Management of the Febrile Infant Risk Minimizers vs. Test Minimizers

1
Management of the Febrile InfantRisk
Minimizers vs. Test Minimizers

• Steven E. Krug, M.D.
• SAEM Annual Meeting
• St. Louis, MO -- May, 2002

2
Fever, Whats The Big Deal ?

• 65 of children 0-2 years visit a physician for a
  febrile illness
• 10 to 20 of all pediatric visits to EDs
• 20 to 30 of pediatric office visits
• Fever without a source accounts for as many as
  50 of these visits
• A self limited illness in the vast majority
• A small percentage will have a SBI

3
Occult Bacteremia in Febrile Children How It
All Began

• McGowan JE, et al NEJM 1973 2881309
• Febrile children at Boston City Hospital
• 24 hour pediatric walk-in clinic
• Temp gt 38.3 0C -- all ages - 3 month study
• Results
• 10,535 visits ? ? 2165 children with fever
• 708 blood cultures ?? 31 (4.4) true positives
• Note 43 cultures (6.1) produced false positives

4
So, Are You a Risk-Minimizer ?Green SM, Rothrock
SG Ann Emerg Med 1999 33211

• Desire to lower the risk of adverse sequelae from
  occult infections - ROWS
• Do not believe that clinical evaluation is
  sufficient to reliably identify ill children
• Use risk stratification to target higher risk
  patient groups for intervention
• Believe that the potential benefit of reducing
  adverse sequelae justifies empiric diagnostic
  testing and treatment

5
If Not, Perhaps a Test-Minimizer ?Green SM,
Rothrock SG Ann Emerg Med 1999 33211

• Believe the occurrence of adverse outcomes is so
  low as to not justify time, expense and
  invasiveness of risk stratification
• Believe that clinical evaluation and follow up
  will serve to identify nearly all ill children
• Believe that parents prefer less testing and
  treatment
• Are willing to accept a greater chance of being
  wrong

6
Well Intentioned Risk Minimizers at Work Can We
Identify High Risk Children ?

• Demographic and clinical parameters
• age, temperature, petechiae
• Lab screening profiles
• CBC, ESR, CRP, UA, etc.
• OB/SBI risk appears to correlate with some
• sensitivity, specificity and positive predictive
  value for OB and SBI were less than ideal
• Clinical scoring systems (McCarthy, et al.)
• initially very promising, ultimately
  disappointing

7
Identification of High Risk PatientsSensitivity
of the Physical Exam

• McCarthy, et al. Pediatrics 1982 70802
• Yale Observation Score (AIOS)
• incidence of SBI related to score
• lt 10 2.7, gt 10 40.2, gt 16 92.3
• sensitivity 88, specificity 77, low PPV
• negative predictive value of normal Hx PE
  findings plus low AIOS was 96
• AIOS was fairly good in its ability to identify
  the sick, but perhaps even better
  in identifying the well ?

8
Failure of Clinical AssessmentTeach SJ, Fleisher
GR J Pediatr 1995 126877

• Prospective application of the Yale Obs. Scale
• children 3 - 36 months with T gt 39.0 0C
• 611 children in study, 192 with bacteremia
• median YOS (6) was the same for both groups
• YOS gt 10
• sensitivity 5.2 specificity 96.7,
• positive predictive value 4.5,
• negative predictive value 97.1
• The YOS has performed similarly in other
  recent studies

9
? Social Smile SBI ? Bass JW, et al Pediatr
Infect Dis J 1996 15541

• Do smiling febrile children have bacteremia?
• 512 children aged 3 to 36 months of age
• T gt 39.5 0C and WBC gt 15,000
• social smile associated with shorter fever
  duration
• smile present in 45 of bacteremic children
• smile present in 49 of non-bacteremic pts.
• no demographic or laboratory differences found
  between the groups
• Yikes !! - So, is that a smile, or is it a
  grimace ?

