Rita Calejo Pereira Machado Ribeiro1, Joana Raquel Pinto de Carvalho Queirós2, Ana Inês Rodrigues Paiva Ferreira1, Inês Isabel Aires Martins2, Fábio David Monteiro Barroso1.
1Pediatrics and Neonatology Department, Centro Hospitalar do Tâmega e Sousa, Penafiel, Portugal,
2Pediatrics Department, Centro Materno Infantil do Norte - Centro Hospitalar Universitário do Porto, Porto, Portugal.
ADDRESS FOR CORRESPONDENCE
Rita Calejo Pereira Machado Ribeiro, Avenida do Hospital Padre Américo 210, 4564-007.
Email: ritacalejor@gmail.com | | Abstract | Introduction: Fever is one of the main complaints in children admitted to the Emergency Department. It can be a symptom of Serious Bacterial Infection (SBI), being the risk greater in children younger than 3 months.
The aim of this study is to evaluate predictive factors of SBI under 3 months.
Materials and methods: We conducted a cohort retrospective study from January 2020 to June 2022 in infants under 3 months who visited the Emergency Department due to fever without a source. Firstly, a bivariate analysis was made to assess the predictive value of clinical and laboratorial parameters. Posteriorly, a backward stepwise model of binary logistic regression was performed, considering the statistical significant variables, with the dependent variable being the presence or absence of SBI.
Results and discussion: A total of 292 patients were included, 160 (54.8%) were male. The mean age was 57.8 days (SD 22.3). 73 (25%) patients were diagnosed with SBI: 58 (79.5%) urinary tract infections, 13 (17.8%) occult bacteriemias, 8 (11.0%) meningitis, 3 (4.1%) pneumoniae and 6 (8.2%) bacterial gastroenteritis. Classification according to the Rochester criteria, white blood cell and neutrophil count, C-Reactive Protein (CRP) and procalcitonin values were statistically significantly different between the groups (p < 0.001). In the multivariate analysis, CRP values (p <0.001; OR 1.045) showed significance (Hosmer-Lemeshow test: x2= 12.043; p=0.149).
Conclusion: According to this study, CRP value is the most reliable parameter to predict SBI in this population and should be integrated in the evaluation of these patients to help in the treatment and follow-up decision. | | | | Keywords | Fever without a source, Serious Bacterial Infection, Approach.
Abbreviations:
- CRP – C-Reactive Protein
- CSF – Cerebrospinal fluid
- ESBL - Extended Spectrum Beta-Lactamase
- HPF – High Power Field
- IQR - Interquartile Range
- Mdn – Median
- OR – Odds ratio
- RSV - Respiratory Syncytial Virus
- SBI - Serious Bacterial Infection
- SD – Standart Deviation
- SPSS - Statistical Package for the Social Sciences
- WBC - White blood cell
- YIOS - Young Infant Observation Scale | | | | Introduction | Fever is one of the main complaints in children admitted to the Emergency Department, accounting for 10-20% of visits.1 Fever without a source stands as an important entity, as the etiology can be a viral infection and have a benign evolution or be a manifestation of a serious illness with potential severe consequences, such as important morbidity and mortality2,3,4 According to age, etiologies vary and consequently the diagnostic and treatment approach differ.2,3,4,5 An important differential diagnosis in a febrile infant under 3 months of age is a Serious Bacterial Infection (SBI), as this group has a higher risk of SBI than older children, affecting up to 10% of patients, particularly newborns.1,2,3,4,5,6
The majority of guidelines concerning fever without a source under 3 months of age recommend a thorough anamnesis, complete physical exam and laboratory workup. Recent international guidelines of management of fever without a source <3 months recommend a more invasive approach under 21 days of life or 1 month (depending on the guideline): full blood count, measurement of C-reactive protein (CRP) and procalcitonin, blood culture, urinalysis, urine culture, stool culture if diarrhea, chest radiography if respiratory symptoms and lumbar puncture, followed by beginning of antibiotherapy and hospitalization. Between 21 days or 1 month and 3 months of age: full blood count, measurement of CRP and procalcitonin, blood culture, urinalysis, urine culture, stool culture if diarrhea and chest radiography if respiratory symptoms should be obtained, reserving lumbar puncture for certain cases. Beginning of antibiotherapy and hospitalization should be weighted.7,8,9
The aim of this study was to evaluate predictive factors of SBI under 3 months of age, in order to identify infants with fever without a source who have a higher risk of having one of these infections and require a close vigilance and early establishment of antibiotherapy and also identify the ones that can follow the “watch and wait” approach without prejudice. | | | | Methods & Materials | Sample and definitions:
We conducted a cohort retrospective study of 2.5 years duration (January 1, 2020 to June 30, 2022) in infants aged 0 to 3 months who visited the emergency department due to fever without a source.
