Medication errors are common during transitions of care, such as discharge from the emergency department (ED) or urgent care (UC). The Joint Commission has identified medication reconciliation as a key safety practice. Our aim was to increase the percentage of patients with completed medication reconciliation at discharge from our pediatric ED and 4 UCs from 25% to 75% in 12 months.
Key stakeholders included ED and UC physicians and nurses, informatics, and quality management. The baseline process for medication reconciliation was mapped and modified to create a standard process for nurses and physicians. An Ishikawa diagram was created to assess potential failures. Electronic health record interventions included adapting an inpatient workflow and using a clinical decision support tool. Educational interventions included just-in-time training, physician education via division meeting presentations, video tutorials, and physician-specific and group feedback using funnel plots. The secondary process measure was the proportion of patients discharged from the ED and UCs with completed home medication nursing review. We used statistical process control to analyze changes in measures over time.
In the UCs, home medication nursing review increased from 91% to 98% and medication reconciliation increased from 35% to 82% within 4 months. In the ED, home medication nursing review increased from 2% to 83% within 8 months and medication reconciliation increased from 26% to 64% within 18 months.
We successfully increased the proportion of UC and ED discharged patients with completed medication reconciliation.
Medication errors are preventable events that can cause harm to patients.1 Reported medication prescribing and dosing errors range from 10% to 34% in pediatric emergency department (ED) and inpatient settings. Many factors impact the ability to prescribe medications correctly in the pediatric ED. Factors associated with medication errors included prescribing medications during overnight hours, using weight-based dosing, prescriptions written by trainees, and interruptions in workflow.2 –5
Medication reconciliation is the process of obtaining a comprehensive and accurate list of a patient’s medications, supplements, and over-the-counter medications. As part of this process, a clinician reviews the medications a patient reports taking, reconciles them with the new medications that are ordered for the patient, and resolves any discrepancies.6,7 Medication reconciliation is recommended during transitions of care, such as from the ambulatory setting to hospital admission to hospital discharge.6 Implementing medication reconciliation has been shown to reduce medication errors.8 –12 Chiewchantanakit et al found a 75% lower risk of medication error in patients who received medication reconciliation in all transitions of care than in those who had not.9 Similarly, Vira et al found that implementation of medication reconciliation resulted in the interception of 75% of inaccuracies in prescribed medications before hospital discharge. In 2005, The Joint Commission identified medication reconciliation as a key safety practice and made it a National Patient Safety Goal.7 However, there are many challenges in completing medication reconciliation, including a patient’s ability to provide the correct information, patients’ knowledge of their medications, provider workflow, and lack of patient health record integration with their medical home.6,7 Given these obstacles, medication reconciliation continues to be a challenge to implement in many clinical settings.
Our institution identified cases where medication errors could have caused significant harm to patients (eg, incorrect antiepileptic drug dosing), which prompted a hospital-wide initiative to improve medication reconciliation. We identified the need to implement a standardized medication reconciliation process in our pediatric ED. Our aim was to increase the overall percentage of patients with completed medication reconciliation at discharge from our pediatric ED and 4 urgent cares (UC) from 25% to 75% in 12 months.
Methods
Context and Setting
This study was conducted in the Rady Children’s Hospital San Diego ED and 4 UC centers, which serve greater than 100 000 children per year. Rady Children’s Hospital San Diego ED is the only freestanding pediatric ED serving San Diego, Imperial, and South Riverside counties. It is part of a health network composed of 220 primary care and 370 specialty physicians, serving 90% of pediatric patients in these counties.
Planning
A multidisciplinary quality improvement (QI) team composed of ED and UC physicians, nurses, informaticists, and quality leadership was assembled. This study was part of a hospital-wide initiative to improve medication reconciliation in inpatient and outpatient settings. We used process mapping (Fig 1) to identify the existing medication reconciliation process, which consisted of 2 critical steps: (1) A nursing (registered nurse)-led step of reviewing a patient’s existing home medications with the patient and family and documenting this information in the electronic health record (EHR) and (2) a physician (Doctor of Medicine)-led step of reconciling the existing home medications and any new medications prescribed during the visit. In the existing overall medication reconciliation process, there was a single process for the home medication nursing review, but 4 different EHR processes for physician medication reconciliation. Additionally, the home medication nursing review was not visible to physicians completing medication reconciliation. An Ishikawa diagram was created to assess potential barriers (Fig 2). We identified that there was confusion regarding which provider (nursing or physician) was responsible for completing home medication review and medication reconciliation, lack of awareness about the importance of medication reconciliation, lack of a standard process to reconcile medications, and an inability for physicians to view the home medication review.
