Improving Emergency Medicine Clinician Awareness of Prehospital-Administered Medications

Abstract Background/problem: Information transfer between emergency medical services (EMS) and emergency medicine (EM) is at high risk for omissions and errors. EM awareness of prehospital medication administration affects patient management and medication error. In April 2020, we surveyed emergency physicians and emergency department nurse practitioners (NPs) and physician assistants (PAs) regarding the EMS handoff process. Emergency physicians and NPs/PAs endorsed knowing what medications were given, or having received direct verbal handoff from EMS “Often” or “Always” only 20% of the time (n = 71), identifying a need to improve the written handoff process. To assess rates of medication error due to lack of awareness of prehospital administered medications, we measured glucocorticoid redosing in the emergency department (ED) following prehospital dexamethasone administration. In 2020, glucocorticoids were redosed 30% of the time, and our aim was to reduce glucocorticoid redosing to 10% by June 2022. Intervention: We developed and implemented a system innovation where prehospital-administered medications documented in a nursing flowsheet during verbal handoff are pulled directly into the triage note where they are more likely to be reviewed by receiving EM clinicians. Results: Shewhart p-charts were used to evaluate for statistical process change in the process measure of triage note documentation of prehospital medication administration and the outcome measure of glucocorticoid redosing. While the frequency of prehospital dexamethasone administration in the triage note increased, no statistical process change outcome measure of glucocorticoid redosing was observed. However, on repeat survey of EM clinicians in July 2022, 50% now indicated they were aware of prehospital medication administration “Often” or “Always” (n = 61, p = 0.003), 87% maintained they use the triage note as the main source of information regarding prehospital medication administration, and 81% “Always” review the triage note. Conclusions: Innovations that improve accessibility of written documentation of prehospital medication administration were associated with improved subjective assessment of EM clinician awareness of prehospital medications, but not the outcome measure of medication error. Effective error reduction likely requires better system integration between prehospital and EM records.


Introduction
Paramedics are responsible for bringing a significant number of patients to the emergency department (ED).During the prehospital phase of care, paramedics perform procedures and administer medications to treat emergent symptoms and conditions.It is important that emergency medicine (EM) clinicians are aware of prehospital medications administered, as these affect patient management, and lack of awareness may increase risk of medication error (1, 2).Previous studies have demonstrated that information transfer between emergency medical services (EMS) clinicians and ED clinicians regarding prehospital interventions is generally poor and at high risk of omissions and errors (2)(3)(4)(5).
The majority of studies of the transition of care from EMS to ED have focused predominantly on the verbal handoff process (3)(4)(5)(6)(7)(8)(9).However, previous work has also demonstrated that handoff between EMS clinicians and physicians is rare in all but the highest acuity patients, suggesting the importance of timely written handoff tools (1, 10).In a survey study of EM clinicians in Australia, only 50% took the EMS chart into consideration during their patient evaluations, in part due to the lack of timely availability of this information (10).
In our own EMS and hospital system, we identified a significant opportunity for improvement in written handoffs and EM clinician awareness of prehospital-administered medications.To assess the outcome of medication error due to lack of awareness of prehospital-administered medications, we used glucocorticoid redosing following prehospital dexamethasone administration as a quality measure.Under standing orders/protocol, paramedics administer dexamethasone for the treatment of asthma exacerbations, chronic obstructive pulmonary disease exacerbations, and allergic reactions.Dexamethasone is a long-acting glucocorticoid and the half-life of dexamethasone typically precludes redosing of this drug during the ED phase of care (11).Thus, glucocorticoid administration during the ED encounter following prehospital dexamethasone administration is representative of a medication error secondary to lack of EM clinician awareness of prehospital medication administration.This report describes the evolution and outcomes of a quality improvement initiative aimed at improving EM clinician awareness of prehospital administered medications and the subsequent effect on medication errors.

