Purpose
Organisms resistant to antibiotics are increasing rapidly resulting in a global public health threat. Increasing rates of antimicrobial resistance (AMR) are a direct consequence of inappropriate antimicrobial use. [
1] Hospitalized children and adolescents have a very high antibiotic exposure, with 60% of them receiving at least one antibiotic per stay. [
2] Furthermore, a great amount of administered antibiotics is unnecessary or inadequately prescribed. [
1,
3]
Therefore, optimization of antimicrobial use and improvement towards rational prescribing of these valuable drugs is critically needed. This will result in a decrease of antibiotic resistance rates, improvement of patient care and reduced hospital stay for inpatients as well as a reduction in costs attributable to the inappropriate use of antibiotics. [
4,
5] To achieve these goals antibiotic stewardship programs (ASP) with different strategies and bundle approaches for rational use of antibiotics are implemented in hospital and ambulatory care. [
1,
6]
The aim of this survey study was to identify current practice and knowledge related to antibiotic stewardship aspects in paediatric junior doctors, middle-grade doctors and consultants in tertiary care paediatric hospitals in a south-eastern region in Germany using a questionnaire. The results of the survey should subsequently help to identify areas for future educational activities.
Discussion
This survey study aimed to identify current practice and knowledge on aspects of antibiotic use, ASP and infection control in tertiary care paediatric hospitals in and around Munich, Germany. Our analysis demonstrates an overall proportion of correct answers for all sub-sections of the survey of only just above 50%, thus elucidating the critical importance of continuing and improving educational activities covering all areas of antibiotic stewardship. In general, practice and knowledge across all categories did not differ significantly between junior doctors, middle-grade doctors or consultants. This is in line with similar studies, such as the survey of Bowes et al. assessing comparable sections in a survey published in 2014. [
7] Alshengeti et al. developed and analysed the effectiveness of an online virtual patient learning module for paediatric residents about antimicrobial stewardship in 2016. A modified version of Bowes et al. survey [
7] was used to measure the residents' knowledge. The overall knowledge score before the implementation of the ASP module was 58.2%, which is quite similar to the 55.0% we saw in our survey study. [
9]The biggest lack of knowledge was observed regarding the local antibiotic resistance pattern. Knowledge on macrolide-resistant group A
Streptococcus (GAS) and penicillin-resistant
Streptococcus pneumoniae isolates yielded an astonishingly low number of correct answers (26.1 and 8.1%, respectively). Of note, the pneumococcal local resistance rate was primarily correctly estimated by consultants (6/9; 66.7%). The vast majority (74.1%) of participating paediatricians did not know that more than 10% of group A
Streptococcus (GAS) are resistant to macrolides and thus underestimated the issue. Only a very small minority (8.1%) was aware of the local penicillin-resistance rate (prevalence of <1%) in
Streptococcus pneumoniae. Therefore, penicillin resistance in pneumococcus is likely to be overestimated leading to less frequent use of penicillins in common conditions such as community-acquired pneumonia (CAP). Bowes et al. found similar results in their study and pointed out, that this overestimation of antibiotic resistance levels can be an important reason for inappropriate prescribing and de-escalation strategies, resulting in increased use of broad-spectrum antibiotics. [
7]
Regarding hospital hygiene and infection control, basic hygiene measures such as hand hygiene appeared to be well known and common practice across all participating hospitals. But only less than half of the participants were aware of the correct isolation scheme for a patient with MRSA colonization of the upper respiratory tract and pointed out that droplet rather than contact precautions are required in this scenario. [
10] Of note, consultants did significantly worse when answering this question compared to middle-grade and junior doctors (
p value = 0.04). Though consultant-grade doctors should certainly be aware of the correct precautions to be implemented on the ward, the middle-grade and junior doctors are practically dealing with this every day. On the other hand, for the scenario of pulmonary tuberculosis the majority of respondents provided the correct answer for the precautions required (72.9%) with no significant difference between grades of paediatricians. [
11]
Two clinical scenarios were assessed in terms of empiric antibiotic choice. It is noteworthy to emphasize that while the choice of antibiotics for appendicitis was very variable, participants were rather uniformly suggesting ampicillin or amoxicillin for CAP. This reflects a direct effect of establishing and communicating internal guidelines because such a document had only recently been established for CAP at the Dr. von Hauner Children’s Hospital in early 2015, whereas no such document was available for a case of appendicitis at that time. Nevertheless, there is a national reference for appendicitis and pneumonia available in all hospitals (DGPI Handbuch [
12]) Hence, this is additional practical proof of how implementing local guidelines is a very effective measure to improve rational antibiotic use. [
13]
Therapeutic drug monitoring is an essential requirement when administering the glycopeptide vancomycin or aminoglycosides such as amikacin [
13] and is a standard laboratory service available for all participating hospitals. Overall, only 37/111 (33.3%) paediatricians correctly identified both antibiotics as requiring TDM. Thus, our results clearly demonstrate that these drugs, though frequently used in paediatric and, in particular, in neonatal care [
14], are most likely inappropriately applied and monitored putting the respective patients at risk for both an ineffective and potentially toxic therapy. Junior, middle-grade and consultant-level doctors were equally uninformed about the critical need for TDM. Still, when a paediatrician thinks about TDM, he/she is more likely to be aware of measuring serum levels when prescribing vancomycin than amikacin. There is a clear need to address TDM in targeted educational activities in the future.
