American Journal of Health-System Pharmacy

Elsasser, Eric, J.;White, Christina, A.;Jones, Marshall, R.

doi: 10.2146/ajhp170485pmid: 30139726

Abstract Purpose A quality of care–focused process improvement initiative undertaken by the pharmacy service at a Veterans Affairs (VA) medical center is described. Summary In September 2016, near the end of VA˴s 2016 fiscal year (FY), pharmacy leaders at the medical center held a strategic planning retreat to develop goals and objectives for FY 2017. The retreat was facilitated through use of principles of lean methodology, including ˵A3 problem solving,˶ and resulted in development of a transformational plan of care (TPOC). After identifying process improvement projects with the highest value-adding potential, retreat participants prioritized those projects in accordance with targeted value streams encompassing 5 areas of pharmacy operations: United States Pharmacopeia chapter 800 compliance, standard work, physical space, technology, and people. Upon retreat completion, tasks were assigned to pharmacy service managers according to their respective areas of expertise. The status of each project and the projects˴ impact on both pharmacy and facility outcome measures were continually assessed throughout FY 2017. Continuous reevaluation of projects within each value stream allowed for accurate outcome tracking and creation of a pharmacy dashboard. In the months after implementation of the pharmacy service TPOC, improvements in a number of performance metrics were documented. Conclusion Use of lean process improvement methodology and the A3 problem-solving process resulted in more efficient generation and implementation of ideas, consistent follow-up, and the ability to continuously reevaluate pharmacy service operations to ensure progress throughout the year. organization and administration, pharmacy, quality improvement, strategic planning In any healthcare system, strategic planning is crucial to the development of economical and sustainable processes to promote the leadership vision. Effective execution of this vision leads to appropriate prioritization of organizational goals and sets a clear direction for the future. Instead of a 1-time event, strategic planning should be a dynamic process, with continuous evaluation of goals and progress over time to ensure that desired outcomes are achieved.1 This article describes the implementation of a transformational plan of care (TPOC) as a strategic planning modality within the pharmacy service at a Veterans Affairs (VA) medical center. Lean process improvement can be used to implement strategic plans in healthcare due to its focus on value-added activities and elimination of waste.2 Value is defined from the perspective of the end customer, the patient.3 A common tool used for lean methodology implementation is ˵A3 problem solving,˶ an approach to process improvement whose name is derived from the idea that all steps of the problem-solving process can be encapsulated on a single worksheet of standard A3 size (297 × 420 mm). This technique serves as a roadmap to guide the decision-making process and application of principles of lean methodology. Typically involving 8 or 9 steps, A3 problem solving allows a process improvement team to thoroughly examine value streams to identify value and waste through use of the following constructs and/or types of activities and analyses: Aim—Statement of reasons for action and purpose, with documentation of goals and objectives Current state—Identification of what a process looks like today Future state—Determination of what a process should look like in its ideal state Gap analysis—Analysis of gaps separating the current and future states Solution approach—Identification of solutions and their impact on objectives Rapid experiments—Project identification and prioritization Completion plan—Identification of overarching goals of initiatives to bridge gaps between the current and future states Confirmed state—Use of metrics to measure and track progress Insights—Identification of what went well, opportunities for improvement, and lessons learned Within a healthcare organization, a TPOC is the highest level of A3 problem solving. It guides construction of a vision and implementation plan for the foreseeable future; this allows for a streamlined and strategic alignment of an organization˴s goals and initiatives. Ultimately, the TPOC serves as an outlet from which value streams and metrics are prioritized.4 The executive leadership team (ELT) at the VA medical center completed a TPOC to develop a set of 4 organizational pillars: people, service, stewardship, and quality. These pillars serve as an organizational framework for medical center employees to use in the provision of excellent, veteran-centered care. All metrics relevant to the pharmacy service TPOC directly correlate with this philosophy. View largeDownload