AbstractI felt good about myself, driving to the free medical clinic that evening. A full professor at a medical school, leaving my warm home on a cold night after a day at the hospital, seeing patients in clinic in the morning and teaching second‐year students medical ethics in the afternoon (autonomy was the theme; we'd covered beneficence and maleficence earlier in the week). Once a month, patients with cardiac problems come to the clinic, and this was the night. Two students presented the patient, a middle‐aged woman with chest pain. Their presentation was disjointed, in part because of their inexperience, in part because she spoke no English and her story was obtained with the help of a translator. Her eyes rarely met mine; she kept glancing at the translator. I auscultated her heart and lungs, letting the students listen through my stethoscope. What to do now? Had she been a patient that morning back in my clinic, I would have thought to myself, “I can't get a good history; why not do a stress test?” I thought, for what seemed like an eternity: If I send her for a stress test, that will deplete the funds needed for other patients. How important is this test? Is it really necessary? If I send her for a stress test, will someone with a suspicious mass on their chest x‐ray not be able to get a CT scan because no money is left?