10
Lessons Learned From the Search for OB/SBI Risk
Factors for OB/SBI

• Age neonates, 28-90 days, 3-36 months
• Fever OB risk increases with temperature
• hyperpyrexia (T gt 40.5 0C) - 8 to 25 OB
• Petechiae - 15 to 20 occurrence of SBI
• Immunodeficiency - e.g. HIV, SCD
• WBC gt 15,000 Bands gt 1000
• 5 fold increased risk for occult bacteremia
• Ill patient or toxic appearence
• The big question -- Are all of these still
  true ??

11
Identification of Low Risk PatientsPhysical
Laboratory Screening Profiles

• Rochester Criteria Dagan et al J Pediatr 1985
  107855
• T gt 380C, term, well appearing, secure follow-up
• peripheral WBC between 5-15,000/mm3
• band count lt 1500 /mm3
• urinalysis with lt 10 WBC/hpf
• no evidence of ear, soft tissue or bone
  infection
• Performance 233 infants 0-2 months of age
• 1 of 144 (0.7) low risk infants had SBI
• 22 of 89 (25) high risk infants had SBI

12
Identification of Low Risk Patients Dagan R, et
al J Pediatr 1988 112355

• Modified Rochester Criteria
• added diarrhea to criteria
• if present lt 25 WBC/hpf on stool smear
• Performance
• 237 infants 0-2 months of age
• 0 of 148 low risk infants had SBI
• 21 of 88 (24) high risk infants had SBI
• So, perhaps we can identify the low risk children
  !

13
Identification of Low Risk PatientsBaker MD, et
al NEJM 1993 3291437

• CHOP Low Risk Criteria
• T gt 38.20C, well appearing, low IOS
• WBC lt 15,000 /mm3 BNR lt 0.2
• urinalysis with lt 10 WBC/hpf
• CSF with lt 8 WBC/mm3 (-) gram stain
• negative chest x-ray
• Performance 747 patients aged 29-56 days
• 64/65 patients with SBI noted as high risk
• 1/287 assigned to low risk had SBI
• OPD assignment saved 3,100/patient

14
CHOP Low Risk Protocol More DataBaker MD, et
al Pediatrics 1999 103627

• Three year study (1994-1996) at CHOP
• Infants 29 to 60 days of age with T gt 38.0 0C
• 422 infants
• 43 (10) with SBI
• UTI (4) OB (2.1) BM (1.2) BGE (1.2)
  cellulitis (1.2)
• 101 (24) identified as low risk
• no SBI in the low risk infants
• Note Over 8 years this protocol has shown a
  nearly perfect 100 negative predictive value for
  gt1200 infants

15
Are Febrile Neonates DifferentBaker M, Bell L
Arch Pediatr Adol Med 1999 153508

• Can laboratory screening profiles reliably
  identify febrile neonates with low risk for SBI ?
• applied CHOP protocol to 254 infants (3 - 28
  days)
• 43 of infants qualified for OPD management
• 32 infants (12.6) with SBI
• 17 UTI (6.7) 8 OB (3.1) 4 BM (1.6), BGE
  (.8)
• 5 low risk infants had serious infections
• would miss 20 infants with SBI per 1,000
• Yes, febrile neonates are indeed different

16
Identification of Low Risk PatientsBaskin M, et
al J Pediatrics 1992 12022

• The BCH Low Risk Criteria
• T gt 380C, low IOS, presence of secure follow-up
• peripheral WBC lt 20,000/mm3
• CSF WBC lt 10/mm3
• urinalysis dip with (-) leukocyte esterase
• all patients treated with ceftriaxone
• Performance 503 patients aged 28 -89 days
• 27 of 503 (5.4) who met the criteria had SBI
• 9 OB (1.8), 8 UTI (1.6), 10 BGE (2.0)
• all were treated and were well at follow-up