slide Eric Elsasser, Pharm.D., is a pharmacy manager with inpatient pharmacy services at Duke University Hospital in Durham, North Carolina. He is a 2015 graduate of Purdue University College of Pharmacy. He completed both a postgraduate year 1 pharmacy practice residency and a postgraduate year 2 health-system pharmacy administration residency at the Richard L. Roudebush VA Medical Center in Indianapolis, Indiana. View largeDownload slide Eric Elsasser, Pharm.D., is a pharmacy manager with inpatient pharmacy services at Duke University Hospital in Durham, North Carolina. He is a 2015 graduate of Purdue University College of Pharmacy. He completed both a postgraduate year 1 pharmacy practice residency and a postgraduate year 2 health-system pharmacy administration residency at the Richard L. Roudebush VA Medical Center in Indianapolis, Indiana. In the fall of 2015, at the start of fiscal year (FY) 2016, the pharmacy leadership team met to develop a TPOC modeled after that of the ELT. This was the first TPOC specific to the pharmacy service, and it resulted in the creation of 5 value streams: inpatient pharmacy, outpatient pharmacy, anticoagulation clinic, pharmacy call center, and people. A variety of projects to streamline pharmacy processes and elevate the positive impacts of pharmaceutical care for patients were identified within each value stream. However, while these value streams encompassed the breadth of pharmacy services within the medical center, they resulted in practice silos that may have hindered intradepartmental collaboration. Thus, the decision was made to revisit the strategic planning process at the beginning of FY 2017. As a VA level 1a facility, the medical center receives the most complex and diverse types of patients within the VA network. The pharmacy service maintains a robust presence throughout the medical center, with roughly 200 full-time employee equivalents. In addition to providing 24-hour inpatient and outpatient pharmacy services, pharmacists and technicians operate an anticoagulation clinic and a pharmacy call center, round on inpatient care teams, and provide medication therapy management in ambulatory care clinics. The 4-fold purpose of the project was to facilitate strategic and tactical planning for the pharmacy service, develop process improvement initiatives, align with the medical center and departmental vision, and build upon the progress and lessons learned from the FY 2016 TPOC. Implementation of the TPOC In September 2016, the pharmacy administration team participated in a strategic planning retreat. The timing was chosen to allow completion of the retreat prior to the beginning of FY 2017 on October 1, 2016. This retreat was organized and facilitated by a health-system pharmacy administration (HSPA) resident. Upon initiation of the strategic planning retreat, the reasons for action were focused on the pharmacy service˴s vision of being a nationally recognized leader in the provision of quality pharmaceutical care across the healthcare continuum. Pharmaceutical care provided to end customers should be patient centered, safe, timely, and cost-effective. The development of confident and competent pharmacy practitioners is ensured by dedication to excellence through research and training. Throughout the retreat, an emphasis was placed on the creation and continuation of process improvement initiatives within the pharmacy service. Participants included a variety of supervisors and managers within the department. As potential projects were developed, applicable managers were tasked to serve as ˵process owners˶ according to their area of expertise. All aspects of the planning and problem-solving activities were based on the use of lean process improvement methodology and A3 process mapping. Completion of the A3 problem-solving process was necessary to lay the framework for identification of appropriate actions going forward. Aim. The initial step in the A3 problem-solving process was to develop a unified view of the reasons for action and purpose related to the pharmacy service TPOC. Identification of overall goals and objectives for the strategic planning retreat was completed at this initial stage. Current and future states. The current and future states were mapped through a strengths, weaknesses, opportunities, and threats (SWOT) analysis. Strengths and weaknesses are often centered on internal issues, while opportunities and threats have an external focus. For the current-state assessment, the pharmacy leadership focused on the strengths and weaknesses of the initial TPOC process that took place throughout FY 2016. Looking ahead to the future state, department leaders developed a list of opportunities and threats relevant to the FY 2017 TPOC. Use of a SWOT analysis provided valuable insights into the current and future states of the pharmacy service. Many of these insights stemmed