Oral Themed Presentation
Background: : Pediatric resuscitations are rare and stressful for trainees. Challenges include the need for rapid assessment and quick actions like ordering medication in doses calculated according to weight. Rates of errors between 10-26% have been reported in simulated studies. Studies on cognitive aids to reduce errors in pediatrics reported lower risks of tenfold errors and errors for continuous infusion medication. An App simulated study showed reduction in continuous infusion medication errors and decreased time to preparation. Using applications could facilitate medication orders and potentially decrease prescribing errors (PE) and improve patient safety.
Research Question: : Does the use of smartphone applications decrease rates of prescribing errors among residents during simulated pediatric resuscitations?
Methodology: : A crossover randomized experimental trial in a tertiary care hospital simulation center. Participants: Residents rotating in the pediatric ED. Intervention: Apps providing pre-calculated medication doses vs. book providing pre-calculated medication doses. Primary outcome: Medication-prescribing errors defined as a drug prescribed in a dose varying by > 20% from the recommended dose, or incorrect route. Residents were involved in two sets of paired scenarios during two sessions [anaphylaxis: ampicillin or nuts and status epilepticus: febrile or non-febrile]. Participants were their own control and randomized to the App or book and to one of each scenario pair as their first session and the alternate during the second session. Two trained independent raters evaluated videotaped scenarios. Primary analysis: Difference in mean medication error proportions between both groups. Our sample size was 40 residents, representing their own control.
Results: : Between December 2016 and August 2017, 40 residents prescribed 1224 medications during 160 scenarios.
Video performances were evaluated by two blinded raters. The inter-rater reliability was good for most variables with ICC's and Kappa scores > 0.60. Apps: PediSTAT (75%), EZDrips (20%) and Safedose (3%).
Median number of medication dosages prescribed were 8 and 8 (anaphylaxis and status epilepticus)
There were 24/80 (30%) scenarios in the App group with at least one error and 16/80 (20%) in the control group, a difference of 10% (95% CI -3.5, 23%). The number of PEs per 100 bolus medications were relatively low and similar. There was no statistical difference in the number of errors between both groups: 4.31 (29/673) versus 3.02 (20/662) (intervention and control groups respectively). PEs most commonly involved saline bolus (n=9), epinephrine (n=7) and diazepam/lorazepam (5).
The App group had 4 scenarios with at least a tenfold error and there were 2 in the control group. Fentanyl was involved in four of these errors while epinephrine was involved in two.
For medications administered by infusion, PEs occurred in 11/40 (28%) scenarios in the application group and 3/40 (8%) in the control group (diff of 20%; 95% CI 3, 36%).
Discussion/Conclusions: : There were no differences in proportions of PE’s between Apps and a book. However, the low rate of PE’s for both groups reported emphasizes the recommendation of accessing dose-calculated tools during stressful pediatrics resuscitations to reduce medication PE. Use of an App was associated to a higher risk of PE for infusion medication.
More studies on Apps are necessary to gather data on types of errors and to explore ways to improve their design and usage in actual pediatric resuscitations to evaluate the impact on patient safety.