text in 
English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
Bacterial meningitis is a severe infection with a high global burden of morbidity and mortality 1. Neonates are particularly susceptible to these infections due to their immature immune system, characterized by deficient humoral and cellular effector responses 2-3. Furthermore, diagnosis is challenging due to the nonspecific clinical manifestations of meningeal involvement 4. Adverse outcomes are common, with high mortality rates (around 40%) and permanent neurological sequelae such as auditory, visual, motor, and language impairments 5-7.
The incidence and etiology of neonatal bacterial meningitis (NBM) exhibit heterogeneous behavior concerning socioeconomic conditions, type of delivery, and adherence or access to prenatal care 8-9. Several authors suggest that the etiology differs between high and low-income countries. In the former, Streptococcus agalactiae (group B streptococcus) and Escherichia coli predominate, while in the latter, bacteria such as Enterococcus spp., Klebsiella spp., and Enterobacter spp. are more common 8,10. In the United States, the majority of cases are attributed to group B streptococci (GBS) followed by Escherichia coli, whereas in African countries, other Gram-negative bacilli, such as Klebsiella pneumoniae, are predominant 11-12. There are exceptional reports of bacteria like Streptococcus pneumoniae and Haemophilus influenzae causing neonatal meningitis, primarily in low-income countries; however, these are more significant in children or adults 13.
Globally, there are approximately 126,000 cases of NBM per year with a mortality rate reaching up to 58% in low-income countries, whereas in high-income countries, it decreases to 10% 5,8.
In Colombia, individual reporting of this infection during the neonatal period is not carried out. Instead, cases occurring among these patients are included within the pediatric group 14-15, a practice that obscures the specific profile of bacterial meningitis, whose etiology varies widely according to age. A study in the United States demonstrates that in NBM, the etiological agents and their frequencies of occurrence in children, both inside and outside the neonatal period, exhibit stark differences 12. Identifying these differences is especially important for adjusting empirical antibiotic therapy 16.
Due to the lack of local knowledge on this issue, the aim of this study was to describe the clinical, microbiological, and epidemiological characteristics of neonatal bacterial meningitis in three hospital centers in the city of Medellín from 2010 to 2019.
Methodology
Electronic records of neonates diagnosed with bacterial meningitis from January 1, 2010 to December 31, 2019 were obtained. These neonates were treated in three reference hospitals: Hospital Universitario San Vicente Fundación, Clínica Universitaria Bolivariana, and Hospital General de Medellín.
Diagnoses were conducted following the codes from the International Classification of Diseases (ICD-10). Patients aged ≤28 days with bacterial isolation in cerebrospinal fluid (CSF), or in blood if there was also an alteration in the former, were included. Cases where no microorganism was isolated or had an additional risk factor for central nervous system infection, such as myelomeningocele, cranial trauma, encephalocele, immunodeficiencies, or central nervous system surgeries, were excluded.
Medical histories and laboratory tests were extracted from the SAP Business Suite® and Servinte Clinical Suite Enterprise® electronic management systems. The information was consolidated into a Google® spreadsheet containing the variables of interest (demographic data, signs and symptoms, diagnoses, microbiological cultures, antibiograms, hemograms, CSF cytology, dates, mother’s data, imaging, molecular and serological tests) defined according to relevant literature.
Subsequently, these variables were adjusted through a pilot test conducted with 10 initial medical records to select the definitive ones considering the reporting frequency of each variable. Two researchers searched for and entered the information independently. Data verification was done in Microsoft Excel® v19.0 (contingent and range verification) 17. The statistical analysis was performed using the IBM Statistical Package for the Social Sciences SPSS®26 software.
In cases where microbial isolation was obtained only from blood samples, the CSF cytology had to show at least two of the following criteria to be included: leukocyte count ≥16 and ≥30 cells/ mm3 (for full-term and premature patients, respectively), glucose ≤24 mg/dL, proteins ≥128 mg/dL, positive CSF Gram stain, or a positive result in CSF serological or molecular tests 18.
