Enhancing the Quality of Prescribing Practices for Older Veterans Discharged from the Emergency Department (EQUiPPED): Preliminary Results from Enhancing Quality of Prescribing Practices for Older Veterans Discharged from the Emergency Department, a Novel Multicomponent Interdisciplinary Quality Improvement Initiative

Abstract

Suboptimal medication prescribing for older adults has been described in a number of emergency department (ED) studies. Despite this, few studies have examined ED-targeted interventions aimed at reducing the use of potentially inappropriate medications (PIMs). Enhancing Quality of Prescribing Practices for Older Veterans Discharged from the ED (EQUiPPED) is an ongoing multicomponent, interdisciplinary quality improvement initiative in eight Department of Veterans Affairs EDs. The project aims to decrease the use of PIMs, as identified by the Beers criteria, prescribed to veterans aged 65 and older at the time of ED discharge. Interventions include provider education; informatics-based clinical decision support with electronic medical record–embedded geriatric pharmacy order sets and links to online geriatric content; and individual provider education including academic detailing, audit and feedback, and peer benchmarking. Poisson regression was used to compare the number of PIMs that staff providers prescribed to veterans aged 65 and older discharged from the ED before and after the initiation of the EQUiPPED intervention. Initial data from the first implementation site show that the average monthly proportion of PIMs that staff providers prescribed was 9.4 ± 1.5% before the intervention and 4.6 ± 1.0% after the initiation of EQUiPPED (relative risk = 0.48, 95% confidence interval = 0.40–0.59, P < .001). Preliminary evaluation demonstrated a significant and sustained reduction of ED-prescribed PIMs in older veterans after implementation of EQUiPPED. Longer follow-up and replication at collaborating sites would allow for an assessment of the effect on health outcomes and costs.

An increasing number of older adults are evaluated in emergency departments (EDs) annually.¹ More than half are discharged without admission and receive a treatment plan, which frequently involves a new prescription medication.² Suboptimal prescribing for older adults has been described in a number of ED-based studies over the past 3 decades.^3,4 A study of veterans discharged from a Veterans Affairs (VA) Medical Center (VAMC) ED suggested that the use of potentially inappropriate medications (PIMs) was common and associated with a trend toward greater risk of repeat ED visits, hospitalization, or death.⁵

Newly released guidelines for geriatric ED care cite interventions to address prescribing quality as a specific goal for high-quality ED care.⁶ Previous studies in the ED and outpatient care settings suggest that electronic decision support and provider audit and feedback improve the quality of prescribing for specific high-risk medications.^7,8 This article describes the design and implementation of Enhancing Quality of Prescribing Practices for Older Veterans Discharged from the Emergency Department (EQUiPPED), a quality improvement initiative⁹ aimed at implementing evidence-based strategies to reduce prescribing of PIMs, and presents preliminary results from the first of eight sites to implement the model.

METHODS

Setting

EQUiPPED is an ongoing quality improvement initiative in eight VA EDs. The design and implementation of EQUiPPED at the initial site, a large, urban VAMC with an average annual ED volume of 41,000 visits of which 40% represent veterans aged 65 and older are described. The ED employs a diverse group of staff providers, including board-certified physicians in emergency medicine (EM) (n = 7), internal medicine (IM) (n = 3), and family practice (n = 2); a dually board-certified EM/IM physician; advanced practice providers (n = 2); and a clinical pharmacist. The ED also uses more than 40 moonlighting physicians and midlevel providers who work from one to 20 shifts per month; these moonlighting providers did not participate in the initial phase of EQUiPPED.

Program Development and Goals

An interdisciplinary team comprising ED physicians, geriatricians, gerontologists, clinical pharmacists, quality improvement nurses, and clinical applications coordinators developed EQUiPPED using the Vision-Analysis-Team-Aim-Map-Measure-Change-Sustain model of process improvement, which was developed within the VA from Lean and Six Sigma principles.¹⁰ The primary goal of the project was to reduce ED prescribing of PIMs for veterans aged 65 and older at the time of ED discharge. PIMs were defined as American Geriatrics Society 2012 Beers Criteria Update¹¹ Category 1 medications to avoid in all older adults. Chronic nonsteroidal anti-inflammatory drug (NSAID) prescriptions were defined as any dose for longer than 30 days. Research in various settings supports the Beers criteria, which are widely used as a marker of prescribing quality.^12,13 Based on literature review and baseline data, the team set a goal of reducing PIM use to 5% or less.⁷

Intervention Components

Provider Education

A geriatrician presented a geriatric pharmacology lecture during three shifts over the course of a month for ED providers (November 2012, Figure 1). Twelve of 16 (75%) staff providers attended one of these sessions. ED staff also reviewed the 2012 Beers Criteria Update during a quarterly journal club held in January 2013. Finally, reminder cards identifying the top five most-frequently prescribed PIMs or PIM categories based on local prescribing trends were provided at computer stations in the ED.