Setting
This project was performed in Rochester, NY and included patients transported by EMS to two health care system hospitals, including a large, tertiary care hospital with 118,000 annual ED visits and an urban community hospital with 51,000 annual ED visits.The hospitals use EPIC as the electronic medical record.The quality initiative included patients transported by 12 agencies who provide advanced life support and operate within the Monroe-Livingston Regional EMS system.Due to hospital crowding and lack of ED bed availability, verbal handoff of EMS patients occurs at EMS triage between the triage nurse and EMS clinician in the majority of instances (Table 1, question 4).Bedside verbal handoff between EMS clinicians and EM clinicians does routinely occur following prehospital activations for trauma, stroke, and myocardial infarction, and for other critically ill patients

Initiative
The initiative entailed development of a system modification to the triage documentation process.An EPIC smartlink was included in the triage note so that any prehospital medications recorded by the ED triage nurse in the triage navigator were automatically displayed in the body of the triage note to facilitate review by EM clinicians.Prior to the initiative, documentation of prehospital medications within the triage note was at the discretion of the triage nurse, and was documented inconsistently.The initiative was developed through a collaborative effort of EM clinicians, ED pharmacists, ED nursing, and physician leadership with IT support.

Study Design
The study design was a quality improvement initiative applying the Model for Improvement as a framework and this report is structured using the Revised Standards for Quality Improvement Reporting Excellence Guidelines (12,13).This project was undertaken as a quality initiative, and per the University of Rochester's guidelines for determining human subject research, it did not meet the definition of research according to 45CFR46.

Data Collection and Quality Measures
Initial data on the scope of the problem was collected via survey of EM clinicians (physicians and NPs/PAs) using a five-point Likert scale distributed in April 2021 by departmental e-mail, which assessed different components of the EMS-ED handoff process (Table 1).The survey was developed by regional EMS medical directors, EM resident physicians, and EMS clinicians who were interested in exploring opportunities to improve the EMS-to-ED handoff process.Survey results were collected in RedCap (14,15) and analyzed using Microsoft Excel.
Given initial survey results, the decision was made to specifically focus on EM awareness of prehospital-administered medications.To more objectively assess medication error secondary to EM clinician lack of awareness of prehospital medication administration, we used the outcome measure of glucocorticoid redosing following prehospital administration of dexamethasone (Table 2).Patients were included (quality measure denominator) who were transported by EMS agencies within the Monroe-Livingston region to participating hospitals and received prehospital dexamethasone.Patients were excluded if they left the emergency department prior to evaluation by EM clinicians, or the triage note indicated only that the patient was taken directly to the critical care bay, where presumably verbal handoff took place.All ages were included.The corresponding ED encounter for each included patient was matched by search of patient name, date of birth, and date of service.Following matching, the ED encounter record was reviewed for whether the patient received any glucocorticoid during the ED encounter.Documentation of the prehospital dexamethasone administration within the triage note for these patients was also collected as a process measure.

Data Analyses
Survey responses were analyzed using descriptive statistics.Before and after survey responses were analyzed using chi-square tests.Data on process and quality measures were tabulated in Microsoft Excel.QI Macros was used to construct Shewhart p-charts to evaluate for statistical process change in both the process measure of documentation of prehospital dexamethasone administration within triage note and the outcome measure of glucocorticoid redosing.

Problem Definition and Baseline Performance
To determine the scope of the local problem, we surveyed EM clinicians within our own health care system about their experiences with the EMS-ED handoff process (Table 1).Following a single e-mail invite, we received 71 survey responses, yielding a 33% response rate from an e-mail distribution list of 111 faculty, 48 NPs/PAs and 56 residents, not all of whom work clinically, as the authors had access to a single listserv for both clinical and non-clinical faculty.Of the respondents, 51 (23 attending physicians, 20 resident physicians, 8 NPs/PAs) self-reported working exclusively at the tertiary care hospital, 17 (9 attending physicians and 8 NPs/PAs) exclusively at the community hospital, and 3 attending physicians work at both hospitals.
Only 20% of responding clinicians reported knowing what medications were given to their patients by EMS "Often" or "Always."However, medications administered by EMS were considered to be amongst the most important information to be passed on at handoff by 96% of survey respondents, and 92% of survey respondents indicated that knowing prehospital medications administered to ED patients directly affects their care "Often" or "Always."Direct verbal handoff from EMS was reported to occur "Often" and "Always" only 18% and 1% of the time respectively, leading us to focus on improving written handoffs within our local system.Baseline performance on the outcome measure of glucocorticoid redosing was measured retrospectively using data from 2020.We found that glucocorticoids were re-dosed 30% (n ¼ 449) of the time in 2020.Given the identification of these system weaknesses around EM clinician awareness of prehospital medications administered, a regional quality improvement project to improve this process was initiated using glucocorticoid redosing as the primary outcome measure, with an aim to reduce glucocorticoid redosing to 10% by June 2022.

