Journal of Gerontological Nursing

Geropharmacology 

Nursing Home Medication Reconciliation: A Quality Improvement Initiative

Monica Tong; Hye Young Oh; Jennifer Thomas, PharmD; Sheila Patel, PharmD; Jennifer L. Hardesty, PharmD, FASCP; Nicole J. Brandt, PharmD, MBA, BCGP, BCPP, FASCP

Abstract

The current quality improvement initiative evaluated the medication reconciliation process within select nursing homes in Washington, DC. The identification of common types of medication discrepancies through monthly retrospective chart reviews of newly admitted patients in two different nursing homes were described. The use of high-risk medications, namely antidiabetic, anticoagulant, and opioid agents, was also recorded. A standardized spreadsheet tool based on multiple medication reconciliation implementation tool kits was created to record the information. The five most common medication discrepancies were incorrect indication (21%), no monitoring parameters (17%), medication name omitted (11%), incorrect dose (10%), and incorrect frequency (8%). Antidiabetic agents in both sites were the most used high-risk medication. This initiative highlights that medication discrepancies on admission are common in nursing homes and may be clinically impactful. More attention needs to be given to work flow processes to improve medication reconciliation considering the increased risk for adverse drug events and hospitalizations. [Journal of Gerontological Nursing and Mental Health Services, 43(4), 9–14.]

Abstract

The current quality improvement initiative evaluated the medication reconciliation process within select nursing homes in Washington, DC. The identification of common types of medication discrepancies through monthly retrospective chart reviews of newly admitted patients in two different nursing homes were described. The use of high-risk medications, namely antidiabetic, anticoagulant, and opioid agents, was also recorded. A standardized spreadsheet tool based on multiple medication reconciliation implementation tool kits was created to record the information. The five most common medication discrepancies were incorrect indication (21%), no monitoring parameters (17%), medication name omitted (11%), incorrect dose (10%), and incorrect frequency (8%). Antidiabetic agents in both sites were the most used high-risk medication. This initiative highlights that medication discrepancies on admission are common in nursing homes and may be clinically impactful. More attention needs to be given to work flow processes to improve medication reconciliation considering the increased risk for adverse drug events and hospitalizations. [Journal of Gerontological Nursing and Mental Health Services, 43(4), 9–14.]

Medication reconciliation refers to:

…the process of creating a complete and accurate list of a patient's medication regimen at each transition of care point by comparing the regimen prior to admission to the new setting of care to the orders at the new setting of care and determining the correct medications at the transitions point. (Institute  for  Healthcare  Improvement, 2017, para. 2) 

The process comprises the following steps: collection and verification of the patient's medication history from various sources; clarification to ensure the medications, doses, and instructions are appropriate; reconciliation and documentation of changes in the orders; and the creation of a new complete and accurate medication list for the patient. Appropriate medication reconciliation across the continuum of care can reduce medication errors, adverse drug events (ADEs) related to the discrepancies, unnecessary hospitalizations, and hospital readmissions. The Joint Commission (2017) introduced medication reconciliation as a National Patient Safety Goal in 2006 (Agency for Healthcare Research and Quality, 2015), and in 2017, it is National Patient Safety Goal 3: Improve the Safety of Using Medications for nursing care centers/long-term care (LTC) settings.

Older adults access the health care system across various settings due to multiple acute or chronic comorbid conditions, polypharmacy, lack of support from family members, and increased rates of cognitive decline (Cumbler, Carter, & Kutner, 2008). Therefore, older adults are likely to experience multiple care transitions across the spectrum of care and are at risk for ADEs due to incomplete transfer of information. A randomized retrospective chart review of electronic health records (EHRs) of older adults across three transition care points (i.e., hospital admission, hospital discharge to a skilled nursing facility [SNF], and SNF to a nursing home) found medication discrepancies between hospital discharge and SNF admission increased from 10.7 to 12.6, and then decreased between the SNF and nursing home from 12.6 to 8.9 (Sinvani et al., 2013). Furthermore, a cross-sectional study found approximately 21.3% (n = 495/2,319 medications) reviewed on admission to a SNF had a medication discrepancy and occurred most commonly for cardiovascular, opioid analgesic, neuropsychiatric, hypoglycemic, antibiotic, and anticoagulant agents (Tija et al., 2009). For types of medication discrepancies noted, dose and route of administration were frequently omitted. Another study that focused on one care transition (i.e., between the hospital and nursing home) examined the mean numbers of medications altered during the transfer and ADEs (Boockvar et al., 2004). The findings noted that on average there were 1.4 medication changes from hospital to nursing home and 3.1 from nursing home to hospital. Approximately 14 ADEs occurred due to medication changes, 12 of which occurred in the nursing home after nursing home readmission (Boockvar et al., 2004). These studies highlight that medication reconciliation can be improved across the spectrum of care to help decrease ADEs.