Figure 1.

Timeline for Enhancing Quality of Prescribing Practices for Older Veterans Discharged from the Emergency Department (EQUiPPED) implementation and evaluation.

Informatics-Based Clinical Decision Support

Mapping the computerized provider order entry (CPOE) process identified multiple steps that could increase prescribing of PIMs in older veterans at the time of ED discharge. Existing discharge order sets consisted of an alphabetized list of commonly used medications without consideration of dosage adjustments for age or renal function. Interviews revealed that providers preferentially used the existing list for prescribing because it streamlined order entry and not because they necessarily felt the listed medications were superior choices.

Subsequently, the EQUiPPED team developed geriatric outpatient pharmacy order sets grouped according to common discharge diagnoses with medication options preferred for use in older adults. These templates were placed alongside existing order sets in the electronic medical record (EMR) order menu for all veterans checked into the ED. Dose adjustments for renal impairment, point-of-prescribing education regarding medications to avoid, and links to synthesized geriatric content were embedded within the order sets. When providers prescribed medications using the order sets, the order entry process was streamlined with prepopulated fields including: “days supply,” “number of pills,” and method for prescription pick-up. Pharmacists, ED providers, and the local VA pharmacy and therapeutics committee reviewed template logic that the group developed before implementation. In addition to expediting the order entry process, the order sets simultaneously encouraged providers to choose safer medications.¹⁴ The order sets were activated in February 2013.

Individual Provider Feedback

All ED providers (n = 16) received at least one face-to-face feedback session using a process of individual practice pattern feedback (academic detailing), which has previously been demonstrated to be an effective method for changing provider prescribing practices alone or in combination with clinical decision support tools.⁸ Provider feedback included audit and feedback of individual prescribing patterns and peer benchmarking comparing individual performance with that of other ED providers at the same site.^15,16 An ED staff physician colleague conducted feedback sessions over 3 months from February to April 2013. Each provider subsequently received standardized monthly reports that included total number of veterans aged 65 and older that they evaluated and discharged from the ED, total number of medications prescribed, and proportion of total medications that were PIMs. The reports also contained a breakdown of the drug classes for PIMs and peer benchmarking for the same time period.

Program Evaluation

As a quality improvement project, the Department of Veterans Affairs Office of Geriatrics and Extended Care deemed the analysis of the EQUiPPED outcomes to be a nonresearch evaluation. The primary outcome of interest was the monthly number of PIMs as defined according to the Beers criteria that ED staff providers prescribed to veterans aged 65 and older at the time of discharge from the ED. Poisson regression was used to compare the number of PIMs prescribed in the 8 months before the first EQUiPPED intervention with 17 months of data from after the initial EQUiPPED intervention. The total number of prescriptions served as the offset term in the model, and a piecewise, nonlinear regression model was used to evaluate the pattern of PIMs prescriptions over time. Analyses were conducted using SAS version 9.2 (SAS Institute, Inc., Cary, NC).

RESULTS

From April 2012 through April 2014, VA ED staff providers treated and discharged 4,042 veterans aged 65 and older. In the 8 months before the EQUiPPED intervention (April to November 2012), the average monthly proportion of PIMs prescribed was 9.4 ± 1.5%. This declined to 4.6 ± 1.0% after the initiation of the EQUiPPED interventions (relative risk = 0.48, 95% confidence interval = 0.40–0.59, P < .001). Figure 2 shows the percentage of PIMs prescribed monthly from April 2012 through April 2014 with the results of a piecewise, nonlinear regression model. The model indicated that the inflection point occurred in April 2013 and that the observed change was sustained for the remainder of the evaluation period.

Figure 2.

Monthly percentage of potentially inappropriate medication (PIM) prescriptions by staff providers for veterans aged 65 years and older discharged from the emergency department (April 2012 to April 2014) (piecewise nonlinear regression).

At the facility, the top five most-frequently prescribed PIMs or categories as defined according to the Beers criteria were diphenhydramine, hydroxyzine, skeletal muscle relaxants, chronic NSAIDs, and promethazine. The proportion of prescriptions decreased for all of these except hydroxyzine and diphenhydramine. The greatest decrease occurred in the category of muscle relaxants, with a 60% relative reduction.

