The management of neonatal and pediatric sepsis remains a significant and challenging public burdensome health issue on a worldwide scale (Ruth et al., 2014; Weiss et al., 2015; Shane et al., 2017). Globally pooled data revealed that the case fatality rate of pediatric sepsis was approximately 31.7% in developing countries and 19.3% in developed countries (Tan et al., 2019), and the mortality of neonatal sepsis ranged from 11% to 19% (Fleischmann-Struzek et al., 2018). Nevertheless, current guidelines and opinions recommended that dosing strategies of antibiotics should be optimized according to pharmacokinetic/pharmacodynamic (PK/PD) principles (Rhodes et al., 2017; Weiss et al., 2020) and encouraged to provide individualized antimicrobial dosing approaches for special populations whenever possible (Hartman et al., 2020; Tu et al., 2021; Cao et al., 2022) given the scarcity of antibiotic agents in the pediatric population and the rise in antibiotic resistance.

In neonatal sepsis, the choice of antibiotic is complex and should be based on clinical syndrome, underlying disease, drug intolerances, and local pathogen susceptibility (Plunkett and Tong, 2015). β-lactams and vancomycin are two of the most frequently utilized antimicrobials due to their broad range of activity and broad therapeutic index (Plunkett and Tong, 2015; Phe et al., 2020; Burgunder et al., 2022). β-lactams, the cornerstone of the empirical antibacterial therapy in neonatal sepsis owing to their rare toxicity with only a few exceptions (Muller et al., 2018), mainly include three categories: penicillins (e.g., amoxicillin, ampicillin, and oxacillin, et al.), cephalosporins (e.g., cefazolin, cefuroxime, and ceftriaxone et al.) and carbapenems (e.g., meropenem, imipenem-cistatin, and ertapenem, et al.). Vancomycin, one of the most frequently prescribed glycopeptides in neonatal infections, is recommended to cover vascular catheter-associated coagulase-negative staphylococci and Meticillin resistant Staphylococcus aureus(Jacqz-Aigrain et al., 2015; Minotti et al., 2021).

Generally, PK refers to the study of concentration changes of a drug over a given time period, and PD describes the relationship between PK exposure and pharmacological effect (Abdul-Aziz et al., 2015). As time-dependent antibiotics, extended infusion (EI) or continuous infusion (CI) administration of β-lactams could theoretically lead to improved PK/PD profiles when the concentration remains above the minimum inhibitory concentration (MIC) of the causative pathogen (fT > MIC) for a longer duration (Rizk et al., 2017; Veiga and Paiva, 2018). Vancomycin could also theoretically better meet PD targets when provided as prolonged administration (DiMondi and Rafferty, 2013; Gwee et al., 2014; Alonso-Moreno et al., 2021). Although EI/CI administration for critically ill patients has been suggested by clinical practice guidelines (Bretonnière et al., 2015; Egi et al., 2021; Evans et al., 2021), there lacks convincing evidence demonstrating its superiority compared with a short-term infusion or bolus dosing in children (Zhou et al., 2021), and pediatricians seem to be not knowledgeable of it or to apply it frequently in clinical practice. There are currently no data available demonstrating the practical use of EI/CI approaches in the management of neonatal sepsis in China.

This study is a subset of a national survey titled “Current status of drug optimization for the application of β-lactams antibiotics and vancomycin in the treatment of neonatal sepsis in China” (OBTAINS study). We aimed to conduct a cross-sectional survey of pediatricians in China to characterize the current implementation of EI/CI-lactams or vancomycin in neonatal sepsis and discover its potential factors.

To the best of our knowledge, this large nationwide survey represents the first cross-sectional study to investigate the implementation of EI/CI of β-lactams and vancomycin in neonatal sepsis in China. The findings could contribute to identify pediatricians’ current knowledge, clinical practice and attitude regarding this antimicrobial optimization strategy.

Generally, the findings of this survey were well represented by 1501 pediatric physicians in mainland China according to their demographic characteristics and clinical experience. It is worth mentioning that only approximately one-third of the respondents had access to EI/CI of β-lactams (most frequently meropenem, imipenem-cistatin, cefoperazone-sulbactam and piperacillin-tazobactam) and vancomycin for 2–3 h in most cases, with selective administration of a loading dose in critically infected neonates. Notably, the prolonged strategy was primarily supported by guidelines and consultations from medical specialists (such as anti-infection specialists, clinical pharmacists), but the attitude and clinical outcomes were of considerable uncertainty and wide variability among the respondents. A large proportion of respondents did not clearly know or provide accurate answers to the stability of nine kinds of β-lactams and vancomycin in our study. Due to unawareness of or uncertainty over the stability of infusion, pediatricians may consequently decide to abandon the EI/CI strategy.

