+ Reply to Thread
Results 1 to 5 of 5

Thread: How do I chose between sensitivity/specificity and PPV/NPV?

  1. #1
    Points: 1,207, Level: 19
    Level completed: 7%, Points required for next Level: 93

    Posts
    7
    Thanks
    3
    Thanked 0 Times in 0 Posts

    How do I chose between sensitivity/specificity and PPV/NPV?




    I have just finished a validation study, where two specialists used the available information to classify patients according to their diseases status as true cases and no cases. I have confirmed diagnoses for a sample of those patients and I now want to see how successfully the specialists evaluated the disease status of those patients.

    I tried to find it out myself but I still don't fully understand the difference between sensitivity and PPV resp. specificity and NPV. Which of those values are relevant for my analysis? Is it necessary to present all four values or is it enough to just present the PPV and NPV?

    Thank you for your help! This is probably a really basic question but I did not really find an answer to it myself.

  2. #2
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: How do I chose between sensitivity/specificity and PPV/NPV?

    People typically present all 4, plus accuracy metric (i.e, TP + TN over TP + TN + FN + FP. if you are just creating a classification tables for each specialist, than you could present those table values or just the numbers for TP, FP, FN, and TN and the reviewer can always calculate them on their own.


    Yes all four metric seem the same for most people until you worked with them enough. The difference is in the descriptions (e.g., of those with the condition..., of those without the condition..., of those testing positive..., of those testing negative...)


    A side note, the SEN and PPV as well as the SPEC and NPV will typically be comparable given the similarities in the denominators. However the prevalence of the condition in your sample can change the PPV and NPV metrics, a nuance, which many don't realize.
    Stop cowardice, ban guns!

  3. #3
    Points: 1,207, Level: 19
    Level completed: 7%, Points required for next Level: 93

    Posts
    7
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: How do I chose between sensitivity/specificity and PPV/NPV?

    Thank you very much! Just one more question...

    Does the fact, that my sample only comprises of patients with a putative diagnosis for the disease in question, have an influence on the assertion of the PPV and NPV? Will they still be relevant (just higher values) or are they only relevant, when the sample displays the true prevalence of a disease?

  4. #4
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: How do I chose between sensitivity/specificity and PPV/NPV?

    putative: based on clinical judgment/gestalt?


    If yes, you can only generalize findings to comparable scenarios, and as you mentioned, be apprehensive with PPV and NPV conclusions.
    Stop cowardice, ban guns!

  5. #5
    Points: 1,207, Level: 19
    Level completed: 7%, Points required for next Level: 93

    Posts
    7
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: How do I chose between sensitivity/specificity and PPV/NPV?


    Sorry that was badly phrased.

    With the putative dignoses I meant the ones which I am testing for validity. I have some gold standard diagnoses which I want to compare the putative diagnoses to.

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats