{"id":895,"date":"2014-03-19T21:00:22","date_gmt":"2014-03-19T20:00:22","guid":{"rendered":"https:\/\/blogs.bmj.com\/adc\/?p=895"},"modified":"2014-09-12T11:45:25","modified_gmt":"2014-09-12T10:45:25","slug":"positive-about-predictions","status":"publish","type":"post","link":"https:\/\/blogs.bmj.com\/adc\/2014\/03\/19\/positive-about-predictions\/","title":{"rendered":"Positive about predictions"},"content":{"rendered":"

\"\"<\/a>In a previous post I muttered about how unhelpful sensitivity and specificity are to practicing clinicians, and how what we really want to know are the predictive values of a test.<\/p>\n

Remembering the Table<\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

\n\n\n\n\n\n\n
<\/td>\nReally diseased<\/td>\nReally not diseased<\/td>\n<\/td>\n<\/tr>\n
Test +ve<\/td>\nA<\/td>\nB<\/td>\n1.. A\/(A+B)<\/td>\n<\/tr>\n
Test -ve<\/td>\nC<\/td>\nD<\/td>\n2.. C\/(C+D)<\/p>\n

5.. D\/(C+D)<\/em><\/td>\n<\/tr>\n

<\/td>\n3… A\/(A+C)<\/td>\n4… D\/(B+D)<\/td>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

<\/p>\n

Now I like to think about % with disease. I want to know about 1.. the predictive value of a positive test (the proportion of patients with a positive test who have the disease) aka positive predictive value (PPV). I also like to know 2.. the predictive value of a negative test (the proportion of patients with a negative <\/em>test who have<\/em> the disease).<\/p>\n

Others will instead want to know the ‘NPV’ (5..) the proportion of patients with a negative <\/em>test who don’t have <\/em>the disease.<\/p>\n

Me \u2013 I don’t think about having a 95% PPV and 98% NPV as my test results, I like to know that 95% patients with a positive test have the disease. But I don’t think “98% of patients with a negative test really are disease free”, but “2% of patients with a negative test actually have this condition”<\/em>. (It’s this type of focusing on the failures that make me a cheery chap to spend an evening with.)<\/p>\n

If the test you’re evaluating is being used in the same sort of population, with the same sort of prevalence of disease as you are working in, you can stop here and believe these numbers. If there’s a difference \u2013 more or less disease, more or less suspicion \u2013 then you need to go an learn about Bayes and likelihood ratios and Fagin, pickpockets and child exploitation.<\/p>\n