Medication reconciliation is a complex, but necessary, process to prevent patient harm from medication discrepancies. Locally, the steps of medication reconciliation are completed consistently; however, medication errors still occur, which suggest process inaccuracies. We focused on removal of unnecessary medications as a proxy for accuracy. The primary aim was to increase the percentage of patients admitted to the pediatric hospital medicine service with at least 1 medication removed from the home medication list by 10% during the hospital stay by June of 2022.
Using the Model for Improvement, a multidisciplinary team was formed at a children’s hospital, a survey was completed, and multiple Plan–Do–Study–Act cycles were done focusing on: 1. simplifying electronic health record processes by making it easier to remove medications; 2. continuous resident education about the electronic health record processes to improve efficiency and address knowledge gaps; and 3. auditing charts and real-time feedback. Data were monitored with statistical process control charts.
The project exceeded the goal, improving from 35% to 48% of patients having at least 1 medication removed from their home medication list. Improvement has sustained for 12 months.
The combination of interventions including simplifying workflow, improving education, and enhancing accountability resulted in more patients with medications removed from their home medication list.
Medication discrepancies can cause significant harm to patients. Transitions of care, especially hospital admissions and discharges, are critically important because studies show 29% to 70% of patients have a discrepancy, and many of these were deemed clinically significant.1–6 Because of the need to prevent adverse drug events, medication reconciliation is a Joint Commission National Patient Safety Goal.7 Completing all of the steps of medication reconciliation decreases harm by correcting discrepancies.8–11 Though the electronic health record (EHR) provided access to previous medication prescriptions within the system, we identified issues with the home medication list often containing duplicate medications, old medications, or inaccurate doses or frequencies through chart review and reported safety events. Because the home medication list stays with the patient from visit to visit and is not always updated, this perpetuates these medication errors throughout the patient’s medical journey.
Medication reconciliation is defined as the process of creating the most accurate list of all the medications a patient is taking and comparing that to the list of ordered medications.12 It is a multiple-step process with collection of the home medication list from the patient/family, updating the medical record, making medical decisions about medications, ordering those medications, and getting the list of medications back to the patient (Fig 1). Though patients are given a list of their home medications with every visit, accuracy is even more important with hospitalizations where every medication needed must be ordered rather than in the ambulatory setting. As with many other systems, our hospital has been consistently completing the medication reconciliation process within the EHR within 24 hours of admission for years (at 97% or above); however, there continued to be inaccuracies in the home medication list and errors in ordering during the hospital stay.13–15
Completing admission medication reconciliation within our EHR involves checking a box that the home medication list has been reviewed and then choosing what to do about each medication during admission from the options of “Order, Don’t order, Replace, Stop Taking.” If a patient has no home medications, simply checking the box that the list has been reviewed completes medication reconciliation. These steps can be done without accurate revisions to the home medication list. Discharge medication reconciliation is a required element before a patient is able to be discharged from the hospital, so it is always done. This is defined as clicking an icon for each home medication with the options of “Continue, Change/New Prescription, or Stop” and each inpatient medication with the options of “Prescribe or Do Not Prescribe.” Updates to the home medication list do not happen as often as they should because of a variety of reasons including time, logistics within the EHR, incomplete information at the time of the visit, and medical knowledge gaps about medications prescribed by others. In pediatric hospitals, the patient is often not able to give an accurate medication history and the parent/guardian is not always present on admission. Reviewing the chart is time-consuming, and it is not always clear why medications have been prescribed or pulled into the chart from other sources. Some physicians and nurse practitioners (NPs) report discomfort removing something that is not prescribed by themselves or their own specialty. Though clinical pharmacists or pharmacy technicians are used in many systems, they are not available for patients on the pediatric hospital medicine (PHM) service within our system.9,16 Much of the literature has focused on adult hospital systems, including the Multicenter Medication Reconciliation Quality Improvement Study MARQUIS1 and then MARQUIS2 with development of toolkits and use of quality improvement methodology, though many of the concepts apply to pediatric hospitals, as well.10,11 We instituted improved processes, as well as education, as validated in other hospital systems to achieve our primary and secondary goals.10,11,17–19