Fever was defined as rectal temperature ≥38oC or axillar temperature ≥37.6oC or tympanic temperature ≥37.8oC.
Fever without source was considered in patients with fever whose etiology remained unknown after a careful history and physical examination, with a duration of less than 7 days.
We contemplated urinary tract infection, occult bacteriemia, meningitis, bacterial gastroenteritis, pneumonia, septic arthritis, osteomyelitis and soft tissue infection as Serious Bacterial Infections. Current and updated definitions of these infections were considered for the diagnosis.
All patients with fever with an identified source, vaccination in the previous 48 hours and septic appearance were excluded.
Data Collection:
Electronic medical records were consulted. We collected demographic (age and gender), clinical, laboratorial and radiological data from all patients.
The clinical data was described using maximum temperature at admission and two classifications of risk of SBI – Rochester criteria and Young Infant Observation Scale (YIOS).
Rochester criteria was analyzed as patients that fulfilled or not all nine low risk conditions:
- Temperature ≥38ºC
- Born at term
- Without perinatal antibiotherapy
- Without underlying pathology
- Without previous hospitalizations
- Without signs of focal infection
- White blood cell (WBC) count between 5000-15 0000/uL
- Neutrophil count <1500/uL
- Urinary sediment analysis ≤10 WBC/hpf
Young Infant Observation Scale was evaluated as low or high risk of SBI: low risk if score <7 and high risk if score ≥7, considering the conditions presented in Table 1.7
Laboratorial data included WBC count, neutrophil count, measurement of CRP and procalcitonin, urinary sediment analysis, Cerebrospinal fluid (CSF) cytochemical analysis and microbiological results of blood, urine, respiratory nasopharyngeal secretions, cerebrospinal fluid and stool.
Radiological data included chest radiography.
Management of patients was divided as medical discharged, observation at a short-stay inpatient unit (at least 12 hours of vigilance) or admitted to a pediatric or neonatology ward.
Antibiotherapy was also assessed.
Statistical analysis:
The Statistical Package for the Social Sciences (SPSS®) version 27 was used for statistical analysis.
Categorical variables were presented as frequencies and percentages and continuous variables as means and standard deviations (SD) or medians and interquartile ranges (IQR) for variables without normal distribution. Normal distribution was checked using Shapiro-Wilk test, visualization of histogram or skewness and kurtosis, as appropriate.
To compare a continuous variable with a categorical variable we used the T-test, after checking for equality of variances with the Levene’s test. When normality distribution was not verified, Mann-Whitney test was performed.
To evaluate associations between two categorical variables we used Chi-square/Fisher test.
Posteriorly, the association of the variables that showed statistical significance in the bivariate analysis was assessed by means of binary logistic regression. A backward stepwise model was performed, with the dependent variable being the presence or absence of SBI.
All reported p values were two-tailed, with a p value of < 0.05 indicating statistical significance.
Table 1. Young Infant Observation Scale.
| Peripheral blood studies |
| Observation |
Normal
(1 point) |
Moderate disease
(3 points) |
Severe disease
(5 points) |
| Appearance |
Smiling/Not irritable |
Irritable, but consolable |
Irritable and inconsolable |
| Respiratory effort |
Without respiratory effort |
Mild to moderate respiratory distress (>60 bpm, retractions or grunting) |
Respiratory distress with inadequate effort (apnea, respiratory failure) |
| Peripheral perfusion |
Pink
Warm extremities |
Marbled skin
Cold extremities |
Pale
Shock |
Legend: bpm – beats per minute
| | | | Results | During the study period a total of 2882 infants under 3 months were observed at our emergency department. 389 (13.5%) because of fever, of which 292 (75.1% of febrile infants) were because of fever without a source.
