Question 5: What kind of study would be better to draw this kind of conclusion? (that 'forehead temperature in excess of 35.6°C is suggestive of fever'.)
They need to know what range forehead temperature has in people who do have fever and in people who do not have fever, as shown by internal temperature. So they need people with fever as well as those without, defined by a gold standard. They need a diagnostic study. They could then calculate the sensitivity and specificity for any given cut-off of forehead temperature. They could draw a ROC curve. For any given prevalence of fever, they could then calculate the positive predictive value and know to what extent forehead temperature in excess of 35.6°C is suggestive of fever.
Back to question.
To Martin Bland's home page.
This page maintained by Martin Bland.
Last updated: 2 March, 2007.
Back to top.