The meaning of positive blood cultures taken from a CVC

What is a clinically significant coagulase-negative staph bacteraemia? This has exam relevance. Question 14 from the second paper of 2017 asked various detailed questions about coagulase-negative staphylococci, and  Question 23.1 from the first paper of 2013 asks the candidate how they would react to a S.epidermidis growing from a culture which was drawn from a central line in the process of insertion. For a coherent answer, the time-poor exam candidate is redirected to the LITFL article on CRBSI and CLABSI. That is what you call a "brief summary". In contrast, what follows here may represent spontaneous prose.

Blood cultures taken from central lines

So, your CVC culture is positive. What could this mean?

Generally speaking, the practice of taking cultures from CVCs is discouraged by guideline-makers (eg. the American College of Physicians).

Arguments against taking cultures from the central line

• Cultures collected during CVC insertion are more likely to yield contaminants than peripheral blood cultures (8% of the time vs. 3% of the time)
• The resulting false positive results can increase laboratory costs by 20% and antibiotic costs by 39%, not to mention the costs involved in resiting the line, and the increased risk of repeated central line cannulation.
• Almost half of the patients with coagulase-negative staphylococci in their CVC cultures end up getting antibiotics (usually vancomycin) which, apart from its expense, can't be good for the development of antibiotic resistance.

However:

• There are no good quality studies to confirm that these concerns are valid; none of the available studies were blinded, and there is thus far no "gold standard" definition of bacteraemia against which they could be held.

Discriminating false positives from true positives

A recent (2006) review article suggests some strategies to help you tell which CVC culture warrants CVC removal, and which does not.

Identity of the organism

• This seems to be the most important feature.
• Audits of blood culture results have revealed a series of organisms which almost always represent a true systemic bacteraemia or fungaemia:

Organisms which Usually Represent "True" Bacteraemia:

Staphylococcus aureus Streptococcus pneumoniae Streptococcus pyogenes Streptococcus agalactiae Escherichia coli Enterobacteriaceae Pseudomonas aeruginosa

Listeria monocytogenes Neisseria meningitidis Neisseria gonorrhoeae Haemophilus influenzae Bacteroides fragilis Cryptococcus neoformans Candida albicans

• Conversely, there is also a selection of organisms which almost always represent contamination:

Organisms which Usually Represent CVC Colonisation:

Corynebacterium sp. Bacillus sp. other than Bacillus anthracis Propionibacterium acnes Micrococcus sp.

Viridans group streptococci Enterococci Clostridium perfringens Coagulase-negative staphylococci.

Coagulase negative Staphylococci:

Coagulase-negative staphylococci represent an area of uncertain twilight between these two groups. For most, these are benign contaminants. For some, they are horrific valve-eating pathogens. Practically speaking, it does not matter to the intensivist whether the blood culture represents systemic bacteraemia or catheter colonisation; rather it is more important to determine whether the clinical scenario warrants concern. If you're concerned, you will remove the line and start antibiotics.

 Question 23.1 from the first paper of 2013 asks the candidate how they would react to a S.epidermidis growing from a culture which was drawn from a central line in the process of insertion.

• Coagulase-negative staph bacteraemia might sound quite rare (7 per 10,000 admissions) but is in fact the most common catheter-related bacteraemia.
• Significance of a coagulase-negative staph in a blood culture is determined by the presence of the clinical features of infection and in those cases one would be obliged to change the line. If it was a true bacteraemia, it will occur again in 20% of patients in whom the catheters remain in situ, compared to 3% in patients whose lines are changed.

Risk factors for clinically significant S.epidermidis bacteraemia:

Anyone at risk of native valve endocarditis Artifical valves History of rheumatic heart disease Immunosuppressed Implated pacemaker, or any other surgical implant Low birth infant Elderly person (>65 years of age)

In other words, people with the above risk factors are much more likely to have a serious problem if they really do have S.epidermidis in their blood; which means that the risk-benefit balance favours the removal of the CVC and the commencement of antibiotics.

Differential time to positivity

• When you draw a positive blood culture from the CVC, you are left guessing: is the CVC really contaminated with that organism, or did I just suck up bacteraemic blood through a nice clean CVC (and the source is elsewhere)?
• Differential time to positivity is a potential means of discriminating "true" CVC -related bacteraemia from systemic bacteraemia.
• One takes two sets of cultures: one from the central line, and another peripherally.
• Basically, if the CVC culture becomes positive earlier than the peripheral culture, the CVC is likely to be the source.
• Differential time to positivity of 120 minutes or more was associated with 81% sensitivity and 92% specificity for short term catheters. It is less specific in long-term lines.
• However, one ends up taking two sets of cultures, which is somewhat counterproductive if one's intention is to save money on laboratory fees.

Number of positive culture sets

• For true bacteraemias, multiple blood culture sets will usually grow the same organism .
• For patients with CVCs and a coagulase negative staph growing from different numbers of cultures, the positive predictive value was:
  • 55% when one of one culture was positive,
  • 20% when one of two cultures was positive,
  • only 5% when one of three cultures was positive.
• The abovementioned multiple culture sets should be taken from different sites. For two positive culture sets, the positive predictive value was:
  • 98% if both samples were obtained through the vein
  • 96% if one sample was obtained through a central line
  • only 50% if both samples were obtained through a catheter.