As with Patel et al. who evaluated antibiotic use according to the CDC 12-step program on a NICU, we used the CDC recommendations as the evaluation tool for our audit [8, 10]. Prolonged empiric antimicrobial treatment without clear evidence of infection, failure to narrow antimicrobial therapy when a causative organism is identified, and no clear documentation about indication and plan for therapy are reported to be common problems, and published data on general inappropriate use of antibiotics have similar findings to our results during phase I [12–16].
The first step in building an intervention for antibiotic surveillance/stewardship is to select the most appropriate strategy that leads to a sustained improvement in a defined area. In the guidelines for developing an institutional program to enhance antimicrobial stewardship there are two core strategies described: 1) prospective audit with feedback and 2) formulary restriction and pre-authorisation . Initiation of antibiotic therapy in the setting of a PICU patient is often empiric and based on apparently worsening clinical condition, with sepsis as a possible factor. During phase I of our audit vancomycin, carbapenems and third generation cephalosporins were not used regularly (not in our top 5 most commonly used antibiotics) and we concluded that the local formulary guidelines appeared to be respected without possible improvement with further formulary restriction and pre-authorisation [16–18]. We therefore selected prospective audit with feedback as a possible intervention, a strategy proposed in a recent review by Newland as the most favourable and efficient . The basis on an audit and feedback intervention is a review of every antibiotic course by a third party (typically an infectious disease physician or a clinical pharmacist) with the audit presented back to the prescriber. For implementation we faced the same barriers reported in other paediatric antibiotic stewardship programs: lack of resources, including funding, time, and personnel [9, 11]. We solved this challenge by developing a new strategy for audit and feedback. Instead of personnel audit and feedback by a third party, we created a checklist and used this as a way of audit and feedback (checklist as mandatory form for self-review and feedback by the treating physician). Recent publications about successful use of simple checklists guided us in developing our intervention [19–21]. Key elements for creating a successful checklist are short, clear and basic questions with reminders of routine care . With our checklist, we aimed to support the communication and documentation of indications for the initiation and continuation of antibiotic therapy, which served to remind the treating physician to review and potentially stop therapy or target the isolated pathogen. There are other publications regarding the use of mandatory prescription forms to start antibiotic therapy . The innovation of our approach is that the clinician is reviewing the prescription with help of a checklist at 2 points after start of antibiotic therapy (at 48 hours and 5 days).
The importance of a high compliance rate (power calculation aimed for a compliance rate of 80%) for a successful intervention is emphasised with the statistically not significant increase of antibiotic courses of empiric therapy < 3 days. In contrast to the aviation industry, there is still scepticism by some physicians about the use of checklists for the safe routine care and emergency situations in medicine, illustrated with a citation from A. Gawande in his recently published book “The Checklist Manifesto”: “It runs counter to deeply held beliefs about how the truly great among us – those we aspire to be – handle situations of high stakes and complexity” . A mandatory checklist in a computerised prescription system offers a possible way to improve compliance and impact of the intervention. Due to the potential risk of premature discontinuation of therapy a computerised checklist has to be mandatory but without an automatic antibiotic stop-order .
The limitations of our audit were the short surveillance time, the low compliance to the intervention (<80%) and the low numbers of proven infection with positive cultures (no conclusions regarding susceptibility patterns possible). The main strength of our audit was that it was a simple and achievable intervention which was low in cost (no additional funding) and resources. The construction of an easy attainable strategy is very important due to the fact that lack of resources and time are major barriers to implementing an antimicrobial stewardship program in paediatrics.