Prevalence of Hospital-Onset Bacteremia Pre- and Post-Implementation of a Needleless Blood Sampling Device From Existing Peripheral Catheters

Abstract

Repeated access of peripheral intravenous (IV) devices theoretically increases the risk of bacterial exposure. PIVO™ (VelanoVascular) is a needleless, single-use device that enables blood sampling from an existing peripheral IV. The goal of this retrospective observational exploratory study was to evaluate the influence of PIVO use on rates of hospital-onset bacteremia and fungemia (HOB) by comparing HOB rates in the year before and after PIVO introduction in hospitals implementing PIVO and over similar time periods in “control” hospitals with no PIVO. Two hospitals implementing PIVO (Hospital 1, a large community hospital; Hospital 2, a tertiary oncology center), and 71 control hospitals were included. During the 1-year period before and after PIVO introduction, HOB rates decreased in hospitals 1 and 2 by 31.9% and 41.8%, respectively. Control hospitals that did not use PIVO had a 12.4% decrease in HOB rates. Multivariable logistic regression analyses found that PIVO was associated with a lower risk (Hospital 1 odds ratio [OR]: 0.63; 95% CI, 0.42-0.94) or no change (Hospital 2 OR: 1.05; 95% CI, 0.72-1.52) in HOB rates. Control hospitals also showed no change in HOB rates between the 2 time periods. These data do not support concerns about increased risk of bacteremia with PIVO.

INTRODUCTION

Bacteremia is a common complication in hospitalized patients and is associated with increased mortality and high costs.1–4 Central line-associated bloodstream infections (CLABSIs) are a mandated quality metric, but bloodstream infections from other sources also result in significant morbidity and cost.1,5 Hospital-onset bacteremia and fungemia (HOB) has been suggested as a more comprehensive quality metric,6–9 and at the time of this writing, has currently passed the National Quality Forum (NQF) Patient Safety Committee review as a recommended metric to the Centers for Medicare & Medicaid Services (CMS).10 Peripheral intravenous (IV) devices are documented sources of bacteremia.11 Accordingly, analyses of HOB cases potentially associated with peripheral IV devices provide an important basis for risk-benefit evaluations in the use of specific devices.

PIVO™ (VelanoVascular, San Francisco, CA) is a needleless, single-use, sterile device that attaches to a peripheral IV catheter system to collect a fresh venous sample without additional venipuncture.12 The PIVO flow tube is designed to extend beyond the suboptimal draw conditions around the IV to reach vessel locations where blood flow is optimal for aspiration. A major advantage of PIVO is that it allows for blood sampling without needle sticks, which may improve patient satisfaction and extend IV dwell time.13 Once blood collection is complete, the device is retracted, removed from the peripheral IV, and discarded. A potential concern is that multiple PIVO access events may result in increased aggregate bacterial exposure contiguous to the intravascular space for existing peripheral IVs.14,15 Additionally, the flow tube could conceivably be a conduit of commensal skin flora and, thus, lead to bacteremia.

Attribution of PIVO to bacteremia events has not been previously studied. This exploratory study compared facility-level bacteremia prevalence during the year before and after PIVO introduction.

METHODS

Study Design

This was a retrospective observational study using data from adults (≥18 years) in the BD Insights Research Database (Becton, Dickinson and Company, Franklin Lakes, NJ), which contains laboratory, medication, and admission/discharge data from US medical facilities.9,16 The study was approved as involving use of a limited retrospective data set for an epidemiology study and exempted from consent by the New England Institutional Review Board/Human Subjects Research Committee (Wellesley, MA) and conducted in compliance with Health Insurance Portability and Accountability Act requirements.

To evaluate the association between PIVO and bacteremia, HOB and commensal event rates at the facility level (the same level of reporting that will be used should HOB become a formal CMS metric) in the 1-year period prior to PIVO implementation were compared to rates in the year following PIVO implementation at 2 hospitals. Hospital 1 was a community hospital with >300 beds that implemented PIVO in January 2021. Hospital 2 had >200 beds and was a tertiary oncology center that implemented PIVO in March 2021. Control (non-PIVO) hospitals consisted of 71 acute care facilities with ≥200 staffed beds. Changes in bacteremia events at facilities were compared using March 1, 2021, the midpoint of PIVO implementation at hospitals 1 and 2, as the cutoff point for the pre- and post-period analysis. All adult patients (≥18 years of age) with an HOB or commensal event, as defined below, were included in these analyses. There were no exclusion criteria and no requirements for use of a particular peripheral or vascular access device.

PIVO Description and Use

PIVO is a needleless blood collection device that attaches to a peripheral intravenous catheter (PIVC) system.12,17 The device consists of an inner tube with a user-activated slider, proximal flexible tube with female luer, outer housing, and a clip-to-connect distal end, which attaches to the PIVC system. The female luer attaches to the blood transfer device or syringe. The inner tube is advanced to collect a blood sample. Once complete, the inner tube is retracted and the device is removed from the PIVC. Discard samples are collected, if indicated, based on the Infusion Therapy Standards of Practice (the Standards).18 Health care staff at hospitals planning to implement PIVO as a peripheral access option underwent robust and systematic training prior to PIVO use. In hospitals where PIVO was implemented, clinical staff determined the need for PIVO use in specific patients based on clinical acumen.

Outcomes

Hospital admissions with an HOB event were identified as the first positive pathogen isolate (as per the Centers for Disease Control and Prevention [CDC] Sepsis Toolkit pathogen list19) on day 3 of admission or after that was not present in blood cultures in the community-onset period. A commensal event (typically, but not always, a “contaminated culture”) was defined as an organism in the CDC commensal list isolated in a blood culture taken during the hospital stay.19

Statistical Analysis

The prevalence of bacteremia events was estimated at the facility level in the 1-year period prior to and after PIVO implementation and was reported per 100 admissions. The association between PIVO and bacteremia events was evaluated using logistic regression models adjusted for age, gender, length of stay, COVID-19 status, readmission status, and blood culture collection during the first 2 days of hospital admission. Analyses were conducted using R (version 4.1.2, R Foundation for Statistical Computing, Vienna, Austria), with RStudio (Boston, MA).

RESULTS

This study included 73 hospitals (2 with PIVO, 71 control [without PIVO]) with over 2.2 million hospital encounters. The overall median patient age was 61 years; 45% were male, and 12% of encounters had an intensive care unit admission. During the study period, there were 253 HOB and 747 commensal events at PIVO hospitals and 6845 HOB and 44 183 commensal events at control hospitals.

Changes in Rates of HOB and Commensal Events

In unadjusted analyses, both the PIVO and control hospitals had a reduced prevalence of HOB and commensal events during the post-PIVO compared to pre-PIVO period (Figure 1), but the relative and absolute reductions were larger at PIVO hospitals. In multivariable adjusted analyses, the HOB prevalence was significantly lower in the post-PIVO compared to the pre-PIVO period at Hospital 1 and were not significantly different at Hospital 2 or control hospitals (Figure 2). Rates of commensal events were significantly lower post-PIVO at both Hospital 1 and Hospital 2. The control hospitals did not have a statistically significant change in rates of commensal events.

Figure 1.

Prevalence of (A) hospital-onset bacteremia and (B) commensal events in unadjusted analyses of PIVO hospitals and controls.

Figure 2.

Multivariable-adjusted association between PIVO and hospital-onset bacteremia or commensal events at Hospital 1, Hospital 2, and control hospitals.

Given that this study spanned months impacted by the COVID-19 pandemic, multivariable analyses included adjustments for COVID-19 status. Hospital 1 had higher COVID-19-positive rates in the pre-PIVO versus the post-PIVO period (9.3% vs 7.6%), while Hospital 2 (1.0% vs 1.6%) and the control hospitals (5.5% vs 5.7%) had lower COVID-19-positive rates in the pre-PIVO versus the post-PIVO period. Despite these differences, unadjusted analyses showed similar trends in HOB and commensal rates for the hospitals involved in these studies (Figure 1).

DISCUSSION

In this observational exploratory study comparing bacteremia in the 1-year periods prior to and after PIVO implementation, one hospital had a significantly lower risk of HOB after PIVO introduction, while another hospital had no change. At both hospitals, there was a significantly lower rate for commensal events in the post-PIVO compared to pre-PIVO period. These data do not support concerns that PIVO may result in excess risk for total bacteremic events.

