446 PICC-Associated Bloodstream Infections at Boston Medical Center

Saturday, March 20, 2010
Grand Hall (Hyatt Regency Atlanta)
Julie Bharucha Trivedi, MD , Boston University Medical Center, Boston, MA
Paschalis Vergidis, MD , Mayo Clinic, Rochester, MN
Carol Sulis, MD , Boston Medical Center, Boston, MA
Tamar Barlam, MD , Boston Med Ctr, Boston, MA
Background: Peripherally inserted central catheters, or PICCs, are often placed in hospitalized patients to obtain additional intravenous access for administration of various medications, nutrition or collection of blood for lab testing.  However, there are limited data on rates and risk factors influencing development of bloodstream infections (BSIs) due to PICCs in the general hospitalized population. Objective: We sought to describe the rates of PICC-associated BSIs and the role of potential risk factors such as patient demographics, comorbid conditions, indications for PICC placement, duration PICC was kept in place as well as location of use of PICC such as hospital, long term care/skilled nursing facility (LTC/SNF), or home.  Methods: We conducted an observational, prospective study of patients with a PICC placed while hospitalized.  Data on patient demographics, medical conditions, PICC characteristics, and positive blood cultures were collected from the electronic medical record and patient interviews.  Results:   We identified 7 BSIs among 189 patients for an overall rate of 1.88 infections per 1000 catheter days (CD). Surgical patients (n=3) had a rate of 3.53 infections per 1000 CD; nonsurgical patients (n=4) had a rate of 1.39 BSI per 1000 CD. The inpatient infection rate (n=2) was 1.07 per 1000 hospital CD; both infections occurred in surgical patients. The remaining 5 BSIs were diagnosed after discharge, for a rate of 2.70 per 1000 non-hospital CD; 2 patients were home and 3 were in a LTC/SNF. Six of 7 PICCs were double lumen. None of the 7 patients were immunosuppressed with HIV-infection, hematologic malignancy, solid organ transplantation or other immunosuppressive treatment; 1 patient each had diabetes, dementia, or was dialysis-dependent. Five of 7 patients were receiving antibiotics for preexisting infections and 3 of the 7 were on TPN. Patients with TPN and subsequent BSI had PICCs in place for an average of 53 days, with a median of 46 days (range 14-100) versus a mean of 19 days (median=16.5 days, range 7-19) in those receiving TPN without infection. Rates of infection in patients receiving TPN were 7.03 per 1000 CD vs 1.21 per 1000 CD in those not receiving TPN. Conclusions: Although overall rates of infection were low, it appears that surgical patients and patients discharged with a PICC in place were more likely to develop a BSI. Additionally, patients receiving TPN tended to have PICC in place for longer durations and had higher rates of BSIs. There were no clear associations between baseline medical conditions and developing a BSI.