In the ongoing effort to improve pharmacy processes, medication-use evaluation (MUE) is a valuable tool for ensuring and improving medication-use processes.¹ MUE is also commonly referred to as drug-use review, drug-use evaluation, target drug, and drug-use program.^1-5 Ideally, MUE is a multi-faceted, multi-disciplinary, evidence-based, standardized quality improvement (QI) approach to review the medication-use process to understand how an intervention (for example, a medication or group of medications) is being used and to continually improve patient care.

MUE is performed to attain several outcome- or process-oriented key objectives^1,3,6:

• Analyze the medication-use process (prescribing, procurement, transcribing/documenting, dispensing, administering, and monitoring)
• Determine how a medication is being used in a specific clinical situation or across patient populations
• Evaluate the effectiveness of several different medications in a particular setting
• Investigate adverse events or medication errors as part of a root-cause analysis
• Evaluate the direct and/or indirect cost impact of different medication-use practices from outcome and/or resource perspectives

More practically, MUEs can serve to evaluate medication-use processes for safety and efficiency, fine tune the formulary, identify off-label uses of medications, measure prescribing compliance with published clinical guidelines or external quality measures, or as an outcome research tool.^1,3,5-7 Developing an effective MUE requires seven systematic steps^1,3,6:

1. Define Responsibilities

First, a guidance team serves to provide organizational authority for conducting the MUE, from gaining buy-in for performing the MUE to implementing the necessary changes gleaned from the findings. Typically, the guidance team includes the pharmacy team or project leader, the clinical coordinator, and the operations leader and/or clinical specialist (who keeps the team on-task with activities and meeting deadlines). The guidance team may also include leaders from other disciplines, such as medicine or nursing, who are stakeholders in the change. Formal, authoritative support for the MUE from administration and clinical staff is critical to success.

Secondly, the MUE project team is responsible for MUE logistics, including establishing the metrics, identifying data sources, and analyzing the findings. This team should be multidisciplinary, with clearly defined roles (particularly for data collection and analysis), with clear expectations for regular meetings and discussions of findings. Pharmacists typically take a lead role in carrying out the MUE process. Organizational support can assist in focusing the team and addressing challenges or barriers encountered during the process.

2. Establish Scope, Objective(s), and QI Indicators

The goal of performing an MUE may be to evaluate a process improvement related to medication use, address a safety concern or clinical challenge, or to measure compliance with an internal or external quality benchmark. Defining the scope of the MUE is an important step that involves the entire team; be sure all members are focused on solving or evaluating the specific challenge identified. For example, an MUE involving evaluation of patient counseling efforts could have the following purpose: To evaluate the percentage of patients receiving medication adherence counseling prior to hospital discharge; however, a better, more focused purpose would be: To evaluate the percentage of patients with congestive heart failure who received medication adherence counseling within 48 hours of hospital discharge. The second scope statement is more targeted and presents a specific measuring point.

Quality organizations and accrediting bodies, such as National Quality Forum, Pharmacy Quality Alliance, Centers for Medicare and Medicaid Services, and The Joint Commission, provide quality measures and a supporting evidence base that can be adapted for MUE development; sample lists of MUE objectives are available in the literature. While it is important to include a balance of quality measures in the MUE, if the number of measures is excessive, the MUE will become too complex. Should the MUE contain multiple objectives, consider breaking it into different phases, or assign smaller work groups within the multidisciplinary team to complete separate components. In this case, schedule regular information-sharing sessions to ensure the work groups are progressing toward a common goal. Metrics must be defined prior to beginning the MUE and may focus on process, economic, or clinical outcome measures, or a combination of these (see SIDEBAR). Early decisions establishing the scope, objectives, and metrics are critical to MUE design, and are likely to save both time and resources in the overall MUE process.⁶

3. Specify the MUE Design

As a QI tool, MUE is most easily carried out using the Plan-Do-Check (Study)-Act (PDCA) methodology.8 This will ensure careful planning of purpose and method (plan), responsibility for carrying out the MUE and data gathering (do), team-based discussion of findings and implications (check), and review of recommendations for change and forward action (act). The MUE can focus on a single patient care or pharmacy practice area, or it can be completed across several areas. Selecting the correct patient population is important to make sure the data set is as complete as possible. To minimize data variation and confounding factors, the team should determine inclusion and exclusion criteria. Determine the length of time for data collection, as well as the targeted number of samples required. Unlike a clinical trial, where sample size is specifically calculated, MUE is less rigid, but typically data collection spans a 6-month to 1-year period and involves anywhere from a few (<50) to a large number (hundreds) of patients, depending on the data source and MUE objective.

The MUE can be completed retrospectively or can be carried out in a concurrent or even prospective manner. Retrospective data review is limited by what was documented in the past, whereas concurrent or prospective data collection helps ensure all relevant data elements are gathered. The MUE can evaluate use of a single medication, several medications, or an overall approach to management of a disease or process.

4. Determine Data Source and Data Management Plan

As part of the MUE planning process, it is critical to determine the best source of data for the measurement indictors and data variables that will be collected. Patient records, claims databases, charge or usage databases, decision-support data (eg, CPOE), or analytics programs (eg, smart pumps) are all potential sources. Data also may be gathered via patient or employee interview or survey. Ideally, data is gathered electronically, but this is not always possible.

Note an important caution concerning patient information contained in medical records: As the lines between QI efforts (such as MUE) and research continue to blur,9-11 MUE data collection is becoming systematic (whether retrospective or prospective) with statistical analysis; thus, patient data confidentiality must be protected at all costs. Most institutions have quality oversight committees that adjudicate the need for research board approval for QI MUE projects. The data sources and facility-specific approval processes for the MUE should be discussed prior to initiating data collection.