from lessons learned in carrying out the FY 2016 TPOC and solidified the pharmacy leadership˴s focus throughout the A3 problem-solving process. Strengths identified in the SWOT analysis included the use of lean processes, the ability to complete a second TPOC, and increased awareness of project timelines. Weaknesses included suboptimal staff engagement, unrealistic expectations, constrained timelines, and the overuse of pharmacy residents to facilitate projects. Many opportunities were noted, including opportunities to reassess value streams, reprioritize goals on the basis of facility metrics, increase frontline staff involvement, and promote proactivity and a focus on strategic rather than tactical planning. Lastly, threats identified by the group included unknown political forces, limited funding, and rising pharmaceutical costs. Gap analysis. This step in the A3 problem-solving process allowed for identification of barriers, or gaps, separating the current and future states. Understanding the gaps in a given process was key in the development of valuable and sustainable solutions. An affinity diagram was then used to categorize all identified gaps into various groups according to their natural relationships. These groups were identified as areas of emphasis, providing a foundation for solution development and implementation throughout the remainder of the strategic planning process. Five areas of emphasis, or value streams, were developed as a result of the natural groupings within the affinity diagram during the gap analysis. Identified value streams for the FY 2017 TPOC were as follows: United States Pharmacopeia (USP) chapter 800 compliance, standard work, physical space, technology, and people. Going forward, all projects and solutions were categorized within 1 of these 5 value streams to streamline efficiency. Solution approach. As a next step, solutions to address the value streams developed in the gap analysis were identified. In addition to creating a solution, the pharmacy leadership projected the outcomes that would result from solution implementation. An ˵if/then˶ table was used to arrange these solutions. These thought processes laid the foundation for the generation of impactful project ideas later in the problem-solving process. Ideas developed through the solution approach correlated with each of the 5 value streams. Expected outcomes of implementing the identified solutions included the following: compliance with USP chapter 800 standards, resulting in enhanced safety for patients and employees; consistency in the delivery of pharmacy services to customers; decreased medication errors and waste; and increased staffing efficiencies, flexibility, and employee satisfaction. Rapid experiments. With the solution approach in place, the pharmacy leadership identified a comprehensive list of all projects related to the process improvement goals. After determination of projects with the highest value-adding potential, focus was placed on the prioritization of projects throughout various value streams created by the administration team. Projects were then assigned a priority level (high, moderate, or low) according to their value and urgency. Additionally, a ˵degree of control˶ classification was established for each project. Projects with a high rating for degree of control were those in which the pharmacy service was the dominant or sole stakeholder; on the other hand, projects rated low for degree of control would require the involvement of additional departments or forces outside the pharmacy service. Upon completion of the retreat, tasks were distributed among team members for execution. For each project, applicable managers were deemed process owners by area of expertise. For example, projects related to inventory management were assigned to the procurement supervisor, those related to inpatient services were assigned to the inpatient supervisor, and so on. When addressing the rapid experiments, multiple projects were discussed and prioritized appropriately within each of the 5 value streams. For example, renovation of the inpatient pharmacy i.v. room was designated as a high-priority project, with a moderate degree of control, within the USP chapter 800 value stream, while acquisition of medication carousels was designated as a moderate-priority project, with a low degree of control, within the technology value stream. An example of a low-priority project with a high degree of control was pharmacist coverage on the medicine team, as this workflow had previously been implemented and required only minor modifications. Of the 28 projects identified, 16 were prioritized as high-, 9 as moderate-, and 3 as low-priority projects. Eight projects were designated as having a high degree of control, with 15 and 5 classified as moderate- and low-control projects, respectively. Completion plan. The completion plan was laid