To the Editor: I found the conclusion by Edward J. Federman, Ph.D., et al. (1) that "climate factors correlated modestly but significantly with length of stay" (p. 1669) to be intriguing and to also be good common sense. To emphasize their finding, Dr. Federman and colleagues reported that climate alone accounted for almost as much variance in length of hospital stay (14%) as did the more typically thought of "diagnoses together with history of previous admissions" (p. 1672) (17%) in a recent study.

I must comment concerning the apparently contradictory finding that the percentage of homeless veterans admitted to psychiatric units was actually higher in hospitals located in warmer climates. Such a finding highlights the importance of other system and patient factors in length of hospital stay. Warmer temperatures "allow" homelessness to exist and thus create a greater pool from which to draw patient admissions. Also, the veterans population may likely be part of this homeless community, as well as participants in an anecdotally recognized phenomenon of seasonal migration of the population that is homeless and mentally ill.

This was an often-discussed phenomenon when I worked in a Maryland Department of Veterans Affairs hospital that was located along the southern route of the fall/winter migration; similarly, there was a return to New England in the summer. This "snowbird" phenomenon in this context was discussed as well in the emergency room of the Maine community hospital in which I also worked. I certainly agree that future studies should examine the effects of such "other contextual variables on health care and specifically on hospital length of stay" (p. 1673).