prescribing phase and those that occurred during the computerised phase. All steps of the process were assessed using a data collection sheet. Statistical analysis was performed by PSPP software. Results Fifty pockets of parenteral nutrition were examined during both phases. This study showed a statistically significant improvement in considering both the sodium and fluid intake of the other drugs prescribed with: OR=0.40, 95% CI: 0.30 to 0.58, p<10-3, OR=0.30, 95% CI: 0.19 to 0.45, p<10-3 , respectively.Regarding the preparation step, the order of components introduced was significantly better when using the software: OR=0.12, 95% CI: 0.04 to 0.35), p<10-3 . The labelling was significantly more respected with computerisation: OR=0.22, 95% CI: 0.06 to 0.74, p=0.017.No impact was detected in the transcription step when using the software with: OR=1.53, 95% CI: 0.53 to 4.42, p=0.424. Likewise, no impact was detected in the administration step with: OR=0.49, 95% CI: 0.04 to 5.58, p=1. Conclusion The implementation of the prescribing software was beneficial in terms of error management, time and traceability. The computerisation of the process, from the prescription to the administration, is necessary to guarantee security and efficiency in the neonatal intensive care unit. Thus, it is recommended to generalise this pilot experiment in the interests of both prescribers and patients.

No abstract

BackgroundFluoroquinolones are frequently used in hospital for many indications. However, overuse or incorrect utilisation may cause resistance to these antibiotics. Furthermore, we had to analyse the consequences of a recently ofloxacine suppression from our hospital antibiotic list.PurposeWe wanted to know the level of relevance of fluoroquinolone prescription including: indication, fluoroquinolone molecule choice, dosage (including CKD-EPI based dose reduction), duration, route, eventual association, compliance with the antibiogram, drug interaction and catch of fluoroquinolone from 6 months before.Material and methodsWe carried out a 3 month transversal retrospective study. Between February and April 2017 each nominal prescription of fluoroquinolone was included using our pharmacy validation software. All services were included except intensives care units and emergencies. Then an intern in the pharmacy processed analyses of the relevance of the previous parameters with the help of senior infectious disease staff. The reference guideline used for relevance and analysis was the 2015 French Spoken Infectious Diseases Society (SPILF) recommendations.ResultsTwo hundred and six patients were included. The most recovered fluoroquinolone was levofloxacine 47.1% (95% CI: 40.3 to 53. 9). The average duration of treatment was 12 days (95% CI: 10.3 to 13.6). The most used route was oral 81.7% (95% CI: 76.2 to 87.2). In a large majority of fluoroquinolone prescription was probabilistic 60% (95% CI: 53.4 to 67.7).Analysis of relevance showed that indication was respected in 84.7% (95% CI: 79.5 to 89.8) of cases. The association of correct duration, dosage, molecule used and route was observed in 63.7% (95% CI: 53.8 to 73.6). 70.8% (95% CI: 58 to 83.7) of renal insufficiency patients (CKD-EPI <60 ml/mn) received the exact dose reduction based on CKD-EPI. We identified the presence of a drug interaction in 13.8% (95% CI: 9 to 18,8) of prescriptions. Antibiotic association was noted in 42,7% (95% CI: 35,7 to 49,7) of prescriptions; only a few, 43.6%, (95% CI: 32.59 to 54.59) were relevant.ConclusionEven in the context of important changes in the fluoroquinolone list in our hospital, prescriptions are mostly relevant for indications and molecule choice. Serious medical information seems to be necessary concerning antibiotics associations. Adequate fluoroquinolone reduction dose must be a priority for renal insufficiency patients. Solutions available may be a special control using the biological software allied with pharmaceutical analysis targeted at low CKD-EPI patients.No conflict of interest