Eur J Pediatr (1990) 149 : 797-799
European Journal of
Pediatrics
9 Springer-Verlag1990
Routine lumbar punctures in the newborn - are they justified? P. M a c M a h o n 1, L. Jewes 2, and J. de Louvois 3 1Department of Paediatrics, Charing Cross Hospital, London W6, UK 2Department of Paediatrics, Bristol Maternity Hospital, Bristol, UK ~Department of Microbiology, Queen Charlotte's and Chelsea Hospital, London W6, UK Received July 4, 1988 / Accepted January 5, 1990
Abstract. The records of 62,107 neonates cared for at three neonatal referral units between 1979 and 1985 were analysed with regard to the n u m b e r of lumbar punctures (LP) performed and the results obtained. The mean perinatal mortality rate was 9.3/1000 live births and the neonatal mortality rate was 5.9/1000. The units were comparible in this respect. Two hundred and sixty three (0.4%) babies weighed less than 1.0 kg at birth and 510 (0.8%) babies weighed 1.0-1.5 kg. A total of 1554 LPs were performed on 1084 babies, 1.7% of the whole population. Seventeen babies were diagnosed as having meningitis following CSF examination; 5 weighed less than 1.5kg (VLBW) (incidence of meningitis 6.5/1000 births) and 12 weighed more than 1.5 kg (incidence 0.2/1000 live births). A n LP was p e r f o r m e d on 44% of V L B W babies and 1.2% of babies weighing more than 1.5 kg. In both groups approximately 1.5% of CSF samples were positive. Eight of the 17 positive results were obtained during the first 3 postnatal days. Only one V L B W b a b y had meningitis on the 1st day of life (1.3/1000). In view of the low incidence of meningitis and the risk factors associated with an LP it is questionable whether this procedure should be a routine investigation in the V L B W newborn. Key words: L u m b a r puncture - Meningitis - N e w b o r n Sepsis
It is c o m m o n practice for microbiological specimens to be collected before an infant commences antibiotics, however, there is no agreement on whether the pretreatment screen should include a lumbar puncture (LP) [2, 5]. Justification for LP in the newborn ranges from inclusion as part of the sepsis work up, to the m o r e conservative approach in which the procedure is used only in the investigation of very sick babies [5]. The possible benefits of routine LP have never been determined, however a n u m b e r of risk factors have been identified. A n LP m a y initiate meningitis in babies who are bacteraemic [12, 13, 16]; induce 'coning' in patients with raised intracranial pressure [6] and result in damage to the spinal cord (Pryse-Davies, unpublished information). If a needle without a stylet is used there is a risk of intraspinal epidermoid tumour [1]. In addition changes in blood gas tensions [3] and major clinical deteriorations [17] have been observed during the performance of an LP. The procedure is clearly painful even for the smallest babies [5] and the frequency of traumatic taps or unsuccessful LPs is high and m a y approach 50% [2, 12]. This study was undertaken in order to document current medical practice regarding LPs in the newborn period and, in addition, to ascertain the justification for this procedure in very low birthweight (VLBW) infants during the first 3 days of life.
Patients and methods Introduction In the United Kingdom there are over 100 cases of neonatal meningitis a year [8], with an overall mortality rate of 2 5 % - 4 0 % [8, 11]; approximately 20% of the survivors have neurological sequelae [9]. Meningitis is particularly likely to occur in low birth weight infants (45% of cases) and when the m e m b r a n e s were ruptured for over 24h (12% of cases) [4, 8]. Offprint requests to: J. de Louvois Abbreviations: ELBW = extremely low birth weight; VLBW =
very low birth weight; LP = lumbar puncture
The records of the neonatal and medical microbiology departments of three major neonatal referral units (Charing Cross and Queen Charlotte's and Chelsea Hospitals, London and the Bristol Maternity Hospital, Bristol) were studied for the years 1979 - 1985. The total numbers of VLBW infants (inborn and outborn included) and their survival rates were analysed for each unit to determine that the units were comparible. The records of all LPs performed during the study period were analysed. The LP result was defined as 'positive' if: (1) a bacterium was isolated from the CSF on culture; (2) bacteria were seen in the Gram film or; (3) the CSF contained a significant number (> 20/gl) of leucocytes, or a discernible excess of leucocytes in bloodstained CSF samples. These definitions of a positive result were used because they represented the results upon which clinical action was taken. In each unit it was the practice to perform and LP only on babies with signs and symp-
798 toms of severe sepsis. An LP was not part of the routine pre-antibiotic screen. Positive findings were therefore considered diagnostic for the purpose of management.
