Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Is lumbar puncture necessary for evaluation of early neonatal sepsis?

Three Part Question

In [a newborn] is [lumbar puncture necessary] to [rule out meningitis in suspected sepsis in the first few days of life (0–3 days)]?

Clinical Scenario

A newborn baby born at 37 weeks is noted to be unwell at 18 h postnatally. The mother gives a history of prolonged rupture of membranes for 36 h. The baby is feeding poorly and is jittery, with a temperature of 38°C. A clinical diagnosis of early sepsis is made and lumbar puncture is suggested on the ward round as a part of sepsis evaluation. Several publications on the use of lumbar puncture in late-onset sepsis, including a recent review article by Malbon et al,1 suggest that lumbar puncture is an important method of investigation and should be considered in babies for >48 h old, with suspected sepsis.
We wonder whether there is sufficient evidence to justify lumbar puncture in early sepsis.

Search Strategy

Search date: September 2005
Cochrane Library: Nil relevant
Medline: 1950–to date; Embase: 1974–to date; Cinhal: 1982–to date via Dialog Datastar
Search terms: (Neonatal ADJ sepsis or Neonatal ADJ septicaemia or Neonatal ADJ meningitis or meningitis and infant–newborn# or Early ADJ sepsis, or Early ADJ septicaemia) and (Lumbar ADJ puncture or LP or Spinal ADJ tap or CSF ADJ examination). Limit to English language and newborn infants from birth to 1 month.
Total number of hits: 51
Cross-references obtained: 6

Search Outcome

Total number: 57, of which 5 studies were eligible

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Visser et al,
Newborn with suspected sepsis evaluated within 72 h of life. Total n = 323Retrospective case notes review, level 2bIncidence of meningitisTotal no of meningitis = 6, with the incidence of meningitis being 18/1000Uncontrolled cohort. Indications of sepsis evaluation are not well described. No mentionwhether all suspected newborns were screened
Eldadah et al,
All infants admitted with RDS were evaluated for sepsis with LP within 24 h of life. n = 203Prospective study, level 1bNo of cases with meningitisNo cases of meningitis were found. BC was positive in 17 infantsStudy does not indicate any long-term follow-up, so no knowledge about missed cases of meningitis. Only included babies with RDS
Hendricks-Munoz and Shapiro,
Newborns <34 weeks admitted with suspected sepsis or those with risk factors for sepsis was evaluated within 6 h. LP was carried out only in BC positive cases. Total no was 1390, of which 32 were BC positive; 15 of them died before CSF evaluation; 12 of these 15 patients had LP after death and 16 of the remaining 17 cases were evaluated with ante-mortem LPRetrospective case notes review, level 2bNo of cases of meningitis, number of missed/partially treated cases of meningitisNo cases of meningitis were found. No missed or partially treated cases were detectedUncontrolled cohort. Small study but only looking at BC positive cases. Babies were treated with antibiotic before CSF was obtained, so the early meningeal seeding could be missed. No mention about the timing of follow-up to detect missed cases
Weiss et al,
All infants admitted with respiratory distress on the first day of life had undergone LP as a part of sepsis screen. n = 1495Retrospective case notes analysis, level 2bFrequency of meningitis. Degree of association between meningitis and sepsis4 cases of meningitis were detected with an incidence of 2.7/1000. BC were positive in 3 of these casesStudy included only babies admitted with respiratory distress on day 1 of life
Ajayi and Mokulu,
Phase 1: Newborns with suspected sepsis and those with risk factors for sepsis were evaluated within 72 h of age with lumbar puncture. n = 263. Phase 2: Newborns within 72 h of life with signs of severe sepsis only were evaluated with lumbar puncture. n = 50Retrospective case notes review level 2bNo of LP done. No of cases of meningitis and no of missed or partially treated cases3 times fewer LP carried out in phase 2 than in phase 1. No cases of meningitis detected in both phases (95% CI 0 to 1.1). No missed or partially treated cases of meningitis foundUncontrolled cohort. No mention about timing of follow-up to detect partially treated or missed cases. Estimated maximum risk was considered to avoid the fallacy inherent to zero numerators and the attendant complacency


