Yusuf S, Goel AK, Khobragade AW, Das P, Manoj GSV, Saha AK, Reddy SP, Keshri S, Shah S. Determinants of infection for antibiotic initiation at pediatric emergency admission: A prospective observational study. World J Clin Pediatr 2025; 14(4): 105820 [DOI: 10.5409/wjcp.v14.i4.105820]
Corresponding Author of This Article
Anil K Goel, Professor, Department of Pediatrics, All India Institute of Medical Sciences, GE Road Raipur, Raipur 492099, Chhattīsgarh, India. akgoel@aiimsraipur.edu.in
Research Domain of This Article
Infectious Diseases
Article-Type of This Article
Observational Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Dec 9, 2025 (publication date) through Nov 3, 2025
Times Cited of This Article
Times Cited (0)
Journal Information of This Article
Publication Name
World Journal of Clinical Pediatrics
ISSN
2219-2808
Publisher of This Article
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Yusuf S, Goel AK, Khobragade AW, Das P, Manoj GSV, Saha AK, Reddy SP, Keshri S, Shah S. Determinants of infection for antibiotic initiation at pediatric emergency admission: A prospective observational study. World J Clin Pediatr 2025; 14(4): 105820 [DOI: 10.5409/wjcp.v14.i4.105820]
Samreen Yusuf, Anil K Goel, Swasti Keshri, Division of Pediatrics Emergency Medicine, Department of Pediatrics, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Ashish Wasudeo Khobragade, Department of Community Medicine, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Padma Das, Department of Microbiology, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Gudipudi Sai Vamsi Manoj, Sai Pratap Reddy, Department of Pediatrics, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Anish Kumar Saha, Department of Nephrology, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Seema Shah, Department of Biochemistry, All India Institute of Medical Sciences, Raipur 492099, Chhattīsgarh, India
Author contributions: Yusuf S and Goel AK contributed to concept design, literature search, manuscript preparation, and final approval; Keshri S, Das P, Shah S, and Khobragade AW contributed to data capture and manuscript editing; Manoj GSV, Reddy SP, and Saha AK contributed to figure and table design and data analysis; All authors read and approved the final manuscript.
Institutional review board statement: The Institute Ethics Committee, All India Institute of Medical Sciences, Raipur (Chhattisgarh) reviewed and discussed above referenced PG thesis proposal in the meeting held on August 14, 2021.
Informed consent statement: All provided informed consent.
Conflict-of-interest statement: All authors report no relevant conflicts of interest for this article.
STROBE statement: The authors have read the STROBE Statement-checklist of items, and the manuscript was prepared and revised according to the STROBE Statement- checklist of items.
Data sharing statement: Data will not be shared with any person or institute or organization and privacy of the data and document will be kept confidential.
Open Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Anil K Goel, Professor, Department of Pediatrics, All India Institute of Medical Sciences, GE Road Raipur, Raipur 492099, Chhattīsgarh, India. akgoel@aiimsraipur.edu.in
Received: February 8, 2025 Revised: March 26, 2025 Accepted: June 9, 2025 Published online: December 9, 2025 Processing time: 266 Days and 6.6 Hours
Abstract
BACKGROUND
Due to non-specific and overlapping features, particularly in infants, and a lack of awareness of important signs and symptoms, early diagnosis of sepsis is difficult.
AIM
To identify the determinants for bacterial infection based on clinical evidence and point-of-care (POC) investigations in children at admission to a pediatric emergency medicine (PEM) unit.
METHODS
This prospective observational study was conducted in a PEM unit of a tertiary care hospital in Central India. A scoring system based on these determinants for initiation of antibiotics in the golden hour (i.e. the first 60 minutes) was developed, termed as the Children’s Antibiotic Requirement Evaluation Score (CARES) Index.
