Antibiotics Chemother., vol. 21, pp. 146-150 (Karger, Basel 1976)
Diagnosis of Neonatal Sepsis via Matemofetal Transmission RELlER
and O.
DE
BETHMANN
Centre de Recherches biologiques neonatales (Directeur: A. Hopital Port-Royal, Paris
MINKOWSKI),
Septicemia in the neonate is a generalized bacterial infection proved by a positive blood culture. Clinical descriptions of neonatal sepsis are available in the literature. This paper will consider only certain hematologic aspects of primary septicemia or that which is diagnosed within 72 h after birth. Between January 1973 and June 1974, 31 patients with this neonatal sepsis thought to be secondary to maternal contamination, were admitted to our intensive care unit (lCU): 13 gram-negative and 18 gram-positive. The maternal infection was characterized by: fever either during delivery or during the 3 weeks before delivery, 11 cases (5 gram-negative, 6 gram-positive); urinary tract infection with the same organism recovered from infant, 3 cases (2 gram-negative, 1 gram-positive); infected vaginal discharge with the same organism recovered, 15 cases (7 gram-negative, 8 gram-positive); placental infection, 10 cases (5 gram-negative, 5 gram-positive). The blood cultures were taken from a peripheral vessel on admission to the ICU. Only one infant blood culture was negative on admission (4 h) but the culture on the 3rd day grew out group D streptococcus which was also found in the placenta. Table I shows the population studied with their times of admission, the number of cases of meningitis in each group and the outcome. Initial clinical signs occurred very early, 2 or 3 h after birth. Only 3 infants out of 31 arrived after 24 h of age. 1 patient with group D streptococcus (48 h) survived and 2 others with group B streptococcus (32 h) and E. coli (48 h) both died within 2 days after admission. The main causes for transfer to our unit were apnea, cyanosis and respiratory distress syndrome. In general in our unit antibiotics are not given until infection is strongly suspected. In addition to the history concerning the end of the pregnancy
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
J.P.
147
RELlER/DE BETHMANN
Table 1. Characteristics of neonatal sepsis
Organism
Group B streptococcus Listeria Group D streptococcus Nontypeable streptococcus E. coli Enterobacter cloacae Ristella distasonis Total
Number
Meningitis Time of arrival at ICU h
Alive
8 4
8 3
3 3
5 3
5 3
13 3
0 3
5 0
9 3 3
9 3 4
3 0 0
3 3
31
6
6
17
Shift to the left
Thrombocytopenia
5 (4 not possible) 4
7
4 (2 not counted)
4
Table II. Hematologic data
Organism
Leukopenia < 4,000
Neutropenia 30,000
3
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
and the delivery, the decision to give antibiotics was based on hematologic data and fibrinogen levels. Table II shows these hematologic data. The white cell count at birth may range from 9,000 to 30,000/mm3, and the premature infant has a lower total white cell count. It was therefore decided to define leukopenia as below 4,000/mm3, neutropenia below 1,500/mm3 and leukocytosis above 30,000/ mm 3. Qualitative changes are important to consider when the ratio of band forms to polymorphonuclear neutrophils is higher than 40%. It should be
148
RELlER/DE BETHMANN
noted that sometimes the measurement of this ratio is not possible when there is marked neutropenia (4 patients with E. coli sepsis). The severity of the hematologic disorders seemed to correspond with the severity of the disease. The 9 infants with E. coli sepsis had either leukopenia and neutropenia or leukocytosis with an abnormal shift to the left during the first day of life. Only the 2 patients with Enterobacter cloacae who died, had hematologic disorders. The 4 with listeriosis had an abnormal shift to the left but only 3 had leukocytosis and thrombocytopenia; these 3 died. Out of 8 cases with group B streptococcus, 3 died; 2 had severe leukopenia and neutropenia. Unfortunately, platelets were not counted on these 2 infants. 4 others had thrombocytopenia, 1 had normal platelets. Anemia was noted only once. Hyperbilirubinemia was not more frequent in this group B streptococcus sepsis than it was in other types. Ristella distanosis and group D streptococcus had no particular hematologic picture. These data suggest that leukopenia, neutropenia or leukocytosis are hallmarks of very severe septicemia while thrombocytopenia and shift to the left regularly seem to have no prognostic value. The increase of fibrinogen level during the first days of life seems to be a reliable finding in neonatal infection. In a previous study, reported at the 4th European Congress of perinatal medicine, we have shown that there is a significant increase in the fibrinogen level in infected infants in the first and the second days of life compared to noninfected infants (fig. 1). The imporp.;0,0001
p.;:O,oOl
5
4
3 0
;t.
