Dedicating Pre-Existing Central Venous Catheters for Parenteral Nutrition: Are Routine Surveillance Blood Cultures Necessary?

Author(s):
Sandra Marie Swoboda, Johns Hopkins Medical Institutions; Patricia Brown, Johns Hopkins Hospital; David Efron, Johns Hopkins Medical Institutions; Lindsay Stander, Johns Hopkins Hospital; Pamela Lipsett, The Johns Hopkins Medical Institutions

Background:

Routine surveillance blood cultures are used to dedicate pre-existing catheters for parenteral nutrition as a standard of care at our institution.   The most recent guidelines state that “catheter cultures should not be obtained routinely”, yet long-term indwelling lines in patients who have been managed out of hospital or at another institution may be at higher risk, especially if the line is to be dedicated for nutrition administration.

Hypothesis:

Routine surveillance cultures are not required prior to dedicating a pre-existing catheter for parenteral nutrition (CPN).

Methods:

This was a retrospective database review of patients from 2006-2014 with pre-existing central venous catheters that were dedicated for parenteral nutrition. Surveillance blood cultures were taken from the lumen of the catheter per protocol. Risk factors for infection included line type, number of lumens, presence of fistula, ostomy, abscess, surgical vs. medical patient, reason for use of catheter prior to parenteral dedication, wbc and Tmax. Descriptive statistics were used.

Results:

Overall, 92/852 patients had a positive blood culture (+ CX)(10.7%). Medicine patients (60/349) were significantly more likely to have +CX vs. surgical patients (15% vs. 7%, OR; 2.22(95% CI:1.4, 3.6). The majority of catheters were PICC followed by Mediport and Hickman Catheters (HC) with HC having the highest infection rate (HC (15%) v. PICC (9%) and Mediport (8%), p=0.02). The majority of cultures were Gram positive (61%) followed by gram negative (23%), yeast (10%) and multi-organisms (6%). Risk factors such as chemotherapy, home CPN, fistulas, ostomy, intra-abdominal abscesses or fever were not significant (p>0.25).The mean Tmax for + CX was 37.3 C. The mean WBC for +CX was 7.9 (+5.5) compared to 9.1 (+5.5) for negative cultures (p=0.05). The regression model showed that medical patients without a history of cancer were more likely to have a positive blood culture.

Conclusions:

The overall rate of positive blood cultures was 10.7%. “General” risk factors in patients for suspicion of infection (fever, leukocytosis, fistula, cancer, surgery, type of line), are not predictive of positive blood cultures in this patient population. Routine surveillance cultures should continue as standard practice for dedication for parenteral use.