Medication Accuracy in Electronic Health Records for Microbial Keratitis

Key Points

Question

Are formal medication lists in electronic health records accurate for ophthalmic medications?

Findings

In this cross-sectional study conducted on clinical encounters of patients with microbial keratitis, almost one-quarter of medications did not match between the clinician’s clinical progress note and the formal electronic health record medication list.

Meaning

Medication lists held in electronic health records may not completely capture up-to-date, accurate clinical information of ophthalmic medications.

Abstract

Importance

Electronic health records (EHRs) contain an abundance of health information. However, researchers need to understand data accuracy to ask appropriate research questions.

Objective

To investigate the concordance of the names of medications for microbial keratitis in the structured, formal EHR medication list and the text of clinicians’ progress notes.

Design, Setting, and Participants

This cross-sectional study, conducted in the cornea section of an ophthalmology department in a tertiary care, referral academic medical center, examined the medications of 53 patients with microbial keratitis treated until disease resolution from July 1, 2015, to August 1, 2018. Documentation of medications was compared between the structured medication list extracted from the EHR server and medications written into the clinical progress note and transcribed by the study team.

Exposure

Medication treatment for microbial keratitis.

Main Outcomes and Measures

Medication mismatch frequency.

Results

The study sample included 24 men and 29 women, with a mean (SD) age of 51.8 (19.6) years. Of the 247 medications identified, 57 (23.1%) of prescribed medications differed between the progress notes and the formal EHR-based medication list. Reasons included medications not prescribed via the EHR ordering system (25 [43.9%]), outside medications not reconciled in the internal EHR medication list (23 [40.4%]), and medications prescribed via the EHR ordering system and in the formal list, but not described in the clinical note (9 [15.8%]). Fortified antimicrobials represented the largest category for medication mismatch between modalities (17 of 70 [24.3%]). Nearly one-third of patients (17 [32.1%]) had at least 1 medication mismatch in their record.

Conclusions and Relevance

Almost 1 in 4 medications were mismatched between the progress note and formal medication list in the EHR. These findings suggest that EHR data should be checked for internal consistency before use in research.

This cross-sectional study investigates the concordance of the names of medications for microbial keratitis in the structured, formal electronic health record medication list and the text of clinicians’ progress notes.

Introduction

The accuracy of a medication list is paramount to continuity of care, patient safety, and sound research. Poor documentation of medications can result in polypharmacy, medication interactions, and even death.¹ The mandated conversion to electronic health records (EHRs) in 2014 was done so partly to increase this accuracy. However, the EHR captures medication data in multiple locations and does not always reliably capture prescribed medications or prevent errors.² The ability for clinicians to record medications in both structured and unstructured data formats in the EHR creates the possibility of discordant information existing within one record.³ For medication data, the structured data of the medication list result from clinicians prescribing medications electronically or reconciling prescribed medications in the EHR. Unstructured data that represent medication occur in the free-text fields of clinical progress notes. Medications are only one example of data that can be captured in multiple formats in the EHR, creating variability in data concordance.⁴

An important clinical tool to check medication accuracy is medication reconciliation. Medication reconciliation has been mandated since 2005. One purpose of medication reconciliation is to provide patients with a clinician-approved postvisit summary.⁵ A prior investigation using EHR data for corneal diseases noted the discrepancies in medications listed in the formal medication list in the EHR and the clinical progress note.⁶ In other medical disciplines, studies have shown that the formal medication list in the EHR likely does not capture medications completely, accurately, and comparably across its sections.⁷ It was unclear the extent to which this error occurs with ophthalmic medications.

The purpose of this study was to explore discrepancies between medication documentation for microbial keratitis (MK), an acute ophthalmic disease with frequent medication changes. Topical medications can effectively treat MK and manage ocular sequelae, but medications can disrupt the corneal surface. Therefore, clinicians carefully consider medication choice and dosing. Many medications are prescribed for MK, and doses are changed frequently. Clinicians and patients need to clearly communicate the name and dose of medications prescribed at each visit. Verbal communication is performed, but written communication generated from formal medication lists in the EHR are provided to patients in the form of postvisit medication summaries. These postvisit summaries must be accurate to ensure patients’ safe and appropriate use of medications to treat MK.