Results
A total of 62,107 infants were delivered in the three units during the study period. The statistical indicators of the overall standards of perinatal care were similar in each unit with a mean perinatal mortality rate of 9.3 per 1,000 (range 8.2-10.4) and a neonatal mortality rate of 5.9 per 1000 (range 5.2-6.2). The survival rates of infants (inborn and outborn included) of less than 1 kg birth weight (extremely low birth weight)(ELBW) and between 1 and 1.5 kg birth weight were also similar for each of the units with overall mean mortality rates of 46% (range 4 3 % - 4 9 % ) and 12.3% (range 10.0%-15.2%) for the groups respectively. A total of 1,554 LPs were performed on 1,084 (1.7% of the total) infants (Table 1). These included multiple taps from hydrocephalic babies if these samples were sent to the laboratory for examination. Two hundred and seventy six LPs (18% of the total) were performed on 122 of the 263 E L B W infants (46%), 364 LPs (23% of the total) were performed on 216 of the 510 babies (42%) with birth weights of l - l . 5 k g and 914 LPs (59% of the total) were performed on 746 of the 61,334 infants of birth weight > 1.5 kg (1.2%). An LP was performed during tile first 3 days of life in 6.5% of E L B W and 7.8% of babies weighing 1.0-1.5 kg (7.4% of all V L B W babies). The threshold for performing an LP, as judged by the
Table 1. Lumbar puncture in the newborn period
No. of babies No. of babies having an LP No. of LPs performed No. of LPs in first 3 days of life (% of total)
< 1 kg
1-1.5 kg > 1.5 kg
All infants
263 122 (46%)
510 216 (42%)
62,107 1,084 (1.7%)
276 17
364 40
(6.5%)
61,334 746 (1.2%) 914 468
(7.8%)
1,554 525
(0.76%)
Birthweight 1-< 1.5kg > 1.5kg
3 2 12
1.00 0.55 1.31
11.4 3.9 0.2
percentage of V L B W infants on whom this procedure was undertaken, was the same of all three units. Seventeen (1.1%) of CSF samples were positive (Table 2) and in 12 the culture grew a significant organism, the commonest isolate being group B streptococci. The remaining five samples contained an excess of leucocytes ( > 20/gl) or showed bacteria in the Gram film. Five of the positive CSF samples were from the 773 V L B W infants. The overall incidence of meningitis in this group was therefore 6.5 per 1,000 (11 per 1000 for babies under 1.0kg and 3.9 per 1000 for those weighing 1.0-1.5kg) compared to 0.2 per 1000 (12 of 61,334) among babies weighing more than 1.5 kg. The CSF findings were positive in 1.5% (5 of 338) of V L B W babies who had an LP and in 1.6% (12 of 746) of those weighing more than 1.5 kg. The inclusion of some outborn babies in the study population does not significantly alter these figures. Eight of the 17 positive samples were collected in the first 3 days of life. Three were from 57 V L B W infants, 5.3% of those sampled. In all of these the CSF white cell count was normal. Only one V L B W infant had a positive LP in the 1st day of life (1.3 per 1,000) and this baby died despite intensive management including antibiotics. Two other infants had positive CSF samples on the 1st day of life, group B streptococci were isolated from both and the babies survived. Three of 17 patients with positive CSF findings died, all were V L B W infants. The overall case mortality rate was therefore 18% and for the sub-group of VLBW infants it was 60%. Bacteria were isolated from the CSF of all the babies who died.