Lumbar puncture has always been an invaluable tool to diagnose meningitis. In the neonatal period, septicaemia can be indistinguishable from meningitis. The overall incidence of neonatal meningitis is 0.25–1.0 per 1000 live births (Wiswell, Shattuck). Practice varies between hospital units as regards early sepsis evaluation. Although blood culture has been regarded as an essential component of sepsis screen, the role of lumbar puncture is debatable especially in the first 72 h of life. Previously published data showed that neonatal septicaemia can coexist with meningitis in up to 30% of patients (Vissar). On the other hand, lumbar puncture can be associated with major risks including hypoxaemia, clinical deterioration and many other hazards in small and sick babies (Weisman, Speidel). Moreover, in about 30% of patients, the cerebrospinal fluid tap could be traumatic or inadequate (Weisman, Schreiner). Although many of the studies did not compare the incidence of meningitis between groups with early-onset sepsis presenting with symptoms and groups with suspected sepsis because of perinatal risk factors without any overt symptoms,Please confirm the changes made in the sentence the published literature shows the incidence of meningitis in asymptomatic newborns undergoing evaluation only because risk factors is virtually nil (Kumar, Fielkow, Prabhakar). The study by Visser et al observed a very high (1.8%) incidence of meningitis in babies within 72 h of life. This study also noted that in 15% of cases, blood culture was negative. Many of the later studies did not show such a high incidence. Kumar, Schwersenski. For example, studies by Ajayi and Mokuolu and Hendricks-Munoz and Shapiro looked at around 1700 babies but found no cases of meningitis. Even their long-term follow-up did not show any case of missed or partially treated meningitis. Two other similar studies, Eldadah, Weiss which looked at babies admitted with respiratory symptoms within 24 h of birth, also found a very low incidence of meningitis (only four cases of meningitis in > 1700 neonates evaluated with lumbar puncture). The statistically estimated maximum risk of meningitis in suspected early sepsis is only 1.1% and that in blood culture proved sepsis is 0–10.3%. It seems that there is no need to carry out lumbar puncture in neonates suspected of early sepsis who are being evaluated purely for perinatal risk factors, or in those presenting with mild symptoms. It should still be undertaken in babies with severe illness or obviously where meningitis is strongly suspected.

Clinical Bottom Line

Overall incidence of neonatal meningitis is 0.25–1.0 per 1000 live births (grade A). Uncontrolled studies suggest that meningitis is very uncommon in asymptomatic babies with only perinatal risk factors for sepsis, so in this group lumbar puncture can be safely omitted from the early sepsis screen (grade B). In strongly suspected cases, lumbar puncture should be included in an examination of sepsis (grade B).


  1. Visser VE, Hall RT. Lumbar puncture in the evaluation of suspected neonatal sepsis. J Pediatr 1980;96:1063–7.
  2. Eldadah M, Frenkel LD, Hiatt IM. et al. Evaluation of routine lumbar punctures in newborn infants with respiratory distress syndrome. Paediatr Infect Dis J 1987;6:243–6.
  3. Hendricks-Munoz KD. Shapiro DL. The role of the lumbar puncture in the admission sepsis evaluation of the premature infant. J Perinatol 1990;10:60–4.
  4. Weiss MG. Ionides SP. Anderson CL. Meningitis in premature infants with respiratory distress: role of admission lumbar puncture. J Paediatr 1991;119:973–5.
  5. Ajayi OA. Mokuolu OA. Evaluation of neonates with risk factor for infection/suspected sepsis: is lumbar puncture necessary in the first 72 hours of life? Trop Med Int Health 1997;2:284–8.
  6. Wiswell TE, Baumgart S. Gan on CM. et al. No lumbar puncture in the evaluation for early neonatal sepsis: will meningitis be missed? Pediatrics 1995;6:95.
  7. Shattuck KE, Chonmaitree T. The changing spectrum of neonatal meningitis over a fifteen year period. Clin Pediatr 1992;31:130–6.
  8. Weisman L, Emrenstein GB. Steenbarger JR. The effect of lumbar puncture position on sick neonates. Am J Dis Child 1983;137:1077–9.
  9. Speidel BD. Adverse effects of routine procedures on preterm infants. Lancet 1978;1:864–5.
  10. Schreiner RL, Kleiman MB. Incidence and effect of traumatic lumbar punctures in the neonate. Dev Med Child Neurol 1979;21:483–7.
  11. Kumar P, Sarkar S, Narang A. Role of lumbar puncture in neonatal infection. J Paediatr Child Health 1995;31:8–10.
  12. Fielkow S, Reuter S. Gotoff SP. Cerebrospinal fluid examination in symptom free infants with risk factor for infection. J Paediatr 1991;119:971–3.
  13. Prabhakar BR, Campe J, Prabhakar K. et al. Lumbar puncture in suspected neonatal sepsis. Children's Hosp Q 1999;11:15–17.
  14. Schwersenski J, Mcintyre L, Bauer CR. Lumbar puncture frequency and cerebrospinal fluid analysis in the neonate. Am J Dis Child 1991;145:54–8.