The sample characterization is described in Table 2.
Table 2. Sample caracterization.
| |
|
All
(n=292) |
Presence of SBI
(n=73) |
Absence of SBI
(n=219) |
| Age (days) - mean + SD |
|
57.8 + 22.3 |
57.0 + 24.5 |
58.0 + 21.6 |
| Age group - n (%) |
= 28 days |
37 (12.7) |
12 (16.4) |
25 (11.4) |
| > 28 days |
255 (87.3) |
61 (83.6) |
194 (88.6) |
| Gender - n (%) |
Feminine |
132 (45.2) |
29 (39.7) |
103 (47.0) |
| Masculine |
160 (54.8) |
44 (60.3) |
116 (53.0) |
| Max. Temp. – median (IQR) |
|
38.4 (0.6) |
38.4 (0.5) |
38.5 (0.7) |
| YIOS – n (%) |
Low risk |
291 (99.7) |
218 (99.5) |
73 (100.0) |
| High risk |
1 (0.3) |
1 (0.5) |
0 (0.0) |
| Rochester – n (%) |
Low risk |
222 (76.0) |
70 (95.9) |
152 (69.4) |
| Not low risk |
41 (14.0) |
3 (4.1) |
38 (17.4) |
| Not classifiable |
29 (10.0) |
0 (0.0) |
29 (13.2) |
| WBC count – median (IQR) |
|
9590.0 (7565) |
14234.2 (6981.0) |
9619.7 (4595.3) |
| Not performed – n (%) |
28 (9.6) |
0 (0.0) |
28 (12.8) |
| Neutrophil count – median (IQR) |
|
3305.0 (4327.5) |
6120.0 (5858.0) |
2680.0 (3070.0) |
| Not performed – n (%) |
28 (9.6) |
0 (0.0) |
28 (12.8) |
| PCR – median (IQR) |
|
7 (28.7) |
43.6 (60.3) |
0 (11.2) |
| Not performed – n (%) |
25 (8.6) |
0 (0.0) |
25 (11.4) |
| PCT – median (IQR) |
|
0.1 (0.3) |
0.5 (3.7) |
0.1 (0.2) |
| Not performed – n (%) |
147 (50.3) |
41 (56.2) |
106 (48.4) |
| Urinary sediment analysis – n (%) |
Altered |
57 (19.5) |
53 (72.6) |
4 (1.8) |
| Not altered |
222 (76) |
20 (27.4) |
202 (92.2) |
| Not performed |
13 (4.5) |
0 (0.0) |
13 (6.0) |
| Urine culture – n (%) |
With significant bacterial growth |
58 (19.9) |
58 (79.5) |
- |
| Without significant bacterial growth |
184 (63.0) |
15 (20.5) |
169 (77.2) |
| Not performed |
50 (17.1) |
0 (0.0) |
50 (22.8) |
| Blood culture – n (%) |
With bacterial growth |
13 (4.5) |
13 (17.8) |
- |
| Without bacterial growth |
202 (69.2) |
56 (76.7) |
146 (66.7) |
| Not performed |
77 (26.4) |
4 (5.5) |
73 (33.3) |
| CSF cytochemical analysis – n (%) |
Altered |
8 (2.7) |
7 (9.6) |
1 (0.5) |
| Not altered |
22 (7.5) |
10 (13.7) |
12 (5.4) |
| Unsuccessful LP |
1 (0.4) |
0 (0.0) |
1 (0.5) |
| Not performed |
261 (89.4) |
56 (76.7) |
205 (93.6) |
| Chest radiograph – n (%) |
Altered |
12 (4.1) |
3 (4.1) |
0 (0.0) |
| Not altered |
3 (1.0) |
4 (5.5) |
8 (3.7) |
| Not performed |
277 (94.9) |
66 (90.4) |
211 (96.3) |
| Stool culture – n (%) |
With significant bacterial isolation |
6 (2.1) |
6 (8.2) |
- |
| Without significant bacterial isolation |
15 (5.1) |
3 (4.1) |
12 (5.5) |
| Not performed |
271 (92.8) |
64 (87.7) |
207 (94.5) |
| Respiratory nasopharyngeal secretions – n (%) |
Not identified |
148 (50.7) |
55 (75.3) |
93 (42.5) |
| SARS-CoV-2 |
75 (25.7) |
4 (5.5) |
71 (32.4) |
| Influenza A |
8 (2.7) |
2 (2.7) |
6 (2.7) |
| RSV |
4 (1.4) |
0 (0.0) |
4 (1.8) |
| Not performed |
57 (19.5) |
12 (16.4) |
45 (20.5) |
