Occult Bacteremia in Infants

1
Occult Bacteremia in Infants

• Current controversies and future developments
• Denise Watt
• Dec. 6, 2001

2
Outline

• background and epidemiology
• management algorithms
• evidence for Abx
• oral vs. parenteral
• antibiotic resistance
• pneumococcal conjugate vaccine

3
Case

• 10 month old girl, previously well
• URI symptoms x 10 days
• drinking well, wetting diapers, no N/V/D
• 2hr hx fever, lethargy, irritability
• O/E 180, 42, T39.2, 95
• looks unwell, moaning/crying, HEENT normal, clear
  BS, some indrawing, CVS normal, abd benign, no
  rash

4
Occult Bacteremia definitions

• FWS
• rectal temp ? 38?C, no focus, no obvious
• virus, non-toxic, no significant underlying
• illness/immunocompromise
• OB
• FWS and ve BC
• 10-20 PED visits for febrile illness
• 20 febrile children lt3yr no source

5
Epidemiology pre-HIB

• prior to early 1990s
• OB incidence 3-12 of FWS
• 60-85 S.pneumo
• 5-20 HIB
• 40 complication rate

6
Epidemiology post-HIB

• incidence of OB (FWS, 3-36 mos, T?39C)
• 1.6-2.8, highest age 1-2 yr (Kupperman 1998, Lee
  1998)
• 90-95 S.pneumo
• 96 ? invasive HIB lt5 yr (Alpern 2000)
• 5 non-typhoid Salmonella
• others Neisseria, GAS, GBS, Moraxella, E.coli,
  S. aureus

7
Implications of OB

• 10 SBI if untreated, 17 persistent bacteremia
  (Harper, Baraff)
• meningitis 1 (Baraff), 2.7 (Rothrock)
• 7.7 mort, 25-30 neuro sequelae
• overall risk of meningitis in untreated FWS
  0.02-0.05
• natural course of OPB?
• 96 resolve without Abx (Alpern 2000)

8
Occult BacteremiaSubsequent development of focus
Dashefsky, J Pediatr 1983., Shapiro, J Pediatr
1986., Woods, AJDC 1990., Baraff, Pediatr Infect
Dis J 1992.
9
Local Microbiology

• S. pneumo bacteremia rates vary widely across
  Canada
• related to rates of BC drawn
• rate in Calgary unknown
• 30 OPB/yr, 10 SBI/yr, 4-5 meningitis/yr
• 20 cases invasive HIB/yr (most adults)
• 139 BC last year age 1-15 (all comers)
• 27 contaminants

10
Predicting OB

• Hx and PE unreliable
• may appear well
• subjective vs. objective toxicity
• YOS gt10 sensitivity 77, specificity 88
• age
• fever
• OPB rare if temp lt39.0 (lt1) 3.7 if gt40
• similar response to defervescence OB (Baker
  1989, Bonadio 1993)

11
Predicting OB

• lab tests insensitive
• U/A most common occult bacterial infection (2
  febrile lt5yr)
• WBC gt15x109 sens 67-80, spec 69
• ANC best predictor (Kuppermann 1998)
• gt10x109 sens 76, spec 78
• band count unhelpful
• one poke most practical (CBC hold BC)

12
Blood Cultures

• 12 ves return for F/U before BC result, 50
  called back (Joffe 1992)
• time to ve 36hr, time to F/U 43hr
• most pathogens ve lt 18hr
• F/U more important than BC
• 76 SBI or PB called back (Bachur 2000)
• BC allow earlier F/U and Rx
• faster lab techniques coming?

13
Approach to FWS

• lt2-3 mos, 3-36mos, gt3 yr treated differently
• lt1980s, all pt lt3mos admitted for septic W/U and
  empiric Abx
• low risk criteria developed to avoid hospital
  admission

14
Low-risk criteria

• Rochester criteria 1985 (lt2 mos)
• NPV 98.9, PPV 12
• Boston criteria 1992 (lt3 mos)
• NPV 95
• Philadelphia criteria 1993 (1-2 mos)
• NPV 99.7, PPV 14

15
Baraff

• expert consensus (Pediatrics 1993)
• 1-3 mos, low risk
• option 1 septic W/U and Abx
• option 2 urine CS and observe
• 3-36 mos, non-toxic septic W/U if Tgt39.0
• update (Annals Emerg Med 2000)
• 3-36 mos
• T?39.0 U/A T?39.5 WBC ? BC (send if gt15)
• if empiric Abx, do LP!