A cross-sectional analysis over 3 years in North Carolina nursing homes found that approximately 11% of errors involved a patient transitioning into a nursing home, and these errors had a higher odds of patient harm compared to errors not involved in transitions (odds ratio = 1.85; 95% confidence interval [1.3–2.63]) (Desai, Williams, Greene, Pierson, & Hansen, 2011). Common errors occurred in staff communication, order transcription, medication availability, pharmacy issues, and name confusion. As most of the studies are descriptive about the rate of discrepancies or classes of medications involved and not the types of discrepancies found, the current study reports common types of discrepancies due to medication reconciliation through monthly retrospective chart reviews of the nursing home admission medication reconciliation. Through the identification of common types of discrepancies, the quality improvement (QI) initiative proposes solutions to address the common discrepancies to improve the medication reconciliation process.

Method

Setting

Two nursing homes within Washington, DC volunteered to participate in this QI initiative (i.e., Site A and Site B). Site A has 230 SNF beds and 200 LTC beds. Charts were reviewed from February 1 to June 30, 2016, and were mainly paper charts but nursing notes were included in an EHR. Site B has 360 SNF beds and 240 LTC beds. Charts were reviewed in April 2016, June 2016, and July 2016, using an EHR.

Intervention

The Capital DC Care Coordination and Medication Safety Coalition started a QI initiative focused on improving the medication reconciliation workflow process in Washington, DC. Members of this coalition included hospitals, nursing homes, hospices, home health agencies, pharmacies, providers, and the QI organization. This portion of the initiative examined the medication reconciliation process from the nursing home admission perspective. The QI initiative was a retrospective chart review of at least 10 patient admission records per month from two different nursing facilities in Washington, DC, over a period of at least 3 months. The evaluation of these data provided preliminary findings of the current status of the medication reconciliation process and types of discrepancies.

Prior to starting the medication reconciliation project, the nursing home's medication reconciliation process was described via a workflow diagram (Figure 1). The purpose of the diagram was to determine the flow of medication reconciliation, key players, and have students become familiar with the process. At both sites, the flow of medication was similar, in which the nurse practitioner was the key player in performing medication reconciliation. Resident records were selected for review if the resident was admitted to the nursing home during the month of the retrospective review.


            Medication reconciliation workflow process example for Site A.

Figure 1.

Medication reconciliation workflow process example for Site A.

Pharmacy students (M.T., H.Y.O.) in their third year of training were selected to review the medication records due to their advanced training. They performed the reviews under the supervision of the consultant pharmacist (S.P.), and any issues were communicated to the consultant pharmacist and pertinent individuals at the facility, such as the nursing unit managers and Director of Nursing. The information was stored in a secure cloud-based site with no patient identifiers. Any paper documents were shredded or kept locked in the facility.

The standardized Microsoft Excel® spreadsheet tool is based on multiple medication reconciliation implementation tool kits, including Johns Hopkins Medical Center and the Arizona Quality Improvement Organization Health Services Advisory Group. The tool documents the patient's demographic data (non-identifiable) and discrepancies in the medication reconciliation. Within the tool, the patient is de-identified using numbers and letters. Demographic information collected included age, race, gender, number of medical conditions, number of medications, and behavioral health issues, if any. The standardized tool documents discrepancies by comparing the hospital discharge list to the Medication Administration Record (MAR) and, when applicable, previous medication lists in patients' charts. At the end of each month, the tool was used to analyze the data by calculating the percentages of common discrepancies.

If major medication discrepancies were identified, the Director of Nursing, consultant pharmacist, and nurse unit manager were informed of the discrepancies, possible adverse effects to monitor, and management if necessary to ensure patients' safety.

