Discussion

This study provides a detailed description of antibiotic prescribing decisions in a tertiary-care PICU and contributes data from a low- and middle-income country setting where stewardship literature remains limited. The overall antibiotic utilization rate in this study was 59.1%, which is consistent with several studies that have reported high rates of antibiotic use in PICUs, ranging from 60% to 80% of admitted children. ^{Reference Weiss, Fitzgerald and Pappachan9,Reference Versporten, Bielicki and Drapier10} Systematic evaluation of antibacterial use provides insight into prescribing trends and supports the development of targeted antimicrobial stewardship strategies in this high-risk environment. ^{Reference Schuts, Hulscher and Mouton11–Reference Gandra, Alvarez-Uria and Turner13}

The predominance of respiratory illnesses as the principal indication of antibiotic use is consistent with findings from pediatric critical-care studies worldwide. ^{Reference Weiss, Fitzgerald and Pappachan9,Reference Versporten, Bielicki and Drapier10} Respiratory presentations often pose a diagnostic challenge as infective and inflammatory conditions of the lung can present with similar clinical and radiological findings. Hence, clinicians frequently initiate empirical antimicrobial therapy while awaiting further diagnostic clarification. ^{Reference Vu, Vu and Hoang14}

An important finding of the present study is the influence of clinical assessment on prescribing decisions. Approximately half of antibiotic initiations were based on clinical evaluation, emphasizing the realities of critical-care practice where immediate treatment decisions must often be made before definitive diagnostic information becomes available. ^{Reference Hersh, Jackson and Hicks15,Reference Phillip, Levy Mitchell and Rhodes Andrew16} This approach is appropriate in unstable patients; however, it reinforces the need for structured reassessment strategies to minimize the use of empirical therapy. ^{Reference Hyun, Hersh and Namtu17}

Inflammatory markers such as CRP, ESR, and procalcitonin contributed to the prescribing decisions but were rarely used in isolation. These biomarkers may help decision-making, but their diagnostic accuracy remains imperfect. ^{Reference Stocker, van Herk, El Helou and Dutta18} The findings support the current stewardship recommendations that advocate interpretation of inflammatory markers with the clinical and microbiological context rather than a standalone indicator for antibiotic therapy. ^{Reference De Waele, Akova and Antonelli19}

The blood culture positivity rate (2.4%) observed in this study is consistent with previous reports from ICUs, where the diagnostic yield of blood cultures is often low despite a high burden of suspected infection. Several factors may contribute to the low yield, including prior antimicrobial exposure and insufficient blood volumes obtained from pediatric patients. ^{Reference Woods-Hill, Koontz and Xie20} In contrast to blood cultures, urine cultures demonstrated a higher diagnostic yield and directly influenced antibiotic modification in genitourinary infections. Culture-guided treatment remains a cornerstone of antibiotic stewardship. Antibiotics were discontinued in a substantial number of children with neurological presentations once CSF culture was sterile, and exanthematous febrile illnesses after serological testing were suggestive of viral etiology. Diagnostic clarity facilitated cessation of unnecessary antimicrobial therapy, highlighting the value of review of antimicrobial prescriptions in response to evolving clinical and microbiological information. ^{Reference Hyun, Hersh and Namtu17–Reference De Waele, Akova and Antonelli19}

Cephalosporin plus aminoglycoside was the most commonly prescribed antibiotic combination (48.3%) among pediatric critical-care patients, reflecting its role as a preferred empirical regimen for severe infections due to its broad-spectrum coverage and synergistic activity. The combination of cephalosporin, aminoglycoside, and glycopeptide (6.9%) was used in patients requiring broader coverage, particularly against resistant gram-positive pathogens, while penicillin plus aminoglycoside (6.4%) was prescribed less frequently. The predominance of β-lactam–based combinations is consistent with evidence supporting the early use of broad-spectrum antibiotics in critically ill children, where prompt and appropriate antimicrobial therapy is crucial for improving clinical outcomes. ^{Reference Hersh, Jackson and Hicks15}

From a stewardship perspective, several practical opportunities emerge from this study. First, implementation of mandatory antibiotic review at 48–72 hours could facilitate earlier de-escalation and discontinuation. Second, strengthening diagnostic stewardship through timely specimen collection before antibiotic administration may improve microbiologic yield. Third, development of system-specific treatment pathways for common PICU conditions, particularly respiratory infections, could reduce variation in prescribing practices and support judicious antimicrobial use. ^{21,Reference Dellit, Owens and McGowan22}

This study adds to the growing international evidence that antimicrobial stewardship in PICUs should move beyond uniform prescribing frameworks toward context-sensitive strategies that integrate clinical assessment with laboratory evidence, radiological data, and microbiological reports when available. ^{Reference Schuts, Hulscher and Mouton11,Reference De Waele, Akova and Antonelli19}

Limitations of the study: Microbiological confirmation was not available for all patients, and culture positivity rates were low. The analysis was descriptive and did not evaluate appropriateness of therapy, guideline concordance, duration of therapy, resistance patterns, mortality, length of stay, or clinical outcomes. It was a single-center study; future multicentre studies incorporating these measures would provide a more comprehensive assessment of antimicrobial stewardship performance in pediatric critical-care settings.