01
~- 2 01 0
oE
.a
u:
!~
p;~
2020 Day 1
1213 Day 2
1
N
7 7 Day 3
Fig. 1. Mean fibrinogen level (g/l) on days 1, 2 and 3 in 20 infected (D) and 20 non(~)
newborn infants. Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
infected
149
Diagnosis of Neonatal Sepsis via Maternofetal Transmission
Table III. Evolution of fibrinogen levels (in gO/oo) according to the organism and outcome
Germes
Day 1
Day 2
Listeria (7)
5.25 4.17 2.45 5.03 6.71 6.90 3.09
3.09 4.25 7.60 3.98 5.80
6.04 2.50 1.87 3.72 2.23
5.71 3.84 2.54 2.46 3.79
2 4.15 6.15 4.30 5.74 1.30 3.02 4.65
3
Group B streptococcus (5)
Group D streptococcus (2) Staphylococcus (1) E. coli (4)
Proteus (1)
Day 3 3.20
3.52
2.88
4.20 dead 4.69 4.30 2.61
dead 3.24 dead dead 4.71
Evolution alive alive alive alive alive alive alive alive alive dead alive alive alive dead alive dead alive dead dead alive
Table IV. Fibrinogen value according to the organism and the clinical severity
E. coli Listeria
Group B streptococcus Enterobacter C1 Group D streptococcus Nontypeable streptococcus Ristella
Fibrinogen,
gO / 00
day 1
day 2
alive dead alive dead alive dead alive dead
4 5 1 3 5 3 1 2
4.40 2.45 6 4.60 3.80 1.75 4.24 2.94
5 2.10 7.60 5.60 4.10 ? 4.80 3.36
alive dead
5 1
2.10 2.20
3.50 3.30
alive
1
5.74
4.20 Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
Number
RELlER/DE BETHMANN
150
tance of this increase may allow one to predict the causative agents. Table III shows that the highest levels are found in gram-positive infections particularly involving Listeria. A similar rise was noted in the 31 infants with neonatal sepsis, i.e. on day 1, mean fibrinogen value was 3.60 gO/oo and on day 2 it was 4.35 gO /00. Table III shows the variations of this increase according to the organism and the clinical status, the more severe cases having no modification or occasionally a decrease of the fibrinogen value. We found the same data when looking separately at the different organisms responsible for neonatal sepsis (table IV). The explanation for this high fibrinogen value is the same as in children or in adults, where increased fibrinogen is part of the serum protein changes observed during the acute phase reaction. The synthesis of acute phase protein starts very early during fetal life, beginning at the 8th to 9th week of gestation. Thus the increased fibrinogen production by the fetal liver can presumably start as soon as the fetus is contaminated by maternal infection. This determination of fibrinogen levels can be very helpful in the early diagnosis of neonatal infection in the same way as measurements of C-reactive proteins, orosomucoid or even haptoglobin levels. In conclusion, in addition to classical clinical and biological data which are sometimes difficult to interpret in the neonate, there are many laboratory determinations available when one suspects neonatal infection. Each of these arguments has to be considered very carefully in order to decide whether or not it is useful to give antibiotics to the distressed infants. The choice of antibiotics can be made before bacteriological results are obtained. Of course, direct examination of the placenta with Gram's stain may yield relevant information. Otherwise, in most cases the combination ampicillin-gentamicin covers a sufficiently broad spectrum. Urinary infections, premature rupture of the membranes if treated with only ampicillin or fJ-Iactamin without culture results, are considered to be indications for giving colimycingentamicin. Predicting a severe outcome, the combination of sclerema neonatorum, leukopenia, neutropenia and low fibrinogen value, is an indication for a new therapeutic approach like repeated exchange transfusions with fresh blood. It seems to us that only this new therapeutic method can improve the survival rate of the severe primary neonatal sepsis which otherwise has a bad outcome even when treated very early after birth.
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
Dr. J.P. RELlER, Centre de Recherches biologiques neonatales, Hopital Port-Royal, 123, boulevard Port-Royal, F-75014 Paris (France)
Diagnosis of Neonatal Sepsis via Matemofetal Transmission RELlER
and O.
