An article in the American Journal of Managed Care neatly illustrates the difficulty in reducing inappropriate care and expense. (AJMC Article) The commentary begins by referring to the Rand Health Insurance Experiment, which found that, in general, more cost sharing of various types by patients led to lower utilization and lower expenses. Rand also found that on average, health outcomes were the same with higher cost sharing, which means that lower utilization did not appear to affect the patients’ health. But Rand and others found that utilization of necessary services was reduced along with that of unnecessary ones, leading to a question about how outcomes could seem to be the same. The answer, Rand and others believe, is that unnecessary services can have adverse effects on health and can add further cost. If a patient has a procedure they didn’t really need, for example, there is the possibility of complications and errors, which add to cost and worsen the patient’s health. Reducing necessary services also has obvious bad effects. So the problem has been to take the generalized finding that cost sharing reduces spending without worsening outcomes, on average, and apply it patient by patient. This is an especially a concern for chronic disease sufferers, who may be most likely to skip needed care, and low income individuals, who may have greater price sensitivity. Two possible approaches are more patient education about choosing appropriate care and being involved in decisions about what care they receive, and greater use of value-based benefit designs, which lessen cost-sharing on high-value services such as preventive care and medications for chronic illnesses. Lessons include that any reform measures have to be carefully examined for unintended consequences and that getting to the goal of providing only necessary care for each patient will require spending more on care management.