Baseline process map (a) and standardized process map (b) for ED and UC.
Ishikawa diagram with factors that may lead to incomplete or inaccurate medication reconciliation.
Ishikawa diagram with factors that may lead to incomplete or inaccurate medication reconciliation.
Based on these barriers, a key driver diagram was created to guide interventions and was updated over the course of the initiative (Fig 3). The quality improvement (QI) team reviewed data from the preceding 3 months to establish a baseline for nursing home medication review and medication reconciliation. We found that the percentages of medication reconciliation completed for patients being discharged were 26% for the ED and 35% for the UC. After project initiation, weekly data were collected prospectively and reviewed by the QI team.
Interventions
Our interventions were implemented over 23 months using Plan-Do-Study-Act cycles. There were 3 categories of interventions: (1) EHR optimization, (2) education and training, and (3) provider specific feedback.
EHR Optimization
In September 2020, a standardized EHR workflow, which was adapted from an existing inpatient EHR workflow for discharge medication reconciliation, was designed for the ED and UC. This novel process allowed for a single EHR access point to reconcile medications, whereas previously there were 4 possible access points to accomplish the same outcome. This workflow also linked the data from the home medication nursing review to the physicians, a process that was previously disconnected. The initial screen of the workflow showed the home medication nursing review process. This included names of medications, doses, frequency, and routes, in addition to if the patient was taking the medication, not taking the medication, taking as needed, not taking as prescribed, or unknown. This screen also allowed for nursing staff to input comments regarding potential discrepancies in how patients were taking medications. The second and final screen of the workflow showed a summary of the home medication review by category of taking, not taking, or not taking as prescribed, including changes made by nursing staff. It also allowed the physician to continue, discontinue, or change the medications. In June 2021, based on physician feedback, the order of the screens was switched such that physicians could take action on medications immediately upon entry to the workflow and view the home medication review details entered by nursing if needed. To track data, an automated structured query language-based report was developed to determine whether home medication review and medication reconciliation had been completed via navigating through the workflow screens. In November 2020, 2 months into the initiative, a Best Practice Advisory (BPA) alert, a clinical decision support (CDS) tool, was created to display for both nursing and physicians upon entry to the chart when home medication review and reconciliation were incomplete. Based on feedback, the physician medication reconciliation BPA alert was relocated within the chart to the disposition section in January 2022 such that physicians could click on a hyperlink, leading them to the standardized medication reconciliation workflow. In December 2021, based on UC physician feedback, the BPA was removed only from the UC setting.
Education and Training
In September 2020, all physicians were educated about the standard medication reconciliation process via e-mail and presentation at division meetings. The physicians received guidance on how to access the new workflow, view the nursing home medication review, and reconcile medications in the EHR. Nursing champions introduced a rounding process, involving just-in-time home medication review training. In November 2020, several interventions took place, including a medication reconciliation tutorial video and tip sheet for physicians, laminated reminders on each computer used by nurses and physicians, and an EHR banner on the ED track board listing weekly nursing and physician average completion percentages. In June 2021, when the order of the physician workflow screens was switched, the tip sheet and video tutorial were updated. In January 2022, just-in-time medication reconciliation training was introduced for all residents, fellows, and new attending physicians.
Physician-Specific Feedback
In December 2020, we initiated physician specific feedback, where each fellow and attending physician received a personalized e-mail with their monthly individual average percentage of completed medication reconciliation in addition to an anonymized, comparative display of group physician performance via a funnel plot. This was continued for 3 subsequent months and was then transitioned to group feedback because of stable improvements. Group feedback differed from individual feedback in that each physician received an e-mail addressed to all faculty members that displayed the group funnel plot but without individual percentages reported. This was transitioned back to individual provider feedback in November 2021 because of a decrease in medication reconciliation completion, which was then sent out quarterly afterward. Additionally, we introduced targeted feedback to lower performing physicians via a second personalized e-mail in April 2022. In November 2021, the top 10 performers of medication reconciliation were offered maintenance of certification small group points and a small financial incentive at the end of the year.
Measures
The primary outcome measure was the proportion of patients discharged from the ED and UCs with completed physician medication reconciliation, determined by the physical EHR-linked medication list printed out at the time of discharge. Our process measure was the proportion of patients discharged from the ED and UCs with completed home medication nursing review. A structured query language based report was developed to track whether medication reconciliation and home medication review was completed for a given patient based on navigation through the standardized workflow. Our balancing measure was the average time required to complete the home medication review and medication reconciliation. Patients included in this study were patients being discharged from the EDs and UCs that had existing medications upon arrival and patients with no medications upon arrival that were discharged with medications. Patients who were excluded were patients being admitted or transferred and patients who both presented with no existing medications and were discharged with no medications. During the study period, the average number of patients discharged per day from the ED was 140 and from the UC was 89.