Cases in which an unusual microorganism was identified (such as coagulase-negative staphylococci), were analyzed by a pediatric infectious disease specialist, and only those approved based on clinical and microbiological criteria were included. The protein concentration and leukocyte count in CSF were corrected in case of traumatic lumbar puncture; proportions of 1000 erythrocytes/mm3 (CSF):1.1 mg/dL protein (CSF) and 1000 erythrocytes/mm3 (CSF):1 leukocyte/mm3 (CSF) were used, respectively 19-20.
Based on reports from other studies regarding the elapsed time between symptom onset and diagnosis 21, all symptoms presented by the newborn starting two days before the lumbar puncture were included. The results are presented through tables and descriptive statistics.
Ethical aspects
Proper disposition, handling, and custody of each medical record were ensured during the study and dissemination of its results. The project received approval from the ethics committees of the three hospital institutions and the Faculty of Medicine of the University of Antioquia.
Results
A total of 125 records of patients diagnosed with meningitis, according to the ICD-10, were analyzed. Only 42 met the inclusion criteria, and 83 were excluded for the following reasons: no microbiological isolation 46, incomplete microbiological information 17, isolation of coagulase-negative staphylococci (CoNS) without correlation with an altered CSF (10), congenital diseases 8, and explicit culture contamination throughout its processing, including collection 2. Organized by the institution from which they originated, the sample consisted of 16 records from Hospital Universitario San Vicente Fundación, 13 from Clínica Universitaria Bolivariana, and 13 from Hospital General de Medellín.
Epidemiological Information
50% of the patients were premature; 21% had a gestational age (GA) between 33 and 36 weeks, and 29% had a GA of ≤32 weeks. 42.9% of the cases were children hospitalized from birth. After diagnosis, four neonates died during their stay (total mortality), three of them with a history of prematurity (Table 1).
Table 1 Epidemiological information and other data related to admission
| Variable | Categories | n* | Fi (%n) | x̄ | SD |
|---|---|---|---|---|---|
| Infant Information: | |||||
| Gestational Age† | - | 42‡ | (100) | 35 | 4 |
| Weight (g) | - | 42 | (100) | 2352.6 | 833.9 |
| Height (cm) | - | 39 | (92.8) | 44.9 | 5.2 |
| Gender | Female | 42 | 14 (33.3) | - | - |
| Male | 28 (66.7) | ||||
| Infection Onset (days) | - | 42 | (100) | 11 | 9 |
| Mother's Information: | |||||
| Mother's Age (years) | - | 36 | (85.7) | 25 | 8 |
| Parity | - | 37 | (88.1) | 2 | 1 |
| Prenatal Check-ups | - | 26 | (61.9) | 6 | 2 |
| Type of Delivery | Vaginal | 41 | 32 (78) | - | - |
| Cesarean | 9 (22) | ||||
| City of Residence | Medellín | 41 | 18 (43.9) | - | - |
| Non-Metropolitan Area§ | 16 (39) | ||||
| About the Disease: | |||||
| Hospital Stay (days) | - | 42 | (100) | 34 | 21 |
| Infection Classification | Early | 42 | 12 (28.6) | - | - |
| Late‖ | 30 (71.4) | ||||
| Death¶ | - | 41 | 4 (9.8) | - | - |
* Represents valid n.
Fi: frequency
x̄: average
SD: standard deviation
† Gestational age in weeks
§ FAM: Outside the metropolitan area
‖ ≥ 4 days after birth
Total deaths during hospital stay
Source: own elaboration
Clinical Information
71.4% of the newborns had a late-onset infection, and 28.6% had one that was early-onset. Fevers and irritability were the most prevalent symptoms (Table 2). Muscle tone alteration was observed in 15 children, hypotonia in 12, and hypertonia in 3. Meningeal signs were not evident in any case. Regarding CSF parameters, the most common finding was hyperproteinorrachia (85.7%), followed by pleocytosis at 82.1%. Of these, 46% had counts between 31 and 1000 cells/mm³, and 20.5% between 1001 and 4000 cells/mm³. Hypoglycorrhachia (<25 mg/dL) was found in 20 cases (47.6%) (Table 2).
