at http://www.batchgeo.com/map/ca5449d55ccb94010dd9dbfbc0fdd870