17
Components of Fever ProtocolsAvner J, Baker MD
EMCNA 2002 2049
18
Empiric Antibiotic TherapyDoes it Work?

• Carroll WD, et al Pediatrics 1983 72608
• Small study (10 patients) - PCN vs. placebo
• Difference between groups was not great
• Jaffe DM, et al NEJM 1987 3171175
• Large multi-center study - amoxicillin vs.
  placebo
• Enrolled 955 children 3-36 months with T gt 39.0
  0C
• 27 (2.8) with bacteremia -- small number of
  cases
• Outcome differences between groups were not
  great

19
Empiric Antibiotic TherapyDoes it Work?

• Bass JW, et al Pediatr Infect Dis J 1993
  12466
• Prospective study - augmentin vs. ceftriaxone
• 519 children aged 3-36 months -- 60 (11.6) with
  OB
• T gt 40.0 0C - or - T gt 39.5 0C and WBC gt 15 K
  
• Both Rx regimens appeared to be adequate
• Fleisher GR, et al J Pediatr 1994 124504
• Multi-center study - ceftriaxone vs. amoxicillin
• 6733 patients -- 195 (2.9 ) with bacteremia
• ...ceftriaxone eradicated bacteremia, had fewer
  focal complications, and less persistent fever

20
Is There a Cost Effective Strategy? Lieu TA, et
al Pediatrics 1992 891135

• Decision analysis, cost-effectiveness model
• 6 strategies for management of febrile infants
• 28 to 90 days with Temp gt 38.0 0C
• used data from literature (e.g. Baskin, Baker)
• Clinical judgment alone appeared to be the least
  effective clinical model and the 2nd least cost
  effective strategy
• Full sepsis W/U and outpatient IM ceftriaxone
  was judged to be the most effective strategy

21
Practice GuidelinesBaraff L, et al Ann Emerg
Med 1993 Pediatrics 1993

• Expert consensus panel recommendations
• Based on meta-analysis of the literature
• Fever is defined as gt 38.0 0C for 0-3 months and
  gt 39.0 0C for 3-36 months
• Infants at greatest risk during 0-3 months
• Rochester criteria selected as screening criteria
  for high vs. low risk
• Note Panel members confessed risk-minimizers

22
Consensus Panel Guidelines

• Toxic-Appearing Infants and Children
• Hospitalize, evaluate and treat for presumed
  sepsis, meningitis, or SBI
• This holds for all age groups
• THIS SHOULD BE A NO BRAINER

23
Consensus Panel Guidelines

• Febrile (low risk) Infants lt 28 days of age
• Despite low probability of sepsis and studies
  showing favorable outcome for outpatient
  observation, the panel recommends SBI evaluation
  and hospital admission for all infants with
  either parenteral therapy or close observation

24
Consensus Panel Guidelines

• Low-Risk Infants 28-90 Days of Age
• Obtain urine culture and provide close follow-up
  
• - OR -
• Full sepsis evaluation (blood, urine, CSF) and
  treat with IM ceftriaxone
• All children who receive presumptive therapy
  should have an LP

25
Consensus Panel Guidelines

• Low-Risk Infants 3-36 Months of Age
• Urine culture for males lt 6 mo females lt 2 yrs
• Stool culture if blood or mucus or gt 5 WBC/hpf
• Chest x-ray if decreased breath sounds or SOB
• Blood culture if T gt 39.0 0C and WBC gt 15,000
• Empiric therapy if T gt 39.0 0C and WBC gt 15,000
• No diagnostic tests or antibiotics if T lt 39.0
  0C

26
The Febrile InfantVariability in Management
Approaches

• Ros SP, et al. Pediatr Emerg Care 1994 10264
• Surveyed members of AAP Section on EM
• Numerous fever and age group definitions
• 74 routinely screen with a CBC
• 45 routinely draw blood cultures
• 36 use clinical appearance as basis for
  culturing
• 53 routinely administer antibiotics
• 44 use lab criteria as basis for antibiotic Rx
• Despite published guidelines, no clear standard
  of care!