Table 2 Symptoms, laboratory parameters, and invasive interventions
| CSF Parameters | n* | Fi (%n) | Median | Q1 | Q3 | |
|---|---|---|---|---|---|---|
| Proteins mg/dL | 42† | (100) | 310.6 | 151.4 | 498.4 | |
| Glucose mg/dL | 42 | (100) | 20.3 | 9.0 | 46.3 | |
| Leukocytes | 39 | (92.9) | 342.0 | 65.0 | 2250.0 | |
| PMN‡ (%) | 36 | (85.7) | 88.0 | 69.8 | 93.6 | |
| MON (%) | 35 | (83.3) | 10.0 | 6.0 | 32.0 | |
| Blood PCR mg/dL | 39 | (92.9) | 7.0 | 3.2 | 14.0 | |
| Turbid appearance | 39 | 34 (87.2) | - | - | - | |
| Xanthochromic | 39 | 25 (64.1) | - | - | - | |
| Gram stain (positive) | 41 | 16 (39) | - | - | - | |
| Culture (positive) | 41 | 23 (56.1) | - | - | - | |
| Symptoms | n* | Fi (%n) | ||||
| Fevers | 20 (48.8) | |||||
| Seizures | 12 (29.3) | |||||
| Bulging fontanelle | 41 | 2 (4.9) | ||||
| Drowsiness | 9 (22) | |||||
| Irritability | 23 (56.1) | |||||
| Inappetence | 7 (17.1) | |||||
| Sepsis§ | 42 | 38 (90.5) | ||||
| Invasive Interventions | n* | Fi (%n) | ||||
| Urinary catheter | 16 (38.1) | |||||
| Epicutaneous catheter | 42 | 22 (52.4) | ||||
| Mechanical ventilation | 13 (31) | |||||
| Umbilical catheterization | 12 (28.6) | |||||
* Represents valid n. Fi: frequency; Median; Q1: quartile 1; Q3: quartile 3
† Numbers in bold represent the absolute frequency for each variable
‡ PMN: polymorphonuclear; MON: monocytes; PCR: C-reactive protein
§ Diagnosis concurrent with or prior to meningitis Source: own elaboration
Regarding the blood count, 70% of the cases were in the range of 5 to 23 x103 cells/µL for leukocyte count, and 26.2% had leukopenia (<5 x 103 cells/µL). The average was 11.8 x 103 cells/ µL (SD 7.9). The mean relative value was 52.1% for neutrophils, 32.7% for lymphocytes, and 11.4% for monocytes. 39% (16 of 41) of the patients had thrombocytopenia (platelet count <150,000/µL).
Microbiological Information
The most frequent isolates were Escherichia coli, present in 17 cases (40.5%), whose susceptibility profile showed resistance to several cephalosporins, including cefepime; moreover, in two of these isolates, extended-spectrum beta-lactamases (ESBL) were expressed. Streptococcus agalactiae was also present in 17 patients. The sensitivity of this microorganism to penicillin was 100% (Table 3).
Table 3 Bacteria isolated in neonates with bacterial meningitis and antibiotic susceptibility
| Etiologic Agent | Fi (n%) | PEN | AMP | OXA | SAM | TZP | FOX | CTX | CAZ | CRO | FEP | ATM | IPM | MEM | ETP | AMK | GEN | CIP | LVX | LZD | VAN | SXT | Isolates | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Blood | CSF | ||||||||||||||||||||||||
| Escherichia Coli | 17 | _ | _ | _ | *6/16 | 14/14 | 10/10 | _ | 5/7 | 3/4 | 14/16 | 6/6 | 8/8 | 7/7 | 7/7 | 17/17 | *11/13 | 9/17 | _ | _ | _ | 0/2 | 9 | 8 | |
| (40,5) | (37,5) | (100) | (80) | (71,4) | (75) | (87,5) | (100) | (100) | (100) | (100) | (100) | (84,6) | (52,9) | (0) | |||||||||||
| Klebsiella pneumoniae | 1 | _ | _ | _ | 1/1 | _ | _ | _ | _ | _ | 1/1 | _ | _ | _ | _ | 1/1 | 1/1 | 1/1 | _ | _ | _ | _ | 1 | ||
| (2,4) | (100) | (100) | (100) | (100) | (100) | ||||||||||||||||||||
| Pseudomonas aeruginosa | 2 | _ | _ | _ | _ | 1/1 | _ | _ | 1/1 | _ | 2/2 | 1/1 | 2/2 | 2/2 | _ | 2/2 | 1/1 | 1/1 | 1/1 | _ | _ | _ | 1 | 1 | |
| (4,8) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | |||||||||||||||
| Pseudomonas fluorescens | 1 | _ | _ | _ | _ | 1/1 | 0/1 | _ | _ | 1/1 | _ | *0/1 | _ | 1/1 | 1/1 | 1/1 | _ | _ | _ | _ | 1 | ||||
| (2,4) | (100) | (0) | (100) | (0) | (100) | (100) | (100) | ||||||||||||||||||
| Streptococcus agalactiae | 17 | 3/3 | 16/16 | _ | _ | _ | _ | 2/2 | _ | 2/2 | _ | _ | _ | _ | _ | _ | _ | _ | 1/1 | 6/6 | 5/5 | 4/4 | 6 | 8 | |
