We are testing a new diagnostic tool and comparing it to the actual gold standard for this diagnosis.
Briefly, we examined 25 patients with the new diagnostic tool (test A) and the gold standard diagnostic tool (test B). Test A gives a positive result or a negative result (no variability or range in numbers, just "positive" or "negative" as outcome). We then performed test B which also gives a "positive" or "negative" results and which is considered the true result since this is the gold standard diagnostic tool.
All patients having a positive result on test A (n=18), had a positive result on test B (n=18).
Of all patients having a negative result on test A (n=7), 5 were negative on test B but 2 were positive on test B.
Overall, 23 patients had the same outcome on test A and test B, 2 were different, which means that our new diagnostic test has a sensitivity of 92% (if we consider test B to have 100% sensitivity).
Can you recommend me any more statistics on this data, to draw conclusions? Any idea to look at this data from another perspective? Any help or insight is appreciated.
Thank you
You mean beyond sensitivity and specificity
(ie. https://www.medcalc.org/calc/diagnostic_test.php)?
Confidence intervals.
You have a small dataset, so there is considerable imprecision to your estimates. Confidence intervals would help to clarify this.
Thank you for your valuable contributions. @Ronán Michael Conroy: can you elaborate on how to measure these confidence intervals on these non-numerical data? Thank you
The calc in the link gives you CI (based on Binomial proportion confidence interval). references are provided too.
https://www.medcalc.org/calc/diagnostic_test.php
If you take the elder test as "truth" and both tests are dichotomous, then sensibility/specificity/positive predictive value/negative predictive value could be enough.
Perhaps you can also see how often the two tests show concordant results by looking at it as if it were a problem of inter-rater agreement (you could use an index like Krippendorff's alpha).
I am just wondering if your sample size is enough (and if the sample is truly representative of the population on which you'd like to use the new test).
if outcome measure is dichotomous go for agreement analysis kappa statistic . If you have gold standard to evaluate new diagnostic test ROC and are UNDER THE CURVE HELPS TO EVALUATE.
(Note: Krippendorff's alpha provides an estimator that includes as particular cases various other commonly used indexes of chance-corrected agreement)
Please apply Cohran Test, or Mac Nemar Test , because your answer are 1 or 0. Thanks
- Badges
- Science method