out after determination of projects and priorities. Unlike the rapid experiments, plan components were not categorized into specific value streams. Rather, these large-scale goals and solutions were identified to bridge the gap between the current and future states. As a result, multiple large-scale solutions for bridging the gap between the current and future states were identified as part of the completion plan. These solutions included the acquisition of grant-funded positions, inpatient inventory optimization, submission of a request for a research pharmacist, outpatient pharmacy workflow modifications, improvement of outpatient pharmacy vault operations, and bridging of the departmental informatics gap. Confirmed state. The confirmed-state analysis emphasized the overall results and progress of the TPOC itself. Analysis of projects developed within each of the value streams was completed to create a pharmacy service dashboard. Relevant metrics were measured and displayed within the dashboard to track TPOC progress. This dashboard was updated monthly, and the pharmacy leadership met on a quarterly basis to reevaluate and update the strategic plan. The status of projects, staff involvement, and the impact on both pharmacy and facility outcome measures were continually assessed throughout the year to determine effectiveness. As a whole, the pharmacy service TPOC generated a vast array of initiatives and goals to enhance the quality of pharmaceutical care provided within the medical center. With regard to the 28 projects identified as rapid experiments, the distribution by value stream was as follows: USP chapter 800 compliance, 3; standard work, 10; physical space, 2; technology, 7; and people, 6. Dashboard metrics were designed for specificity to the pharmacy service as well as the 4 organizational pillars of the medical center. Both the status of projects and their impact on metrics were updated throughout the year to determine effectiveness. Plan implementation and dashboard maintenance were overseen by the HSPA resident. Table 1 displays a summary of notable metrics used. Table 1 Select Pharmacy Service Metrics Used in Process Improvement Initiative Organizational Pillar and Metric Initial State Target State Confirmed State People: staff with introductory lean training (%) 33.8 >85.0 87.0 Service: rate of abandoned calls (%) 13.35 <5.00 1.01 Quality: ward check compliance (%)a 75 >90 100 Stewardship: annual inventory turns (no.) 12.9 13.0 15.0 Organizational Pillar and Metric Initial State Target State Confirmed State People: staff with introductory lean training (%) 33.8 >85.0 87.0 Service: rate of abandoned calls (%) 13.35 <5.00 1.01 Quality: ward check compliance (%)a 75 >90 100 Stewardship: annual inventory turns (no.) 12.9 13.0 15.0 aAs measured via monthly technician-driven inspections of medication storage areas for removal of expired medications. View Large Table 1 Select Pharmacy Service Metrics Used in Process Improvement Initiative Organizational Pillar and Metric Initial State Target State Confirmed State People: staff with introductory lean training (%) 33.8 >85.0 87.0 Service: rate of abandoned calls (%) 13.35 <5.00 1.01 Quality: ward check compliance (%)a 75 >90 100 Stewardship: annual inventory turns (no.) 12.9 13.0 15.0 Organizational Pillar and Metric Initial State Target State Confirmed State People: staff with introductory lean training (%) 33.8 >85.0 87.0 Service: rate of abandoned calls (%) 13.35 <5.00 1.01 Quality: ward check compliance (%)a 75 >90 100 Stewardship: annual inventory turns (no.) 12.9 13.0 15.0 aAs measured via monthly technician-driven inspections of medication storage areas for removal of expired medications. View Large Insights. As the final step in the problem-solving process, a critical evaluation of the strategic planning retreat was performed. Specifically, this evaluation focused on identification of what went well, opportunities for improvement, and lessons learned. A variety of strengths and limitations were associated with the problem-solving process. Strengths identified included use of a SWOT analysis, prioritization of projects, and avoidance of an arbitrary 1-year time limit for project completion. Additionally, honest, open, and positive team dialogue was evident throughout the retreat. Process limitations included varied levels of engagement among team members, time constraints resulting from the coordination of multiple pharmacy supervisors˴ schedules, timing of the retreat prior to receipt of a finalized budget for FY 2017, limited wall space for process mapping, and unknown political forces. The political factors included shifts in personnel and in the priorities of the local ELT as well as the potential for modifications to the Veterans Health Administration resulting from the 2016 presidential election. Discussion Process improvement is a continually evolving cycle, and experiences thus far with the TPOC have shed light on opportunities for positive change. The pharmacy service constantly strives to improve outcomes for the end customers, or veterans. Over the past few years, the pharmacy leadership has modified its approach to best fit the needs of patients as well as the ever-changing landscape of the healthcare system. In a typical A3 problem-solving process, the current and future states are represented with process maps to outline successive actions required to get from ˵point A˶ to ˵point B.