Analysis
Our study was granted a QI exemption by our institution’s institutional review board. Data analysis was conducted via statistical process control charts created in Microsoft Excel with QI Macros. Shifts in the centerline were created when sustained special cause variation was established, which was defined by 8 consecutive points above or below the centerline. Special cause variation in the form of single data points outside of the control limits were investigated for possible causes.13
Results
In the 4 UCs, home medication nursing review increased from 91% to 98% within 4 months of project implementation with sustained improvement for 19 months after the last centerline shift and 21 months after the last intervention (Fig 4). UC medication reconciliation increased from 35% to 87% with 3 centerline shifts within the first 4 months of the initiative (Fig 5). This increase was sustained for 15 months from the last shift and 7 months from the last intervention. After removal of the BPA in December 2021 in the UCs, there was an increase in special cause variation with single data points outside of the control limits and a downward shift to 82%, which was sustained for 4 months.
In the ED, home medication nursing review increased from 2% to 84% (Fig 6) within 8 months of project implementation with sustained improvement 13 months after the last shift. ED medication reconciliation increased from 25% to 52% within 5 months, with 3 shifts within the first 4 months, followed by a downward shift in July 2021 likely secondary to increases in ED volume attributed to the corona virus disease 2019 pandemic, after which the centerline was sustained at 48% until November 2021 (Fig 7). Since November 2021, the emergency department had 2 upward shifts to a medication reconciliation percentage of 64% sustained for 5 months. The average time required for nurses to complete home medication review was 5 minutes (range 0.5–7 minutes) and the average time required for physicians to complete medication reconciliation was 2.5 minutes (range 0.5–5 minutes).
Discussion
Using QI methodology, we increased the total percentage of completed medication reconciliation in the ED from 26% to 64% and from 35% to 82% in the UCs. Interventions were still implemented after our 12 month goal up until 23 months because there was still potential for improvement, particularly in the ED. The UCs began with home medication nursing review completion above target, which is likely because of having patients with fewer baseline home medications. Additionally, the task of home medication review was assigned to a single nurse for a patient, which may facilitate completion of the task; in the ED, multiple nurses may be assigned to a patient. The UC physicians advocated to remove the BPA in December 2021 because of the significant sustained improvement; however, this resulted in a downward shift from 87% to 82% but remained above our target of 75%. During September and October 2021, the UCs saw a larger volume of patients than at any other time and thus had lower home medication review at that time. The ED setting started with home medication nursing review at 2%. This low baseline was largely because of nursing staff verbally reviewing medications visualized in the EHR workflow screen without physically selecting the button in the EHR that would reflect completed home medication review. ED home medication review saw the largest increase in completion and reached above the target of 75%.
In our study, EHR CDS had the largest effect on medication reconciliation completion in both the ED and UCs. By introducing the BPA CDS tool, ED home medication nursing review and medication reconciliation for both ED and UC saw significant increases in completion of their respective processes with shifts in all 3 measures. The largest shift from implementing the BPA was in ED home medication nursing review, where there was an initial increase from 4% to 66%. ED medication reconciliation had an additional shift from 55% to 59% after the BPA was moved to the disposition screen, the area of the chart that most physicians interact with at discharge. Challenges with CDS tools include alert fatigue, where providers are exposed to a large volume of CDS tools and either ignore or become accustomed to seeing them, which dampens the effect of the alert over time.14 The BPA was made functional in January 2022 by adding a hyperlink to the medication reconciliation workflow. Adding functionality to the BPA beyond acting solely as a reminder is a more effective use of CDS. Furthermore, CDS tools are more successful when aligned with other motivational interventions and when evaluating the effect of the CDS tool regularly.15 We did this by incorporating other interventions simultaneously and by evaluating the impact on our aim after implementation of the BPA. Our improvement in medication reconciliation supports other studies that demonstrate CDS as an effective tool in improving prescribing practices.16 –18
Another effective intervention was physician-specific feedback. For both measures in the ED, provider-specific feedback was associated with shifts in the centerline toward improved completion of both home medication nursing review and medication reconciliation. This feedback started a conversation regarding barriers and challenges to completing medication reconciliation. Audit and feedback has been a prevalent and effective strategy in changing clinical practice in QI initiatives, including in emergency department settings.19 –22 Feedback is most effective when given by a senior colleague in a verbal and written format with specific delineated goals.19 In our study, ED physician specific feedback was originally provided monthly with individual data, starting in December 2020 and was continued for 3 months. At this point, because of steady improvement, the feedback was then provided monthly without individual data, showing group performance for 3 months. In July 2021, this feedback was transitioned to quarterly for 12 months because of increasing demands on physicians because of increasing ED volumes and staffing shortages related to the corona virus disease 2019 pandemic. During this time of less frequent and less individual feedback, 2 downward shifts were seen in completion of mediation reconciliation. This trend is supported by studies that show that feedback can be optimized by providing frequent (monthly or more frequent) and individual performance data.23