A total of 292 patients were included in our study, 160 (54.8%) of whom were male and 132 (45.2%) female. The mean age was 57.8 days, with a standard deviation of 22.3. Patients were divided in two groups: 37 (12.7%) newborns and 255 (87.3%) with age > 28 days.
The median maximum temperature at admission was 38.4oC (IQR 0.6).
A total of 263 patients were classified considering the Rochester criteria, 222 (84.4%) didn’t fulfill low risk criteria.
YIOS (n=292) classified all patients with low risk, except one (99.7% low risk).
73 (25.0%) patients were diagnosed with SBI: 58 (79.5%) urinary tract infections, 13 (17.8%) occult bacteriemias, 8 (11.0%) meningitis, 6 (8.2%) bacterial gastroenteritis and 3 (4.1%) pneumonias.
Urine analysis was performed in 279 patients and was altered (>10 WBC/hpf or nitrituria) in 57 (20.4%). An urine culture was obtained in 242 patients, with significant bacterial growth in 58 (24.0%). The most frequent pathogen was E. coli (n=50, 86.2%).
Blood culture was obtained in 215 patients, 13 (6.0%) had bacterial growth and 202 (94.0%) were sterile. Considering the positive cultures, were identified 3 (23.1%) Staphylococcus aureus, Staphylococcus epidermidis and E. coli; 1 (7.7%) Neisseria meningitidis, Streptococcus pneumoniae, Staphylococcus saprophyticus and Staphylococcus hominis.
Lumbar puncture was performed in 31 patients, CSF cytochemical analysis was normal in 23 (74.2%) and 8 (25.8%) had meningitis. Of the later patients, in one patient (12.5%) was identified Neisseria meningitidis, E. coli, E. coli with extended spectrum beta-lactamase (ESBL+) and Streptococcus pneumoniae and no agent was identified in 4 (50.0%).
Stool samples were collected in 21 patients, 6 (28.6%) had isolation of Campylobacter jejuni and 15 (71.4%) had no bacterial identification.
Cultural exams are described in Table 3.
Table 3. Cultural exams.
| Serious Bacterial Infection – n (%) |
73 (25.0) |
| Urinary tract infection – n (%) |
58 (19.8)
(79.5% of SBI) |
| E.coli |
50 (86.2) |
| Proteus mirabilis |
3 (5.2) |
| Klebsiela oxytoca |
3 (5.2) |
| Klebsiela pneumoniae ESBL + |
1 (1.7) |
| E. coli ESBL + |
1 (1.7) |
| Occult bacteriemia – n (%) |
13 (4.5)
(17.8% of SBI) |
| Staphylococcus aureus |
3 (23.1) |
| Staphylococcus epidermidis |
3 (23.1) |
| E. coli |
3 (23.1) |
| Neisseria meningitidis |
1 (7.7) |
| Streptococcus pneumoniae |
1 (7.7) |
| Staphylococcus saprophyticus |
1 (7.7) |
| Staphylococcus hominis |
1 (7.7) |
| Meningitis – n (%) |
8 (2.7)
(11.0 of SBI) |
| Without identification of agent |
4 (50.0) |
| Neisseria meningitidis |
1 (12.5) |
| E. coli |
1 (12.5) |
| E. coli ESBL+ |
1 (12.5) |
| Streptococcus pneumoniae |
1 (12.5) |
| Pneumoniae – n (%) |
3 (1.0)
(4.1 of SBI)
|
| Bacterial gastroenteritis – n (%) |
6 (2.1)
(8.2% of SBI) |
| Campylobacter jejuni |
6 (100) |
Chest radiography was taken in 15 patients and 3 (20.0%) were diagnosed with pneumonia.
