Positive Predictive Value of MycoPlasma Assays

[Buster]

In post http://www.latitudes.org/forums/index.php?showtopic=9948&view=findpost&p=83817 , Fixit asked

 

 

Buster..does the same hold true for myco p..

 

a reading in mayish was 1.60 another draw 3 weeks ago was 1.79 (above .90 igg could be positive)

dan doc said no worry...margin or error..and these numbers can fluctuate...

however..in last week we have motor tics which were 90% other than times of stress or a specific trigger..

how much time should go by before another draw...??

 

One of the better references on mycoplasma assays is http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1153783/pdf/1937-04.pdf where they compared 10 methods for analyzing Igg and Igm for mycoplasma pneumonia. These assays were from

• AniLabsystems
  
• Biotest
  
• CFT
  
• Diagnosys
  
• ImmunoCard
  
• ImmunoWell
  
• Novum
  
• Platelia
  
• Ridascreen
  
• Serion classic 4
  
• SerodiaMycoII
  
• SeroMP
  
• Virotech
  

 

Of these tests, most had an agreement between the Igg test and the Igm test of ~80% (meaning the two tests the same on the sample 80% of the time). Notably the Novum test only agreed between Igg and Igm tests 54% of the time.

 

In terms of positive predictive value (how often a positive is a real positive) these are all over the map. Novum and Immunocard has a positive predictive value of ~30%, whereas others were around 80%. This means that a lot of the tests had a lot of false positives. The negative predictive value (i.e., how often is a negative a real negative) was more around 80-90%. So you should read this that negatives are more informative than positives.

 

A very nice part of the paper deals with when to run the Igg and Igm tests. figure 2 shows the sensitivity of the test was best at >=16 days after onset. If tested within first week, most did not have results.

 

I do not follow mycoplasma enough to know the significance of rises in test results, but it looks like the tests are exquisitely sensitive to running the same test twice and given the low specificity and low positive predictive value, you should find out which assay was run and look at the negatives more than the positives.

 

Buster

[Fixit]

In post http://www.latitudes.org/forums/index.php?showtopic=9948&view=findpost&p=83817 , Fixit asked

 

 

Buster..does the same hold true for myco p..

 

a reading in mayish was 1.60 another draw 3 weeks ago was 1.79 (above .90 igg could be positive)

dan doc said no worry...margin or error..and these numbers can fluctuate...

however..in last week we have motor tics which were 90% other than times of stress or a specific trigger..

how much time should go by before another draw...??

 

One of the better references on mycoplasma assays is http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1153783/pdf/1937-04.pdf where they compared 10 methods for analyzing Igg and Igm for mycoplasma pneumonia. These assays were from

• AniLabsystems
  
• Biotest
  
• CFT
  
• Diagnosys
  
• ImmunoCard
  
• ImmunoWell
  
• Novum
  
• Platelia
  
• Ridascreen
  
• Serion classic 4
  
• SerodiaMycoII
  
• SeroMP
  
• Virotech
  

 

Of these tests, most had an agreement between the Igg test and the Igm test of ~80% (meaning the two tests the same on the sample 80% of the time). Notably the Novum test only agreed between Igg and Igm tests 54% of the time.

 

In terms of positive predictive value (how often a positive is a real positive) these are all over the map. Novum and Immunocard has a positive predictive value of ~30%, whereas others were around 80%. This means that a lot of the tests had a lot of false positives. The negative predictive value (i.e., how often is a negative a real negative) was more around 80-90%. So you should read this that negatives are more informative than positives.

 

A very nice part of the paper deals with when to run the Igg and Igm tests. figure 2 shows the sensitivity of the test was best at >=16 days after onset. If tested within first week, most did not have results.

 

I do not follow mycoplasma enough to know the significance of rises in test results, but it looks like the tests are exquisitely sensitive to running the same test twice and given the low specificity and low positive predictive value, you should find out which assay was run and look at the negatives more than the positives.

 

Buster

 

 

Thanks buster...i will try to digest this better during the week when i have a little time