Provider education was another intervention associated with improvement in both home medication review and medication reconciliation, though to a lesser degree when compared with the BPA and physician specific feedback. Education interventions are a lower reliability process when compared with reminders (such as the BPA).24 Both nurse rounding, which involved bedside instruction of nursing staff, and introduction of just-in-time training for physicians were associated with a shift, from 4% to 66% for ED home medication nursing review and a shift from 55% to 59% for ED medication reconciliation. Just-in-time training has been an emerging teaching method in medicine and has been demonstrated to be valuable in teaching for QI initiatives.25 Various educational tools were used to maximize success for providers that might respond to different learning formats. Education included demonstrating the EHR workflow in meetings, providing tip sheets, laminated reminders, and a video tutorial. These educational interventions were more essential in the ED compared with the UC. In the ED, a single patient may have multiple nurses and physicians (including trainees) in a single encounter, which adds to the difficulty of completing medication reconciliation given that more people require education for the medication reconciliation of a single patient. This differs from the UC setting where a single nurse and physician are assigned to a patient and roles are more clearly delineated.
Standardization of medication reconciliation has been shown to improve patient safety.11 Though anecdotally reported at our institution, initiation of standardized medication reconciliation resulted in good catches that otherwise would have gone unnoticed and had risk for potential harm to a patient, particularly for behavioral health patients and patients with known seizures. Before this initiative, there was no standardized workflow for ED medication reconciliation at our institution. This was the first and essential step to assess and measure improvement in medication reconciliation. There are limited studies regarding medication reconciliation in ED settings, and the existing studies in EDs that demonstrate successful medication reconciliation largely cite using pharmacist-led interventions.26 –29 Our study illustrates a medication reconciliation process in a setting without ED-specific pharmacy resources. This process relies on obtaining history from patients and families who do not always know exact doses or medications, which is challenging in a fast-paced setting where there are interruptions in workflow.30 Thus, our defined processes could be applied to other institutions that lack pharmacists or pharmacy staff in the ED, which can be challenging with patients who have longer lists of medications.
There were several limitations to our study. Near misses and safety reports related to medication reconciliation were tracked during the study and were found to be too low frequency of an event to draw meaningful conclusion, as they are reporter dependent. The quality of medication reconciliation was also difficult to examine. For example, physicians who were unsure of how to address a medication and deferred to discuss with a patient’s subspecialist did not get credit for completed medication reconciliation. Despite a scenario such as this being a “good faith effort” consistent with guidelines set by the Joint Commission, there was no automatable way to give completion credit for this with the current EHR infrastructure.7 Another limitation is that our current workflow is not generalizable to institutions not using the same EHR (Epic). Yet, this process of standardization could be applied to institutions without an EHR, as home medication review and medication reconciliation could be completed with paper charts as well.
Further studies are needed to assess the consequence of these interventions on patient outcomes and to assess the effect of medication reconciliation on higher risk ED populations (behavioral health, patients with high-risk medications, eg, antiepileptic medications, cardiac medications, patients with large volume medication lists).
Conclusions
Using QI methodology, we successfully increased the proportion of patients discharged from the ED and UCs with completed medication reconciliation, from 26% to 64% within 18 months and from 35% to 82% within 4 months, respectively. This initiative is ongoing with the aim to reach our goal in the ED and sustain improvement in the UC centers.
Dr Sheth conceptualized the study, designed quality improvement interventions, validated data, supervised data collection, analyzed the data, and drafted the initial manuscript; Dr Bryl conceptualized the study, designed quality improvement interventions, and supervised data analysis; Dr Bialostozky designed the data collection instrument, designed quality improvement interventions, collected data, and validated data; Drs Mishra, Billman, Selah, and Langley, and Ms Santiago and Ms Heitzman conceptualized the study and designed quality improvement interventions; Dr Hollenbach supervised data analysis; Mr O’Crump designed the data collection instrument and collected data; and all authors critically reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no conflicts of interest relevant to this article to disclose.
Comments