To the Editor.
—In the Editorial by Drs Denke and Winker,1 there is the unfortunate implication that our article2 from the New Haven, Conn, site of the Established Populations for Epidemiologic Studies of the Elderly (EPESE) was published prematurely and should not have preceded a multicenter analysis. The comments reveal a misunderstanding of the nature of the EPESE collaboration and the special features of this observational study.Unlike many multicenter trials, the EPESE sites are a confederation of studies with distinct sampling designs and features. Although the EPESE was a collaboration contract program, individual sites had the option to publish results separately, where unique data were available. For this reason, single-site publications from EPESE far outnumber the pooled analyses.There are unique features of the New Haven site that led us to pursue our analysis separately. New Haven is the only site in which diagnoses of hospitalized acute
Krumholz HM, Vaccarino V, de Leon CFM, Seeman TE, Berkman LF. Cholesterol and Coronary Heart Disease Mortality in Elderly Patients. JAMA. 1996;275(2):110. doi:10.1001/jama.1996.03530260024018