A retrospective study of antibiotic regimens and hospital length of stay among children with typhoid fever in Pakistan

Typhoid fever remains one of the major healthcare problems in low- and middle-income countries, especially Pakistan, where extensively drug-resistant (XDR) strains have emerged significantly. The pediatric population is highly vulnerable, but evidence describing how commonly used antibiotic regimens are associated with hospitalization outcomes remains limited. This research aimed to describe antimicrobial resistance patterns and examine the association between monotherapy and dual antibiotic therapy with hospital length of stay (LoS). This retrospective analysis reviewed medical records of culture-confirmed pediatric typhoid patients (aged < 17 years) admitted between January 2019 and May 2022 at Indus Hospital and Health Network (IHHN), Karachi, Pakistan. This study analyzed sample characteristics, clinical presentation, antibiotic resistance patterns, and treatment regimens. We used chi-square, Mann-Whitney U, Kruskal-Wallis, and multivariable log-linear regression tests, with significance set at p ≤ 0.05. Of 319 patients included, 84.6% had XDR Salmonella typhi , with high resistance to first-line antibiotics and 100% resistance to Ciprofloxacin, while Imipenem, Meropenem, and Azithromycin retained 100% susceptibility. Dual antibiotic therapy was more common among XDR cases and was generally selected for patients considered clinically severe or resistant by treating clinicians and was associated with significantly longer LoS [7 days (IQR 5–9)] than monotherapy [5 days (IQR 4–7), p < 0.001]. This association persisted after multivariable adjustment for strain type, age, gender, residence, and symptom burden (adjusted geometric mean ratio 1.28, 95% CI 1.12–1.46, p < 0.001), although residual confounding by unmeasured disease severity cannot be fully excluded. XDR typhoid remains highly prevalent among pediatric patients in Karachi, with substantial resistance to first-line antibiotics. Dual therapy was generally selected for patients considered clinically severe or resistant by treating clinicians, and it remained independently associated with longer hospital stay even after multivariable adjustment for strain, demographic, and symptom-burden variables, suggesting the association is not fully explained by the confounders captured in this dataset. As disease-complication and severity data were not systematically recorded, residual confounding cannot be excluded, and this finding should be interpreted as an association rather than a causal treatment effect. These findings underscore the need for strengthened antimicrobial stewardship and prospective studies incorporating standardized severity measures.