A prospective study of Hickman/Broviac catheters and implantable ports in pediatric oncology patients.

We prospectively studied the continuous function and complication rates of 286 central venous catheters consecutively placed in 264 children and young adults at a single institution over a 19-month period (median follow-up, 376 days). Externalized catheters (91 Hickman [H], 113 Broviac [B]) and implantable ports (n = 82) were compared for complications, including infection and thrombosis. The most frequent major complication of all catheters was infection, although the rates of infection varied with the duration of catheter use and were generally lower than reported by others. Overall, when catheter failures (removal) for infection, obstruction, or dislodgement were considered, ports had a significantly longer failure-free duration of use (P = .0024) than did externalized catheters. Likewise, ports had a significantly longer infection-free (P less than .01) duration of use than H and B catheters. However, differences in patient age and clinical characteristics among the three catheter groups may have affected the outcome. In analysis of pairs matched for diagnosis, therapy, and age, ports had lower infection rates than did B catheters after 100 days (P = .053). This difference became significant at 400 days of catheter use (P = .029). Although there was a trend toward lower rates of infections for ports v H catheters, this difference was not significant. In view of our results in matched pairs, selection of catheter type based on clinical characteristics and patient preferences remains a reasonable therapeutic approach despite the apparent advantages of ports. The superiority of ports for long-term use (greater than 100 days) needs to be confirmed in a large randomized clinical trial.