DISCUSSION

The EQUiPPED program used provider education, clinical decision support for improved workflow in a CPOE or EMR, and provider feedback with audit and peer benchmarking in a busy urban VAMC. This intervention achieved the target reduction in the proportion of PIMs prescribed to older adults discharged from the ED (<5%) and a greater than 50% total reduction in the prescribing of PIMs.

Given that there will be a growing number of older adults who will be seen in EDs, guideline committees have recommended models of care incorporating a multidisciplinary approach for improving ED care.⁶ Development of EQUiPPED involved a multidisciplinary group of providers with expertise in geriatrics, pharmacology, behavioral psychology, emergency medicine, and health services research. After seed funding from a university affiliate intramural award at the first site, VA funding provided support to expand the team and incorporate additional sites. Collaboration between Geriatric Research, Education, and Clinical Centers provided an infrastructure for rapid expansion within the VA.

Implementation of the EQUiPPED program was not designed to assess the effect of each intervention, but in keeping with a model for dissemination and implementation science, EQUiPPED addresses several potential leaks along the knowledge translation pipeline of “1) awareness—2) agreement—3) adoption—4) adherence.”¹⁷ EQUiPPED implementation increases awareness using a multipronged educational program, facilitates agreement and adoption through streamlined order entry, and encourages adherence with regular feedback to providers regarding prescribing patterns.

Feedback from providers was important in developing order sets that facilitated workflow and were not perceived as burdensome, which is a crucial characteristic of successful clinical decision support systems.¹⁴ Although the EQUiPPED order sets were designed to encompass common discharge diagnoses, providers wanted more-focused information on specific drugs to avoid. This type of reminder was added to the program in the form of a 3- by 5-inch card placed on the edge of provider workstation computer screens to serve as a point-of-prescribing reminder. In the top five medications or categories, significant reductions were noted in three (muscle relaxants, promethazine, chronic NSAIDs). The reduction in PIMs was sustained for 12 months, suggesting a culture change with respect to prescribing patterns for older adults discharged from the ED.

There are limitations to this initiative that merit acknowledgment. PIM use in this project was defined based on the 2012 update to the Beers criteria. Some have argued that Beers-defined PIMs divert focus from the smaller group of medications inappropriately dosed or monitored (antihyperglycemics, anticoagulants, pain medications) that account for the majority of ED visits or hospitalization in elderly adults.^18,19 By framing prescribing organized according to discharge diagnoses, providers received information regarding appropriate dosing of antihyperglycemic medications, warnings and reminders about antibiotic interactions with anticoagulants, and alternative medications for pain. Others have questioned whether the Beers criteria apply to short-duration prescriptions of less than 1 week, which ED providers often prescribe.²⁰ The Beers PIM framework appeared to be understood and manageable when a multidisciplinary team used it as part of a multicomponent intervention. The Beers criteria serve as a marker of prescribing quality to evaluate the effect of EQUiPPED on provider prescribing behavior,¹² yet it is only known that the number of PIMs declined. Additional analyses of broader health outcomes will allow insights into the potential effect on other areas. A capitated-payment model, a fully integrated EMR with CPOE, and close pharmacy and medical center oversight make the VA setting unique. National implementation of meaningful-use criteria by the Centers for Medicare and Medicaid Services make these results potentially replicable outside a VA facility.

As an implementation project, the before-and-after design limits inferences about the EQUiPPED program’s effect on prescribing patterns, although the long-term outcomes presented are unlikely to have occurred by chance. It is likely that local factors related to facility and clinical leadership and staff dynamics affected the results, which may not be generalizable to all settings of care or other ED sites. For example, coincident interventions, such as the hiring of a clinical pharmacist in the ED, could have affected the results. Data from expansion sites will allow assessment of the potential for broader dissemination.

CONCLUSIONS

EQUiPPED, an initiative that combines provider education, clinical decision support for improved workflow in a CPOE or EMR, and provider feedback with audit and peer benchmarking, represents a unique quality improvement initiative within the ED setting to affect the care of older adults. EQUiPPED was successfully implemented in a busy, urban VAMC and was associated with a sustained reduction in PIM prescription to older veterans. Implementation is under way at seven additional VA sites to assess the potential for broader dissemination.

Supplementary Material

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online version of this article:

Appendix S1. ED pharmacy quick orders in place prior to EQUiPPED.

Appendix S2. EQUiPPED pharmacy quick order initial screen.

Appendix S3. Examples of EQUiPPED order sets.