Intervention Development and Implementation
Three change ideas were developed and tested during the course of the improvement project.The first change idea was the development of a short, written hand-off tool for EMS clinicians.This written handoff tool was developed through initial knowledge-building Plan-Do-Study-Act cycles involving a small group of EMS clinicians, where initial feedback indicated possibility of adoption by (1) making it small enough to fit in the monitor, (2) simple enough with checkboxes containing common medications, and (3) printed on brightly colored adherent pads so that the paper could be stuck to the patient's ED paper chart, where patient stickers, EKGs, and other prehospital documentation (e.g.transfer paperwork, advance directives) are kept.Prototypes were printed and distributed at two EMS agencies where feedback was nearly universally negative and the tool rarely used in practice.The predominant critique of the tool was that it duplicated the work of the EMS clinician.
Following failure of adoption of the written handoff tool, one team member spent time observing the verbal handoff process between EMS and the ED triage nurse.It was noted that nearly universally, EMS clinicians included prehospital-administered medications in their verbal report.These medications are often recorded within the ED triage nurse navigator portion of the electronic medical record as part of a prehospital flowsheet.However, prior to the intervention, these data were not included in any physician or NP/PA workflows and often overlooked, as it was at the far bottom of the EPIC triage tab.A retrospective chart review of a sample of charts from January 2020 to April 2021 was performed (n ¼ 262).We found that prehospital administration of dexamethasone was recorded variably within the medical record [triage note only (12%), flowsheet only (30%), both the triage note and flowsheet (41%), and neither location (18%)].We also found that documentation in the triage note was associated with a decrease in the probability of redosing compared to the prehospital flow sheet alone (9% vs. 21%, p ¼ 0.006).The improvement team therefore met with nursing and physician ED leadership and brainstormed ways to include prehospital medications as part of the triage note without increasing triage time or increasing ED triage nurse tasks.An EPIC smartlink was added to the triage note so that any prehospital medications recorded by the ED triage nurse in the triage navigator is automatically displayed in the body of the triage note, which is regularly reviewed by the treating physician or NP/PA (Figure 1).This change was rolled out by nursing leadership who provided training and guidance that the prehospital medication flowsheet should be completed prior to the triage note.The change went live on November 4, 2021.
Between January 2020 and June 2022, 1232 patients who received prehospital dexamethasone by a regional EMS agency were transported to an included health system hospital.One hundred eleven were excluded as they were documented to go directly to the critical care bay.Eight hundred (71%) of the patients were transported to the tertiary care hospital, and 321 (29%) were transported to the community hospital.The median patient age was 56 years (range 2-95 years).The triage note change implementation was associated with statistical process change with significant improvement in triage documentation of prehospital dexamethasone administration, from 62% prior to 91% after smartlink implementation (Figure 2(A)).However, there was no statistical process change in the outcome measure of glucocorticoid redosing (Figure 2(B)).Charts were reviewed to determine if one subset of EM clinicians was more likely to redose than another (e.g.off-service resident, EM resident, EM attending, EM NP/PA).Redosing errors were distributed equally among all groups.Education about the quality improvement initiative was provided to EM attendings and EM residents during faculty meeting and resident conference in March 2022, with no further statistical process change in the outcome measure.EM clinicians were surveyed to assess whether there was an improvement in their subjective awareness of prehospital administered medications.Sixty-one EM attending physicians, residents, and NPs/PAs answered the post-survey administered in July 2022: 50% endorsed knowing what prehospital medications were administered to their patients "Often" or "Always," which represented a significant improvement from baseline (chi-square ¼ 18.05, p-value ¼ 0.001) (Figure 3).Improvement was reported at both the tertiary care hospital (from 16% to 44%) and the community hospital (from 35% to 83%).Of note, 87% of survey respondents maintained they use the triage note as the main source of information regarding prehospital medication administration, and 81% "Always" review the triage note prior to seeing the patient (supplemental Table 1).