16
Bachur 2001

• Recursive partitioning model
• U/A first step
• WBC lt4 or gt20
• T gt 39.6
• age lt 13d
• 82 sensitive
• admit 28 (vs. 53 with Rochester)

17
Cost-Effectiveness of FWS strategies

• 1990s BC and empiric Abx for all
• Lee (Pediatrics 2001)
• FWS, age 3-36 mos, OPB (1.5)
• meningitis 1 outcome
• incl. health care and societal costs
• CE CBC selective BC Rx if WBC ?15
• 30,800 / life-year saved
• if rate OPB?, less aggressive aproach

18
Why guidelines need re-evaluation

• controversy among experts
• lower incidence of OB
• elimination of HIB
• cost and complications of tests and Rx
• pen-resistant S. pneumo
• not followed anyway (Finklestein 2000)
• vaccine..

19
Antibiotics and FWS

• Only 2 prospective RCTs with placebo
• both small, pre-HIB
• Jaffe 1987
• no change in SBI
• Abx ? fever, improved appearance
• large, retrospective study (Harpur 1995)
• more focal infection, admissions w/o Abx
• Abx and meningitis (meta-analysis Baraff)
• no Abx 5.8 oral or parenteral Abx 0.4

20
Rothrock 1997 Meta-analysis

• not all RCTs, underpowered
• no significant ? meningitis
• significant ? SBI (OR 0.35 p0.003)
• NNT to prevent 1 meningitis 651
• NNT to prevent 1 SBI 2190
• NNH with Abx for every meningitis prevented 567
• no prospective studies post-HIB

21
Oral vs. Parenteral Antibiotics

• Fleisher (1994)
• no sign difference in focal infections
• ? persistent fever with Ctx
• not blinded, not intention-to-treat, pre-HIB
• Rothrock (1997) meta-analysis
• meningitis OR0.67 (oral vs. parenteral)
• SBI OR1.48
• closer F/U with parenteral

22
Risks of Empiric Antibiotics

• cost (tests, Rx, F/U, hospitalization)
• side effects
• discomfort of tests, treatment
• altered presentation (Rothrock 1992)
• development of resistant strains
• missed/partially Rx focal infections
• parental preference?
• will accept small risk of SBI vs. discomfort of
  tests Rx (Kramer, Oppenheim)

23
Penicillin-resistant Pneumococcus

• Castillo
• San Diego 1991-8 18 pen resistance
• 14 int. resistance 1991, 42 in 1998
• no difference in mortality
• NS increased resistance with prior Abx use

24
Pen and Cephalosporin resistance

• Silverstein
• 11 year review 8 resistance
• no diff in outcome, LOS in pen-resistant
• Ceftriaxone-resistant more focal infection, more
  LPs, more febrile at F/U, more admitted (NS), ?
  HR and temp at presentation

25
Antibiotic resistant Pneumococcus in Calgary

• 15 pen resistance
• lt2 amoxicillin resistance
• 10 Cefuroxime resistance
• 3-4 Ceftriaxone resistance
• need higher MIC for CNS
• clinically, has not been an issue

26
Conjugate Pneumococcal Vaccine

• heptavalent, 4 doses 2,4,6,12-15 mos
• FDA approval Feb 2000 (Prevnar)
• 3 RCTs of safety and immunogenicity
• Rennels (1998)
• Shinefield (1999)
• Black (2000)
• efficacy 97, intention-to-treat 94
• including ALL S.pneumo serotypes 89
• similar SE as DPTP/HIB, none severe

27
Pneumococcal Vaccine

• significantly ? OM
• Black ongoing trial on herd immunity
• long-term efficacy?
• strain selection?
• Bottom line
• will significantly decrease burden of S.pneumo
  disease
• likely lag time to change practices

28
Impact Of Prevnar in N. California33,000 with
1 dose Feb 2000-Mar 2001
Shinefield et al. 3rd Intl PID Conference
Monterey, 2001
29
Pneumococcal VaccineCost Effectiveness

• Lieu (JAMA 2000)
• cost lt savings if each dose lt46 (US)
• present 56 (US) 278,000/life-yr saved
• gt2x savings for society vs. health payer
• ? 760 million/3.8M infants/yr in US
• most from parental work loss, ? productivity
• Calgary 110/dose (84 at ACH)
• current immunization budget 17M/yr
• cost of SP vaccine 13M/yr

30
Occult Bacteremia Summary

• age, temp, appearance important
• dont forget U/A
• save labs for unwell
• faster BC techniques in distant future
• F/U most important tool
• empiric Abx have very limited role
• no clear evidence favouring parenteral

31
Occult Bacteremia Summary II

• antibiotic-resistance is rising impact small in
  Calgary
• vaccine WILL change the face of FWS
• Its viral!
• until then, the controversy continues!
• Are you a risk-minimizer or test-minimizer?
• (Green, Rothrock. Annals Emerg Med. 1999)

32
Case revisited

• WBC 14.9
• ANC 8.3
• BC ve S.pneumo in 24hr (pen I)
• R/A looks well, T 38.5
• Mgt?

33
Case cont.

• Ceftriaxone IV
• F/U ID clinic
• well-looking
• Ctx IV x 3 days, then Amoxil x 7 days

34
QUESTIONS?