Decreasing errors in pediatric continuous intravenous infusions

From Clinfowiki
Jump to: navigation, search

Decreasing errors in pediatric continuous intravenous infusions


When medicating children and infants through an intravenous infusion, there is a great likelihood of one or more errors occurring. The majority of these errors are likely to happen during medication ordering. More pharmacologic knowledge and accuracy is required for calculating pediatric intravenous dosages because they are based on age, weight, body surface area. Also when compared to adults, children have a lower tolerance for dosage errors and different physiological responses.

This study aimed to 1) assess the type and frequency of errors in prescriber ordering and pharmacy dispensing of intravenous drugs and 2) evaluate the effect an introduction of a web-based calculator and decision support system would have on the error rate in pediatric intravenous drug orders. The two phases observed through this study were the first phase where all dosages were calculated manually and orders were hand-written. In the second phase, the medical staff had an opportunity to use the web-based system, but its usage was not mandatory.


The primary results of the study were that the second phase saw a 50% reduction in orders with errors (27% phase one vs. 14% phase two). If a 50% decrease in error rates weren’t impressive enough, when only considering the calculator-based prescriptions in phase two compared to phase one, there was an 83% reduction in orders with at least one error (33% phase one vs. 6% phase two) and a complete elimination of high-risk prescribing errors (26% phase one vs. 0% phase two). Of the phase two orders that were hand-written, the errors increased from 27% to 70%, but the reasons for this were not understood. In addition, where hand-written orders may have had up to 4 errors per order, the highest number of errors in any calculator-based order was two.

Another impressive result is that 88% of the phase two orders were calculator-based. The high adoption rate was attributed to benefits over paper-based work including automated computation, order form generation, fewer steps, decreased time required and ease of use.

It was noted that there were no pharmacy preparation errors during either phase. This might have been attributed to observational effect, even though the pharmacy staff was blinded to the actual purpose of the observations.

One type of error that increased between the two phases was missing patient demographic information. During the study, the web-based system required that information to be entered in free-text. After the study, the system was modified to integrate with the hospital patient information system, providing a drop-down list of patients and automatically populating fields.

The findings of this study resulted in the academic medical center mandating the use of the calculator for all pediatric continuous infusions.


It’s important to note that the quality of the system likely had a direct effect on the adoption rate and error rate. The software itself could introduce additional errors, not prevent enough errors, or be so difficult to use that the adoption rate is low. The paper reported that none of the current major CPOE systems appear to provide a similar level of support for pediatric infusion calculations. After reading the results of this study, this is disturbing as building this type of functionality does not appear to be a monumental task, and the benefits would be significant.

It would be interesting to study what the error rates were after the use of the calculator became mandatory.