27
What Do Parents PreferOppenheim PI, et al Ann
Emerg Med 1994 24836

• Interviewed parents regarding management options
  for febrile infant/child scenarios
• Parents successfully identified the strategies
  associated with a higher probability for an
  adverse outcome
• 71 chose options with less testing and treatment
  (and greater risk!)
• So, perhaps parents are test minimizers??

28
What do Parents PreferBennett JE, et al Arch
Pediatr Adol Med 2000 15443

• Survey of parent utilities for outcomes of OB
• convenience sample, single urban PED
• 94 subjects interviewed
• provided with 8 possible outcomes
• blood drawing viewed to be of minimal risk and
  concern
• Parents were intolerant of adverse outcomes
• Okay, so maybe parents are risk-minimizers

29
Parents, Physicians AntibioticsBauchner H, et
al Pediatrics 1999 103395

• Survey of AAP general pediatricians
• 610 responses (67)
• 40 indicated that parents frequently ask for
  antibiotic when the MD feels it is not needed
• 48 stated parents often pressure them to
  prescribe antibiotic therapy
• nearly 1/3 stated they occasionally or
  frequently comply with that pressure
• parental pressure viewed as 1cause of the
  unnecessary use of antibiotics

30
Risk Minimizers vs. Test Minimizers
- Published guidelines - ROWS - Risk of OB
sequelae - Parental preferences
- Cost of care - Changing Hx of OB/SBI - Risk
of testing - Risk of Rx - Parental preferences
To Test/Treat - or - Not To
Test/Treat
Attorneys, Payors, and Other Predators
31
Management of the Febrile InfantWhats
Controversial, Whats Changed?

• Eradication of Hemophilus influenzae
• what is the current risk of OB and SBI
• what is the natural history of pneumococcal OB
• do the 1993 consensus guidelines make sense
• Continuous blood culture monitoring systems
• True efficacy of empiric antimicrobial therapy
• The febrile infant with a viral infection
• OB/SBI risk with hyperpyrexia petechiae

32
Disappearance of H. influenzae

• Prior to introduction of Hib vaccine (1987)
• 10-15 of OB and majority of OB related SBI
• 12,000 cases/yr invasive H. flu in children lt 5
  yrs.
• 33-60 develop focal infection, 15-25 develop
  BM
• 12 times more likely than pneumococcus
• Currently about 300 cases per year (94/95)
• no longer a leading cause of sepsis/meningitis
• incidence now greatest in children lt 5 months
• Median age for BM 1986 15 mo. ?? 1995 25
  yrs.

33
Risk of Bacteremia in the Post-Hib EraLee GM
Arch Pediatr Adol Med 1998 152624

• Three year study (1993-1996) at BCH
• Children aged 3 - 36 months with T gt 39.00C
• 11,911 patients
• 75 received CBC, 74 had blood cultures
• 149 positive blood cultures (1.6)
• 92 pneumococcal
• no H. influenzae isolates!!
• Is BCH the center of the risk-minimizer universe
  ?

34
Bacteremia in Boston Lee GM Arch Pediatr Adol
Med 1998 152624

• Prevalence greatest in 6 - 24 month age group
• WBC and absolute neutrophil counts were the most
  accurate predictors for bacteremia
• WBC gt 15 x 109 -- Sens. 86, Spec. 77
• attributed to higher WBC with pneumococcal OB
• OB risk associated with temperature
• OR gt 40.0 1.9 gt 40.5 2.6 gt 41.0 3.7
• Lower OB rate not explained by Hib vaccine

35
Outcome of Pneumococcal BacteremiaBachur R,
Harper MB Pediatrics 2000 105502

• Re-evaluation of children in ED with OPB
• Nine year study at BCH (1987-96)
• 548 episodes of OPB
• 40 (7) with PB or new focal infection
• 14 PB(2.5) 8P(1.5) 8M(1.5)
  6C(1.0) 4 PC(0.7)
• patients not initially Rx, and those treated who
  remained febrile were are greatest risk for PB
• majority with OPB can be managed as outpatients
• Three diagnoses/cases of BM were
  controversial