˶ Given the nature of the project described here, it was appropriate to approach the current and future states through use of a SWOT analysis; this allowed the pharmacy leadership to more accurately identify areas for improvement within the TPOC. To mitigate any ˵silo effect˶ resulting from the value streams in the FY 2016 TPOC, a different set of 5 value streams (USP chapter 800, standard work, physical space, technology, and people) was created in the FY 2017 TPOC. These new value streams allowed for a continued focus on the strategic needs of the pharmacy service and instilled greater cohesiveness in the planning process. In contrast to the FY 2016 TPOC, the FY 2017 TPOC listed project ideas that were not limited to those that could be completed within a 12-month timeframe. Throughout the year, the pharmacy leadership periodically revisited the list to update project statuses, add new ideas, and remove projects as necessary. As seen in the rapid-experiments phase, most projects were given high priority. With regard to degree of control, however, most projects fell into the moderate-control category. These results are to be expected, as many high-level undertakings within a medical center require the input and/or expertise of professionals in multiple departments. Additionally, after reevaluation of the TPOC throughout the year, 4 of the 10 large-scale goals identified for plan completion were placed on hold due to shifting priorities within both the pharmacy service and the medical center. Evaluation of the TPOC development process and TPOC effectiveness was primarily focused on metrics identified within the confirmed state characterized through the A3 problem-solving process. Positive strides were seen with regard to a variety of metrics, including an increase in the proportion of pharmacy staff with introductory lean training to 87%. This served as a positive feedback loop, as an increased number of trained staff members led to an enlarged capacity for participation in process improvement initiatives. Another notable metric was a decrease in the rate of abandoned calls within the pharmacy call center, which resulted in improved customer service and efficiency. Presentation of these data to the medical center leadership led to an increase in allotted resources and personnel to promote the departmental vision. Ultimately, some of the most applicable insights gained from the TPOC resulted from the confluence of ideas and strategies brought forth by dedicated staff within the pharmacy service, all of which served to improve the value provided to the patient population of veterans. Future applications for the TPOC include the continued alignment of facility metrics with pharmacy service metrics, additional involvement of frontline staff, and an increased frequency of value stream reassessment. Conclusion Use of lean process improvement methodology and the A3 problem-solving process resulted in more efficient generation and implementation of ideas, consistent follow-up, and the ability to continuously reevaluate pharmacy service operations to ensure progress throughout the year. Additional information The contents of this article do not represent the views of the U.S. Department of Veterans Affairs or the U.S. government. Disclosures This material is a result of work supported with resources and the use of facilities at Richard L. Roudebush VA Medical Center, Indianapolis, IN. The authors have declared no potential conflicts of interest. References 1 Sanborn M . Developing a meaningful strategic plan. Hosp Pharm. 2009 ; 44 : 625 – 9 . Crossref Search ADS 2 Going lean in healthcare. IHI innovation series white paper . Cambridge, MA: Institute for Healthcare Improvement; 2005 . 3 Toussaint JS and Berry LL. The promise of lean in health care. Mayo Clin Proc. 2013 ; 88 : 74 – 82 . Crossref Search ADS PubMed 4 Toussaint JS and Gerard RA. On the mend: revolutionizing healthcare to save lives and transform the industry . Cambridge, MA: Lean Enterprise Institute; 2010 . Copyright © 2018, American Society of Health-System Pharmacists, Inc. All rights reserved.

Kang,, Amy;Thompson,, Ashley;Rau,, Johnny;Pollock,, Allison

doi: 10.2146/ajhp170465pmid: 30139727

