• TacoNot
    arrow-up
    4
    arrow-down
    0
    ·
    6 months ago
    link
    fedilink

    SOLVD Health said in its application that AvertD had demonstrated a sensitivity of about 82% and a specificity of about 79%.

    Those numbers suggest that roughly 1 in 5 results would be false negatives and roughly 1 in 5 would be false positives.

    That doesn’t seem very good

    • jadero
      arrow-up
      1
      arrow-down
      0
      ·
      6 months ago
      link
      fedilink

      Total error rate is about 2/5? That’s barely better than a coin flip.

      “Oh yes, I just got tested for my potential to become an opioid addict. Give me a second to flip a coin: heads I read the report, tails I toss it in the recycle bin.