RESULTS
Out of a total of 1419 children presenting to the PEM, 802 children were enrolled. The best predictors of infection were found to be abnormal color, fever, features of fluid overload, altered sensorium, cellulitis, use of antibiotics, and C-reactive protein. These parameters were amalgamated to form the CARES Index, which showed moderate accuracy for determinants of infection for initiation of antibiotics in a pediatric emergency. Among the POC investigations, the sensitivity of the erythrocyte sedimentation rate (80.00%) was maximum, while specificity was highest for procalcitonin (75.31%).
CONCLUSION
We highlighted the association of clinical parameters, symptoms, and POC investigations in pediatric bacterial infections. Some parameters emerged as strong predictors of sepsis, and no single factor was sufficient for a definitive diagnosis.
Core Tip: Diagnosing pediatric sepsis is challenging in a pediatric emergency medicine setting due to nonspecific signs and symptoms. This can lead to both overutilization and underutilization of antibiotics. The Children’s Antibiotic Requirement Evaluation Score Index is a composite of seven easy-to-determine clinical parameters, i.e. abnormal color, fever, fluid overload, altered sensorium, cellulitis, prior antibiotic use, and C-reactive protein levels, which can guide the emergency physician to use antibiotics rationally during the golden hour of care (i.e. the first 60 minutes). This should help to develop a standard and objective approach for the judicious use of antibiotics in pediatric emergency medicine setting.
Citation: Yusuf S, Goel AK, Khobragade AW, Das P, Manoj GSV, Saha AK, Reddy SP, Keshri S, Shah S. Determinants of infection for antibiotic initiation at pediatric emergency admission: A prospective observational study. World J Clin Pediatr 2025; 14(4): 105820
Infectious diseases represent a major cause of morbidity and mortality in India and account for 63% of all hospital admissions in pediatric emergency[1,2]. Serious bacterial infections such as bacteremia, pneumonia, meningitis, and urinary tract infections, have an incidence rate of 7%-11% in young infants[3]. Infectious disorders that present with fever are responsible for almost 50% of the deaths of children under the age of 5 years worldwide (5.4 million)[4]. The common presentations of bacterial infection are breathlessness (80%), irritability (75%), and poor feeding (75%)[5]. India has the highest prevalence of bacterial infections worldwide[6], justifying the role of antibiotics. On the other hand, the irrational use of antibiotics is rampant, and the prevalence of resistance has increased[7]. The World Health Organization has declared antimicrobial resistance as one of the top 10 global public health threats facing humanity. It can be argued that most children admitted to the pediatric emergency medicine (PEM) unit, irrespective of the clinical presentation, receive at least one shot of antibiotic at admission.
Current scoring systems for assessing sepsis in pediatric patients, such as pediatric sequential organ failure assessment (pSOFA) and pediatric risk mortality (PRISM)[8], primarily focus on survival prediction and do not effectively recognize sepsis. pSOFA is designed to track organ dysfunction in critically ill children over time, primarily in the intensive care unit rather than for early infection detection in emergency settings. If pSOFA is used as a threshold for antibiotics, children with early sepsis without organ dysfunction may be missed, leading to delayed treatment and increased morbidity. PRISM is a mortality prediction tool that evaluates overall illness severity but does not specifically help in diagnosing infection or determining the need for antibiotics. The focus of PRISM on overall severity rather than infection markers may result in unnecessary antibiotic use in non-infectious conditions, contributing to antimicrobial resistance.
There is no established scoring model specifically designed for antibiotic initiation in children. This study aimed to develop a simple, clinically applicable scoring system to identify potential infection parameters, facilitating timely antibiotic administration while minimizing irrational use in PEM.
MATERIALS AND METHODS
This prospective observational study was conducted in a PEM unit of a tertiary care hospital in Central India after obtaining approval from the Institutional Ethical Committee. The objective of the study was to identify the determinants for bacterial infection based on clinical evidence and point-of-care (POC) investigations in children at admission to the PEM unit. A scoring system based on these determinants for initiation of antibiotics in the so-called ‘golden hour’ (i.e. the first 60 minutes) was developed to rationalize its use and termed the Children’s Antibiotic Requirement Evaluation Score.