36
Persistent Bacteremia/Meningitis in OPB Bachur
R, Harper MB Pediatrics 2000 105502
37
Prevalence/Outcome of Occult BacteremiaAlpern
ER, et al Pediatrics 2000 106505

• Three year retrospective study (1993-1996)
• 5900 children aged 2-24 months, T gt 39.0
• Prevalence of OB 1.9
• 83 pneumococcal H. influenzae not isolated
• Focal bacterial infections in 17 (0.3)
• pneumonia (8), cellulitis (4), osteo (2), others
  (3)
• Serious adverse outcome in 2 (0.03)
• meningitis (1), sepsis/death (1)
• Note 96 OB with spontaneous resolution without
  Rx

38
Occult Bacteremia in PhiladelphiaAlpern ER, et
al Pediatrics 2000 106505

• Mean time to culture shorter for true positives
• mean times true () 14.9 hrs false ()
  31.1
• lt 18 hours 13x more likely to be true pathogen
• Nearly all true positives re-evaluated in ED
• average time from notification 10.6 ( 9.7) hrs
• 33 were still febrile
• 53 admitted to the hospital
• 4.8 found to have persistent bacteremia
• oral antibiotics Rx at 1st visit did not affect
  rate of PB
• Perhaps blood cultures can be an effective
  screen?

39
Prevalence/Outcome of False () Blood Cultures
Alpern ER, et al Pediatrics 2000 106505

• Overall contamination rate was 2.1
• 85 were re-evaluated in ED
• 35 were still febrile and were admitted
• 1.9 of repeat cultures also contaminated!
• At least in Philadelphia, the risk of a
  contaminated blood culture equals or exceeds that
  of a true positive !
• Remember the data from McGowan, 1973?

40
Prevalence of False () Blood Cultures Alpern
ER, et al Pediatrics 2000 106505
41
Use of Antibiotics to Prevent SBIBulloch B, et
al Acad Emerg Med 1997 4679

• Meta-analysis of published RCCTs
• 4 studies Carroll, Jaffe, Fleisher, Bass
• Antibiotic use trended to ? risk for SBI
• odds ratio 0.60 (P.O.) 0.38 (I.M.)
• need to treat 414 kids to prevent 1 SBI case
• no significant effect of antibiotic therapy
• Concluded that widespread antibiotic use should
  not replace clinical judgement

42
Outcomes in Occult BacteremiaBulloch B, et al
Acad Emerg Med 1997 4679
Does empiric therapy truly reduce the risk for
SBI?
43
Predictors of Pneumococcal BacteremiaKuppermann
N, et al Ann Emerg Med 1998 31679

• With invasive H. influenzae infections out of the
  picture, are there unique predictors for OPB
• Multivariate analysis - 6,500 children 3-36
  months
• 164 children (2.5) with OPB
• Three variables retained association with OPB
• ANC OR of 1.15 for each 1,000 cells/mm3 ?
• if ANC gt 10,000 -- OPB rate 8.2
• temp OR of 1.77 for each 10 C ?
• age lt 2 years OR of 2.43 vs. 2-3 years of age

44
Band Counts in Young Febrile ChildrenKuppermann
N, et al Arch Pediatr Adol Med 1999 153261

• Compared CBC findings in febrile children with a
  documented SBI (bacteremia or UTI) versus a
  proven respiratory viral infection
• Children with SBI had a greater mean ANC
• 11.3 x 109 vs 5.9 x 109
• No differences in percentage band count or
  absolute band count between the groups

45
Identification of Children with UMDKuppermann N,
et al Pediatrics 1999 103e20

• Clinical/hematologic features of children with
  unsuspected meningococcal disease (UMD)
• retrospective, four center study,1985-96
• 381 children with meningococcal disease
• 45 (12) with UMD discharged home !!
• compared to 6400 culture negative children
• no difference in Temp, WBC, ANC
• significantly higher band counts in UMD
• predictive value of band count was low (PPV
  0.06)
• Bad news There is still no crystal ball for UMD

