In April's Journal of Family Pracitce, this paper concludes that "Doctors often misinterpret patient expectations. Improving communications between patient and doctor may be central to reducing patients019 demand for antibiotics. " I agree. And I support the authors' efforts to determine if delayed prescriptions ("Take this in 3 days if you don't feel better") but I wonder about the mixed message that this sort of practice sends to the patient. Since we know that patients can be treated safely with only symptomatic therapy, the delayed antibiotic prescription sends the message that antibiotics may in fact be indicated — even though we have already said that they are not. Shouldn't we be consistent in our message? My preference is to have the patient call or return to the office if their symptoms persist. But since the average cold lasts 6 to 7 days … the patients will still have symptoms after 3 days … will take the antibiotics .. and then will feel better. Patients (and perhaps even the physicians) will then assume that the temporal relatiohnship of the intervention and the resolution of the symptoms = causality. But of course (as I discussed in the entry on circumcision a few weeks ago .. ) correlation does not = causality.