The multidisciplinary team surveyed pediatric residents and developed a key driver diagram (Fig 2) to understand the biggest barriers to accurate medication reconciliation. We also obtained qualitative feedback from the other members of our medication reconciliation committee. Tracking the true accuracy of the details of the home medication list is difficult, requiring intensive chart review and discussion with families and staff. Thus, we focused our primary metric on tracking removal of medications from the home list as a proxy for ensuring accuracy of the home medication list. We felt that removing medications was a sign that the ordering physician or advanced practice provider took time to focus on the accuracy of the list rather than just quickly clicking through boxes. Knowledge gaps both about general medication reconciliation concepts, as well as the process within the EHR, were recognized as areas to focus. The primary measure aim was to increase the percentage of patients admitted to the PHM service with at least 1 medication removed from the home medication list by 10% while in the hospital. We recognized that many patients already had an accurate medication list at admission, and many physicians and NPs were already doing this, which is why a 10% increase was chosen rather than choosing an absolute percentage or average number of medications to be removed. All patients admitted to the hospital medicine service at our primary site were included and medications removed at any point in their admission were included. Our secondary aim was to increase the average number of medications removed per patient by 10% by June of 2021, which was extended to June of 2022 when we felt we could improve even further.
Methods
Context
This was done at a freestanding teaching quaternary care children’s hospital, focusing on patients on the acute care units on the PHM service. Care is provided primarily by 4 resident teams, with some patients also cared for by 1 NP team with 1 to 2 NPs and 1 attending. Resident rotations are 4 weeks in duration, with 1 to 2 senior residents, 2 to 4 interns, and 1 attending per team, with census varying widely because of seasonality and the coronavirus disease 2019 pandemic, with a range of 128 to 672 (average 400) patients discharged per month. Epic has been the inpatient EHR since 2012 and has been the EHR throughout the organization for multiple years, with different start times for different areas. The medication reconciliation process involves multiple steps with history-taking, updating the EHR, medical decision-making, placing orders, and giving the list to the family (Fig 1). Clinical pharmacy/pharmacy technicians are not available to help with this process.
A core team was formed, consisting of the PHM attending who was chairperson of the hospital’s medication reconciliation committee, multiple pediatric residents, and a pharmacologist. This core team worked closely with the rest of the medication reconciliation committee. This committee is composed of pharmacists, nurses, physicians, an advanced practice provider, an EHR support representative, and a data analyst. Using the Model for Improvement as the quality improvement methodology, a survey was completed to assess the biggest barriers to completing accurate medication reconciliation for pediatric residents because they are the team members doing most of the medication reconciliation work.20 A Pareto chart compiled the results, demonstrating that time was the biggest barrier, followed by issues related to getting a medication history from the family, and EHR issues, with medical knowledge only rarely being reported. Multiple Plan–Do–Study–Act (PDSA) cycles were performed, focusing on simplifying the EHR workflow, education about the processes within the EHR to improve efficiency and understanding, addressing knowledge gaps about the rationale for medication reconciliation, and auditing charts to provide feedback. The team met monthly to review data on completion of admission medication reconciliation, completion of discharge medication reconciliation, number of patients with at least 1 medication removed, and average number of medications removed per patient, as well as discuss interventions and feedback. Balancing measures were also monitored monthly with previously established chart audits by clinical pharmacology and safety event review. The Standards for QUality Improvement Reporting Excellence 2.0 guidelines were used to describe our project.21
PDSA #1: Improving EHR Usability With Simplification
Before this project, when a medication was removed from the home medication list, >30 options were available as reasons for why the medication was discontinued from the home medication list. The first intervention was to reduce these options to 2, either “Stop Taking (On after-visit summary [AVS]),” which would populate the AVS discharge instructions, or “Remove (Not on AVS),” which would take the medication off the home medication list but would not show on the AVS (Fig 3). The “Remove” option was used mainly for old medications or duplicates so that the medication instructions are clear for families. Simplifying the options made it more efficient to remove medications while having clear instructions for the family. This usability improvement was made by the hospital informatics team and the medical staff was informed of the change by e-mail.