46
Febrile Children with BronchiolitisKuppermann N
Arch Ped Adoles Med 1997 1511207

• Evaluated risks of bacteremia and UTI in febrile
  children with/without bronchiolitis
• 432 children aged 0-24 months
• Children with bronchiolitis had significantly
  fewer positive cultures
• blood 0 vs. 2.7 urine 1.9 vs. 13.6
• none of the children lt 2 months of age with
  bronchiolitis (36) had bacteremia or UTI

47
SBI Risk in Children With Recognizable Viral
Syndromes

• Greene DS, Harper MB Pediatr Infect Dis J
  199918258
• Five year retrospective study (1993 -1998)
• Children aged 3-36 months with T gt 39 0C
• 1347 children diagnosed with a RVS
• croup, varicella, bronchiolitis, stomatitis
• blood cultures obtained in 65
• 2 of 876 (0.2) blood cultures were positive
• Office-based physicians have known this for a
  very long time.

48
Bacteremia in Fever PetechiaeMandl KD, et al
J Pediatr 1997 131398

• Prior studies suggest a high risk for bacteremia
• 7 to 11 incidence of meningococcemia
• Enrolled 411 children -- (58 3-36 mo.)
• 8 (1.9) with bacteremia or clinical sepsis
• six with serious invasive bacteremia
• none of 357 well-appearing children had OB
• toxic appearance had sensitivity of 100
• WBC gt 15 K or lt 5K had sensitivity of 100
• all children with meningococcemia had purpura

49
Occult Pneumonia in Febrile Children Bachur R,
et al Ann Emerg Med 1999 33166

• What is the incidence of occult pneumonia in
  febrile children with high WBC ?
• Prospective cohort ED study
• age lt 5 years, T gt 39 0C, WBC gt 20,000
• radiographs obtained in 225 of 278 patients
• positive radiographic findings in
• 40 of those with a suggestive clinical exam
• 26 of those without clinical evidence for
  pneumonia
• recommends empiric chest radiography

50
UTIs in Febrile InfantsShaw KN, et al
Pediatrics 1998 102e16.

• UTI is by far the most frequent SBI
• fever may be only presenting sign of UTI
• What is the prevalence of UTI in febrile infants
• 2400 febrile infants -- overall 3.3
• gender -- male 1.8 female 4.3
• race -- white 10.7 AA 2.1 others 5.7
• other source -- yes 2.7 no 5.9
• temperature -- lt 39.0 2.2 gt 39.0 3.9
• Should we screen all febrile children for UTI ?

51
UTIs in Febrile Children Gorelick, Shaw Arch
Ped Adol Med 2000154386.

• Developed clinical decision rule ?
• T gt 39.0 0C ? fever gt 2 days
• white race ? age lt 1 year
• absence of another potential source
• All with UTI had at least one risk factor
• Presence of any two factors
• sensitivity 95, specificity 31
• Rule eliminated 30 of unneeded cultures

52
Risk of SBI in Febrile Seizures Trainor J, et
al Clin Pediatr Emerg Med 1999 113

• Multi-center study of ED management of simple
  febrile seizures (Chicago, 1998)
• 455 children with febrile seizure
• 1.3 with bacteremia
• 5.9 UTI
• 12.5 with abnormal chest x-ray
• normal CSF in all who had an LP (135)

53
Meningitis Risk in Simple Febrile Seizures
Whats Been Reported ?

• Literature review of reported cases of febrile
  seizures and meningitis
• 2,870 cases of febrile seizures with LPs
• 1.7 with bacterial meningitis
• 17 of those with meningitis described as
  clinically inapparent
• Is occult bacterial meningitis a significant
  clinical entity ?