Discussion
(0.85%)
Table 2. Positive CSF findings in 1554 samples
Organism
Table 3. Positive CSF samples
Total > 1.5kg
E. coli E. cloacae
1 0
0 1
1 2
2 3
Gp B streptococci Gp D streptococci
1 0
1 0
3 1
5 1
L. monocytogenes
0
0
1
1
No pathogen isolated
1
0
4
5
Total
3
2
12
17
The incidence of neonatal meningitis is estimated to be 0.2-0.5 per 1,000 in term and 3 per 1,000 in preterm infants [7]. This is consistant with the incidence of 11.4 per 1000 in E L B W babies, 3.9 per 1000 in babies weighing 1.0-1.5kg and 0.2 per 1000 in babies weighing more than 1.5 kg found in this study. Neonatal meningitis is a serious but potentially treatable condition and many authoritative texts recommend that an LP should only be omitted in infants suspected of being septic when infection is clearly clinically localized to one site [10]. Such an approach does not consider the potentially harmful effects of the procedure nor the fact that up to 50% of LPs may be unsuccessful or result in a traumatic tap. Such an approach is not supported by the results of this study. Furthermore during the first few hours of life many minor procedures regarded as part of routine intensive care can produce hypoxia in preterm infants [15]. Perfor-
799 m a n c e of an L P in the flexed position m a y lead to an increase in the t r a n s c u t a n e o u s PCO2 [12]. T h e knee-chest position can also cause the b l o o d pressure to rise [3, 14] and the sudden clinical deterioration which can accomp a n y this p r o c e d u r e m a y necessitate mechanical ventilation [16]. These considerations suggest that LPs are m o r e likely to lead to significant complications when p e r f o r m e d in n e w b o r n p r e t e r m infants. All three units in this study applied a policy of only u n d e r t a k i n g an L P w h e n there were clinical indications of severe sepsis thus in the five babies whose CSF cultures were negative the CSF cell counts were considered clinically to justify a diagnosis of meningitis. Because clinical indications of meningitis are so non-specific the suspicion of meningitis is high in babies weighing less than 1.5 kg at birth and as a result 44% of such infants in this study had an L P c o m p a r e d to 1.2% of their heavier contemparies. H o w e v e r , C S F findings were positive (1.5% of samples) with equal f r e q u e n c y f r o m babies weighing less than 1.0 kg or m o r e than 1.5 kg. T h e practice of routine L P in all infants prior to the administration of antibiotics would seem questionable given that only o n e of 773 V L B W babies had meningitis on the 1st day of life. T h e three centres in this study operated a selective policy with regard to l u m b a r p u n c t u r e yet only 1.5% of CSF samples were positive, even those f r o m V L B W babies. T h e 3 V L B W babies with culture positive CSF during the first 3 days of life all h a d n o r m a l CSF cell counts ( 0 - 2 leucocytes/gl) confirming the p o o r prognostic value of a CSF cell count in low birth weight babies suspected of having meningitis. T h e findings of this study strongly suggest that an L P should only be carried out during the n e w b o r n period w h e n there are clinical signs and s y m p t o m s of severe sepsis and n o t as a routine on all low birth weight babies and that it should not be a routine investigation on all babies prior to receiving antibiotics. A n L P is a p o t e n tially hazardous p r o c e d u r e in V L B W infants and we believe that a m o r e selective a p p r o a c h to its use is not only justified but correct.
Acknowledgements. We thank Dr. I. Kovar, Charing Cross Hospital, Dr. DR. Harvey, Queen Charlotte's and Chelsea Hospital and Dr.P.Dunn and Dr. P. Fleming, Bristol Maternity Hospital for permission to study their patients.