Methods

Medication data were collected from the EHR-generated formal medication list and the text in the EHR of clinical progress notes of 53 participants with MK receiving treatment from July 1, 2015, to August 1, 2018. Patients were under the care of 1 of the 5 cornea specialists at the University of Michigan. The names and doses of medications to manage MK or ocular sequelae were collected, and data were reviewed by a cornea specialist (M.A.W.). The formal medication list was extracted from the EPIC EHR data warehouse of the University of Michigan Sight Outcomes Research Collaborative (SOURCE) repository over the same period of active treatment of MK. Excluded medications were nonpharmacy-prescribed medications (ie, over-the-counter) or medications prescribed for surgery (eg, intravenous hydration). The University of Michigan Medical School Institutional Review Board approved this study. This study adhered to the Declaration of Helsinki.⁸ Patients provided written consent for participation in the MK study.

Results

Fifty-three participants with MK were identified from July 1, 2015, to August 1, 2018: 24 men and 29 women with a mean (SD) age of 51.8 (19.6) years. Forty-six patients (86.8%) identified as white, 5 (9.4%) identified as black, 1 (1.9%) identified as Asian, and 1 (1.9%) identified as other race/ethnicity. After excluding over-the-counter medications (n = 15) and perisurgical medications (n = 63), a total of 247 medications were identified, including antimicrobials (fortified and nonfortified), corticosteroids, mydriasis eyedrops, and glaucoma eyedrops. Nearly one-third of patients (17 [32.1%]) had at least 1 medication mismatch in their record. More than three-quarters of medications (190 [76.9%]) could be reconciled between the progress note and formal medication list (Table). Among the 23.1% (n = 57) of differing medications, the causes included: medications not prescribed through the EHR ordering system (ie, telephone order) (25 [43.9%]), unreconciled medications (ie, prescribed by an outside clinician) (23 [40.4%]), and medications prescribed in the EHR, but not written in the clinical note (9 [15.8%]). The reason a clinician did not document a medication in the progress note could not be ascertained from the retrospective data. Fortified antimicrobials represented the largest medication category for disagreement (17 of 70 [24.3%]). Medication dosing documentation was inconsistent between modalities, but was not analyzed in detail.

Table. Medication Reconciliation.

Characteristic	Value
Unique patients, No.	53
Medications used by patients, total No.	325
Medications, excluding those used in OR and OTC medications, total No.	247
Matching medications, No./total No. (%)	190/247 (76.9)
Unmatched medications, No./total No. (%)	57/247 (23.1)
Reasons for unmatched medications, No./total No. (%)
Not prescribed through EHR system	25/57 (43.9)
Prescribed by an outside clinician, not reconciled in EHR	23/57 (40.4)
Prescribed through EHR, not documented in the progress note	9/57 (15.8)

Abbreviations: EHR, electronic health record; OR, operating room; OTC, over-the-counter.

Discussion

This study indicates that almost one-quarter of medications prescribed to patients for an acute ophthalmic condition requiring powerful medications and frequent changes of medications could not be reconciled between the clinicians’ management plans and the formal list in the EHR. This mismatch affected one-third of the patients in the study. This disconnect potentially places patients at risk for avoidable medication errors, such as the use of an inappropriate medication or adverse drug events, if they rely on postvisit summaries. These findings are consistent with other studies, including a study of ambulatory medications reporting that only 79% of medications were reconciled correctly between progress notes and the EHR.⁹ Treatment for MK requires frequent medication changes, compounding from specialty pharmacies, and telephone ordering after clinical hours, so it is unclear if results of this study should be applied to ophthalmic topical medications for chronic diseases such as glaucoma.

Double documentation of medications can be a substantial challenge for health care professionals.¹⁰ Clinicians managing acute MK likely focus on an accurate clinical notes and patient communication rather than updating the formal medication list in the EHR. Dose accuracy was not analyzed, but discrepancies occurred in unreported data. Qualitative and prospective work should be conducted in eye clinics to further understand specific causes of discordance in the context of the high documentation burden placed on clinicians. Reconciling medication names and updating medication dosing may represent 2 overlapping, but unique, problems for clinicians. Electronic health record developers should create software solutions to ease documentation burden and facilitate medication reconciliation. A promising potential strategy is natural language processing to capture and export medication names, doses, and directions from the free-text progress note.¹¹ If it is validated, natural language processing could be applied to help with clinical reconciliation or to function in large data sets to generate accurate information for research studies.

Limitations

In large EHR-based registries, researchers should analyze medication data with caution. Extrapolating our results to large databases with greater patient, clinician, and EHR system diversity and variability may not be appropriate. Our study was conducted at a tertiary care, academic facility with a specific group of clinicians, so the results may have site-specific causes. However, research in other medical fields does corroborate the occurrence of medication mismatch between EHR sections.¹²

Conclusions

This study found that almost 1 in 4 medications were mismatched between the progress note and formal medication list in the EHR. Researchers should validate the medication accuracy in their sample or estimate the percentage error of medications before including medications in large analyses.