Results

The Table summarizes the differences in data collected between the two sites. Overall, a total of 154 discrepancies were found (Figure 2). From the combined data, the five most common medication discrepancies were incorrect indication (21%), no monitoring parameters (17%), medication name omitted (11%), incorrect dose (10%), and incorrect frequency (8%). From examining the summary data, no conclusions can be drawn whether having a paper file or EHR made a difference in medication reconciliation.


            Site Overview of Findings

Table:

Site Overview of Findings


            Medication discrepancies identified (N = 154).

Figure 2.

Medication discrepancies identified (N = 154).

Indications must be listed to determine the need for the medication; nursing homes are required to have one indication listed for each medication. Common reasons for incorrect indication include: the incorrect indication for the medication was written in the chart, the class of the medication was noted (i.e., instead of the indication), or incorrect transcription. Examples of incorrect indication include buspirone (Buspar®) for dementia, allopurinol (Zyloprim®) for heart failure, and loratadine (Claritin®) for antihistamine. Potential sources that increase the rates of these discrepancies are the printed hospital discharge lists, which commonly do not list the indications for the medications, and a patient's medical history is listed in various sources or may be incomplete.

Regarding no monitoring parameters listed, the top classes were antihypertensive agents and pain medications. For instance, antihypertensive agent hold parameter for systolic blood pressure or heart rate was a common omission. For patients on multiple pain medications, such as acetaminophen and opioid drugs, the differentiation between severity of pain (i.e., mild, moderate, and severe) was reviewed for documentation. Differentiation of pain medication severity was the second most common monitoring omission.

Medication name omitted was the third most common discrepancy. This omission is different from missing medications because medication name omitted accounts for new admissions who were missing medications on the final MAR, whereas missing medications referred to readmission cases.

The final two most common discrepancies, which are actual errors, were incorrect dose and incorrect frequency, and these usually occurred during transcription. Major examples are included below.

  • Diltiazem (Cardizem®)—Hospital discharge list stated to administer diltiazem twice daily but due to a transcribing error, the patient received diltiazem three times daily for 5 days. The medication was discontinued because the hypertension regimen changed. Fortunately, the patient did not experience any major adverse events from this error.
  • Prednisone—Hospital discharge list stated to administer prednisone 10 mg for 3 days and then 5 mg for 3 days. Staff at the LTC facility transcribed it as prednisone 10 mg for 3 days. Patient was given 10 mg of prednisone for 4 days; therefore, the stop date was exceeded and wrong instructions for the medication were written.
  • Haloperidol (Haldol®)—Given around the clock instead of as-needed. Patient had only three medications, two of which were antipsychotic agents (i.e., haloperidol and quetiapine [Seroquel®]). Both antipsychotic medications were given around the clock. The doctor in charge of neuropsychiatric services was called and asked to review the medication list the next day.

For the other common medication discrepancies with rates >5%, the following examples are presented.

  • Missing medications in readmissions (7%)—A patient was admitted to the hospital for shortness of breath, but upon return to the nursing home, the chronic obstructive pulmonary disease inhalers were not restarted.
  • Stop date omitted (6%)—Usually included anticoagulant agents and occasionally antibiotic agents.
  • No indication (6%)—Occurred when no indication was listed on the MAR and was more prevalent at the nursing home with paper records (i.e., Site A).

Discussion

Medication reconciliation in nursing homes in Washington, DC can be improved, as approximately 64% to 90% of charts reviewed every month contained a medication discrepancy. The most common discrepancies were incorrect indication, no monitoring parameters, medication name omitted, incorrect dose, and incorrect frequency. Analysis of possible reasons for medication discrepancies yielded four strategies to improve the medication reconciliation process: Responsibilities, Standardization, Education, and Monitoring.

Responsibilities should be identified for each member of the health care team in the medication reconciliation process as detailed in the flow chart (Figure 1). For example, the responsibility of the nurse or consultant pharmacist is to perform the initial medication reconciliation, and the responsibility of the physician would be to verify the indications of the medications. The responsibility of medication reconciliation should be shared and communicated efficiently and effectively to reduce errors.