| (40,5) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | (100) | |||||||||||||||||
| Staphylococcus epidermidis | 1 | _ | _ | 0/1 | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | 1 | ||
| (2,4) | (0) | ||||||||||||||||||||||||
| Staphylococcus aureus | 1 | _ | _ | 1/1 | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | 1 | 1 | |
| (2,4) | (100) | ||||||||||||||||||||||||
| Staphylococcus hominis | 1 | _ | _ | 0/1 | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | 1/1 | 1/1 | 1/1 | 1 | 1 | |
| (2,4) | (0) | (100) | (100) | (100) | |||||||||||||||||||||
| Staphylococcus warneri | 1 | _ | _ | 0/1 | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | _ | 1/1 | 1/1 | 0/1 | 1 | ||
| (2,4) | (0) | (100) | (100) | (0) | |||||||||||||||||||||
| Total | 42 | ||||||||||||||||||||||||
Numbers are presented as susceptible bacteria/tested bacteria (% susceptible). Antibiotic acronyms according to WHONET: PEN: penicillin; AMP: ampicillin; OXA: oxacillin; SAM: ampicillin-sulbactam; TZP: piperacillintazobactam; FOX: cefoxitin; CTX: cefotaxime; CAZ: ceftazidime; CRO: ceftriaxone; FEP: cefepime; ATM: aztreonam; IPM: imipenem; MEM: meropenem; ETP: ertapenem; AMK: amikacin; GEN: gentamicin; CIP: ciprofloxacin; LVX: levofloxacin; LZD: linezolid; VAN: vancomycin; SXT: trimethoprim-sulfamethoxazole. Fi: frequency. Intermediate sensitivity: E. coli, 7 SAM and 1 GEN; S. fluorescens 1 MEM.
The antibiotics shown for each microorganism are those suggested by CLSI M100-ED30, approved for clinical use by the FDA.
Source: clsi.org free resources
Empirical Antibiotic Therapy
The combination of ampicillin and an aminoglycoside was used in 64.1% of the patients (25/39) (amikacin in 20 of 25 and gentamicin in 5 of 25), ampicillin plus a cephalosporin in 17.9% (cefotaxime 5 of 39; cefepime 2 of 39), and cefepime plus vancomycin or meropenem in 17.9%.
Discussion
This research describes the clinical, microbiological, and epidemiological characteristics of neonatal meningitis in three reference institutions in the city of Medellín. Local data are limited, and this knowledge is required for the correct and timely approach to empirical treatment.
This study identified Streptococcus agalactiae and Escherichia coli as the predominant agents, revealing an etiological similarity with high-income countries and a difference concerning studies carried out in South America. In Peru, Guillen et al. 22 describe Escherichia coli and Listeria monocytogenes as the most prevalent (17% and 13% respectively), while Streptococcus agalactiae was less common (3%). On the other hand, in Brazil 23, a wide variety of enterobacteria is reported with a significant presence of Staphylococcus aureus (11.7%) and Neisseria meningitidis (8.8%). These differences reaffirm the need to study the local etiological profile and the inconvenience of defining empirical therapy by extrapolating data from other countries, even if they may be similar socioeconomically.
Based on the onset of clinical manifestations, neonatal bacterial meningitis (NBM) is classified into early meningitis (which occurs in the first 72 hours of life) and late meningitis (which occurs after the first three days of life). In the early form, the most common agents are those that colonize the maternal genital tract, GBS and Escherichia coli, primarily 8. GBS has been described in 35% of screening cultures in asymptomatic pregnant women 24. In contrast, Staphylococcus aureus, coagulase-negative staphylococci (CoNS), Pseudomonas aeruginosa, and other Gram-negative bacilli found in the hospital environment cause late meningitis 4.