54
Meningitis Risk in Febrile Seizures Green SM,
et al Pediatrics 1993 92527

• Studied children with meningitis -- how many
  presented solely with seizures?
• 486 children with bacterial meningitis
• complex seizures present in 79
• 93 of those with seizures were obtunded
• of the few with normal LOC, 78 had nuchal
  rigidity
• the two patients without meningismus had
  other straightforward indications for LP
• Occult meningitis is more myth than fact

55
What is the Cost Effective Strategy?Yamamoto LG,
et al Am J Emerg Med 1998 16193

• Updated decision analysis which considered
• low incidence of H. influenzae infections
• emergence of resistant S. pneumoniae
• negative consequences of unnecessary Rx
• Assuming zero or low Rx consequences -- empiric
  therapy associated with best outcomes
• Assuming realistic Rx consequences - no testing
  and no treatment option may be best

56
Cost Effectiveness Post-Vaccine?Lee GM, et al
Pediatrics 2001 108835

• Updated prior decision analysis, considering
• Elimination of H. influenzae
• Lower rate of occult bacteremia (1.5)
• Published efficacy of empiric Rx
• Negative consequences of unnecessary Rx
• At current rate of OB, CBC plus selective blood
  culture and treatment is still best
• If OB rate lt 0.5, strategies employing empiric
  testing treatment should be eliminated

57
Serotyping of Pneumococcal OBAlperin ER, et al
Pediatrics 2001 108e23

• What is potential efficacy of pneumococcal
  vaccine in the prevention of OB
• S pneumoniae accounts for the vast majority (83)
  of pathogens in children with OB
• Eight serotypes isolated
• 6A, 9V, 19F, 18C, 4, 6B, 23F, 14
• 98 of serotypes would be covered by the
  currently licensed vaccine -- all except 6A
• Good newsWe may soon erradicate OPB

58
Food For Thought And A Little Math

• Current risk of OB 1.5 to 1.9
• 92 pneumococcal
• Risk of meningitis in OPB 1 to 2
• Risk of adverse sequelae in BM 33 - 50
• Need to treat 2500 febrile kids to prevent one
  case of BM 5000-7500 per adverse sequelae
• remember that antibiotics may not prevent BM!
• incidence of ADRs 150 - 600(?)/ case of BM
• alarming growth rate of antibiotic resistance

59
Evolving Pneumococcal ResistanceKaplan, et al
Pediatrics 1998 102538.

• Prospective surveillance study of invasive
  pneumococcal infections
• three year (1993-1996), eight center study
• 1291 systemic pneumococcal infections
• Proportion of non-susceptible isolates (PCN,
  ceftriaxone) increased annually
• nearly doubled over the three year period
• penicillin resistance 21
• ceftriaxone resistance 9.3

60
So, What Do We Actually Know

• Extremely common presenting complaint
• Much concern (phobia?) regarding fever
• Fairly effective strategies to identify low risk
  infants - these do not apply to neonates
• 1 bad actor (H. flu) effectively erradicated
• UMD - pediatricians nightmare - still out
  there
• Risk of OB, now under 2, primarily OPB
• Can apply risk stratification to OPB
• 93-96 spontaneous resolution of OPB

61
So, What Do We Actually Know

• No consensus regarding optimal approach to the
  febrile infant
• Not entirely clear what parents want
• Empiric Rx does not prevent sequelae
• Rising rates of antimicrobial resistance
• UTI remains the most common occult SBI
• RVS are a reasonable explanation for fever
• Pneumovax may make this all a moot point

62
Some Friendly Advice

• Keep abreast of the literature
• Discuss this with colleagues mentors
• local practice variations
• institutional practice guidelines
• antimicrobial resistance rates
• Both approaches (RM TM) are defensible
• Choose the best strategy for you
• Be consistent
• Always treat the ill appearing child with fever

63
Notable Quotes

• Unfortunately, many practitioners have become
  reluctant to rely on clinical judgement,
  preferring diagnostic tests and frequent use of
  antibiotics..
• We should resist the urge to use antibiotics
  empiricially, especially in a patient who looks
  well, for whom antibiotics have not been shown
  clearly to be beneficial
• -- JK Stamos, ST Shulman Lancet 1997
• Antibiotics are not antipyretics
• -- SE Krug Overheard many evenings in CMH ED