Screenshot of EHR modification reducing the number of options for discontinuation reason.
Screenshot of EHR modification reducing the number of options for discontinuation reason.
PDSA #2: Addressing Knowledge Gaps
The second intervention focused on addressing knowledge gaps surrounding the medication reconciliation definition and the processes within the EHR. We felt that this would enable us to address many of the barriers identified in the survey, including time, medication history from families, and EHR issues. This was done in PDSA ramps. Education was done verbally with slides to define medication reconciliation, demonstrate what the expected workflow should be with hands-on demonstration, and what needed to be done within the EHR and functionality within the EHR to make the process more efficient, including how to sort by pharmaceutical class and adding the column to patient lists (PDSA #3) when that was available. Education was initially trialed with 1 of the inpatient resident teams by the resident members of the improvement team. Feedback was obtained, it was modified on the basis of the residents’ suggestions for improvement, and then continued every 4 weeks when new residents began their rotation. Printed examples were left in the resident team workroom. This continued for over a year until PDSA #4 was well established. After this was established as effective, similar education was done with the PHM division for both attending physicians and NPs at a division meeting and by e-mail. Education allowed for discussion to address concerns about medications prescribed by others and the ability to answer specific questions from individual physicians and NPs. Medication reconciliation was incorporated into the new pediatric intern boot camp education each year, highlighting many of the same concepts in a session that allowed the interns to practice within the EHR, which has continued yearly. This was followed by a yearly intern noon conference session with role-playing of a patient scenario with the thought process that the intern should be using to reinforce concepts after interns had started seeing patients. Brief medication reconciliation reminders were done at a few morning report sessions to keep this as a quality focus and remind residents who to reach out to with educational questions.
PDSA #3: Further EHR Usability With a Visual Cue
Functionality within the EHR was further enhanced by adding a column to patient lists so that physicians/NPs had the ability to see when outside medications were available to be reconciled into the local chart. Though this already showed in the admission navigator, this new visual cue made it easier for all members of the team to address outside medications, providing more opportunities for improving accuracy. This was incorporated into the education that was given to improve utilization of this functionality.
PDSA #4: Chart Audits for Real-Time Feedback
The last major intervention was doing real-time chart audits once within every 4-week rotation, and was done for 15 months until the project formally stopped in June of 2022. A member of the improvement team reviewed all the patient charts on the PHM service to look at the home medication list to look for opportunities to make the home medication list more accurate. Specific audit criteria included looking to see which medications were duplicated and which medications were not ordered for admission but remained on the home medication list, and then sent an e-mail to the senior residents/NPs and the attending for each team about the potential need to remove these medications from the list. The e-mail also included a reminder of how to do that with screenshots, the time and clicks saved when medications were removed, and who to contact if questions. This was initially done by the chair of the medication reconciliation committee and then transitioned to the resident members of the team after it was established that the audit criteria was well defined and the feedback was well received. This was initially sent to the senior resident or NP and attending caring for the patient at the time of the e-mail who would be able to act upon it before the patient was discharged. The audits generated discussion and allowed for clarification about details with specific people. This was expanded to send the audit to the admitting (often overnight) team, as well as all the current team members caring for the patient to reach a greater audience. This greater audience allowed for changes for both the patients currently in the hospital, as well as changes to how physicians and NPs approached medication reconciliation in the future. This replaced the education done in PDSA cycle #2 once it was well established.
Nursing PDSA Cycle
The majority of the work of the project focused on interventions related to ordering physicians and NPs, though some work was also done with the admission, discharge and transfer nurses to help with flagging prescriptions for removal and removing patient-reported medications from the home medication list. Engagement with a representative happened first, followed by education with this small group of nurses and tracking of medications they were able to flag for removal were trialed. Because of nursing staffing shortages and competing priorities, this was not expanded during this project.