All children aged 3 months to 15 years admitted to the PEM unit between December 2021 and June 2023 were included in the study after providing written consent. Chronic bacterial illnesses [e.g., tuberculosis, and those admitted on a daily-care basis (e.g., thalassemia, leukemia etc.)] were excluded from the study. All pertinent clinical features (such as age, nutritional status, fever, cough, rash, loose stools, respiratory distress, shock, etc.), and all POC investigations [e.g., C-reactive protein (CRP), complete blood count/differential leukocyte count, lactate] were noted in a structured format.
Since the study was conducted during the pandemic and due to a lack of similar studies to predict the presence or absence of bacterial infection during the golden hour, a convenient time-bound, purposive, sampling method was adopted to recruit the subjects. Hence all children admitted to the PEM during the study period who fulfilled the inclusion criteria were included in the study and totaled 802 participants. We used the formula, n = 100 + 50 × i, where i is the number of independent variables a sample size of 800 would impart 80% power to the study [9]. Hence our sample size was sufficient to provide adequate power.
All enrolled patients were triaged (including primary and secondary assessment), and emergency management was initiated and evaluated for admission as per the unit protocol. The triage, physical examination, and POC investigations of the consenting patients were documented. Children with infection were classified into four categories: (1) No infection; (2) Suspected infection; (3) Probable infection; and (4) Confirmed infection based on the definition for subgroup analysis. All the patients were monitored and noted for clinical evidence of infection in either the PEM/ward/intensive care unit based on parameters within the first hour of admission and new signs/symptoms noticed in the first 24 h of admission. The presence or absence of infection was confirmed based on: (1) Clinical evidence; and (2) Microbiological/radiological data. All children were followed up to discharge or death (outcome) and subdivided into children with infection (confirmed, probable, and suspected infection) and children without infection. Patients of each group were further divided into two categories: (1) Those who received antibiotics in the golden hour; and (2) Those who did not receive antibiotics in the golden hour.
Data was analyzed using SPSS software Version 21.0. (IBM Corp., Armonk, NY, United States: IBM Corp.) and R software version 4.3.1 (The R Foundation for Statistical Computing, Vienna, Austria). Data was presented using frequency and percentage in tabular and graphical forms. The normality of data was assessed using the Shapiro-Wilk test. mean ± SD and median (interquartile range ± SD) were used to measure the central tendency for normal and skewed data. The χ2 test was performed to assess the significance of the association between different clinical features and types of infection. Simple logistic regression was used to determine the potential bacterial infection predictors. Those with a P value < 0.05 were selected for multivariable logistic regression. To adjust for confounders a multivariate binomial logistic regression model was used. Adjusted odds ratio [95% confidence interval (CI)] was determined.
The cutoff point was calculated for each POC investigation [total leukocyte count (TLC), absolute neutrophil count, CRP, procalcitonin, serum albumin, lactate)] and the area under the curve (AUC) was graphically shown using the receiver operating characteristic (ROC) curve with 95%CI. Sensitivity, specificity, positive predictive value, negative predictive value, and positive and negative likelihood ratio were determined to assess the diagnostic accuracy of each investigation. Variables from the multivariate analysis were used in the final model to find the best predictors of infection. The final model was fitted by the backward elimination method. The model with the lowest Akaike Information Criteria and Bayesian Information Criteria value was selected as the final fitted model. The Hosmer-emeshow test was used as a goodness of fit test. The goodness of fit was also assessed by pseudo r2 (Cox and Snell and Nagelkerke Pseudo r2). P value of < 0.05 was considered statistically significant.
Operational definition
Confirmed infection: Culture-proven infection[10] [blood, urine, cerebrospinal fluid (CSF), pleural fluid, ascitic fluid] excluding fungal infection or serological evidence demonstrating the presence of antibody or antigen against organisms like scrub typhus, leptospirosis, Rickettsia, or microscopic detection of the organism in a sterile clinical specimen (CSF for Haemophilus influenzae, pneumococcus, meningococcus, etc.).
