It has much less to do with rheumatic fever and much more to do with the realities of an interaction with the patient in front of you than the numbers on paper. Antibiotics are still widely prescribed for many clear viral etiologies because it’s a lot easier to make an argument for a possibility of bacterial etiology and acquiesce than it is to have to take 10-15 minutes to educate a patient who 1) feels like crap 2) came in and paid their copay and 3) wants to get better. Add in institutional/systemic bureaucracy and patient satisfaction ratings and it will often push to favor prescribing an antibiotic for a viral respiratory infection. In theory one can be a purest, but where the rubber meets the road you can’t full-court-press on every patient every day because you will crash and burn.
So if someone has strep they feel bad enough to come in, you’re going to throw an incredibly cheap antibiotic at them even if it only shows 24 hour improvement above placebo, because those 24 hours can be valuable not just for symptoms but return to work. In addition, 24 hours is an average, meaning some will have >24 hours and again its compared to placebo, not compared to doing nothing (which is too often what people assume when they think about and discount a placebo effect, which often accounts for real and sizable components of true treatment effects).