Abstract Purpose Results of an evaluation of therapeutic interchange (TI) program outcomes with use of prescriber alerts alone or in combination with pharmacist prescribing are reported. Methods A retrospective single-center study was conducted to compare TI outcomes before incorporation of prescriber alerts encouraging formulary agent use into the electronic medical record (period 1), after alert implementation (period 2), and after implementation and expansion of TI protocols including pharmacist prescribing authority (period 3). The evaluation focused on TI orders for 3 drug classes: angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), and 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (HMG-CoA RIs). The primary outcome was formulary medication utilization. Results In total, 2,881, 2,700, and 3,088 prescriptions for medications in the ACEI, ARB, or HMG-CoA RI class were ordered during periods 1, 2, and 3, respectively. Overall formulary adherence improved from 78.3% in period 1 to 97.6% in period 2 and 99.2% in period 3 (p < 0.001 for both comparisons with period 1). The percentages of inappropriate dosing conversions were 51.6%, 37.2%, and 2.4% in periods 1, 2, and 3, respectively; the corresponding percentages of inappropriate discharge medications were 64.5%, 16.3%, and 2.4%. Conclusion Percentages of formulary medications in 3 medication classes, considered separately and together, increased with the implementation of TI alerts for prescribers and the addition of TI-related pharmacist prescribing authority. Over the same study periods, percentages of inappropriate dosing conversions and inappropriate discharge medications decreased. clinical decision support, pharmacist-driven, therapeutic interchange Medication formularies are used to promote effective and safe medication therapy while minimizing healthcare costs.1,2 Therapeutic interchange (TI) is the substitution of a medication with one that is therapeutically equivalent but chemically different. When administered in therapeutically equivalent dosages, drug products within the same pharmacologic class are expected to yield similar clinical outcomes and have similar adverse-reaction profiles. Therefore, TI, when properly designed and implemented, can achieve clinically appropriate, safe, and cost-effective drug therapy.3 When medications are inappropriately interchanged, however, it can lead to inadequate clinical effectiveness as well as potential harm.4,5 A built-in clinical decision support system within an electronic medical record (EMR) system, including alerts regarding nonformulary medication orders along with appropriate dosage conversions for formulary medications, can be used to maximize the effectiveness of TI and minimize the potential for errors.6,–8 At University of California, San Francisco Medical Center (UCSFMC), the pharmacy and therapeutics (P&T) committee established a TI protocol for commonly prescribed hospital formulary medications. The P&T committee selected medications for TI that are the most appropriate for the needs of the medical center˴s patient population from the perspectives of safety, effectiveness, and cost. In 2012, UCSFMC implemented an EMR and, as part of the TI initiative, incorporated into the order-entry system ˵prescriber alternative alerts˶ for 3 commonly used classes of medications that account for high volumes of inpatient orders: angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), and 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (HMG-CoA RIs). If a nonformulary medication within the selected medication classes was ordered, a pop-up prescriber alternative alert proposed a therapeutic equivalent along with appropriate dosing conversion ratios (Figure 1). Existing TI programs at UCSFMC solely based on prescriber alerts were expanded to incorporate pharmacist prescribing authorities that enabled pharmacists to authorize TI for eligible classes of medications when nonformulary medications were requested by ordering providers. Figure 1 View largeDownload slide Pop-up alert in electronic medical record indicating nonformulary status of a desired medication (fosinopril) and suggesting a formulary alternative (lisinopril) and corresponding dosage. Monopril, Prinivil, and Zestril are trademarked brands of Bristol-Myers Squibb, Merck, and AstraZeneca, respectively. Figure 1 View largeDownload slide Pop-up alert in electronic medical record indicating nonformulary status of a desired medication (fosinopril) and suggesting a formulary alternative (lisinopril) and corresponding dosage. Monopril, Prinivil, and Zestril are trademarked brands of Bristol-Myers Squibb, Merck, and AstraZeneca, respectively. The study described here was performed to evaluate the effectiveness of using clinical decision support in the EMR for TI programs at UCSFMC, as well as associated medication error rates, and to compare the effectiveness of prescriber alternative alerts alone and in combination with pharmacist-driven TI. Additionally, the pharmacoeconomic impact of TI was analyzed. The