New cardiac surgery programs continue to open across the United States, and it is not known how new programs deal with potentially low volumes during their start-up period. We compared patient, procedure, and physician characteristics and short-term mortality at established cardiac surgery programs, new programs in general hospitals, and new specialty cardiac hospitals.

We used Medicare Provider Analysis and Review, part B physician claims, and denominator files to evaluate established and new programs performing coronary artery bypass graft surgery (CABG) from 1994-2003. Short-term mortality was defined as death in-hospital or within 30 days.

From 1994-2003, 257 new programs in general hospitals and 20 new specialty hospitals opened; and 884 established programs were in operation. New programs in general hospitals had much lower CABG volume than established programs and performed fewer concomitant valves and reoperations. New specialty hospitals had high CABG volume from inception, similar valve and reoperation rates to established programs, and conducted more elective procedures. Short-term mortality was significantly lower at new programs in general hospitals.

Start-up strategies used by new specialty hospitals and new programs in general hospitals differed markedly. By choosing to conduct safer procedures on low-risk patients, new general programs may have offset potential concerns about operating at low volume. Neither type of new program exhibited an increased risk of short-term mortality. The high volume at specialty hospitals may reassure patients and policy makers, although the high proportion of elective procedures and the new program's effect on surrounding hospitals require further consideration.