Methods Simulation of electronic prescription (10 physicians), preparation/administration (10 paediatric/8 adult intensive care nurses) of 5 drugs (midazolam, fentanyl, noradrenaline, ketamine, furosemide) for 15 fictive patients (different dosage/ weight). Two-sessions study (VarC vs StdC, each 150 prescriptions/270 preparations). Issues: time (mean ± SD in seconds); precision (target deviation in%, mediane [IQR]) of drug concentration (quantitative analysis), dose and rate (calculated by

more »

... s). Results With StdC, prescription time was significantly longer (72 ± 36 vs 86 ± 32, p < 0.001) and preparation/administration time shorter (286 ± 98 vs 216 ± 93, p < 0.0001). Precision of drug concentration was increased (4.4% [2.0 to 11.5] vs 4.1% [1.6 to 8.4], p = 0.004) with a reduction of concentrations >20% (44/270 (16.3%) vs 23/270 (8.5%), p = 0.005). Precision of dose was decreased 4.4% [2.0 to 11.1] vs 11.8% [5.1 to 23.3], p < 0.0001) with an increase of dose >20% (42/270 (15.6%) vs 83/270 (30.7%), p < 0.0001). Precision of rate was decreased (0.0% [0 to 0] vs 6.8% [3.2 to 20.6], p < 0.0001). No association with age, years of experience, number of worked hours before study, paediatric or adult ICU nurses was observed on precision. Conclusions Preparation time and drug concentration precision was strongly improved with StdC. Strategies to deal with prescription time and poor dose and rate precision should be considered before moving to StdC. Poster symposium A158 Arch Dis Child 2014;99(Suppl 2):A1-A620