Our data shows that 3 of 17 meningitis cases due to Escherichia coli and 8 of 7 due to GBS started before 72 hours of life. Together, they constitute 91.7% of the total cases of early-onset infection. It is also observed that there are almost three times more cases where the infection had a late onset compared to those with an early onset. The importance of these microorganisms in late sepsis is highlighted, where intrapartum antibiotic prophylaxis has no impact. Even if mothers have a negative culture in prenatal screening, the literature reports late meningitis by GBS transmitted by breast milk or cross-contamination. It also warns that not all meningitis cases due to Escherichia coli are exclusively of vertical transmission, as this bacterium is highly frequent in neonatal infections associated with health care 25-28.
CNS are more frequently isolated in late neonatal infections acquired in the hospital, mainly bacteremia associated with the use of invasive devices 29-30. Although they are not usually found in neuroinfection, up to 12% of neonatal meningitis cases can be due to CNS 30. In this study, three (7.1%) were identified with clinical correlation in CSF cytochemistry. All patients were premature, with a birth weight less than 1500 g and a late onset of the disease. In addition, they had an epicutaneous central catheter, and 2 of 3 of the neonates underwent umbilical catheterization. Despite their low frequency (but great importance) in patients with prolonged hospital stays, each case should be carefully evaluated not to underestimate these bacteria as causal agents of meningitis in the neonatal stage 31.
This research identified three infections by Pseudomonas spp., a frequent opportunistic bacterium in hospital-acquired infections 29. However, cases of meningitis have been described by vertical transmission of this bacterium when mothers are vaginal carriers 32. One of the newborns infected with Pseudomonas aeruginosa was not associated with a hospital stay, as it had an early presentation of the disease, and this microorganism is unusual at this stage. Although no microbiological information was available from the mother, given the characteristics of this case, it is feasible to think that it could have been an infection due to vertical transmission.
Meningitis in the neonatal period presents with an insidious clinic and is usually secondary to bacteremia with clinical manifestations of late sepsis 33. The most common clinical manifestations are fevers, irritability, and respiratory problems, while the least common are seizures and a bulging fontanel 22.
The findings in this study correlate with what is mentioned in the medical literature regarding irritability and fevers as the main symptoms in NBM. It is highlighted that in 17% of the patients no symptoms were evidenced, and a similar behavior has been found in other studies 34-35. The absence of symptoms reaffirms the importance of the patient’s comprehensive evaluation through the analysis of the hemoleukogram, vital signs, maternal screening, membrane rupture time, and other risk factors suggestive of systemic infection that justify the lumbar puncture and CSF analysis.
Low birth weight (LBW) and premature birth are described as risk factors for the development and fatal outcome of NBM 36,37. We found that 50% of the children were preterm babies, and some NBM studies report prematurity in the order of 39% 21. The average gestational age in the four patients who died was 33 weeks, and the birth weight was 2214 g.
CSF analysis is one of the most critical tools in the early diagnosis of meningeal infection. However, it should not be the only decision-making criterion regarding the type of intervention provided to the newborn.
The negative predictive value of pleocytosis is not 100%. In this regard, Garges et al. 38 describe that up to 5% of neonates with meningitis can have a white blood cell count (WBC) in CSF between 0 and 1 WBCs/mm3; and in 10%, it is ≤3, a finding consistent with what was found in this research, where there were cases without pleocytosis in CSF. A study that included 9,111 newborns reported that the use of 20 WBCs/mm3 as a cut-off point in CSF caused a false diagnosis in 13% of neonates with confirmed meningitis (38). Something similar happens with CSF glucose and protein values that overlap in some cases between patients with and without NBM 39-40.
In this research, 14.3% of the children did not have alterations in proteins, and 40.5% did not have alterations in glucose; fatal cases showed an average glucose of 21.2 mg/dL and proteins of 749.1 mg/dL, which correlates with what is described regarding patients with very high protein values and hypoglycorrhachia having a higher risk of mortality or sequelae 40-42.