Medication Reconciliation Committee Engagement
Engagement with the larger medication reconciliation committee was used for feedback and tracking of balancing measures and aligning with any related safety concerns.
Measures
The primary process measure chosen for this project was the percentage of patients on the PHM service at the primary campus with at least 1 medication removed from their home medication list at any time during their hospital stay, typically done at admission or at discharge. All patients were included in the data, regardless of if medications were listed on their home medication list upon admission. This information was obtained through a clinical data analyst’s report of the home medication list within the EHR. This was chosen as a proxy for the outcome measure of an accurate home medication list because there is no electronic method to assess accuracy for such a large volume of patients. The secondary process measure chosen was the average number of medications per patient removed from the home medication list for patients on the PHM service at the Akron campus.
The balancing measure for this project used a chart audit system that was functional before the start of this project. These audits were performed by pharmacologists on roughly half of the patient charts that were on the resident team at the primary campus. The audits were performed every weekday, looking for missing medications, dose adjustments, and medication questions. The medication reconciliation committee reviewed all safety events reported within our system related to all types of medication reconciliation and monitored for events related to removal of medications from the home medication list. We also obtained qualitative feedback during the education and chart audit interventions about how the process was working from residents and PHM faculty.
A P chart was used for the primary measure and a U chart for the secondary measure. Standard rules for defining a shift were 8 points above or below the mean.22
The institutional review board screened this project and determined it was exempt from formal review because it was not considered human subject research. There were no conflicts of interest.
Results
The baseline percentage of patients with at least 1 medication removed from their home medication list during their admission to the PHM service before beginning the project was 35%. This improved to 48% by the end of this project, exceeding our goal of a 10% improvement (Fig 4). There were 2 significant shifts in the primary process measure during the project time frame. Monthly education sessions were associated with the first improved shift and feedback audits were associated with the second shift. Multiple points outside the control limits show that the system is not highly reliable, with some variation possibly attributable to physicians that were part of this project being on the inpatient service for points above the control limits and patient volume and seasonality possibly contributing to points below. This improvement has been sustained for several months.
Control chart P chart of percentage of patients with at least 1 medication removed from the home medication list while admitted to the PHM service.
Control chart P chart of percentage of patients with at least 1 medication removed from the home medication list while admitted to the PHM service.
The secondary measure of the average number of medications removed from the home medication list per patient admitted to PHM also exceeded our goal of a 10% shift. The baseline average measurement of 0.8 medications removed per patient increased to 1.4 medications removed per patient at the end of the project (Fig 5). This was also sustained for several months.
Control chart U chart of average number of medications removed per patient from the home medication list while admitted to the PHM service.
Control chart U chart of average number of medications removed per patient from the home medication list while admitted to the PHM service.
Our balancing measure showed that there were no clinically significant changes in the number of medications missed or the amount of medication questions that were asked with the medication chart audits done by the pharmacology team as part of a separate initiative with the medication reconciliation committee. No safety events were reported involving medications that were removed from the home medication list during hospital admission.
