Clinical decision support for atypical orders: detection and warning of atypical medication orders submitted to a computerized provider order entry system

Abstract

The specificity of medication-related alerts must be improved to overcome the pernicious effects of alert fatigue. A systematic comparison of new drug orders to historical orders could improve alert specificity and relevance. Using historical order data from a computerized provider order entry system, we alerted physicians to atypical orders during the prescribing of five medications: calcium, clopidogrel, heparin, magnesium, and potassium. The percentage of atypical orders placed for these five medications decreased during the 92 days the alerts were active when compared to the same period in the previous year (from 0.81% to 0.53%; p=0.015). Some atypical orders were appropriate. Fifty of the 68 atypical order alerts were over-ridden (74%). However, the over-ride rate is misleading because 28 of the atypical medication orders (41%) were changed. Atypical order alerts were relatively few, identified problems with frequencies as well as doses, and had a higher specificity than dose check alerts.

Keywords: clinical decision support systems; clinical pharmacy information systems; computer-assisted; drug therapy; medical order entry systems; pharmacy service.

Figure 1

A repeatable process for analyzing medication order data from a computerized order provider entry system database is shown. Captured medication order data are queried, and the number of unique order sentences is determined. Orders are grouped by generic medication name and route of administration into ‘medication-route dyads’. The largest dyads, those with the most frequently ordered items, are identified. Analysis of the variability of order sentences used to characterize the orders in the largest dyads identifies high, medium, and low variability dyads. Low (<6 order sentences needed to characterize 80% (phase I) or 99% (phase II) of the orders in a dyad) and medium (6–15 order sentences needed to characterize 80% (phase I) or 99% (phase II) of the orders in a dyad) variability dyads allow for the determination of ‘atypical’ or ‘unusual’ orders by comparison. Five dyads were selected for an ‘atypical order alert’.

Figure 2

An atypical order alert (AOA) for ‘clopidogrel oral 325 mg four times a day’. The first sentence displays the entire order, as entered, in blue with descriptions of the atypical field types in bold underline (eg, ‘Dose and Frequency’). The second sentence repeats the order noting the atypical field values, ‘325’ and ‘four times a day’ (highlighted in a red font). The AOA informs the recipient that less than 1% of past orders for this medication given by this route of administration matched the current order being entered in CPOE. A reevaluation of the order is suggested.