Clinical research study
PICC-associated Bloodstream Infections: Prevalence, Patterns, and Predictors

https://doi.org/10.1016/j.amjmed.2014.01.001Get rights and content

Abstract

Background

Growing use of peripherally inserted central catheters (PICCs) has led to recognition of the risk of PICC-associated bloodstream infection. We sought to identify rates, patterns, and patient, provider, and device characteristics associated with this adverse outcome.

Methods

A retrospective cohort of consecutive adults who underwent PICC placement from June 2009 to July 2012 was assembled. Using multivariable logistic and Cox-proportional hazards regression models, covariates specified a priori were analyzed for their association with PICC-associated bloodstream infection. Odds ratios (OR) and hazard ratios (HR) with corresponding 95% confidence intervals (CI) were used to express the association between each predictor and the outcome of interest.

Results

During the study period, 966 PICCs were inserted in 747 unique patients for a total of 26,887 catheter days. Indications for PICC insertion included: long-term antibiotic administration (52%, n = 503), venous access (21%, n = 201), total parenteral nutrition (16%, n = 155), and chemotherapy (11%, n = 107). On bivariate analysis, intensive care unit (ICU) status (OR 3.23; 95% CI, 1.84-5.65), mechanical ventilation (OR 4.39; 95% CI, 2.46-7.82), length of stay (hospital, OR 1.04; 95% CI, 1.02-1.06 and ICU, OR 1.03; 95% CI, 1.02-1.04), PowerPICCs (C. R. Bard, Inc., Murray Hill, NJ; OR 2.58; 95% CI, 1.41-4.73), and devices placed by interventional radiology (OR 2.57; 95% CI, 1.41-4.68) were associated with PICC-bloodstream infection. Catheter lumens were strongly associated with this event (double lumen, OR 5.21; 95% CI, 2.46-11.04, and triple lumen, OR 10.84; 95% CI, 4.38-26.82). On multivariable analysis, only hospital length of stay, ICU status, and number of PICC lumens remained significantly associated with PICC bloodstream infection. Notably, the HR for PICC lumens increased substantially, suggesting earlier time to infection among patients with multi-lumen PICCs (HR 4.08; 95% CI, 1.51-11.02 and HR 8.52; 95% CI, 2.55-28.49 for double- and triple-lumen devices, respectively).

Conclusions

PICC-associated bloodstream infection is most associated with hospital length of stay, ICU status, and number of device lumens. Policy and procedural oversights targeting these factors may be necessary to reduce the risk of this adverse outcome.

Section snippets

Methods

Using records obtained from our vascular access nursing team, we assembled a cohort of consecutive adult, hospitalized patients who underwent insertion of a PICC from June 1, 2009 to June 30, 2012 at our 145-bed, academic Veterans Affairs (VA) Medical Center. Clinical data such as indication for insertion, number of attempts, vein and arm of insertion, and type of PICC were abstracted from electronic medical records. Information about patient comorbidities, medications and dosages, disposition,

Results

Between June 2009 and July 2012, 966 PICCs were inserted in 747 unique patients, accounting for a total of 26,887 catheter days. The majority of patients who underwent insertion were male (98%), with a median duration of PICC use of 21 days (95% CI, 19-23 days). Most PICCs were placed by vascular access nurses (n = 823; 85%); only 15% (n = 143) were placed by interventional radiology when bedside insertion was not successful or appropriate. The most common indications for PICC insertion

Discussion

Although national data suggest marked reduction in the rate of central line-associated bloodstream infection,7 this hospital-acquired infection remains an important, preventable cause of patient harm in hospitalized settings.8 Notably, little is known about the risk of PICC-associated bloodstream infection in hospital and critical care settings.9 Using a combination of clinical and administrative data and informed by a conceptual model for PICC complications, we found that this adverse event is

Conclusion

We believe this work has important policy and patient safety implications. First, placement of PICCs in critically ill patients appears unwise and should be avoided given the inherent risk of infection in this population. When clinically necessary, the placement of antimicrobial catheters or chlorhexidine-impregnated dressings may be appropriate to prevent PICC-associated bloodstream infections in this context.25 Further, placement of multi-lumen PICCs appears to significantly increase the risk

References (25)

  • M.A. Ugas et al.

    Central and peripheral venous lines-associated blood stream infections in the critically ill surgical patients

    Ann Surg Innov Res

    (2012)
  • P. Chittick et al.

    Early onset versus late onset peripherally inserted central venous catheter infections: an analysis of risk factors and microbiology

    Infect Control Hosp Epidemiol

    (2013)
  • Cited by (137)

    • Risk of VTE associated with PORTs and PICCs in cancer patients: A systematic review and meta-analysis

      2022, Thrombosis Research
      Citation Excerpt :

      It can be utilized for total parenteral nutrition, prolonged chemotherapy or antibiotic treatment. It is inserted by a trained health worker by ultrasound, chest X-ray or fluoroscopy [8]. Long-term use of PORTs is recommended, while short- and medium-term use of PICCs is recommended [9–11].

    View all citing articles on Scopus

    The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.

    Funding: This work was supported by Locally Initiated Project (LIP) Award #41-127 from the Center for Clinical Management Research, VA Ann Arbor Healthcare System. VC is supported by Early Career Awards from the Society of Hospital Medicine and the Research Career Development Core of the Claude D. Pepper Older Adults Independence Center at the University of Michigan.

    Conflicts of Interest: None.

    Authorship: All authors had access to the data and played a role in writing this manuscript.

    View full text