References 1. Choremis C, Economos D, Papadatos C, Gargoulas A (1956) Intrapsinal epidermoid tumours (cholesteatomas) in patients treated for tuberculous meningitis. Lancet II : 437 2. Eldadah M, Fremkel LD, Hiatt IM, Hegyi T (1987) Evaluation of routine lumbar punctures in newborn infants with respiratory distress syndrome. Paediatr Infect Dis 6: 243-245 3. Gleason CA, Martin R J, Anderson JV, Waldemar AC, Sanniti KJ, Fanaroff AA (1983) Optimal position for a spinal tap in preterm infants. Pediatrics 71 : 31-35 4. Goldacre MJ (1976) Acute bacterial meningitis in childhood: incidence and mortality in a defined population. Lancet I : 2831 5. Halliday HL (1989) When to do a lumbar puncture in a neonate. Arch Dis Child 64:313-316 6. Harper JR, Lorber J, Hillas-Smith G, Bower BD, Eykyn SJ (1985) Timing of lumbar puncture in severe childhood meningitis. Br Med J 291 : 651-652 7. Klein JO, Marcy SM (1983) Bacterial sepsis and meningitis. In: Remington JS, Klein JO (eds) Infectious diseases of the fetus and newborn infant. WB Saunders Co, Philadelphia, pp 679-735 8. Louvois J de, Blackbourn J, Hurley R, Harvey DR (1988) Meningitis during the first year of life - a two year prospective study. British Paediatric Association, programme for the 60th Annual Meeting, abstract no P24, p 51 9. Mulhall A, Louvois J de, Hurley R (1983) Efficacy of chloramphenicol in the treatment of neonatal and infantile meningitis: a study of 70 cases. Lancet I : 284-287 10. Pearce PG, Roberton NRC (1986) Textbook of neonatology. Churchill Livingstone, London, pp 725-781 11. PHLS report (1985) Neonatal meningitis: a review of routine national data 1975-83. Br Med J 290 : 778-779 12. Schreiner RL, Kleiman MB (1979) Incidence and effect of traumatic lumbar puncture in the neonate. Dev Med Child Neurol 21 : 483-487 13. Smith KM, Deddish RB, Ogata ES (1986) Meningitis associated with serial lumbar punctures and post haemorrhagic hydrocephalus. J Pediatr 109 : 1057-1060 14. Spahr RC, Mueller-Huebach E (1981) Knee-chest position and neonatal oxygenation and blood pressure. Am J Dis Child 135 : 79-80 15. Speidel BD (1978) Adverse effects of routine procedures on preterm infants. Lancet I : 864 16. Teele DW, Dashefsky B, Rakusan T, Klein JO (1989) Meningitis after lumbar puncture in children with bacteremia. N Engl J Med 305 : 1079-1081 17. Weisman LE, Merenstein GB, Steenbarger JR (1983) The effect of lumbar puncture position in sick neonates. Am J Dis Child 137 : 1077-1079
European Journal of
Pediatrics
9 Springer-Verlag1990
Routine lumbar punctures in the newborn - are they justified? P. M a c M a h o n 1, L. Jewes 2, and J. de Louvois 3 1Department of Paediatrics, Charing Cross Hospital, London W6, UK 2Department of Paediatrics, Bristol Maternity Hospital, Bristol, UK ~Department of Microbiology, Queen Charlotte's and Chelsea Hospital, London W6, UK Received July 4, 1988 / Accepted January 5, 1990
Abstract. The records of 62,107 neonates cared for at three neonatal referral units between 1979 and 1985 were analysed with regard to the n u m b e r of lumbar punctures (LP) performed and the results obtained. The mean perinatal mortality rate was 9.3/1000 live births and the neonatal mortality rate was 5.9/1000. The units were comparible in this respect. Two hundred and sixty three (0.4%) babies weighed less than 1.0 kg at birth and 510 (0.8%) babies weighed 1.0-1.5 kg. A total of 1554 LPs were performed on 1084 babies, 1.7% of the whole population. Seventeen babies were diagnosed as having meningitis following CSF examination; 5 weighed less than 1.5kg (VLBW) (incidence of meningitis 6.5/1000 births) and 12 weighed more than 1.5 kg (incidence 0.2/1000 live births). A n LP was p e r f o r m e d on 44% of V L B W babies and 1.2% of babies weighing more than 1.5 kg. In both groups approximately 1.5% of CSF samples were positive. Eight of the 17 positive results were obtained during the first 3 postnatal days. Only one V L B W b a b y had meningitis on the 1st day of life (1.3/1000). In view of the low incidence of meningitis and the risk factors associated with an LP it is questionable whether this procedure should be a routine investigation in the V L B W newborn. Key words: L u m b a r puncture - Meningitis - N e w b o r n Sepsis
It is c o m m o n practice for microbiological specimens to be collected before an infant commences antibiotics, however, there is no agreement on whether the pretreatment screen should include a lumbar puncture (LP) [2, 5]. Justification for LP in the newborn ranges from inclusion as part of the sepsis work up, to the m o r e conservative approach in which the procedure is used only in the investigation of very sick babies [5]. The possible benefits of routine LP have never been determined, however a n u m b e r of risk factors have been identified. A n LP m a y initiate meningitis in babies who are bacteraemic [12, 13, 16]; induce 'coning' in patients with raised intracranial pressure [6] and result in damage to the spinal cord (Pryse-Davies, unpublished information). If a needle without a stylet is used there is a risk of intraspinal epidermoid tumour [1]. In addition changes in blood gas tensions [3] and major clinical deteriorations [17] have been observed during the performance of an LP. The procedure is clearly painful even for the smallest babies [5] and the frequency of traumatic taps or unsuccessful LPs is high and m a y approach 50% [2, 12]. This study was undertaken in order to document current medical practice regarding LPs in the newborn period and, in addition, to ascertain the justification for this procedure in very low birthweight (VLBW) infants during the first 3 days of life.
