Severe and critical COVID-19 in children during the Omicron BA.5-predominant and post-BA.5 calendar periods: clinical characteristics and risk factors

As severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to evolve, an updated clinical characterization of infected children is needed. This study analyzed the clinical characteristics of symptomatic children hospitalized for coronavirus disease 2019 (COVID-19) during the Omicron BA.5-predominant and post-BA.5 calendar periods and the risk factors for severe and critical disease. We retrospectively reviewed the clinical data of pediatric COVID-19 inpatients admitted to the Fifth Hospital of Xiamen between December 2022 and December 2025. Clinical classifications, underlying comorbidities, symptoms, laboratory test results, co-detection rate, and complications were compared between the BA.5-predominant and post-BA.5 calendar periods. Multivariable logistic regression identified factors independently associated with severe/critical disease. Firth penalized regression sensitivity analyses were performed to address definitional circularity and quasi-complete separation. Of the 519 hospitalized patients, 345 (66.4%) were infants or toddlers. Patients were categorized into mild (192 cases, 37.0%), moderate (227 cases, 43.7%), and severe/critical (100 cases, 19.3%). All the patients improved or achieved clinical recovery. The post-BA.5 calendar period was associated with a lower proportion of severe/critical disease among hospitalized patients, lower oxygen requirement, shorter median hospital stay, and fewer complications compared with the BA.5-predominant calendar period (all P < 0.05). Neurological symptoms were more frequent, and the viral co-detection rate was lower during the BA.5-predominant calendar period ( P < 0.05). Multivariable logistic regression suggested associations between severe/critical pediatric COVID-19 and neurological symptoms, the BA.5-predominant calendar period and young age. However, Firth penalized sensitivity analyses revealed that the extreme estimate for neurological symptoms was driven by definitional circularity and quasi-complete separation, and the associations between the two periods and age groups were unstable across models. Only hyponatremia and elevated neutrophil count remained consistently associated with severe and critical disease in both sensitivity analyses. Hyponatremia and elevated neutrophil count were consistently and independently associated with severe and critical disease. Primary multivariable regression identified associations with neurological symptoms, the BA.5-predominant calendar period and young age; however, sensitivity analyses revealed unstable or biased estimates. Thus, neurological symptoms should be regarded as clinical red flags rather than independent predictors. Routine monitoring of serum sodium levels may be considered.