DE
BETHMANN
Centre de Recherches biologiques neonatales (Directeur: A. Hopital Port-Royal, Paris
MINKOWSKI),
Septicemia in the neonate is a generalized bacterial infection proved by a positive blood culture. Clinical descriptions of neonatal sepsis are available in the literature. This paper will consider only certain hematologic aspects of primary septicemia or that which is diagnosed within 72 h after birth. Between January 1973 and June 1974, 31 patients with this neonatal sepsis thought to be secondary to maternal contamination, were admitted to our intensive care unit (lCU): 13 gram-negative and 18 gram-positive. The maternal infection was characterized by: fever either during delivery or during the 3 weeks before delivery, 11 cases (5 gram-negative, 6 gram-positive); urinary tract infection with the same organism recovered from infant, 3 cases (2 gram-negative, 1 gram-positive); infected vaginal discharge with the same organism recovered, 15 cases (7 gram-negative, 8 gram-positive); placental infection, 10 cases (5 gram-negative, 5 gram-positive). The blood cultures were taken from a peripheral vessel on admission to the ICU. Only one infant blood culture was negative on admission (4 h) but the culture on the 3rd day grew out group D streptococcus which was also found in the placenta. Table I shows the population studied with their times of admission, the number of cases of meningitis in each group and the outcome. Initial clinical signs occurred very early, 2 or 3 h after birth. Only 3 infants out of 31 arrived after 24 h of age. 1 patient with group D streptococcus (48 h) survived and 2 others with group B streptococcus (32 h) and E. coli (48 h) both died within 2 days after admission. The main causes for transfer to our unit were apnea, cyanosis and respiratory distress syndrome. In general in our unit antibiotics are not given until infection is strongly suspected. In addition to the history concerning the end of the pregnancy
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
J.P.
147
RELlER/DE BETHMANN
Table 1. Characteristics of neonatal sepsis
Organism
Group B streptococcus Listeria Group D streptococcus Nontypeable streptococcus E. coli Enterobacter cloacae Ristella distasonis Total
Number
Meningitis Time of arrival at ICU h
Alive
8 4
8 3
3 3
5 3
5 3
13 3
0 3
5 0
9 3 3
9 3 4
3 0 0
3 3
31
6
6
17
Shift to the left
Thrombocytopenia
5 (4 not possible) 4
7
4 (2 not counted)
4
Table II. Hematologic data
Organism
Leukopenia < 4,000
Neutropenia 30,000
3
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
and the delivery, the decision to give antibiotics was based on hematologic data and fibrinogen levels. Table II shows these hematologic data. The white cell count at birth may range from 9,000 to 30,000/mm3, and the premature infant has a lower total white cell count. It was therefore decided to define leukopenia as below 4,000/mm3, neutropenia below 1,500/mm3 and leukocytosis above 30,000/ mm 3. Qualitative changes are important to consider when the ratio of band forms to polymorphonuclear neutrophils is higher than 40%. It should be
148
RELlER/DE BETHMANN
noted that sometimes the measurement of this ratio is not possible when there is marked neutropenia (4 patients with E. coli sepsis). The severity of the hematologic disorders seemed to correspond with the severity of the disease. The 9 infants with E. coli sepsis had either leukopenia and neutropenia or leukocytosis with an abnormal shift to the left during the first day of life. Only the 2 patients with Enterobacter cloacae who died, had hematologic disorders. The 4 with listeriosis had an abnormal shift to the left but only 3 had leukocytosis and thrombocytopenia; these 3 died. Out of 8 cases with group B streptococcus, 3 died; 2 had severe leukopenia and neutropenia. Unfortunately, platelets were not counted on these 2 infants. 4 others had thrombocytopenia, 1 had normal platelets. Anemia was noted only once. Hyperbilirubinemia was not more frequent in this group B streptococcus sepsis than it was in other types. Ristella distanosis and group D streptococcus had no particular hematologic picture. These data suggest that leukopenia, neutropenia or leukocytosis are hallmarks of very severe septicemia while thrombocytopenia and shift to the left regularly seem to have no prognostic value. The increase of fibrinogen level during the first days of life seems to be a reliable finding in neonatal infection. In a previous study, reported at the 4th European Congress of perinatal medicine, we have shown that there is a significant increase in the fibrinogen level in infected infants in the first and the second days of life compared to noninfected infants (fig. 1). The imporp.;0,0001
p.;:O,oOl
5
4
3 0
;t.