Probable infection: Clinical evidence with either: (1) Exudative pleural effusion: As per Light’s criteria[11]; (2) Spontaneous bacterial peritonitis (> 250 polymorphonuclear leukocytes in ascitic fluid[12] in nephrotic syndrome: > 100 neutrophils[13]); (3) Significant pyuria (pus cells > 5 cells/high-powered field[14]), leukocyte esterase, and nitrite positivity; (4) CSF suggestive of bacterial meningitis (> 10 cells with neutrophilic leukocytosis)[15]; and (5) Chest X-ray showing lobar consolidation.
Suspected infection: Clinical evidence alone [10]: (1) Fever. Fever was defined as a rectal temperature ≥ 38 °C (100.4 °F)[16]. The axillary temperature is 0.5 °C lower than rectal temperature [17]; (2) Hypothermia. Hypothermia was defined as rectal temperature < 36 °C[18]; (3) Abnormal mean arterial pressure (MAP). MAP < 5th percentile for age[19]; (4) Tachycardia. Heart rate (HR) above the upper limit normal for age[19]; (5) Bradycardia: HR below the lower limit of normal for age[19]; (6) Tachypnea. Respiratory rate (RR) above the upper limit of normal for age[19]; (7) Bradypnea. RR below the lower limit normal for age[19]; (8) Abnormal saturation. Defined as spO2 < 94%[20]; (9) Abnormal capillary time is defined as > 3 s or flushed (< 1 s)[21]; (10) Random blood sugar abnormality. height Z-score < -3 SD, mid-upper arm circumference < 11.5 cm in children aged between 6-59 months, and bilateral pitting edema [27]; and (11) Immunodeficiency.
RESULTS
This was a prospective observational study carried out over 18 months and included all children aged 3 months-15 years. Signs, symptoms, and POC investigations associated with the evidence of infection were evaluated. After excluding patients with chronic bacterial illnesses like tuberculosis and daily-care admissions, a total of 802 participants in the age group of 3 months to 14 years were included (Figure 1).
The baseline characteristics of the study population have been detailed in Table 1. The mean age of study participants was 4.81 years (SD ± 4.27). The respiratory tract was the most common system involved in our study population (49.13%). Out of 802 cases, there were 131 (16.30%) cases of confirmed infection, 114 (14.20%) had probable infection, 209 (26.10%) had suspected infection, and 348 (43.40%) had no infection.
The symptoms, general physical examination, comorbidities, anthropometry, and other features of subjects in each group have been summarized in Table 1 Fever was more frequently observed in confirmed (62.50%) and probable infection (66.67%). Fever was found to have a significant association with infection (P = 0.000) as compared with HR, RR, MAP, capillary refill time, blood sugar, and saturation. The mean temperature in the confirmed infection group was found to be 37.50 ± 0.92 °C. All the parameters of the PAT [appearance (P = 0.009), breathing (P = 0.007), and color (P < 0.001)] were found to be statistically significant along with the initial physiological categorization (P < 0.001). (Table 1). Table 2 shows mean/median of quantitative variables in the study population. Out of 131 children with bacterial infections that were confirmed by culture, 15 (11.45%) had low procalcitonin levels (< 0.5). Among these 15 children, 13 (86.66%) of them had an abnormal leukocyte count and 11 (73.33%) had an underlying chronic illness. Children with underlying chronic illness (n = 11, 73.33%) were more likely to have a bacterial infection.
Table 2 Mean/median of quantitative variables of the study population.
The diagnostic characteristics of all POC investigations associated with bacterial infection have been listed in Table 3. The highest positive predictive value was for albumin (85.54%) followed by lactate (76.38%) and erythrocyte sedimentation rate (ESR) (74.07%). The highest negative predictive value was found to be for CRP (80.51%). The ROC curve of albumin, procalcitonin, absolute neutrophil count, ESR, CRP, lactate, and TLC along with AUC have been depicted in Figures 2 and 3. Based on these AUCs, the sensitivity, specificity, and accuracy at CRP cutoffs of ≥ 60mg/dL was 68.00% each. ESR of > 42 mm/hour had the highest sensitivity (80.00%) followed by albumin with a cutoff of ≤ 3.7 g/dL (sensitivity: 78.00%). The most specific POC investigation was found to be procalcitonin with a cutoff of 0.72 ng/mL (75.00%). Logistic regression of signs and symptoms have been summarized in Table 4.
