That means that we use the inverse of the formula explained by Patrick Bossuyt in an article from 2011 (1). We do this as having the outcome stated as an event is more common than a non-event. Hazard rates are calculated assuming your outcome is an event (e.g mortality), and not a non-event (e.g survival). Absence of confounding does not correspond to. the average HR of 1.8 during year 1 means that the one-year risk was about 0.49 in the treatment group and 0.28 in the placebo group rather than, say, 49 versus 28. Then the effect on risk is 62 fewer per 1000: The hazard ratio (HR) is the main, and often the only, effect measure reported in many epidemiologic studies. The calculation from the Cochrane handbook uses the baseline risk (called 'assumed control risk (ACR) in the calculation):Īs an example, suppose the odds ratio is OR = 0.73, and a control risk of ACR = 0.3 is assumed. It calculates the number fewer per 1000 (the difference), and we calculate the absolute intervention risk by adding the difference (whether more or fewer) to the baseline risk. We have taken the calculation we use from the Cochrane handbook, chapter 12.5.4.3 . To find the absolute risk in the in intervention group using the below formulas.įrom the intervention risk we calculate the absolute difference by a simple subtraction.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |