By Carol A. Kemper, MD, FACP

Clinical Associate Professor of Medicine, Stanford University, Division of Infectious Disease, Santa Clara Valley Medical Center

Dr. Kemper reports no financial relationships relevant to this field of study.

Smarter Use of PICC Lines

SOURCE: Paje D, Rogers MAM, Conlon A, et al. Use of peripherally inserted central catheters in patients with advanced chronic kidney disease. A prospective cohort study. Ann Intern Med 2019;171:10-18.

Of the different types of vascular access required for hemodialysis, the use of an autogenous arteriovenous fistula (AVF) has been shown to have the fewest long-term complications and reduced mortality. As a result, maintaining vascular access in patients with chronic kidney disease (CKD) for possible future AVF placement should be an important goal of clinical care. National guidelines recommend thoughtful placement of peripherally inserted central catheters (PICCs) in patients with CKD, only when PICC use is truly necessary and appropriate and when it carries clear advantages over other types of vascular access. I call it maintaining their vascular “real estate,” since every PICC line placed has a potential risk for venous thrombosis and stenosis. And yet, PICC lines are being used more often, and many physicians may not be aware of the premonitory warning in patients with CKD.

In this longitudinal, prospective study, the use of PICC lines in 52 Michigan hospitals from 2013 to 2016 was examined, including the risk of complications and their use in patients with CKD. During the study period, 20,545 patients had PICC lines placed, including 62% placed on the wards, 32% in an intensive care unit (ICU), and 1.5% in the emergency department. The median age was 65.1 years, and half of the patients were female. The most common indications for PICC placement were the use of parenteral antibiotics (37.6%), difficult venous access (21.5%), and medications requiring central vascular access (12.4%). Two-thirds were placed by a vascular access nurse, and 20% by interventional radiology; 69% of the devices were 5 French or larger, and 63% were multilumen. The median duration of use in hospital was 11 days. One in four PICCs were used for short duration (< 5 days).

Nearly one in four (23.1%) PICC lines were placed in patients with diminished glomular filtration rate (GFR < 45 mL/min/1.73 m2), including 32.1% of all PICCs placed in the ICU and 18.9% of those placed in patients on the wards. The proportion of PICCs placed in patients with diminished GFR varied considerably between hospitals (from 23.7% to 37.8% in the ICU, and 12.8% to 23.7% on the wards). At the time of PICC placement, 699 (3.4%) patients already required hemodialysis. On the wards, a higher proportion of PICCs were placed in women, and PICC line placement was associated with increasing body mass index.

Complications occurred in 17.8% of patients with PICC lines, including, in decreasing order of frequency, catheter occlusion, venous thrombosis, and catheter-associated bacteremia. PICC complications were more frequent in the ICU setting, and occurred more often with multilumen devices.

Acute care facilities should consider methods to alert physicians to the overuse of PICCs in patients with CKD, either by the use of embedded guidelines in the electronic medical record (EMR), automatic consultation with a nephrologist, or the use of vascular access specialty teams or a trained vascular specialist nurse, who can provide clinical assessment and alternate recommendations. Based on the increased risk of complications compared with single-access devices, the routine use of multilumen PICCs also should be reconsidered. Rather than automatic ordering of a PICC line, central lines remain the recommended choice for short-term vascular access (< 5 days), especially in patients with hemodynamic instability or those requiring pressor support. 

Urine Cultures: A Gateway to Antibiotic Overuse

SOURCE: Brown KA, Daneman N, Schwartz KL, et al. The urine-culturing cascade: Variation in nursing home urine culturing and association with antibiotic use and C. difficile infection. Clin Infect Dis 2019. doi: 10.1093/cid/ciz482. [Epub ahead of print].

Inappropriate use of antibiotics in nursing homes across North America continues to drive antibacterial resistance and the risk of Clostridioides difficile infection (CDI) in the elderly. In my experience, one of the greatest offenses is the misinterpretation of urinalysis and culture results (a problem also common to the clinic and the acute care setting). These authors conducted a retrospective review of nursing home residents in Ontario, Canada, from 2014 to 2017. An assessment of nursing home residents was made once quarterly, documenting the proportion with a urine culture within 14 days of the assessment, the receipt of antibiotics within 30 days, and the incidence of CDI within 90 days. The researchers examined specific antibiotic use and noted those antibiotics commonly used for urinary tract infection (UTI), including ciprofloxacin, nitrofurantoin, and trimethoprim-sulfamethoxazole (TMP-SMX). Given the national healthcare system, information about diagnosis and antibiotic use was available for about 91% of all nursing home residents in the province. CDI was based on both a diagnosis and the use of medication prescribed for CDI.

The analysis included 131,218 residents from 591 nursing homes. Of these, 71% were female, and 55% were older than 85 years of age. Of these, 7.9% had a urine culture obtained within 14 days of the quarterly assessment (ranging from 3.4% for the lowest 10th percentile of facilities and up to 14.3% for the 90th percentile). Seventeen percent of residents received an antibiotic within 30 days, and 5.4% received an antibiotic commonly used for UTI. Antibiotic use within 30 days of assessment included, in decreasing order, ciprofloxacin (2.7%), cephalexin (2.6%), TMP-SMX (2.5%), nitrofurantoin (2.2%), augmentin (2.1%), and other fluoroquinolones combined (2.6%).

Urine culturing strongly predicted an increased use of antibiotics and explained 40% of the variability in antibiotic use between facilities, especially the use of nitrofurantoin. With every doubling of urine culturing, there was a 1.22-fold increase in total antibiotic use and a 1.36-fold increase in the use of antibiotics for UTI. CDI was diagnosed in 2,181 residents (1.6%) within 90 days. After adjusting for various factors, the rate of urine culturing at a facility was associated with a risk of CDI (for every doubling of culturing, the IRR for CDI was 1.18). Within 30 days of the quarterly assessment, 2.6% of residents died. Urine culturing was associated weakly with mortality.

These data suggest that different nursing home facilities have “a culture” of culturing urine, which drives the use of antibiotics at that facility. It’s amazing to me that 17% of nursing home residents received an antibiotic every three months. The use of interventions and education to ensure appropriate collection of urine specimens and the interpretation of those results would go a long way to reducing inappropriate antibiotic use in nursing homes. For one thing, I always maintain that obtaining a clean catch in an elderly person is virtually impossible. Imagine being older than 85 years of age, having to balance yourself while you hold your labia open with one hand, clean yourself with the other hand, and then catch the midstream urine without toppling over. My guess is that most of these specimens represent simple contamination. Furthermore, asymptomatic bacteriuria in elderly patients is not uncommon and does not require treatment.

How Can Vancomycin Dosing Be Improved?

SOURCES: Lodise TP, Rosenkranz SL, Finnemeyer M, et al. The emperor’s new clothes: Prospective observational evaluation of the association between initial vancomycin exposure and failure rates among adult hospitalized patients with MRSA bloodstream infections (PROVIDE). Clin Infect Dis 2019. doi: 10.1093/cid/ciz460. [Epub ahead of print]. Rodvold KA. 60 plus years later and we are still trying to learn how to dose vancomycin. Clin Infect Dis 2019. doi: 10.1093/cid/ciz467. [Epub ahead of print].

Despite 60 years of experience, vancomycin dosing remains problematic, nephrotoxicity continues to occur, and clinical outcomes are not always positive, especially with increasing bacterial minimal inhibitory concentrations (MICs). Our facility provides experienced pharmacy specialists to dose vancomycin, based on daily assessments of trough concentrations, and despite their best efforts, they still cannot get it right.

How can we do better? The answer is it’s possible but not easy. There are improved methodologies, but most facilities continue to base vancomycin dosing on serial trough concentrations. The key to successful vancomycin dosing is the sufficient amount of drug without risking nephrotoxicity. It sounds straightforward, but the problems are as follows:

  • Vancomycin MICs vary depending on methodology (broth microdilution vs. etest) and are not always consistent;
  • Generally, vancomycin treatment is most successful when bacterial MIC < 1; alternate treatments are recommended for infections with MIC > 2 mcg/mL. The in-between MICs are most challenging;
  • While higher MIC values for certain organisms can affect the success of therapy, the type of infection, the bacterial load, and the presence of a device and/or biofilm also influence the success of therapy;
  • Higher concentrations of vancomycin (i.e, troughs 15-20 mcg/mL) are associated with improved outcomes in some patients, especially those with higher MICs (i.e., methicillin-susceptible Staphylococcus aureus and methicillin-resistant S. aureus);
  • This success comes at the expense of nephrotoxicity, which occurs more often at trough concentrations > 15 mcg/mL, and definitely > 20 mcg/mL;
  • Because of variable protein binding and other physiologic factors, trough concentrations do not always correlate with drug exposure; since protein binding varies considerably (10-55%), true drug exposure is difficult to assess;
  • Area-under-the-curve (AUC) dosing may provide the best clinical success and can minimize nephrotoxicity; unfortunately, trough concentrations are a poor marker for AUC.

The authors of these two articles argue that the use of 24-hour AUC was linked more often to clinical success with lower rates of nephrotoxicity, and presently it is the most optimal approach to dosing vancomycin. Maintaining AUC1-24 between 400 and 600 mcg-hr/mL was associated strongly with clinical success, especially for infections with bacterial MIC < 1 mcg/mL. The two most common methods for determining AUC include Bayesian computer models (using as little as 1 or 2 vancomycin concentrations) and log-linear PK computations. Several Bayesian computer programs are available for use, but some training and experience are required. For this reason, while a few academic institutions have switched to AUC-based dosing, it is unlikely that many academic and acute care facilities will make the switch from trough-based dosing, despite its flaws, to the more optimal AUC dosing.