Given the high patient burden and financial costs associated with HOB,1–4 the impact of PIVCs on HOB rates is a key concern. A recent study estimated that 63% of HOB cases in US hospitals were potentially preventable.6 Although HOB is not currently a reportable metric and, therefore, longitudinal data are limited, overall HOB rates tend to trend in concert with CLABSI rates.20 In response to focused efforts from various organizations,14,15,21,22 CLABSI rates decreased by 8% between 2015 and 2021 in US hospitals.23 However, more recent reports indicate a 7% increase in CLABSI rates between 2020 and 2021.24 These data indicate that continued efforts to decrease bacteremia in hospitals are required. At the time of writing, the NQF Patient Safety Committee had reviewed and recommended the CDC proposal to incorporate the HOB measure into a CMS metric.10 If this comes to pass, novel ways of mitigating total bacteremia events will be needed in nursing repertoire and workflows, particularly during surge capacity times such as cold and flu season or a full pandemic.

As an ecological study designed to look at the association to a facility-level metric (eg, HOB), the evaluations in this study were not limited to patients with PIVO. However, it is plausible that the absence of an increased risk for bacteremia in this study can be partially attributed to PIVO, as it reduces the number of venipunctures for laboratory testing. Fewer venipunctures may mitigate clinical risk of iatrogenic introduction of pathogens into the perivascular space, which can sometimes evolve to a true bloodstream infection.25 One study found the risk of CLABSI in neonates increased by 16% for each peripheral venipuncture in the presence of an indwelling peripherally inserted central catheter.26 Patil et al27 also noted increased rates of central venous catheter-related infections in the setting of increased numbers of placement attempts. Notably, the oncology hospital using PIVO did not see a statistically significant change in HOB. This finding is important, given that oncology patients are more likely to require frequent venipuncture for laboratory testing and to be more susceptible to infection and suggests that PIVO may be a potentially safer solution in this high-risk population. Needleless peripheral access has other potential benefits that may be well suited to the hospital environment. In particular, PIVO is comparable to venipuncture in providing high-quality samples12 and is associated with decreased preanalytical errors,28 reduced peripheral IV catheter replacement rates,28 and lower hemolysis rates29 compared with other collection methods. The majority of patients report no pain during PIVO use and a strong preference for PIVO access compared with needle-based collections.12,13

The time frame of this study spanned months impacted by the COVID-19 pandemic, and recent studies have reported higher rates of hospital-acquired infections, including bloodstream infections, associated with increases in the number of hospitalized COVID-19 patients during this time.30–33 These increases may be related, in part, to higher device utilization and longer periods of hospitalization in patients with COVID-19.33 To mitigate this impact, the multivariable analyses used in this study adjusted for COVID-19-positive rates at the hospitals. It should be noted, however, that unadjusted (observed) analyses of HOB and commensal rates showed similar pre- and post-PIVO period trends in the 2 PIVO hospitals, despite differing COVID-19-positive trends, suggesting that pandemic-related effects are unlikely to fully account for the changes in bacteremia rates observed in these hospitals.

LIMITATIONS

A limitation of this analysis is that it is based on only 2 hospitals, and the 2 PIVO hospitals had a higher baseline prevalence of bacteremia events compared to control hospitals prior to the implementation of PIVO. Future studies of hospitals with similar baseline burdens and characteristics may improve the generalizability of the current findings. In addition, this study evaluated the hospital-level effects (given that HOB has been proposed by the CDC as a CMS facility-level reportable metric) of PIVO use rather than specific infections or infection rates in patients utilizing PIVO. Additional studies designed to assess specific PIVO-related risks should be considered.

CONCLUSION

PIVO was not associated with a higher risk for HOB or commensal events and may be a valuable option for hospitals seeking fewer patient venipunctures for laboratory testing. Clinical experience leveraging new technology with patient satisfaction and outcomes has become important in regulated health care environments. As bacteremic and fungemic events cause morbidity and mortality and may become potential national quality metrics, programs that decrease risk while improving patient satisfaction may be of value.