primary outcome was utilization of formulary medications in the ACEI, ARB, and HMG-CoA RI classes before and after the implementation of prescriber alternative alerts and pharmacist-driven TI. Secondary outcomes included the percentage of inappropriate conversions, the percentage of inappropriate discharge medications, and estimated annual cost savings. It was hypothesized that the implementation and expansion of pharmacist-driven TI programs would decrease the use of nonformulary medications without compromising patient safety and would reduce healthcare costs. To the best of our knowledge, the study was one of the first to evaluate not only the effectiveness of pharmacist-driven TI but also the associated medication error rates. View largeDownload slide Amy Kang, Pharm.D., is an assistant professor at Chapman University. She completed a postgraduate year 1 pharmacy practice residency program at University of California, San Francisco, and a postgraduate year 2 infectious diseases (ID) pharmacy residency at University of Southern California. Her current practice site is at Harbor–UCLA Medical Center, where she serves as an ID clinical pharmacist, optimizing antimicrobial regimens for patients with complicated ID problems and creating and implementing institutionwide protocols for antimicrobial usage. She received a B.S. degree in microbiology, immunology, and molecular genetics from University of California. Los Angeles. in 2012 and earned her Pharm.D. from Loma Linda University School of Pharmacy in 2016. View largeDownload slide Amy Kang, Pharm.D., is an assistant professor at Chapman University. She completed a postgraduate year 1 pharmacy practice residency program at University of California, San Francisco, and a postgraduate year 2 infectious diseases (ID) pharmacy residency at University of Southern California. Her current practice site is at Harbor–UCLA Medical Center, where she serves as an ID clinical pharmacist, optimizing antimicrobial regimens for patients with complicated ID problems and creating and implementing institutionwide protocols for antimicrobial usage. She received a B.S. degree in microbiology, immunology, and molecular genetics from University of California. Los Angeles. in 2012 and earned her Pharm.D. from Loma Linda University School of Pharmacy in 2016. Methods The study was a 3-phase retrospective study conducted at UCSFMC. Data collection covered 3 periods: (1) a 3-month period before implementation of prescriber alternative alerts (November 1, 2013–January 31, 2014), (2) a 3-month period after alert implementation (November 1, 2014–January 31, 2015), and (3) a 3-month period after expansion of pharmacist-driven TI to include pharmacist authorization of TI without physician approval (November 1, 2016–January 31, 2017). The study was approved by the UCSFMC institutional review board. Primary data collection was accomplished by extracting an EMR report of all medication orders for ACEIs, ARBs, and HMG-CoA RIs during the specified time periods. Secondary data collection was performed via chart review of randomly selected formulary medication orders (400 orders per study period) in order to identify orders that were conversions from nonformulary orders before admission. Orders that fit the criteria for TI were then evaluated to determine whether inpatient medications and discharge medications were appropriately converted. To derive estimated cost savings, cost per dose was calculated by first dividing (1) the combined drug acquisition cost (through UCSFMC˴s group purchasing organization) for all formulary and nonformulary medications within the targeted drug classes during the specified time periods by (2) the total number of doses dispensed during each period. The total number of doses for each 3-month period was extrapolated to estimate the annual number of doses and then multiplied by the calculated cost per dose to estimate differences in annual medication costs. Statistical analysis was carried out with GraphPad Prism, version 6.0 (GraphPad Software, San Diego, CA) and SPSS (IBM Corporation, Armonk, NY). Power analysis was not performed. Case and control variables were compared with a chi-square test. The a priori level of significance was 0.05, with a standard 95% confidence threshold for rejecting the null hypothesis. Results In total, 2,881, 2,700, and 3,088 inpatient orders for medications within the ACEI, ARB, or HMG-CoA RI class were written during periods 1, 2, and 3, respectively. Utilization comparisons were performed for individual therapeutic classes as well as the 3 classes in composite for each pair of study periods. Utilization of formulary medications within the 3 drug classes, both individually and in composite, increased from period 1 to period 3 (Table 1); all differences across all periods were significant and showed improvements in formulary adherence after the implementation