The cellular parameters of the hemogram are not very useful for the diagnosis of meningitis 38, but they can be predictors of a poor prognosis. Klinger et al. 43 mention that thrombocytopenia in patients with neonatal meningitis is an essential predictor of complications. Since in some cases it was explicitly expressed in the patient’s clinical history that a traumatic lumbar puncture had occurred and the protein and cellular quantification in CSF showed abnormally elevated data, we decided to apply the correction based on the correlation of these analytes in CSF with their values in the blood. This study found thrombocytopenia in 39% of cases. Additionally, 26% of the children had leukopenia (<5x103 WBCs/µL in blood), a finding referred to by Garges et al. 38 as a predictor of poor prognosis in meningitis.
According to what was observed in this research, the results of the CRP (C-reactive protein) had great variation (Table 2), and we believe that the wide dispersion of the values is due to it being measured at different times for each case and that it is an acute phase reactant. Its increase will be proportional to the patient’s condition. It has been reported that the utility of CRP in NBM is reduced 44. Some authors mention that peripheral blood laboratory parameters help distinguish whether an isolate is due to contamination, especially in CNS cases, an aspect that was taken into account in this research to determine if it was indeed an infection 30.
The maternal information that could be collected was very limited. The low socioeconomic stratum was predominant, which is expected given the type of population served in the hospitals included in the study. Regarding prenatal controls, 4 of 26 mothers attended less than four times, a fact that is a risk factor for neonatal sepsis and NBM 45, and that could be a reflection of the untimely nature of the Colombian health system in providing complete assistance and prenatal follow-up to pregnant women in vulnerable conditions 46. It is proposed for future research to compare the findings of this study with results in non-vulnerable populations treated in other medical centers.
There are very few NBM studies that include the antibiotic susceptibility of the isolated microorganisms. Some strains of Escherichia coli showed resistance to third-generation cephalosporins, such as ceftazidime in 2 of 7 isolates. This finding in such a small group of isolates is a limitation, but it is worrisome because cephalosporins are part of empirical treatment. Other authors declare the increase in resistance to these antimicrobials and the expansion of ESBL as a growing problem in the pediatric population 47-48. Even more important was the finding of two strains resistant to cefepime, the most commonly used empirical antibiotic in late neonatal sepsis for coverage of central nervous system infections.
The data obtained on empirical antibiotic therapy show that in clinical practice there is adherence to the protocols and recommendations established by clinical management guidelines for neonatal sepsis in Colombia since the suggested combination of ampicillin-aminoglycoside/ cephalosporin was used in 82% of the cases; the other antibiotic regimens used, according to the literature, are also used in NBM under medical criteria 49-50.
It is advisable to periodically review antibiotic susceptibility, as resistance to third-generation cephalosporins may compromise the effectiveness of empirical therapy.
Conclusions
The causative agents of NBM were Streptococcus agalactiae and Escherichia coli, both in early and late-onset sepsis. Although the most common symptoms are irritability and fevers, a significant number of patients do not show clinical infectious manifestations and are only diagnosed as part of the comprehensive approach to late neonatal sepsis.
The most frequent alterations in cerebrospinal fluid were hyperproteinorrachia and pleocytosis; however, the absence of these does not rule out an infection in the central nervous system. This highlights the role of microbiological, molecular, and serological studies for diagnosis. The antibiotic regimen of empirical therapy is in line with Colombian and international clinical guidelines. Constant surveillance of local etiology is recommended independently for the neonatal stage.
Limitations
Given that only cases with positive isolation in just three hospitals were included, this work might not account for the actual etiological prevalence of NBM in Medellín. However, these institutions were strategically chosen as reference centers in the city, and the information presented is a first approach to the local situation of neonatal bacterial meningitis.
The fact that electronic medical records are relatively new in Colombia limited the time span that the study could cover. Information was not found on maternal conditions during pregnancy, duration of membrane rupture, prophylaxis, and screening for Streptococcus agalactiae. It was not feasible to follow up on the patient’s medical history to capture information on sequelae. Additionally, no further verification of the finding of ESBL-producing organisms was carried out because susceptibility results for some antibiotics were not available. By establishing such rigorous inclusion criteria, there might be a selection bias and valuable data for the characterization of NBM could be overlooked.
