Background: Central line-associated bloodstream infection (CLABSI) is a morbid and potentially lethal complication. National policies related to CLABSI mandate public reporting of this adverse event, with hospitals receiving penalties based on their CLABSI rates. Contemporary data suggest that peripherally inserted central catheters (PICCs) placed outside critical care settings are a large contributor to hospital CLABSI burden. However, how PICC-related CLABSI are documented in the medical record and thus ultimately reported is not well known. Using prospectively collected data from medical records across diverse Michigan hospitals, we examined documentation related to PICC-CLABSI to better understand this issue.
Methods: We used data from the Michigan Hospital Medicine Safety (HMS) Consortium to perform this analysis. Trained abstractors at each hospital review medical records for PICCs placed in medical patients in wards and critical care settings. Detailed abstraction of medical record data for patient (e.g., demographics, comorbidities), device (e.g., indication, dwell time, lumens) and outcomes (e.g. CLABSI) is performed and data are entered into a central registry. Because HMS is interested in capturing care quality, CLABSI is tracked using four methods: physician documentation in the medical record of suspected PICC-CLABSI; PICC removed with the documented reason for removal being suspected CLABSI; documentation of “line sepsis” or “line-bacteremia”; and CLABSI based on the National Healthcare Safety Network (NHSN) definition. For this analysis, we compared differences in frequency and types of CLABSI documentation, hypothesizing that most PICC CLABSI cases would not meet NHSN criteria.
Results: Of 34,779 PICCs, 419 (1.2%) met one of our criteria for CLABSI; 130 met more than one criteria for CLABSI (i.e., multiple documentation aspects were observed [102 with 2 criteria, 27 with 3 criteria and 1 observation with 4 criteria]). The mean time to CLABSI across all definitions was 16.97 days (median=14). The majority (80%) of all PICC CLABSI events occurred within 40 days of catheter insertion. In order of frequency, 31 cases (7.6%) of physician documented CLABSI were noted in our review (mean 16.9, median 14). CLABSI as defined by NHSN criteria was the next most common category, with 108 cases meeting the criteria (25.8% [mean 15.4, median 12]). A total of 157 cases (36.6%) of PICC CLABSI were captured based on documentation of “line sepsis or line bacteremia” in the medical record [mean 17.8, median 15]).” The most frequent category of CLABSI were cases where the PICC was removed for suspected CLABSI, which was indicated for 282 of the 419 cases (67.3% [mean 17.0, median 14]). In all of these cases, no microbiological data or testing was sent to confirm the diagnosis.
Conclusions: In this multi-hospital study employing diverse ways to track catheter outcomes, the majority of PICC-associated CLABSI occurred in the context of PICC removal without confirmatory workup or testing. This practice would not meet definitions for national reporting mandates, as no microbiologic or confirmatory testing would exist to support the definition. Thus, national estimates of PICC CLABSI may under-represent the true burden of disease. Policies aimed at better quantifying practices related to catheter removal or treatment of suspected CLABSI may lead to better reporting of PICC-CLABSI and accountability of hospitals.