Less than 48 Hours: Earlier Time to Positivity for Central Line Cultures

1
Less than 48 HoursEarlier Time to Positivity
for Central Line Cultures

• Kenneth Chen, M.D.
• PL-2, Department of Pediatrics
• UT-Southwestern Medical Center at Dallas

2
Abstract

• One of the liabilities of the widespread use of
  central venous catheters is the risk of
  catheter-associated infection. Febrile children
  with such catheters are routinely admitted for 48
  hours for observation until bacteremia can be
  safely ruled out. Current microbiological
  techniques have resulted in earlier time to
  positivity in bacterial cultures. This review
  presents the rationale for safely declaring
  central line cultures negative in stable children
  at 24 to 36 hours.

3
Case Scenario

• 2 yo boy born at 24 weeks, with short gut
  syndrome, and now TPN-dependent, presents with a
  one-day history of fever to 101.6 and URI
  symptoms.
• Slightly dehydrated, but tolerating G-tube feeds
• Admission labs benign
• Start broad-spectrum antibiotics
• Admit and await culture results

4
Case Scenario

• At 36 hours
• Child looks well
• Fever has broken
• Cultures have shown no growth so far
• Clinical Question
• Is it safe to discontinue antibiotics in stable
  febrile children with central lines in less than
  48 hours?

5
Background

• Before 1980s, culture bottles were manually
  checked once a day
• Antibiotics kept for at least 72 hours
• The first automated culture systems checked
  cultures twice a day
• Empiric antibiotics shortened to 48-72 hours
• Now blood cultures are monitored continuously,
  every 10 minutes

6
Time to Positivity

• All previous studies on time to positivity in
  bacterial cultures were done in healthy patients
• Patients with central lines excluded
• Outpatient settings
• Most cultures grew S. pneumoniae
• Generally concluded that positive cultures are
  detected much earlier than 48 hours

7
McGowan et al. (2000)

• Time to positivity of cultures taken in
  ED/outpatient clinics
• Prospective study (n10,200)
• 258 cultures grew pathogens
• 87 of pathogens turn positive by 24 hours
• 92 by 36 hours
• 95 by 48 hours

8
Alpern et al. (2000)

• Compared mean time to positivity in pathogens vs
  contaminants
• Retrospective cohort study (n5901)
• Pathogens 14.9 h
• Contaminants 31.1 h
• 93.7 of pathogens grew in less than 18 hours
• Bacteria that grew in less than 18 h were 13
  times more likely to be a pathogen

9
Neuman and Harper (2001)

• Looked for factors that might influence time to
  positivity
• Retrospective chart review of 178 cultures which
  grew S. pneumoniae
• Mean time to positivity 11.5 h (range 4.4-25.9)
• 90 positive by 14 h
• No factors associated with faster time to
  positivity (including immune status)

10
Searching PubMed

• Keywordscentral culture children
• Shah SS, Downes KJ, Elliott MR, Bell LM, McGowan
  KL, Metlay JP. How long does it take to "rule
  out" bacteremia in children with central venous
  catheters? Pediatrics. 2008 Jan121(1)135-41.

11
Shah et al. (2008)

• Retrospective cohort study 2000-2003
• Goal to look at time to positivity in children
  with central lines and identify associated
  factors
• Inclusion criteria
• A common pathogen isolated from one culture
  (e.g., S. aureus), OR
• Symptoms of systemic illness with the same
  commensal bacteria isolated in two cultures
  (e.g., coagulase-negative staphylococci)

12
Methods

• 1,987 episodes of suspected line infection
• 387 had positive cultures
• 315 met inclusion criteria
• 200 episodes randomly selected by computer
• Number required to detect factors associated with
  time to positivity with 80 power, a0.05
• Looked for associated factors
• demographics, underlying condition, presenting
  symptoms, TPN, steroids, recent antibiotics

13
Methods

• Regression analysis
• Used data points to generate a best-fit line to
  predict chance of positivity at different times
• Hazard ratio (Cox proportional hazards model)
• Odds that a culture with a given factor will have
  a faster time to positivity than a culture
  without that factor
• Calculated for each associated factor

14
Most Common Isolates
15
Time to Positivity by Isolate
16
Results

• Median time to positivity 14.0 h (interquartile
  range 11.1-20.4 h)
• Gram negatives 12.8 h
• Gram positives 19.0 h
• 10 cultures became positive at gt48 h
• 5 skin commensals
• 3 mycobacteria
• Vomiting was the only factor associated with
  faster time to positivity (HR 1.9, CI 1.4-2.6)

17
Predicted Time to Positivity (Using Linear
Regression, 95 CI)
18
Early Discharge?
If children were discharged at 36 hours, only 1
of 1,987 children would be called back for
gram-negative bacteremia
19
Study Conclusions

• Central line infection can be safely ruled out in
  stable children at 24 to 36 hours
• gt96 detected within 36 hours

20
Caveats

• Study limitations
• 84 of gram positives grew within 36 hours
• Did not look at blood culture volume
• Larger blood culture volumes, drawn in larger
  children, are known to be more sensitive
• Age, used as proxy, was not found to be a
  significant factor
• Most patients did not have a simultaneous
  peripheral blood culture
• Retrospective analysis

21
Recommended Culture Volume
22
In Practice

• When is it safe to stop antibiotics early?
• Stable, afebrile
• Otherwise healthy
• Alternate explanation for fever (e.g., otitis,
  URI)
• Good follow-up

23
Reminders

• Its a lot easier to rule out a contaminant if
  you have two cultures
• Some pathogens always grow slow (e.g., Brucella)

24
References

• Shah SS, Downes KJ, Elliott MR, Bell LM, McGowan
  KL, Metlay JP. How long does it take to "rule
  out" bacteremia in children with central venous
  catheters? Pediatrics. 2008 Jan121(1)135-41.
• McGowan KL, Foster JA, Coffin SE. Outpatient
  pediatric blood cultures time to positivity.
  Pediatrics. 2000 Aug106(2 Pt 1)251-5.
• Alpern ER, Alessandrini EA, Bell LM, Shaw KN,
  McGowan KL. Occult bacteremia from a pediatric
  emergency department current prevalence, time to
  detection, and outcome. Pediatrics. 2000
  Sep106(3)505-11.
• Neuman MI, Harper MB. Time to positivity of
  blood cultures for children with Streptococcus
  pneumoniae bacteremia. Clin Infect Dis.
  2001331324-8.
• Revell, P. Personal correspondence.