Tager, I. B. (Channing Laboratory, Boston, MA 02115), M. B. Ginsberg, S. E. Ellis, N. E. Walsh, I. Dupont, E. Simchen, G. A. Falch and the Rhode Island Nosocomlal Infection Consortium. An epidemiologic study of the risks associated with peripheral Intravenous catheters. Am J Epidemiol 1983;118:839–51. As part of a standardized, multi-hospital prospective surveillance system for nosocomlal infections in Rhode island, an analysis was undertaken in March 1980-February 1982 to determine the interplay of factors that contribute to the risk of phlebitis in peripheral, non-steel, non-butterfly intravenous catheters. The authors studied 3094 patients with 5161 total episodes of peripheral intravenous catheters from day of admission until day of discharge. The overall rate of phlebitis was 2.3% (118 episodes), and the rate of intravenous catheter-associated bacteremia was 0.08% (1 definite episode, 3 possible episodes). Factors significantly associated with the occurrence of phlebitis were: underlying risk for any nosocomlal infection, duration of the catheter episode, chronological order of the episode and an interaction between the latter two variables. Analysis of day-specific risk of phlebitis indicated that, for patients with low risk diagnoses, initial peripheral intravenous catheters might be left In place with relative safety for up to 96 hours. Over this time period, the day-specific risk for such patients ranged between 0.8% and 1.4%, exclusive of the first day. In all other circumstances, the current recommendation of 48–72 hours seems appropriate.