Figure 2 Receiver operating characteristic curve for serum albumin, procalcitonin, C-reactive protein and absolute neutrophil count for differentiating different outcome groups.
ANC: Absolute neutrophil count; AUC: Area under the curve; CRP: C-reactive protein.
Figure 3 Receiver operating characteristic curve for total leukocyte count, lactate, and erythrocyte sedimentation rate for differentiating different outcome groups.
AUC: Area under the curve; ESR: Erythrocyte sedimentation rate; TLC: Total leukocyte count.
Table 3 Diagnostic characteristics of different variables associated with a bacterial infection.
The multivariate model was further simplified using the backward elimination method. The best predictors of childhood bacterial infection included symptoms like features of hypoperfusion (abnormal color), fever, features of capillary leak, altered sensorium, other associated factors like cellulitis and prior antibiotic use, and investigations like CRP. Among these, the most significant association was found with features of capillary leak, followed by CRP. CRP was the only POC investigation that fulfilled the model (Table 5).
Table 5 Predictors of childhood bacterial infection: Final model fitted by backward elimination method.
The final model (Table 6) was selected by considering the lowest value of the Akaike Information Criteria and Bayesian Information Criteria. The Hosmer-Lemeshow test showed there was no evidence of lack of fit of this model. Cox and Snell and Nagelkerke pseudo r2 value also gave the same evidence.
Table 6 Scoring system for initiation of antibiotics in golden hour in pediatric emergency medicine.
Based on the best predictors of bacterial infection derived from the study, we devised a scoring system (Children’s Antibiotic Requirement Evaluation Score Index) that would help clinicians in the emergency department promptly identify bacterial infections with reasonable efficacy (Table 6). Points were assigned to each of the best predictors with a significant association per odds ratio. Moreover, two independent risk factors namely severe malnutrition and immunocompromised state were also included in the score owing to their well-known association with bacterial infection[5]. These parameters were not part of the best predictors probably because of a lower sample size in our study. The cutoff score was obtained from the ROC curve.
The AUC for this scoring system was 0.71 (95%CI: 0.66-0.75) for discriminating definite bacterial infection from no infection. The AUC in our study was of moderate accuracy. This study suggested the following cutoff values of best possible prediction of having bacterial infection in children attending PEM at admission: (1) A cutoff of > 4 had a sensitivity of 83.20% and specificity of 48.70%; (2) A cutoff score > 6 had a sensitivity of 58.80% and specificity of 69.00%; and (3) A cutoff score > 7 had a sensitivity of 48.10% and specificity of 77.90%.
DISCUSSION
Despite extensive research, there still exists multiple caveats regarding the identification of bacterial infection, leading to delays in the diagnosis as well as initiation of appropriate antimicrobial therapy contributing to poor outcomes. To the best of our knowledge, there are no similar studies to justify the use of antibiotics during the golden hour. During the study period a total of 802 patients who fulfilled inclusion criteria were enrolled out of which 131 (16.30%), 114 (14.20%), and 209 (26.10%) were confirmed, probable, and suspected infections, respectively, while 348 (43.39%) had no infection[28]. Fever, cough, features of capillary leak, dyspnea, lower urinary tract symptoms (P = 0.001), seizures (P = 0.045), altered sensorium (P = 0.018), oliguria, chest pain, and ear pain/discharge were some symptoms that were found to have a significant association with infection, which is in accordance with the study by Kiemde et al[29]. Covino et al[30] showed an association between infection and a variety of clinical symptoms, such as cough, vomiting, fever, dyspnea, diarrhea, etc.