Finally, the statistical analysis approach should be specified for the MUE, including the method for addressing missing data, the choice of test for comparisons, models to use for confounding variables, specifying statistical significance, and reporting results.⁶ Experienced clinicians or statisticians can be quite useful to the MUE team in assessing results.

5. Collect Data

Those who are collecting the data should record it into a database, ideally with answers to yes/no questions (1=yes and 0=no) for easy summation. To minimize variation, limit the number of team members assigned to data collection and provide training on where to find the data. When using multiple team members to gather data, consider starting with each data collector evaluating the same three to five patients as a validation step and then discuss the findings together, before the full data collection process begins.

In the final analysis, there may be a need to combine collected data variables into an end point. Data measured can come from a combination of sources, so the person(s) with the data collation responsibility will need to be familiar with the various databases and ensure consistent data input from the collection tools. Collection should not disrupt usual patient care or health care processes, particularly if data are being collected prospectively or involve patient interviews.

The data collection step can be time-intensive, but it is critical to ensure that the MUE effects change.

6. Analyze Findings

Once data are gathered, analysis can begin. Experience with statistics and data analysis is preferred, as this will allow the team to review the initial findings and interpret the results. The data should be evaluated from both the statistical perspective and from a clinical viewpoint.6 The PDCA process calls for team-based analysis of findings, brainstorming implications of data, and discussion of barriers. In addition to reviewing the results, the team is advised to discuss the implications of change, identify what additional input and perspectives are necessary, and consider the need for cycles to test process change. Open discussion of the timetable for change, establishing which changes will occur and when, developing communication and education methods, and ongoing tracking of results are critical to ensure the MUE is useful.

7. Plan for Improvements

Challenges with MUE often include a lack of organization, poor communication, failure to follow through, and incomplete implementation.¹ To countermand this, the team must dedicate time to discussing not only the data findings, but also any process changes that can be efficiently implemented and an associated timetable.

If a change in clinical practice is warranted, clinician education is critical. This education should be focused and efficient, and any new guidelines must be evidence-based, easy to use, and readily accessible. Be sure to consider the increased role of technology in health care as well as the blended roles of the health care team, when developing education. Online educational content, computer reminders, and reference cards may be useful. Education should be multidisciplinary, targeting all key disciplines.

Small cycles of change, such as pilots, are encouraged, prior to implementing a large-scale change. The team should plan to meet post-implementation at regular intervals to discuss challenges and successes, review the process to maintain the gains realized, and expand upon results.

Conclusion

MUE is an important QI tool for evaluating the safe, appropriate use of medications, health care processes, and emerging research ideas (eg, off-label use of medications). Using a systematic, evidence-based, data-driven, multidisciplinary approach, the MUE can be an efficient strategy for determining current practices and setting a course for improvement. Successful MUEs require careful definition of roles, defined metrics, a data management plan, meaningful interpretation of findings, education supporting change, and methods for measuring and maintaining gains.

References

1. Phillips MS, Gayman JE, Todd MW. ASHP guidelines on medication-use evaluation. Am J Health Syst Pharm. 1996;53(16):1953-1955.
2. Tyler LS, Cole SW, May JR. et al. ASHP guidelines on the pharmacy and therapeutics committee and the formulary system. Am J Health Syst Pharm. 2008;65(13):1272-1283.
3. Fanikos J, Jenkins KL, Piazza G, et al. Medication use evaluation: Pharmacist rubric for performance improvement. Pharmacotherapy. 2014;(34 suppl 1):5S-13S.
4. Skledar SJ, Hess MM. Implementation of a drug-use and disease-state management program. Am J Health-Syst Pharm. 2000;(57 suppl 4):S23-S29.
5. Schiller DS. Medication use evaluations as a research tool. Pharmacotherapy. 2014;(34 suppl 1):3S-4S.
6. Faley B, Fanikos J. Best practices for medication use evaluations in postsurgical pain management. Curr Emerg Hosp Med Rep. 2017;5(1):33-40.
7. Vermeulen LC, Beis SJ, Cano SB. Applying outcomes research in improving the medication-use process. Am J Health Syst Pharm. 2000;57(24):2277-2282.
8. Plan-Do-Check-Act Cycle. Agency for Healthcare Research and Quality. https://healthit.ahrq.gov/health-it-tools-and-resources/evaluation-resources/workflow-assessment-health-it-toolkit/all-workflow-tools/plan-do-check-act-cycle. Accessed November 2, 2017.
9. Finkelstein JA, Brickman AL, Capron A, et al. Oversight on the borderline: Quality improvement and pragmatic research. Clin Trials. 2015;12(5):457-466.
10. Bellin E, Dubler NN. The quality improvement-research divide and the need for external oversight. Am J Public Health. 2001;91(9):1512-1517.
11. Casarett D, Karlawish JH, Sugarman J. Determining when quality improvement initiatives should be considered research: Proposed criteria and potential implications. J Am Med Assoc. 2000;283(17):2275-2280.

Susan Skledar, RPh, MPH, FASHP, is Director of Experiential Learning and Continuing Professional Development, and Professor, Department of P&T, at the University of Pittsburgh School of Pharmacy. She received her Bachelor’s and Master’s of Public Health degrees from the University of Pittsburgh. Sue served as Director of the Drug Use and Disease State Management Program for UPMC Presbyterian Shadyside from 1996-2012, and Formulary Management and Drug Use Policy Clinical Specialist across the UPMC health system network of hospitals from 2012-2016. Her professional interests include implementation science, health care quality metrics, protocol design, and medication safety.