In his Counterpoint editorial, Dr Veatch1 argues that physicians cannot presume to know what is best for patients, that recommendations can sway patients’ decisions about medical treatments, and, therefore, recommendations may result in decisions that are not in patients’ best interests. Not surprisingly, we share considerable ground on each of these points. I, too, acknowledge that there are challenges determining what is best for patients, and I recognize the risk of misleading surrogates with recommendations. However, when a patient lacks decisional capacity, has no advance directive, and is in the process of dying, clinical decisions require thoughtful integration of medical facts and patient values. The physician has access to the medical facts by virtue of his training, and he comes to know the patient’s values through discussion with a surrogate. The physician is the most appropriate person to integrate these factors into a recommended plan of care.