MSA! Articles

Highlights from a Study of Residents' Electronic Medication Prescribing Errors

In academic teaching hospitals, medical residents typically enter most of the medication orders for patients using electronic prescribing (e-prescribing) systems. However, little is known about the association between the residents’ level of training and the frequency of medication prescribing errors or when they occur. Some studies have revealed the July effect, suggesting that errors increase in July as new residents begin their training.1,2 However, these studies have not specifically analyzed medication errors as the sole adverse outcome. Also, most studies of errors by residents were conducted before widespread adoption of e-prescribing systems and electronic health records (EHRs).2

The results of a recent retrospective cohort study by Ari Garber and colleagues on medication e-prescribing errors made by 335 internal medicine residents in an academic medical center were just published in the first 2019 issue of the Southern Medical Journal.2 The large study of more than 1.7 million inpatient electronic medication orders during a 4-year period (2011-2015) is among the first to specifically analyze resident medication e-prescribing errors. The objectives of the study were to describe the types and frequency of resident medication e-prescribing errors and to analyze their association with the post graduate year (PGY) of residency training, and the time of day and month that the errors occurred. The study did not identify all types of medication prescribing errors and instead focused only on problems detected by pharmacists that were classified into 5 categories: prescribing a drug that the EHR identified as an allergy; a drug interaction; duplicate therapy; an unclear or incomplete order that needed clarification; and a failure to adjust doses and/or monitor patients with renal impairment. The research team hypothesized that resident medication e-prescribing errors would decrease as they gained experience and would be highest at night, when the residents had less supervision.2

Highlights of the study are provided below;2 however, we recommend that academic teaching hospitals with medical residents read the full results, discussion, and conclusions of this study to gain further insight into resident medication e-prescribing errors and to assist in planning effective strategies to reduce the risk of their occurrence.

Study Results2

Frequency and Harm

  • Despite some error detection functionality (e.g., duplicate therapy and allergy alerts) with the e-prescribing system, pharmacists identified an error in approximately 4% of the residents’ medication orders.

  • None of the resident medication e-prescribing errors detected in this study resulted in patient harm because pharmacists identified and corrected them before reaching patients.

Types

  • Overall, and for each PGY level (1, 2, 3), the most common type of error was a failure to adjust dosing or monitor for renal impairment (40%), followed by unclear or incomplete orders that needed clarification (27%), duplicate therapy (25%), drug interaction (5%), and prescribing a drug to which a patient may be allergic (4%).

Medications

  • Medication classes associated with the highest rates of pharmacy-detected errors were antimicrobials (14%), anticoagulants (9%), colony-stimulating factor agents (8%), biologicals (8%), and antidotes (6%). Among these medications:

    • Errors with antimicrobials were most often associated with lack of renal dose monitoring/adjustments (69%), unclear or incomplete orders (17%), and allergies (5%).

    • Errors with anticoagulants were most often associated with lack of renal dose monitoring/adjustments (65%), duplicate therapy (18%), and unclear or incomplete orders (14%).

    • Errors with colony-stimulating factor agents, biologicals, and antidotes were most often associated with unclear or incomplete orders (89%, 77%, 83%, respectively) and duplicate therapy (9%, 18%, 11%, respectively).

  • With the exception of antimicrobials and anticoagulants, medications prescribed infrequently by residents had the highest rates of prescribing errors.

Timing

  • Resident errors were highest during the day (peaking in the morning), not at night as hypothesized, which the researchers believed may be due to the volume and type of daytime orders and multitasking.

  • Resident errors were less frequent than expected during transition periods (7-9 a.m., 5-7 p.m.), which the researchers believed may be due to the use of handoff tools that limit such errors (e.g., SBAR [Situation, Background, Assessment, Recommendation]), or a failure to detect errors that originate during transition periods because they may not manifest right away.

  • Evidence of the July effect was not found. Errors were most frequent in August and among the least frequent in July. The researchers thought that lower error rates in July may be due to heightened supervision during the first month of residency training, and higher error rates in August may be due to the residents’ growing confidence and realization that all medication orders are verified by a pharmacist.

Training Level

  • The highest frequency of medication e-prescribing errors occurred during PGY 1. The researchers believed that the decrease in errors observed between PGY 1 and PGY 2 may be due to better medication knowledge and familiarity with the EHR.

  • The lowest frequency of medication e-prescribing errors occurred during PGY 2. PGY 1 and PGY 3 residents committed more errors than PGY 2 residents. The higher error rate of PGY 3 residents compared to PGY 2 residents was puzzling, as PGY 3 residents ordered the fewest medications. Possible explanations suggested by the researchers included an increase in patient and therapy complexities, fewer consultations with others before placing orders, and knowledge decay.

  • Resident errors declined during the course of the academic year, with the odds of an error decreasing by 16% throughout the year. 

Study Take-Aways2

The researchers made the following recommendations based on their conclusions regarding the study results.

Additional Resident Supervision

  • Although autonomy fosters resident learning, do not withdraw resident supervision prematurely after the first month of training. The timing of errors suggests the need for increased supervision in August and September, not just in July.

Continued Pharmacy Support

  • The frequency of resident e-prescribing errors underscores the need and value of ongoing pharmacy review of all residents’ medication orders, particularly given widespread alert fatigue that often leads to bypassing EHR error detection functionality.

  • Establish safeguards when dispensing medications infrequently prescribed by residents.

Resident Education

  • Educate residents about the specific kinds of errors that are common when ordering certain types of medications, particularly the 5 classes of medications most often involved in resident e-prescribing errors: antimicrobials, anticoagulants, colony-stimulating factor agents, biologicals, and antidotes.

  • Educate residents about the differing types of errors seen with commonly versus less commonly prescribed medications.

Encourage Consultation

  • Encourage PGY 3 residents to consult with other healthcare professionals when caring for complex patients or ordering medications prescribed infrequently.  

Strengthen Renal Dosing/Monitoring Capabilities

  • Establish a reliable plan to ensure medication dose adjustments and baseline/ ongoing monitoring of patients with renal impairment occurs, particularly when certain anticoagulants and antimicrobials are prescribed. Until the EHR can integrate measures of creatinine clearance with drug prescribing, this may be best accomplished with a pharmacy renal dosing protocol that targets at-risk patients and medications.

References

  1. van der Leeuw RM, Lombarts KM, Arah OA, Heineman MJ. A systematic review of the effects of residency training on patient outcomes. BMC Med. 2012;10:65.
  2. Garber A, Nowacki AS, Chaitoff A, et al. Frequency, timing, and types of medication ordering errors made by residents in the electronic medical records era. South Med J. 2019;112(1):25-31.

More MSA! Articles

The US Food and Drug Administration (FDA) approved LUMOXITI (moxetumomab pasudotox-tdfk) injection in September 2018. Lumoxiti is a CD22-directed cytotoxin indicated for the treatment of adult patients with relapsed or refractory hairy cell leukemia who received at least two prior systemic therapies...