NBME 22 Answers ↦
the prevalence of the exposure and the health outcome are measured at the same time. You are basically trying to figure out how many people in the population have the disease and how many people have the exposure at one point in time.
Case Control would determine ODDS ratio
Cohort would determine Relative Risk
For the case control question, it’s taking that principle that you can’t use case control studies to calculate relative risk and applying it to prevalence.
Basically with case control studies we start by saying okay, I’m going to find 200 people with sinusitis and 400 without. Then, you go back and look at the number exposed/unexposed and calculate the odds ratio. So you can’t use case controls to calculate prevalence because it all depends on how many cases you picked in the first place.
Might make more sense to think about it with a rare cancer like craniopharyngioma or something- let’s say you chose 10 cases and 10 controls and wanted to look at how many people smoked. It wouldn’t make sense to then say the prevalance of craniopharyngioma is 10/20 = 50%.
Thank you, that makes a lot of sense! So we can use relative risk (cohort studies) to calculate prevalence?
@trazobone Hmm I think the wording would be key, you could use a prospective cohort to calculate incidence, but you wouldn’t be able to find prevalence of the gen population unless you had more info. I think the concept is that really to calculate prevalence you need a proper ecologic study looking at population-level data. The way it was worded in the question was tricky though lol since when has “cannot be determined from the info given” ever been a right answer.
@sacredazn thank you! this was the best explanation to use the rare disease comparison. Made everything make so much sense and hopefully I'll actually just remember it now, instead of learning the factoid and failing to recall it all the time.
i guess this makes sense but i don't understand why we are asked to calculate it from tables like this then? is there more info in those?
@hyperfukus The table was given because that a 22 table is typically what you do see regarding data for case-control studies. If the 22 table wasn't include, then literally everytone would pick Choice "E" as the correct answer b/c you can't calculate something without being provided numbers. The difference in including the data-table is that 1. again, you need to report a 22 table because that is typically what you will see regarding data for a case-control study and 2. by including the 22 table, it actually tests if the test-taker realized that the data in the 2*2 table does not help at all with calculating prevalence-- because case-control studies NEVER report on prevalence.
I also want to add that the Q is asking about sinusitis in the GENERAL population, meanwhile the Q-stem discusses a case-control study that is studying the relationship of smoking (an exposure) to development of sinusitis. To then ask "what is the prevalence of sinusitis in the general population" totally disregards taking into account the exposure of smoking which was the entire reason for the study. In other words, asking prevalence would be very non-sensical.