Patients and methods Introduction In the United Kingdom there are over 100 cases of neonatal meningitis a year [8], with an overall mortality rate of 2 5 % - 4 0 % [8, 11]; approximately 20% of the survivors have neurological sequelae [9]. Meningitis is particularly likely to occur in low birth weight infants (45% of cases) and when the m e m b r a n e s were ruptured for over 24h (12% of cases) [4, 8]. Offprint requests to: J. de Louvois Abbreviations: ELBW = extremely low birth weight; VLBW =
very low birth weight; LP = lumbar puncture
The records of the neonatal and medical microbiology departments of three major neonatal referral units (Charing Cross and Queen Charlotte's and Chelsea Hospitals, London and the Bristol Maternity Hospital, Bristol) were studied for the years 1979 - 1985. The total numbers of VLBW infants (inborn and outborn included) and their survival rates were analysed for each unit to determine that the units were comparible. The records of all LPs performed during the study period were analysed. The LP result was defined as 'positive' if: (1) a bacterium was isolated from the CSF on culture; (2) bacteria were seen in the Gram film or; (3) the CSF contained a significant number (> 20/gl) of leucocytes, or a discernible excess of leucocytes in bloodstained CSF samples. These definitions of a positive result were used because they represented the results upon which clinical action was taken. In each unit it was the practice to perform and LP only on babies with signs and symp-
798 toms of severe sepsis. An LP was not part of the routine pre-antibiotic screen. Positive findings were therefore considered diagnostic for the purpose of management.
Results
A total of 62,107 infants were delivered in the three units during the study period. The statistical indicators of the overall standards of perinatal care were similar in each unit with a mean perinatal mortality rate of 9.3 per 1,000 (range 8.2-10.4) and a neonatal mortality rate of 5.9 per 1000 (range 5.2-6.2). The survival rates of infants (inborn and outborn included) of less than 1 kg birth weight (extremely low birth weight)(ELBW) and between 1 and 1.5 kg birth weight were also similar for each of the units with overall mean mortality rates of 46% (range 4 3 % - 4 9 % ) and 12.3% (range 10.0%-15.2%) for the groups respectively. A total of 1,554 LPs were performed on 1,084 (1.7% of the total) infants (Table 1). These included multiple taps from hydrocephalic babies if these samples were sent to the laboratory for examination. Two hundred and seventy six LPs (18% of the total) were performed on 122 of the 263 E L B W infants (46%), 364 LPs (23% of the total) were performed on 216 of the 510 babies (42%) with birth weights of l - l . 5 k g and 914 LPs (59% of the total) were performed on 746 of the 61,334 infants of birth weight > 1.5 kg (1.2%). An LP was performed during tile first 3 days of life in 6.5% of E L B W and 7.8% of babies weighing 1.0-1.5 kg (7.4% of all V L B W babies). The threshold for performing an LP, as judged by the
Table 1. Lumbar puncture in the newborn period
No. of babies No. of babies having an LP No. of LPs performed No. of LPs in first 3 days of life (% of total)
< 1 kg
1-1.5 kg > 1.5 kg
All infants
263 122 (46%)
510 216 (42%)
62,107 1,084 (1.7%)
276 17
364 40
(6.5%)
61,334 746 (1.2%) 914 468
(7.8%)
1,554 525
(0.76%)
Birthweight 1-< 1.5kg > 1.5kg
3 2 12
1.00 0.55 1.31
11.4 3.9 0.2
percentage of V L B W infants on whom this procedure was undertaken, was the same of all three units. Seventeen (1.1%) of CSF samples were positive (Table 2) and in 12 the culture grew a significant organism, the commonest isolate being group B streptococci. The remaining five samples contained an excess of leucocytes ( > 20/gl) or showed bacteria in the Gram film. Five of the positive CSF samples were from the 773 V L B W infants. The overall incidence of meningitis in this group was therefore 6.5 per 1,000 (11 per 1000 for babies under 1.0kg and 3.9 per 1000 for those weighing 1.0-1.5kg) compared to 0.2 per 1000 (12 of 61,334) among babies weighing more than 1.5 kg. The CSF findings were positive in 1.5% (5 of 338) of V L B W babies who had an LP and in 1.6% (12 of 746) of those weighing more than 1.5 kg. The inclusion of some outborn babies in the study population does not