Furthermore, the multivariate analysis findings demonstrated several potential factors, including the education level, the hospital grade, and the department or ward pediatricians worked in, that could be connected to the implementation of the EI/CI strategy in neonatal sepsis. Additionally, we discovered that an increase in the use of the EI/CI approach was linked to the appropriate time pediatricians worked in the neonatal department/ward, the well-developed MDT team for neonatal sepsis, and more frequent MDT discussion on antibiotic use. Several published studies reported that MDT intervention could benefit anti-infection management in the NICU (Hamza et al., 2022). Neonatologists are encouraged to develop and implement antibiotic utilization strategies in collaboration with surgeons (surgical prophylaxis, the management of surgical infections) (Robinson et al., 2020), infectious disease specialists (MIC and drug resistance), pharmacists (PK/PD theories, toxicities, interactions, and drug information), nurses (EI/CI care), and other NICU providers (Katz et al., 2021). Moreover, it is imperative to note that autonomous learning about antimicrobial stewardship in neonatal sepsis through lectures, guidelines and consultations was not substantially correlated with a higher proportion of EI/CI use in our analysis. Consequently, future quality improvement studies or antimicrobial stewardship practices should deliberately arise awareness, optimize processes, and utilize multidisciplinary resources for neonatologists (Li et al., 2022).

Previously, research on EI/CI strategies administered in intensive care units for adults demonstrated that even though optimized administration of β-lactams and vancomycin were universally acknowledged by intensive care specialists, limited fraction of them actually implemented these principles in clinical practice owing to the lack of knowledge and difficulties in accessing information (Dulhunty et al., 2011; Buyle et al., 2013; Tabah et al., 2015; Charmillon et al., 2016; Liebchen et al., 2020). However, only a few studies have focused on pediatric populations with severe infections and optimized antimicrobial dosing strategies (Knoderer et al., 2014). To our knowledge, few large sample-sized cross-sectional surveys have been carried out in China. It is of irreplaceability to further disseminate the knowledge about the PK/PD theory of prolonged strategy in time-dependent antibiotics, current best clinical evidence on efficacy and safety and implementation experience of EI/CI approaches (Tabah et al., 2015).

The studies indicated that the majority of neonatologists could recognize that antibacterial agents’ PK parameters (such as volume of distribution and drug clearance) are significantly altered in critically infected children (Hartman et al., 2020). However, the optimal antibiotic optimization in accordance with PK/PD theory remained multifaceted and poses particular challenges to neonatal pediatricians (Tängdén et al., 2017; Phe et al., 2020). The potential barriers to its implementation might be mainly attributed to the lack of pharmacological knowledge, the absence of an MDT team, and the lack of confidence in clinical practice owing to limited efficacy and safety evidence in pediatrics. Recent published systematic reviews have demonstrated that EI/CI had generally improved clinical outcomes and safety profiles compared with intermittent infusion strategy, with a greater probability of achieving the target (Costenaro et al., 2020; Alonso-Moreno et al., 2021; Zhou et al., 2021). It is undeniable that the EI/CI strategy has emerged as a promising option (De Waele et al., 2014), even though PK of antibiotics may be affected in critically ill neonates undergoing significant physiological alterations and the appropriateness of EI/CI administration remains further PK/PD or clinical evaluation data (Wang et al., 2020; Saito et al., 2021; Yonwises et al., 2021). A clinical practice guideline describing the methods and procedures for using EI/CI of β-lactams and vancomycin in neonatal sepsis management is consequently urgently required based on our study findings.

To further overcome these obstacles, specific implementation strategies should be developed accordingly (Correa et al., 2020). First, guidelines, research or clinical experience need to be disseminated to department administrators, pediatricians and patients’ families (Boland et al., 2019a; Boland et al., 2019b). Second, it is necessary to establish an MDT team regarding antibacterial agents, with at least anti-infection specialists, clinical pharmacists and microbiologists on board (Whalen et al., 2021). Third, the department’s policies, guidelines, or a trustable clinical decision support system may aid pediatricians in standardizing their EI/CI strategies and boosting their self-assurance (Euteneuer et al., 2019; Persad et al., 2021; Daphtary and Baloglu, 2022).

Nonetheless, there are several limitations in this study. Owing to resource constraints, we used a convenience sampling method to ensure that a specific number of results were available for individual province from China rather than conducting a strict stratified random sample across the nation. The questionnaire results might therefore be assumed representative. Additionally, the questions in the survey concerning clinical experience only reflect the impressions and attitudes towards the respondents other than the actual daily clinical practice. Finally, as a cross-sectional study, its causal association should be determined cautiously. The non-significant results in certain domains do not indicate a lack of association but may be due to limited sample size.

In conclusion, this national survey demonstrated that the EI/CI strategy in neonatal sepsis was only recognized by one-third of Chinese neonatologists. To improve the implementation of antibacterial optimization, a solid MDT team should be developed with regularly collaborates. Guidelines regarding prolonged infusion management in neonatal sepsis would be beneficial for clinical practice and should be developed in the near future.