Discussion
This quality improvement initiative had 2 key aims: To increase the percentage of patients admitted to the PHM service with at least 1 medication removed from the home medication list by 10% and to increase the average number of medications removed per patient by 10%. We surpassed our goal for both aims, resulting in almost half of our patients having a medication removed from their medication list and removing an average of 1.4 medications per patient by the end of the project because of the combination of interventions. Removing old or duplicated medications from the home medication list is important for accurate ordering of medications while in the hospital, and then for the family to accurately administer the correct medications after discharge. Having only the medications that the patient is actually taking results in a home medication list that is concise and easily understood. Because there is little literature published about improving medication reconciliation in pediatric hospital settings, this demonstrates success in working with pediatric patient families, utilizing education strategies similar to those developed for adult hospitals in combination with EHR changes and feedback.12,23
Assessment of the actual home medication list accuracy would be the optimal measure to assess proper utilization of medicine reconciliation. However, this was not feasible for the resources and scope of this project, which is why measuring removal of medications was used as a proxy that could be obtained by electronic data analysis. It is the authors’ belief that removal of medications requires the physician/NP to make an active decision about the medication list, and thus signifies a more accurate home medication list, which suggests better attention to all details of the list. Our hospital already had a high rate of completing the steps of medication reconciliation, so this was not able to be used in this project. Knowing that many patients have an accurate medication list before coming to the hospital and many physicians and NPs already were removing medications from the home medication list influenced the goal for the project, with a 10% increase as the goal which we were able to surpass. Having a more accurate home medication list helps decrease harm to patients because medication discrepancies are known to cause medication errors.4,6,14,24
Although clinical pharmacy input has been shown to be beneficial, it is not available in our clinical care model to assist with medicine reconciliation. The lack of clinical pharmacy input to aid in medication reconciliation at our hospital meant that the physicians and NPs were the focus of our process improvements. On the basis of qualitative feedback, empowering physicians/NPs with further information about all the EHR details about medication reconciliation made them more willing to remove medications that they had not prescribed themselves, which is necessary with shared home medication lists in the EHR. Qualitative feedback received revealed that this knowledge gap had previously made physicians/NPs hesitant because many did not understand how removing medications would impact others seeing the EHR, especially when it came to reconciling medications from outside of our local EHR. Many of the usability tips were helpful in empowering physicians/NPs to update the home medication list more accurately because they improved efficiency. Electronic medication reconciliation tools need to be designed to be usable, and education made the simplification within our system more accepted.19 If clinical pharmacy input was available, some of this work could be transitioned to them because it has been effective in other projects and training and responsibilities for physicians/NPs would be lessened.9
Using education as a primary agent of change may be considered 1 of the limitations of this study. Education is historically not considered to be as reliable as other methods for impacting change; however, addressing knowledge gaps is part of the medication reconciliation toolkits that are available from multiple sources.12,23 Education enabled the physicians and NPs to report more comfort removing medications they had not ordered, and this led to culture change where physicians and NPs became more active in the process of medication reconciliation. Other similar studies support the use of education as a primary intervention to decrease medication discrepancies.25,26 Though education required more time up front and is not always reliable, it has resulted in a culture shift in our institution that enabled sustained improvement for several months after the formal conclusion of the project. Because education addressed the multifactorial issues related to medication reconciliation, we believe that it was integral to the success of the project with improving workflows, understanding, and buy-in, and the education included many of the points addressed in the MARQUIS study and did not find any major difficulties using similar methods in the pediatric setting compared with what has been done in adult hospitals.10 Given that multiple types of health care workers can do some of the steps of medication reconciliation, hospitals with clinical pharmacy may have different workflows to be effective in updating the home medication list. Because of staffing constraints, nursing-related interventions were not felt to be impactful with this project. As mentioned above, using removal of at least 1 medication for the primary measure may be considered a limitation because it is a proxy, and some patients may not have any medications to remove but it demonstrates attention to accurately update the home medication list. The final limitation is that this study was performed at 1 institution and resources to perform medication reconciliation are different in every institution.
Conclusions
The combination of the interventions of simplifying the workflow process, enhanced process education, and physician and NP accountability resulted in more patients having medications removed from their home medication list, simplifying the list, and likely leading to improvements in the accuracy of the home medication list. These same concepts can be used in other settings where medication reconciliation is used.
Acknowledgment
We thank Akron Children’s Hospital’s Project to Publication coach Dr Prabi Rajbhandari for her invaluable guidance and support throughout the manuscript writing process.
Dr Jamerino-Thrush’s current affiliation is Division of Emergency Medicine, Children’s Hospital of Michigan, Detroit, MI.
Deidentified individual participant data will not be made available.
Dr Gunkelman conceptualized and designed the project, led data collection, analysis, and interpretation, drafted the initial manuscript, and reviewed and revised the manuscript; Drs Jamerino-Thrush, Genet, and Blackford contributed to the conceptualization and design of the project, contributed to data analysis and interpretation, and critically reviewed and revised the manuscript; Drs Jones and Bigham contributed to the design of the project, contributed to data analysis and interpretation, and critically reviewed and revised the manuscript; and all authors 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