significantly alter these figures. Eight of the 17 positive samples were collected in the first 3 days of life. Three were from 57 V L B W infants, 5.3% of those sampled. In all of these the CSF white cell count was normal. Only one V L B W infant had a positive LP in the 1st day of life (1.3 per 1,000) and this baby died despite intensive management including antibiotics. Two other infants had positive CSF samples on the 1st day of life, group B streptococci were isolated from both and the babies survived. Three of 17 patients with positive CSF findings died, all were V L B W infants. The overall case mortality rate was therefore 18% and for the sub-group of VLBW infants it was 60%. Bacteria were isolated from the CSF of all the babies who died.
Discussion
(0.85%)
Table 2. Positive CSF findings in 1554 samples
Organism
Table 3. Positive CSF samples
Total > 1.5kg
E. coli E. cloacae
1 0
0 1
1 2
2 3
Gp B streptococci Gp D streptococci
1 0
1 0
3 1
5 1
L. monocytogenes
0
0
1
1
No pathogen isolated
1
0
4
5
Total
3
2
12
17
The incidence of neonatal meningitis is estimated to be 0.2-0.5 per 1,000 in term and 3 per 1,000 in preterm infants [7]. This is consistant with the incidence of 11.4 per 1000 in E L B W babies, 3.9 per 1000 in babies weighing 1.0-1.5kg and 0.2 per 1000 in babies weighing more than 1.5 kg found in this study. Neonatal meningitis is a serious but potentially treatable condition and many authoritative texts recommend that an LP should only be omitted in infants suspected of being septic when infection is clearly clinically localized to one site [10]. Such an approach does not consider the potentially harmful effects of the procedure nor the fact that up to 50% of LPs may be unsuccessful or result in a traumatic tap. Such an approach is not supported by the results of this study. Furthermore during the first few hours of life many minor procedures regarded as part of routine intensive care can produce hypoxia in preterm infants [15]. Perfor-
799 m a n c e of an L P in the flexed position m a y lead to an increase in the t r a n s c u t a n e o u s PCO2 [12]. T h e knee-chest position can also cause the b l o o d pressure to rise [3, 14] and the sudden clinical deterioration which can accomp a n y this p r o c e d u r e m a y necessitate mechanical ventilation [16]. These considerations suggest that LPs are m o r e likely to lead to significant complications when p e r f o r m e d in n e w b o r n p r e t e r m infants. All three units in this study applied a policy of only u n d e r t a k i n g an L P w h e n there were clinical indications of severe sepsis thus in the five babies whose CSF cultures were negative the CSF cell counts were considered clinically to justify a diagnosis of meningitis. Because clinical indications of meningitis are so non-specific the suspicion of meningitis is high in babies weighing less than 1.5 kg at birth and as a result 44% of such infants in this study had an L P c o m p a r e d to 1.2% of their heavier contemparies. H o w e v e r , C S F findings were positive (1.5% of samples) with equal f r e q u e n c y f r o m babies weighing less than 1.0 kg or m o r e than 1.5 kg. T h e practice of routine L P in all infants prior to the administration of antibiotics would seem questionable given that only o n e of 773 V L B W babies had meningitis on the 1st day of life. T h e three centres in this study operated a selective policy with regard to l u m b a r p u n c t u r e yet only 1.5% of CSF samples were positive, even those f r o m V L B W babies. T h e 3 V L B W babies with culture positive CSF during the first 3 days of life all h a d n o r m a l CSF cell counts ( 0 - 2 leucocytes/gl) confirming the p o o r prognostic value of a CSF cell count in low birth weight babies suspected of having meningitis. T h e findings of this study strongly suggest that an L P should only be carried out during the n e w b o r n period w h e n there are clinical signs and s y m p t o m s of severe sepsis and n o t as a routine on all low birth weight babies and that it should not be a routine investigation on all babies prior to receiving antibiotics. A n L P is a p o t e n tially hazardous p r o c e d u r e in V L B W infants and we believe that a m o r e selective a p p r o a c h to its use is not only justified but correct.