Respiratory nasopharyngeal secretions were collected in 235 patients for the detection of respiratory viruses (panel of Influenza A virus, Influenza B virus, Respiratory Syncytial Virus and SARS-CoV-2 virus). In 75 (31.9%) SARS-CoV-2 was identified, in 8 (3.4%) Influenza A virus, in 4 (1.7%) Respiratory Syncytial Virus (RSV) and in 148 (63.0%) no virus was identified. Considering the patients with SBI, 6 (8.2%) had co-infection with one of the mentioned virus.
Admission to a Pediatric or Neonatology ward was needed in 94 (32.2%) cases, observation at a short-stay inpatient unit in 86 (29.5%) and 112 (38.4%) were discharged.
85 (29.1%) patients of all infants observed were treated with antibiotics.
We found a statistical significant difference of classification according to the Rochester criteria, WBC count, neutrophil count, measurement of C-Reactive Protein and procalcitonin between the groups. The classification according to the Rochester criteria demonstrated that there was a significant statistical association between not fulfilling the low risk criteria and having SBI, p=0.001. The median value of WBC count was significantly superior in the presence of SBI (Mdn=12720.0; IQR=9305.0) when compared to the group without SBI (Mdn=8670.0; IQR=3070.0), p <0.001. The median value of neutrophil count was significantly superior in the presence of SBI (Mdn=6120.0; IQR=5858.0) when compared to the group without SBI (Mdn=2680.0; IQR=3070.0), p <0.001. The median value of CRP was significantly superior in the presence of SBI (Mdn= 43.6; IQR=60.3) when compared to the group without SBI (Mdn 0.0; IQR 11.2), p <0.001. The median value of procalcitonin was significantly superior in the presence of SBI (Mdn= 0.5; IQR=3.7) when compared to the group without SBI (Mdn 0.1; IQR 0.2), p <0.001.
Age, gender and maximum temperature didn’t present statistical significant results. It was not possible to calculate an association between YIOS classification and the presence or absence of SBI.
In a multivariate analysis for evaluating predictive factors to have SBI, CRP values (p <0.001; OR 1.045) showed statistical significance (Hosmer-Lemeshow test: x2= 12.043; p=0.149). This result means that an elevation of one unit (mg/L) in CRP values, elevates the risk of SBI in 1.045.
Bivariate analysis is described in Table 4 and multivariate analysis is described in Table 5.
Table 4. Bivariate analysis.
| |
Presence of SBI |
Absence of SBI |
p-value |
| Gender – n (%) |
Female |
29(39.7) |
103(47.0) |
0.277 |
| Male |
44(60.3) |
116(53.0) |
| Age (days) – Median (IQR) |
56(42) |
61(34.0) |
0.881 |
| Age group – n (%) |
= 28 days |
12(16.4) |
25(11.4) |
0.264 |
| > 28 days |
61(83.6) |
194(88.6) |
| Maximum temperature (ºC) – Median (IQR) |
38.5(0.7) |
38.4(0.5) |
0.053 |
| YIOS – n (%) |
Low risk |
73(100.0) |
218(99.5) |
- |
| High risk |
0(0.0) |
1(0.5) |
| Rochester criteria – n (%) |
Low risk |
3(4.1) |
38(20.0) |
0.001 |
| Not low risk |
70(95.9) |
152(80.0) |
| WBC count – Median (IQR) |
12720(9305) |
8670(5910) |
<0.001 |
| Neutrophil count – Median (IQR) |
6120(5858) |
2680(3070) |
<0.001 |
| CRP value – Median (IQR) |
43.6(60.3) |
0.0(11.2) |
<0.001 |
| Procalcitonin value – Median (IQR) |
0.5(3.7) |
0.1(0.2) |
<0.001 |
Table 5. Multivariate analysis - final model.