Among the vital signs and general physical examination, only fever and pallor had significant association with infection (P = 0.000), which is comparable to findings from research by Kiemde et al[29] Our study demonstrated a significant association of decreased air entry with infection (P = 0.000), but Covino et al[30] found crepitations as a significant association. The difference in our findings could be explained based on the type of lung involvement or disease enrolled in both the studies. Organomegaly (P = 0.000) and abdominal tenderness (P = 0.005) had a significant association with infection, which is similar to the study by Covino et al[30]. Pupillary abnormalities (P = 0.0041), abnormal Glasgow coma scale (P = 0.001), and meningeal signs (P = 0.000) were all significantly associated with CNS infection[31].
In PAT assessment initial physiological classification (P = 0.001) was found to have a significant association with infection. However, multivariate regression analysis does not reveal any statistical association. Contradictory to our study, an Italian study by Covino et al[30] failed to demonstrate a correlation between infection and initial physiological classification. The probable reason may be delayed referrals to tertiary care centers apart from a lack of awareness among parents. In our study immunodeficiency (P = 0.000) and chronic infection (P = 0.003) were significantly associated with acute bacterial infection. Various studies have shown an association between a spectrum of infections and immunodeficiency and chronic illness[32-34]. In our study a history of prior antibiotic use was found to have a statistically significant association with infection (P = 0.000). A reduction in the culture positivity rate in our study can be explained due to the previous history of antibiotic use. The findings observed in our study are comparable to cross-sectional research by Nyeko et al[34], which also had a 39.5% (n = 83) use of antibiotics prior to the hospital visit.
The mean procalcitonin level in children with no infection was 0.20 ng/mL (0.06-0.78), while in the confirmed infection subgroup it was 2.00 ng/mL (0.33-7.74), which was comparable with the study conducted by Maniaci et al[35]. When comparing individuals who had no infection with those with confirmed infection, the mean CRP was 113.0 mg/dL (0.6-624.0) and 219.0 mg/dL (69.9-538.4), respectively[29,30].
In studies by Maniaci et al[35], Nijman et al[36], and Andreola et al[37], the AUC for procalcitonin ranged between 0.75 and 0.82 and for CRP between 0.77 and 0.85, although the cutoff values determined were different from our study. The higher cutoff in our study population was likely a reflection of the delayed referral and severe cases. Moreover, a delay in presentation leading to sample collection at a later point in the natural history of the disease may also lead to a higher cutoff.
As per the meta-analysis in NICE guidelines, CRP was found to have a positive likelihood ratio of 3.1 and a negative likelihood ratio of 0.27 for bacterial infection in children which is comparable to our study[38]. Procalcitonin had a positive and negative likelihood ratio of 2.85 and 0.39, respectively. In our study, the positive and negative likelihood ratio for TLC was 1.55 and 0.73 respectively, which was similar to the results in a study by De et al[39] (LR+: 1.93, LR-: 0.70). Hypoalbuminemia and high serum lactate as a determinant of infection as previously demonstrated in different studies was re-established in our study. The AUC for ESR in our study was different from studies conducted by Cuello García et al[40] and Park et al[41], which may be explained by the referral of more difficult-to-treat infections to our center.
Combined model for predictors of bacterial infection
By using a fitted model by backward elimination method in our study we established a set of variables as best predictors of bacterial infection. These variables included a change in color (hypoperfusion) on initial triage assessment, fever, features of capillary leak, altered sensorium, cellulitis, prior antibiotic usage, and CRP. There is currently no scoring system that can be used to predict bacterial infection in children. The newly developed scoring system offers a more practical, infection-focused, and early-detection approach compared with existing scoring models like pSOFA and PRISM. The existing scoring systems primarily assess organ dysfunction and mortality risk, which may delay antibiotic initiation in children with early-stage infection. They also include parameters that may not be directly related to bacterial infection, such as arterial blood gases, coagulation markers, and pupillary response. In addition, they require extensive lab investigations, making them impractical for real-time decision-making in emergency situations. Therefore, the model developed by us can serve as a guide to the start of antibiotic treatment at admission (Table 6). The AUC for this scoring system was 0.71 (95%CI: 0.66-0.75) The AUC is of moderate accuracy for discriminating definite bacterial infection from no infection. A cutoff score > 7 has a specificity of 77.90% and a sensitivity of 48.10%. The suggested cutoff values may be a stepping stone for developing a model with a study having adequate sample size and multicentric in design to generalize the study population.
