Background: Peripherally inserted central catheter (PICC) use in patients with chronic kidney disease (CKD) is associated with non-functioning arteriovenous fistulae. Existing guidelines recommend avoiding PICC placement in these patients but how often this occurs and how best to implement this recommendation is not known.

Methods: Within 42-hospitals participating in the Michigan Hospital Medicine Safety Consortium (a collaborative quality initiative), a specific goal to reduce PICC insertion among patients with Stage IIIb chronic kidney disease (eGFR <45 ml/min) was set. Hospitals were provided with resources including National Kidney Foundation Guidelines, appropriateness criteria and relevant toolkits for implementation of interventions to reduce the use of PICCs in patients with CKD. Hospitals were encouraged to implement educational initiatives, policies and consultation with nephrology prior to PICC use in patients with CKD. As well, changes to hospital ordering practices within the electronic medical record (EMR) were recommended. Rate of PICC placement in patients with CKD before and after implementation across HMS hospitals were examined.

Results: Among 29,061 patients with available eGFR data, 5,887 patients with CKD (eGFR <45ml/min) received PICCs. Of these, 842 patients were on hemodialysis when they received a PICC. The most common indication for PICC insertion in patients with CKD were antibiotic administration (42.6%), difficult access (33.2%), and medications requiring central access (22.6%). Of the PICCs inserted, 33.2% were single lumen, 47.6% double and 18.8% triple lumen devices. The median dwell time of PICC use was 9 days.
Baseline rates of PICC use in patients with CKD varied from 9.2% to 33.3% in hospitals before implementation (median = 21.1%, IQR = 17.5-26.9%). Following implementation, rates of PICC use in patients with CKD decreased (median=18.0%, IQR=13.0-21.1%, median of absolute difference = -3.7%, p=<0.001). The greatest decreases (though not statistically different) were observed in hospitals that implemented EMR changes (median of difference: -1.4% vs. -4.0%, p=0.11).

Conclusions: Use of PICCs in patients with CKD is common and problematic. A multi-modal approach can improve appropriate PICC use and EMR changes may represent the single best intervention. Further evaluation of strategies that limit use of PICCs in hospitalized patients with CKD appear necessary.