| Nagelkerke R2 |
X2 |
df |
p-value |
| 0.540 |
12.043 |
8 |
0.149 |
| |
p-value |
OR |
Confidence interval 95% |
| WBC count |
0.082 |
1.000 |
1.000-1.000 |
| CRP value |
<0.01 |
1.045 |
1.020-1.069 |
| Procalcitonin value |
0.05 |
1.336 |
1.000-1.785 |
| | | | Discussion | In our study we found a prevalence of observations due to fever under 3 months of age in the emergency department of 13.5%, which is coherent with the literature.1 Furthermore, in this group in particular, there is a preponderance of fever without a source as a first diagnostic impression for febrile infants (75.1% of febrile infants), which is also consistent with descriptions from other populations.2
In our study, 25.0% of infants under 3 months of age with fever without a source turned out to have a SBI, which is a value superior to that described in some published studies5, although some recent investigations have reported similar values.1 Urinary Tract Infection was the most frequent SBI, with E. coli being the most common causative pathogen, which is similar to that described in the great majority of articles concerning this theme.1,2,3,4,5,6 The other Serious Bacterial Infections diagnosed - occult bacteriemias, meningitis, pneumonias and bacterial gastroenteritis - are described as the other infections more frequently found in comparable populations.6
In 6 cases there was a SBI and a viral co-infection, that means the diagnosis of a viral infection didn’t reject the hypothesis of a SBI.
Risk classifications such as Rochester criteria and YIOS used in this study have been widely used, in order to try to establish a stratification of risk that allows clinicians to optimize their approach. The main objective is to balance a less invasive approach with the intention and necessity to detect all cases of SBI and initiate proper treatment in due course.6 Rochester criteria showed that there was an association between not fulfilling the low risk criteria and having SBI in the bivariate analysis, but not after including it in the final model. In order to increase the accuracy of this classification, to assess if it would improve our model, a study with more patients categorized with this criteria should be performed. In our study it was not possible to evaluate the existence of an association between YIOS and the presence or absence of SBI, which limits the ability to rely on this scale and requires further studies of application of YIOS.
The laboratory evaluation of WBC count, neutrophil count, CRP and procalcitonin showed a statistically significant difference between groups in the bivariate analysis, which is consistent with the literature, that indicates that inflammatory markers give an important diagnostic information in this patients.2,4,6,10 However, in the multivariate analysis, WBC and neutrophil count didn’t present a statistically significant difference between the group with SBI and the group without SBI, which is consistent with the fact that their diagnostic value is limited, as other studies have proved.10 On the other hand, CRP maintains its statistical value and procalcitonin has a p value that is near statistical significance (p=0.05). CRP and procalcitonin measurements have a much higher diagnostic value than WBC and neutrophil count.2,4,6,10 Some investigations suggest that procalcitonin is a better predictor of SBI4 and others advocate that none of these variables is better than the other10, but in our study CRP values have a better result in predicting SBI. Procalcitonin diagnostic value might be underestimated in our population due to lack of data in an important proportion of patients.
An association between high fever (particularly >40oC) and an increased risk of SBI has been reported in some studies11, but in this one there was no relevant association between maximum temperature and the presence or absence of SBI. This is a data provided by the infant’s caregiver, which can be a possible cause to this discrepancy, as a superior temperature might not be objectified and antipyretic might be administered in an early phase. A superior temperature might also be noticed in a posterior time to the first observation at the emergency room, which can underestimate the importance of this parameter.
Although SBI is more prevalent in the group with <28 days than the older patients1,2,3,4,5,6, in our study there was no statistically significant difference considering the age of the patients. The asymmetric sample size of the two groups, with a much higher number of patients with more than 28 days (87.3%), can influence this result.
