Infectious Complications of 393 Peripherally Implantable Venous Access Devices in HIV-Positive and HIV-Negative Patients

Abstract

To compare and investigate the rate of infection in patients with and without human immunodeficiency virus (HIV) who have implantable venous access devices placed by interventional radiologists. Three hundred ninety-one patients undergoing radiologically guided placement of peripheral arm ports were grouped according to their HIV serologic status. Findings were prospectively reviewed in 393 peripherally placed arm ports that were implanted in the basilic, cephalic, or brachial vein under fluoroscopic or sonographic guidance over a 4-year span. Infectious complications were categorized according to severity (local or systemic) and time (periprocedural or late). Three hundred ninety-three ports have been indwelling for a total of 97,256 patient days (range, 1-694; mean duration, 247 days). Among the 30 catheter placements in 29 HIV-positive patients with a total exposure time of 7,242 days, five (one local and four systemic) infections occurred, resulting in a 16.6% overall infection rate, yielding 0.069 infections per 100 catheter days at risk (95% confidence interval [CI], 0.032-0.127). In the remaining 362 HIV-negative patients, 27 (14 local and 13 systemic) infectious complications (7.4%) occurred, translating into 0.030 infections per 100 catheter days (95% CI, 0.021-0.042). The odds ratio of getting an infection from the implantable arm ports in the HIV-positive group was 2.5 times higher than that of the HIV-negative group. The relative risk was similar and was calculated to be 2.3. The P value was .084 (P < .05 required to be considered significant). These results suggest a significant difference in the infectious complication rate encountered in HIV-positive patients compared with the general population. However, the HIV-positive peripheral arm port infection rate compares favorably with the surgically placed catheters and ports. Many more arm ports in HIV-positive patients must be evaluated for the data to achieve an acceptable level of statistical significance.