The limitation of our study was a significant number of patients had received antibiotics prior to admission, which could affect the culture positivity rates. A large sample size with a heterogenous cohort involving different levels of health care facility and a longer duration of enrollment would have resulted in a conclusion.
CONCLUSION
This study highlighted the significant association of various clinical parameters, symptoms, and POC investigations with bacterial infection in children. While individual parameters such as fever, abnormal color, altered sensorium, and CRP emerged as strong predictors, no single factor alone was sufficient for a definitive diagnosis. Among laboratory markers, ESR demonstrated the highest sensitivity, procalcitonin had the highest specificity, and CRP had the best negative predictive value, reinforcing their potential role in infection assessment. Given the variability in diagnostic accuracy, a composite scoring system integrating multiple clinical and laboratory parameters is essential to enhance early and appropriate antibiotic initiation while minimizing unnecessary use.
Footnotes
Provenance and peer review: Invited article; Externally peer reviewed.
Peer-review model: Single blind
Corresponding Author’s Membership in Professional Societies: Indian Academy of Pediatrics, L/2005/G/939; Emergency Medical Association, 116.
Specialty type: Pediatrics
Country of origin: India
Peer-review report’s classification
Scientific Quality: Grade B, Grade C, Grade D
Novelty: Grade B, Grade B
Creativity or Innovation: Grade B, Grade C
Scientific Significance: Grade B, Grade B
P-Reviewer: Kaya B; Mai DN S-Editor: Liu H L-Editor: Filipodia P-Editor: Zheng XM
Saha SK, Schrag SJ, El Arifeen S, Mullany LC, Shahidul Islam M, Shang N, Qazi SA, Zaidi AKM, Bhutta ZA, Bose A, Panigrahi P, Soofi SB, Connor NE, Mitra DK, Isaac R, Winchell JM, Arvay ML, Islam M, Shafiq Y, Nisar I, Baloch B, Kabir F, Ali M, Diaz MH, Satpathy R, Nanda P, Padhi BK, Parida S, Hotwani A, Hasanuzzaman M, Ahmed S, Belal Hossain M, Ariff S, Ahmed I, Ibne Moin SM, Mahmud A, Waller JL, Rafiqullah I, Quaiyum MA, Begum N, Balaji V, Halen J, Nawshad Uddin Ahmed ASM, Weber MW, Hamer DH, Hibberd PL, Sadeq-Ur Rahman Q, Mogan VR, Hossain T, McGee L, Anandan S, Liu A, Panigrahi K, Abraham AM, Baqui AH. Causes and incidence of community-acquired serious infections among young children in south Asia (ANISA): an observational cohort study.Lancet. 2018;392:145-159.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 164][Cited by in RCA: 155][Article Influence: 22.1][Reference Citation Analysis (0)]
Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME, Townsend SR, Reinhart K, Kleinpell RM, Angus DC, Deutschman CS, Machado FR, Rubenfeld GD, Webb S, Beale RJ, Vincent JL, Moreno R; Surviving Sepsis Campaign Guidelines Committee including The Pediatric Subgroup. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012.Intensive Care Med. 2013;39:165-228.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 3146][Cited by in RCA: 3179][Article Influence: 264.9][Reference Citation Analysis (0)]
World Health Organization.
Pocket Book of Hospital Care for Children: Guidelines for the Management of Common Childhood Illnesses. Severe acute malnutrition. 2nd ed. 2013. Available from: https://www.ncbi.nlm.nih.gov/books/NBK154454/.
[PubMed] [DOI]
National Collaborating Centre for Women's and Children's Health (UK).
Feverish Illness in Children: Assessment and Initial Management in Children Younger Than 5 Years. London: Royal College of Obstetricians and Gynaecologists (UK), 2013.
[PubMed] [DOI]