Standardization of the medication reconciliation process through an intake form or polices should be considered. The creation of a standardized intake form for medication reconciliation detailing categories that result in common medication discrepancies should be used. In addition, the form should include a section detailing reasons why certain medications were not continued. Currently, the study sites use a physician order sheet (POS) that contains multiple crucial fields to send to the pharmacy; this information is then transcribed on a MAR, which allows for more free text. Monitoring parameters are usually not carried over to the MAR and transcription errors may occur. Standardization of the MAR to contain crucial fields, such as the POS, may reduce transcribing errors and encourage the completion of crucial fields, such as monitoring parameters.

Education about the importance of medication reconciliation, real world examples of common medication discrepancies (e.g., haloperidol/dosing frequency/indication) coupled with the implications of these discrepancies, and tips on how to avoid common discrepancies should be implemented. Within new nurses' orientation and training, medication reconciliation should be a topic discussed and stressed. The discrepancy of incorrect indication could be reduced if nurses are educated on documentation of proper indications, possible sources for the indication, and encouraged to ask if they do not know. Presentations about common discrepancies and discussions on how to reduce them is another strategy to raise awareness and improve the medication reconciliation process.

Monitoring as a quality assurance and performance improvement measurement for each quarter should be considered to determine common medication discrepancies and their reasons and raise internal awareness about the importance of medication reconciliation. Monitoring could be done by a variety of key individuals, such as nurses or consultant pharmacists. In addition, the use of outside resources, such as pharmacy or nursing students on didactic learning experiences, could also be considered to aid monitoring.

It is also important to consider that medication reconciliation is impacted by the transition of care point before the nursing home, such as the hospital. Through the Capital DC Care Coordination and Medication Safety Coalition, key points were communicated with regard to how hospital staff could reduce medication discrepancies, such as listing indications and creating a standardized report of the final medication list in an easy to read format.

Limitations of the current study include small sample, limited duration, and variations between the two sites and reviewers (i.e., based on amount of experience reviewing charts for medication discrepancies). Future directions include the implementation of the four strategies (i.e., responsibilities, standardization, education, and monitoring) at the sites and evaluation of trends in the improvement of the medication reconciliation process.

Conclusion

On September 18, 2014, Congress passed the Improving Medicare Post-Acute Care Transformation (IMPACT) Act of 2014 (Centers for Medicare & Medicaid Services [CMS], 2017). The IMPACT Act requires, among other significant activities, the reporting of standardized patient assessment data with regard to quality measures, and patient assessment instrument categories. It further specifies that the data

…be standardized and interoperable so as to allow for the exchange of such data among such post-acute care providers and other providers and the use by such providers of such data that has been so exchanged, including by using common standards and definitions in order to provide access to longitudinal information for such providers to facilitate coordinated care and improved Medicare beneficiary outcomes… (CMS, 2017, para. 2) 

One area of focus is medication reconciliation in the nursing home setting. Health care teams must work collaboratively to evaluate their processes, especially considering the frequency in medication discrepancies that were found in the current QI initiative.

References

Site Overview of Findings

n (%)
Variable Site A Site B
Charts reviewed 50 30
Charts reviewed with no medication discrepancies 18 (36) 3 (10)
Medications 452 373
Medication discrepancies 89 65
High-risk medications 46 (10) 39 (10)
Authors

Ms. Tong and Ms. Young Oh are PharmD Candidates, Dr. Thomas is Clinical Assistant Professor, and Dr. Brandt is Professor, Geriatric Pharmacotherapy, Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore; Dr. Patel is Consultant Pharmacist, and Dr. Hardesty is Chief Clinical Officer and Corporate Compliance Officer, Remedi SeniorCare, Towson, Maryland. Dr. Brandt is also Executive Director, Peter Lamy Center on Drug Therapy and Aging, Baltimore, Maryland.

Dr. Patel and Dr. Hardesty are employees of Remedi SeniorCare. The remaining authors have disclosed no potential conflicts of interest, financial or otherwise.

Address correspondence to Nicole J. Brandt, PharmD, MBA, BCGP, BCPP, FASCP, Professor, Geriatric Pharmacotherapy, Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, and Executive Director, Peter Lamy Center on Drug Therapy and Aging, 20 North Pine Street N529, Baltimore, MD 21201; e-mail: nbrandt@rx.umaryland.edu.

10.3928/00989134-20170313-04

Sign up to receive

Journal E-contents