Gender is usually not a criteria to consider as a risk factor according to recent guidelines and studies7,8,9, which is consistent with the results found in this investigation.
| | | | Conclusion | To conclude, the evaluation of these parameters has a great potential to integrate a model to predict SBI in infants under 3 months of age with fever without a source. This prediction intends to promptly treat the ones with high risk of SBI and to not expose the ones with low risk to more invasive investigations and exposure to antibiotics. According to this investigation, CRP values should be assessed in each patient’s context and should help in the treatment and follow-up decision.
Limitations:
Our study has some limitations that should be evaluated, in order to enhance them and possibly replicate this study with more accurate data. Firstly, the approach to the febrile infants was not guided by a protocol, which leads to a variable methodology according to each clinician that consults the patient and a lack of some data needed in this study. Secondly, it was a retrospective analysis, with all limitations associated with this type of report. Finally, this study was conducted using medical records from a single hospital, this population might not be representative of others.
| | | | Compliance with Ethical Standards | | Funding None | | | | Conflict of Interest None | | |
- Pinto EV, Sánchez-Bayle M. Construction of a diagnostic prediction modelo f severe bacterial infection in febrile infants under 3 months old. An Pediatr (Barc). 2017; 87(6): 330-336. [CrossRef]
- Cisneros BG, Benítez PJC. Identificación de los factores de bajo riesgo em el lactante pequeño com fiebre sin foco evidente. ACTA MÉDICA GRUPO ÁNGELES. Volumen 16, No.3, julio-septiembre 2018.
- Aldrige P, Arjun R, Sethumadavan et al. Fever under 3 months and the full septic screen: Time to think again? A retrospective cohort study at a tertiary-level paediatric hospital. Journal of Paediatrics and Child Health (2017). [CrossRef] [PubMed]
- Lee TG, Yu ST, So CH. Predictive value of C-reactive protein for the diagnosis of meningitis in febrile infants under 3 months of age in the emergency department. Yeungnam Univ J Med 2020;(37)2:106-111. [CrossRef] [PubMed] [PMC free article]
- Greenhow TL, Hung Y, Pantell RH. Management and Outcomes of Previously Healthy, Full-Term, Febrile Infants Ages 7 to 90 Days. Pediatrics. 2016;138(6):e20160270. [CrossRef] [PubMed]
- Hubert-Dibon G, Danjou L, Feildel-Fournial C et al. Procalcitonin anc C-reactive protein may help to detect invasive bacterial infections in children who have fever without a source. Acta Paediatr. 2018 Jul;107(7):1262-1269. [CrossRef] [PubMed]
- Pantell RH, Roberts KB, Adams WG et al. Evaluation and Management of Well-Appearing Febrile Infants 8 to 60 Days Old. Pediatrics. 2021;148(2):e20211052228. [CrossRef] [PubMed]
- Férnandez-Cuesta Valcarce MA. Pathways Guide in Primary Care Pediatrics. Fever without a source in children younger than 36 months. AEPap 2017 (on line). Available in: algoritmos.aepap.org.
- National Institute for Health and Care Excellence. Fever in under 5s: assessment and initial management. NICE guideline. Published: 7 November 2019. Available in: www.nice.org.uk/guidance/ng143.
- Van den Bruel A, Thompson MJ, Haj-Hassan T et al. Diagnostic value of laboratory tests in identifying serious infections in febrile children: systematic review. BMJ 2011;342:d3082. [CrossRef] [PubMed]
- Rosenfel-Yehoshua N, Barkan S, Abu-Kishk I. Hyperpyrexia and high fever as a predictor for serious bacterial infection (SBI) in children - a systematic review.
DOI: https://doi.org/10.7199/ped.oncall.2024.45
|
| Cite this article as: | | Ribeiro R C P M, Queirós J R P d C, Ferreira A I R P, Martins I I A, Barroso F D M. Fever without a source under 3 months of age: any predictive factors of Serious Bacterial Infection?. Pediatr Oncall J. 2023 Aug 28: 144-150. doi: 10.7199/ped.oncall.2024.45 |
|