Acknowledgements. We thank Dr. I. Kovar, Charing Cross Hospital, Dr. DR. Harvey, Queen Charlotte's and Chelsea Hospital and Dr.P.Dunn and Dr. P. Fleming, Bristol Maternity Hospital for permission to study their patients.
References 1. Choremis C, Economos D, Papadatos C, Gargoulas A (1956) Intrapsinal epidermoid tumours (cholesteatomas) in patients treated for tuberculous meningitis. Lancet II : 437 2. Eldadah M, Fremkel LD, Hiatt IM, Hegyi T (1987) Evaluation of routine lumbar punctures in newborn infants with respiratory distress syndrome. Paediatr Infect Dis 6: 243-245 3. Gleason CA, Martin R J, Anderson JV, Waldemar AC, Sanniti KJ, Fanaroff AA (1983) Optimal position for a spinal tap in preterm infants. Pediatrics 71 : 31-35 4. Goldacre MJ (1976) Acute bacterial meningitis in childhood: incidence and mortality in a defined population. Lancet I : 2831 5. Halliday HL (1989) When to do a lumbar puncture in a neonate. Arch Dis Child 64:313-316 6. Harper JR, Lorber J, Hillas-Smith G, Bower BD, Eykyn SJ (1985) Timing of lumbar puncture in severe childhood meningitis. Br Med J 291 : 651-652 7. Klein JO, Marcy SM (1983) Bacterial sepsis and meningitis. In: Remington JS, Klein JO (eds) Infectious diseases of the fetus and newborn infant. WB Saunders Co, Philadelphia, pp 679-735 8. Louvois J de, Blackbourn J, Hurley R, Harvey DR (1988) Meningitis during the first year of life - a two year prospective study. British Paediatric Association, programme for the 60th Annual Meeting, abstract no P24, p 51 9. Mulhall A, Louvois J de, Hurley R (1983) Efficacy of chloramphenicol in the treatment of neonatal and infantile meningitis: a study of 70 cases. Lancet I : 284-287 10. Pearce PG, Roberton NRC (1986) Textbook of neonatology. Churchill Livingstone, London, pp 725-781 11. PHLS report (1985) Neonatal meningitis: a review of routine national data 1975-83. Br Med J 290 : 778-779 12. Schreiner RL, Kleiman MB (1979) Incidence and effect of traumatic lumbar puncture in the neonate. Dev Med Child Neurol 21 : 483-487 13. Smith KM, Deddish RB, Ogata ES (1986) Meningitis associated with serial lumbar punctures and post haemorrhagic hydrocephalus. J Pediatr 109 : 1057-1060 14. Spahr RC, Mueller-Huebach E (1981) Knee-chest position and neonatal oxygenation and blood pressure. Am J Dis Child 135 : 79-80 15. Speidel BD (1978) Adverse effects of routine procedures on preterm infants. Lancet I : 864 16. Teele DW, Dashefsky B, Rakusan T, Klein JO (1989) Meningitis after lumbar puncture in children with bacteremia. N Engl J Med 305 : 1079-1081 17. Weisman LE, Merenstein GB, Steenbarger JR (1983) The effect of lumbar puncture position in sick neonates. Am J Dis Child 137 : 1077-1079