When assessing the practical use of local, national or international guidance, consultants appeared to be the group most frequently turning to advice published in guidelines while junior and middle-grade doctors were more likely to directly consult their superiors when choosing an empiric or targeted antibiotic therapy. To a certain extent, this finding reflects an interesting difference between the Anglo-Saxon and the German medical system. While frequent rotation between workplaces (i.e. tertiary-care centres, district general hospitals) is an established standard in UK paediatric training, the majority of German junior and middle-grade doctors are spending their entire training period in the same hospital. Hence, inter-collegial bonds and influences of superiors, as well as an “in-house common practice” mode of action, is more common in Germany than in the UK system where NHS hospitals follow a more national guidelines-oriented and evidence-based-medicine approach (
https://www.nice.org.uk/guidance;
https://www.rcpch.ac.uk/resources/clinical-guidelines-evidence-reviews). However, only recently and after this survey study was conducted, a first national paediatric antibiotic stewardship guideline has been published. [
13] In addition, recent years have seen more national guidelines being developed on topics such as paediatric community-acquired pneumonia. [
15] A national paediatric guideline on appendicitis is currently under development (
https://www.awmf.org/leitlinien/detail/anmeldung/1/ll/006-003.html). These activities clearly reflect that the importance and need for coherent and evidence-based guidance for paediatric ASP have been recognised by the responsible scientific bodies and societies.
The easiest and most difficult ASP measures to be implemented in one’s own clinical working environment were identified. While using less broad-spectrum antibiotics as part of empirical antibiotic therapy was considered the most difficult measure, de-escalating a broad empirical therapy after receiving the pathogen differentiation and antibiogram was considered the easiest measure to achieve. Of note, switching from IV to oral therapy was regarded as the second easiest measure to implement, while reducing the duration of therapy or stopping an antimicrobial in the context of missing signs for infection were identified as rather difficult. These results might be mirroring the activities of the ASP at the Dr. von Hauner Children’s Hospital since 2015 where an initial focus was laid on de-escalation of broad-spectrum antibiotic therapies in the light of microbiological results. [
16] Similarly, participants in Bowes et al. survey considered discontinuation of antimicrobials in cases with no documented infection as most difficult. However, half of the trainees were in accordance with our findings and considered the empiric use of narrow-spectrum antibiotics versus broad-spectrum antibiotics as most difficult to achieve. [
7] Various studies, such as Levy et al. [
3], concluded that the failure of discontinuation and de-escalation of therapy was the most common reason for inappropriate antibiotic use. This is consistent across many publications and reflects the dilemma any clinician is facing when having to decide on continuation or discontinuation of antibiotics. There is an obvious need for better diagnostics and biomarkers to help in the decision making process towards de-escalation or discontinuation of antibiotic therapies. [
13] Unfortunately, the highest number of missing entries was observed in this section of the questionnaire, again reflecting the dilemma illustrated above. Nevertheless, these discontinuation and de-escalation aspects need to be a clear focus of future educational activities of local ASPs.
Given that at the Dr. von Hauner Children’s Hospital an ASP was initiated in 2015, the results of this survey were rather sobering. [
16] There was no significant difference in the knowledge of ASP measures between the Dr. von Hauner Children’s Hospital and other participating institutions.
This may be due to the rather short period of time that the ASP at the Dr. von Hauner Children’s Hospital had been in place. Furthermore, the ASP was not focused on teaching individual doctors but rather consisted of infectious diseases consultation service, development and provision of internal guidelines on empiric antibiotic therapy and clinical ward rounds of an antibiotic stewardship team formulating recommendations to assist paediatric colleagues on ward. [
16] While this approach improves the quality of patient care, it may not have the same effect on the knowledge of antimicrobial stewardship principles amongst paediatricians. Thus, the implementation of a structured teaching program would be a key measure to address this aspect of improving knowledge on the individual doctor’s level.
Our results indicate that the knowledge of hospital-based paediatricians of the south-eastern region of Germany regarding the different areas of ASP is only moderate and clearly needs further improvement to optimise the clinical care in paediatric hospitals.
In contrast to the Anglo-Saxon medical system where infectious diseases training and ASP aspects have long been integrated into the medical training curricula, starting at the medical student’s level only a very small minority (5/111, 5.6%) of all clinicians participating in our study had previous training in the respective areas. Specialist training in this area is an established component of medical postgraduate training in other European countries, such as the UK, or in the USA. Unfortunately, Germany does not yet have an equivalent focus on training in infectious disease, antibiotic stewardship and hospital hygiene. Thus, the results of this study strongly suggest this area as a critical focus for university and post graduate training in paediatrics. Only sustainable educational efforts in ASP and infectious diseases will be able to tackle this evident lack of knowledge in almost all areas covered by our survey and to help us improve patient care in the years and decades to come. That ASP training improves the knowledge was shown in a recent publication analysing knowledge scores of Canadian residents. A significantly higher knowledge score (71.6%) was found four months after implementation of an ASP module compared to before the pre-implementation period (58.2%). [
9]
Our study has several obvious limitations, some of which are due to the simple fact that this was only a survey study. Since we analysed data of German paediatric physicians in and around Munich (south-eastern region of Germany) and the response rate was quite low (42%) our results are not fully generalizable and may not be representative for other paediatric hospitals. We could not perform a non-responder analysis and therefore we do not know to what extent the non-responders differ from the participating clinicians.