Discussion
Improving the transfer of care process between EMS clinicians and receiving facilities is a critical component of improving patient safety.The majority of current literature focuses on improving the process of verbal handoff (4,(6)(7)(8).However, as noted in a 2014 joint position statement of major EMS and EM organizations, verbal handoff alone may lead to inaccurate and incomplete transfer of information to subsequent treating clinicians at receiving facilities who were not present at the time of verbal communication (1).In addition, while previous work focused on missing information as part of the handoff process or in subsequent documentation, it did not document the effects of failure of information transfer on patient care.
In this quality improvement project, we sought to improve a single but important aspect of the transfer of care process: EM clinician awareness of prehospital-administered medications, which was subjectively poor at the beginning of our project.By measuring redosing of glucocorticoids in the ED following prehospital dexamethasone administration, we were able to demonstrate a significant burden of medication error due to weaknesses in the transfer of care process.While we are unable to universally assume glucocorticoid redosing in the ED to be inappropriate and acknowledge the  existence of situations in which glucocorticoid redosing in the emergency department is clinically warranted (16), it is a fitting outcome measure given the once daily dosing for most indications, particularly those included in EMS protocol.
Prior work has advocated for a handoff tool in written or electronic form (1). We found that a handwritten handoff tool was poorly accepted despite co-development with EMS clinicians.Our key intervention of automated incorporation of verbally conveyed prehospital administered medications into the triage note was associated with more than doubling of self-reported EM clinician awareness of prehospitaladministered medications, with the percentage of EM clinicians reporting that they knew what prehospital medications were administered "Often" or "Always" increasing from 20 to 50%.However, there was no change in the outcome measure of medication error as measured by glucocorticoid redosing.This was particularly interesting in that the overwhelming majority of repeat survey respondents indicated that they reviewed the triage note prior to making patient care decisions.This highlights the importance of a metric that measures errors that reach the patient (e.g.redosing) rather than subjective improvement alone.The survey responses suggesting that triage notes were reviewed prior to patient care suggest that the medication error of redosing may not be uniquely due to failure of information transfer, but also how this information is integrated into EM clinician workflow.The persistence of repeat dosing despite the increased availability of EMS medication administration data further highlights the importance of system integration between prehospital and hospital documentation, so that inhospital safety mechanisms, such as computer-based and pharmacist safety checks, are further engaged to enhance patient safety.

Conclusions
Information transfer between EMS and EM clinicians is at high risk for omissions and errors.We identified a system need to improve written handoff regarding prehospital medication administration.Using glucocorticoid-redosing following prehospital dexamethasone administration as a measure, we identified a high rate of baseline medication error due to failure of information transfer.Implementation of an innovation that improved written documentation of prehospital medication administration was associated with improved subjective assessment of EM clinician awareness of prehospital medications, but not the outcome measure of medication error.Effective error reduction will likely require better system integration between prehospital and EM records.

Figure 1 .
Figure 1.Sample triage note with addition of the Prehospital-Administered Medications EPIC V R Smartlink.The Smartlink portion of the note is highlighted with a dotted line.

Figure 2 .
Figure 2. Process control charts for project quality measures (A) Proportion of patients over time who received repeat dose of glucocorticoids.(B) Proportion of patints over time where prehospital administration of dexamethasone was documented in ED triage note.

Table 1 .
Initial survey of EM clinician perception of the EMS to emergency department handoff process.

Table 2 .
Quality measures and associated operational definitions for EM-clinician awareness of prehospital-administered medications.