01
~- 2 01 0
oE
.a
u:
!~
p;~
2020 Day 1
1213 Day 2
1
N
7 7 Day 3
Fig. 1. Mean fibrinogen level (g/l) on days 1, 2 and 3 in 20 infected (D) and 20 non(~)
newborn infants. Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
infected
149
Diagnosis of Neonatal Sepsis via Maternofetal Transmission
Table III. Evolution of fibrinogen levels (in gO/oo) according to the organism and outcome
Germes
Day 1
Day 2
Listeria (7)
5.25 4.17 2.45 5.03 6.71 6.90 3.09
3.09 4.25 7.60 3.98 5.80
6.04 2.50 1.87 3.72 2.23
5.71 3.84 2.54 2.46 3.79
2 4.15 6.15 4.30 5.74 1.30 3.02 4.65
3
Group B streptococcus (5)
Group D streptococcus (2) Staphylococcus (1) E. coli (4)
Proteus (1)
Day 3 3.20
3.52
2.88
4.20 dead 4.69 4.30 2.61
dead 3.24 dead dead 4.71
Evolution alive alive alive alive alive alive alive alive alive dead alive alive alive dead alive dead alive dead dead alive
Table IV. Fibrinogen value according to the organism and the clinical severity
E. coli Listeria
Group B streptococcus Enterobacter C1 Group D streptococcus Nontypeable streptococcus Ristella
Fibrinogen,
gO / 00
day 1
day 2
alive dead alive dead alive dead alive dead
4 5 1 3 5 3 1 2
4.40 2.45 6 4.60 3.80 1.75 4.24 2.94
5 2.10 7.60 5.60 4.10 ? 4.80 3.36
alive dead
5 1
2.10 2.20
3.50 3.30
alive
1
5.74
4.20 Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
Number
RELlER/DE BETHMANN
150
tance of this increase may allow one to predict the causative agents. Table III shows that the highest levels are found in gram-positive infections particularly involving Listeria. A similar rise was noted in the 31 infants with neonatal sepsis, i.e. on day 1, mean fibrinogen value was 3.60 gO/oo and on day 2 it was 4.35 gO /00. Table III shows the variations of this increase according to the organism and the clinical status, the more severe cases having no modification or occasionally a decrease of the fibrinogen value. We found the same data when looking separately at the different organisms responsible for neonatal sepsis (table IV). The explanation for this high fibrinogen value is the same as in children or in adults, where increased fibrinogen is part of the serum protein changes observed during the acute phase reaction. The synthesis of acute phase protein starts very early during fetal life, beginning at the 8th to 9th week of gestation. Thus the increased fibrinogen production by the fetal liver can presumably start as soon as the fetus is contaminated by maternal infection. This determination of fibrinogen levels can be very helpful in the early diagnosis of neonatal infection in the same way as measurements of C-reactive proteins, orosomucoid or even haptoglobin levels. In conclusion, in addition to classical clinical and biological data which are sometimes difficult to interpret in the neonate, there are many laboratory determinations available when one suspects neonatal infection. Each of these arguments has to be considered very carefully in order to decide whether or not it is useful to give antibiotics to the distressed infants. The choice of antibiotics can be made before bacteriological results are obtained. Of course, direct examination of the placenta with Gram's stain may yield relevant information. Otherwise, in most cases the combination ampicillin-gentamicin covers a sufficiently broad spectrum. Urinary infections, premature rupture of the membranes if treated with only ampicillin or fJ-Iactamin without culture results, are considered to be indications for giving colimycingentamicin. Predicting a severe outcome, the combination of sclerema neonatorum, leukopenia, neutropenia and low fibrinogen value, is an indication for a new therapeutic approach like repeated exchange transfusions with fresh blood. It seems to us that only this new therapeutic method can improve the survival rate of the severe primary neonatal sepsis which otherwise has a bad outcome even when treated very early after birth.
Downloaded by: Université de Paris 193.51.85.197 - 1/28/2020 8:02:52 PM
Dr. J.P. RELlER, Centre de Recherches biologiques neonatales, Hopital Port-Royal, 123, boulevard Port-Royal, F-75014 Paris (France)
![[Diagnosis of maternofetal infections].](/assets/img/hold.png)
