The Epidemiology of Medication Prescribing Errors in the Emergency Department

Abstract

Medication errors are a major concern in the Emergency Department (ED). We examined the epidemiology of medication prescribing errors among consecutive adult ED patients during two 10-day periods in the adult ED prior to implementing computerized provider order entry. 479 patients had no medication orders and 1,301 had 3,393 orders. Per 100 orders there were 3.7 potential adverse drug events, 337.1 medication prescribing errors, and 24.2 rule violations.

INTRODUCTION

The Institute of Medicine report “To Err is Human: Building a Safer Health System” stated that the Emergency Department (ED) setting accounts for a large number of medical errors. The ED is vulnerable to medication errors due to clinicians providing episodic care to patients in emergent or urgent situations. Overcrowding episodes, nurse shortage, and scarcity of hospital beds further intensify the high volume of information exchange in a dynamic, high stress setting that is characterized by frequent handovers, multi-tasking, and frequent interruptions.

The implementation of computerized physician order entry (CPOE) has shown potential to reduce medication errors. In the ED, however, there is limited information about the epidemiology of medication errors prior to implementing CPOE. The goal of the study was to examine the epidemiology of medication prescribing errors in the adult ED prior to implementing CPOE.

METHODS

The medication orders for all consecutive adult ED patients were obtained from the medical chart during two 10-day periods (period 1: June 2, 2003 to June 11, 2003: period 2: January 26, 2004 to February 4, 2004). Chart retrieval rate was 97.2% after 3 attempts. Following established error classification methodology^1,2, a clinical pharmacist categorized medication errors into potential adverse drug events (pADE), medication prescribing errors (MPE), and rule violations (RV). Multiple MPEs and RVs within an order were counted separately allowing for more than one MPE or RV per order.

RESULTS

479 patients had no medication orders. The 3,393 orders (range: 1–12 orders) from 1,380 patients resulted in a total of 125 pADEs (3.7/100 orders), 11,439 MPEs (337.1/100 orders), and 820 RVs (24.1/100 orders). The Table displays medication prescribing errors for the two study periods. Allergies to medications had the highest rate of pADEs (3.0/100 orders), missing information (e.g., date/time or interval) was the most frequent MPE (210.8/100 orders), and incorrect abbreviations were the most frequent RV (23.8/100 orders).

Table.

Overall Medication Analysis in the adult ED

	Period 1 (n=1 ,678)		Period 2 (n=1,715)
	Total Number	Number per 100 orders	Total Number	Number per 100 orders
PADE	57	3.4	68	4.0
Duplicate therapy	3	0.18	5	0.29
Inappropriate dose	7	0.42	1	0.06
Inappropriate interval	1	0.06	0	0.00
Inappropriate route	1	0.06	3	0.17
Wrong drug	1	0.06	0	0.00
Wrong unit	1	0.06	0	0.00
Drug interaction	0	0.00	0	0.00
Allergy to medication	43	2.56	59	3.44
MPE	5832	347.6	5607	326.9
Missing information	3670	218.7	3483	203.1
No weight	1671	99.6	1640	95.6
Illegible	491	29.3	484	28.2
RV	430	25.6	390	22.7
Abbreviation	421	25.1	387	22.6
Trailing zero	0	0.0	0	0.0
No leading zero	9	0.5	3	0.2

DISCUSSION

Compared to a study in the pediatric intensive care unit of the same institution², the ED had 1.5 times more pADEs, 10 times more MPEs, and 3.5 times more RVs. The number of medication prescribing errors in the adult ED during two different study periods was high.