Frequency, Timing, and Types of Medication Ordering Errors Made by Residents in the Electronic Medical Records Era

Ari Garber, MD, EdD; Amy S. Nowacki, PhD; Alexander Chaitoff, MPH; Andrei Brateanu, MD; Colleen Y. Colbert, PhD; Seth R. Bauer, PharmD; Zubin Arora, MD; Ali Mehdi, MD; Simon Lam, PharmD; Abby Spencer, MD; Michael B. Rothberg, MD, MPH

Disclosures

South Med J. 2019;112(1):25-31. 

In This Article

Results

Between July 1, 2011 and June 30, 2015, a total of 1,772,462 orders were placed by 335 unique internal medicine residents. The mean number of orders placed per resident was 5291, with a median of 4980 and a range of 14,590. Fewer orders were placed by PGY3 (27%) compared with PGY1 (37%) or PGY2 (36%) residents, and the mean, median, and range of orders stratified by training year appear in Supplemental Digital Content, Appendix Table 2 (http://links.lww.com/SMJ/A128). The classes of medications that were most commonly prescribed were fluids and electrolytes (23%), cardiovascular medications (14%), analgesics (12%), and gastrointestinal medications (10%; Supplemental Digital Content, Appendix Table 3, http://links.lww.com/SMJ/A129), which remained consistent across PGY levels.

In total, 68,545 (3.9%) of residents' orders triggered pharmacist intervention via the iVent system and were subsequently classified as medication order errors (Table 1). In unadjusted comparisons, errors were slightly more frequent in the PGY1 year, during the day, and in the summer months. There was no change in error rates following the lifting of duty hour restrictions. The renal dose monitoring/adjustment required category comprised the greatest number of errors overall (40%) and within each PGY level, followed by order needing clarification (27%), duplicate therapy (25%), drug interaction (5%), and allergy (4%; Table 2). The classes of medications associated with the highest rates of errors were antimicrobials (14%), anticoagulants (9%), colony-stimulating factor agents (8%), biologicals (8%), and antidotes (6%; Table 3). These most common medication order error categories were similar throughout all 3 PGY levels. In general, high rates of errors occurred for medications that were infrequently prescribed, with the exception of antimicrobials and anticoagulants. For the more commonly prescribed medications, the majority of errors were attributed to renal dose adjustment, accounting for 68.6% and 64.9% of the errors for antimicrobials and anticoagulants, respectively. For the less frequently prescribed medications, the majority of errors were attributed to the order needing clarification, which accounted for 89.2%, 77.0%, and 82.8% of the errors for colony-stimulating factor agents, biologicals, and antidotes, respectively. The plurality of errors occurred during the day (9:00 AM–4:59 PM) and peaked in the morning (Figure 1). During the course of the calendar year, the highest rate of errors occurred in August (Figure 2).

Figure 1.

Total number of orders and iVENT rate for 24 hours.

Figure 2.

Total number of orders and iVENT rate for 1 year.

In the multivariable model adjusting for shift, therapeutic class of medication, and calendar year, both postgraduate resident training level and months of training (within a given year) were associated with medication order errors. Specifically, PGY2 residents had the lowest adjusted rate of errors. Compared with PGY2 residents, orders placed by PGY1s were associated with a greater odds of an error (adjusted odds ratio [aOR] 1.06, 95% confidence interval [CI] 1.03–1.10), as were orders placed by PGY3s (aOR 1.07, 95% CI 1.03–1.10). Orders placed in July and August were more likely to contain errors compared with other months, although this relation was driven entirely by error rates in August (aOR 1.09, 95% CI 1.03–1.14). Throughout the course of any single year of training, the odds of a medication order error decreased by 16% (aOR 0.84, 95% CI 0.80–0.89). Finally, orders placed during the night (7:00 PM–6:59 AM) and transition periods (7:00–8:59 AM and 5:00–6:59 PM) versus daytime (9:00 AM–4:59 PM) were less likely to contain errors than those placed during the day (aOR 0.93, 95% CI 0.91–0.96, and aOR 0.93, 95% CI 0.90–0.95, respectively).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....