To the Editor.
—The elegant study by Dr Bindman and colleagues1 provides strong evidence that preventable hospitalizations are a useful marker for monitoring access to care. Nevertheless, there are many unanswered questions about why the poor are hospitalized more frequently. Although this study looked at "propensity to seek health care," the symptoms used to measure this were serious (bleeding, shortness of breath) and may not reflect the decision-making process for most patients with ambulatory care—sensitive conditions. It may be that the more dramatic the symptom, the less likely there will be a difference in care-seeking behavior according to access barriers. If participants had been asked questions about less serious symptoms, such as a cough with purulent sputum production for 2 days, differences in health care-seeking behavior may have been significant in predicting preventable hospitalizations.Differences in case mix between poor and nonpoor patients also deserve further consideration. Although adjustments