of prescriber alternative alerts and the addition of pharmacist prescribing authorities. Analysis of the 400 orders screened per period for secondary data collection revealed that 31, 43, and 42 orders met the criteria for TI in periods 1, 2, and 3, respectively. The percentages of inappropriate dosing conversions and inappropriate discharge medications resulting from the identified TI orders were compared between study periods (period 1 versus period 2, period 2 versus period 3, and period 1 versus period 3). The percentage of inappropriate dosing conversions significantly decreased with use of pharmacist-driven TI (in period 3), as compared with use of prescriber alternative alerts alone (in period 2) (Table 2). Similar between-period differences were observed in the percentages of inappropriate discharge medications. The primary reason for inappropriate conversions was incorrect dosage, and the primary reason for inappropriate discharge medications was continuation of hospital formulary medications instead of resumption of patients˴ home medications upon discharge. Other reasons for inappropriate discharge medications included inappropriate dosage and inappropriate discontinuation of medications upon discharge. Table 3 shows the cost per dose for each medication class and estimated annual costs and savings for the 3 drug classes individually and in composite. Cost per dose was low across all classes of medications, not exceeding $0.60. The yearly estimated cost savings resulting from reduced direct medication costs within the 3 drug classes combined after implementation of TI interventions (in period 3 versus period 1) were $1,773. Annual savings were larger in period 3 versus period 2 ($1,382) than in period 2 versus period 1 ($391). Table 1 Ordering of Formulary Medications During 3 Study Periodsa Drug Class Fraction (%) Orders for Formulary Medications Period 1 Period 2 Period 3 ACEI 627/749 (83.7) 653/665 (98.2)b 796/798 (99.7)b,c ARB 375/415 (90.4) 409/425 (96.2)b 375/377 (99.5)b,c HMG-CoA RI 1,254/1,717 (73.0) 1,574/1,610 (97.8)b 1,892/1,913 (98.9)b,d  Total 2,256/2,881 (78.3) 2,636/2,700 (97.6)b 3,063/3,088 (99.2)b,e Drug Class Fraction (%) Orders for Formulary Medications Period 1 Period 2 Period 3 ACEI 627/749 (83.7) 653/665 (98.2)b 796/798 (99.7)b,c ARB 375/415 (90.4) 409/425 (96.2)b 375/377 (99.5)b,c HMG-CoA RI 1,254/1,717 (73.0) 1,574/1,610 (97.8)b 1,892/1,913 (98.9)b,d  Total 2,256/2,881 (78.3) 2,636/2,700 (97.6)b 3,063/3,088 (99.2)b,e a ACEI = angiotensin-converting enzyme inhibitor, ARB = angiotensin II receptor blocker, HMG-CoA RI = 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. b p < 0.0001 for comparison with period 1. c p = 0.002 for comparison with period 2. d p = 0.007 for comparison with period 2. e p < 0.0001 for comparison with period 2. View Large Table 1 Ordering of Formulary Medications During 3 Study Periodsa Drug Class Fraction (%) Orders for Formulary Medications Period 1 Period 2 Period 3 ACEI 627/749 (83.7) 653/665 (98.2)b 796/798 (99.7)b,c ARB 375/415 (90.4) 409/425 (96.2)b 375/377 (99.5)b,c HMG-CoA RI 1,254/1,717 (73.0) 1,574/1,610 (97.8)b 1,892/1,913 (98.9)b,d  Total 2,256/2,881 (78.3) 2,636/2,700 (97.6)b 3,063/3,088 (99.2)b,e Drug Class Fraction (%) Orders for Formulary Medications Period 1 Period 2 Period 3 ACEI 627/749 (83.7) 653/665 (98.2)b 796/798 (99.7)b,c ARB 375/415 (90.4) 409/425 (96.2)b 375/377 (99.5)b,c HMG-CoA RI 1,254/1,717 (73.0) 1,574/1,610 (97.8)b 1,892/1,913 (98.9)b,d  Total 2,256/2,881 (78.3) 2,636/2,700 (97.6)b 3,063/3,088 (99.2)b,e a ACEI = angiotensin-converting enzyme inhibitor, ARB = angiotensin II receptor blocker, HMG-CoA RI = 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. b p < 0.0001 for comparison with period 1. c p = 0.002 for comparison with period 2. d p = 0.007 for comparison with period 2. e p < 0.0001 for comparison with period 2. View Large Table 2 Inappropriate Conversions and Discharge Medications During 3 Study Periods Type of Inappropriate Order Fraction (%) Inappropriate Ordersa Period 1 (n = 31) Period 2 (n = 43) Period 3 (n = 42) Inappropriate conversion 16/31 (52) 16/43 (37)b 1/42 (2)c,d  Incorrect dosage 15/16 (94) 16/16 (100) 1/1 (100)  Incorrect formulation 1/16 (6) 0/16 0/1 Inappropriate discharge medication 20/31 (65) 7/43 (16) 1/42 (2)  Duplication of therapy 1/20 (5) 0/7d 0/1d,e  Formulary drug continued 12/20 (60) 6/7 (86) 1/1 (100)  Other reason 7/20 (35) 1/7 (14) 0/1 Type of Inappropriate Order Fraction (%) Inappropriate Ordersa Period 1 (n = 31) Period 2 (n = 43) Period 3 (n = 42) Inappropriate conversion 16/31 (52) 16/43 (37)b 1/42 (2)c,d  Incorrect dosage 15/16 (94) 16/16 (100) 1/1 (100)  Incorrect formulation 1/16 (6) 0/16 0/1 Inappropriate discharge medication 20/31 (65) 7/43 (16) 1/42 (2)  Duplication of therapy 1/20 (5) 0/7d 0/1d,e  Formulary drug continued 12/20 (60) 6/7 (86) 1/1 (100)  Other reason 7/20 (35) 1/7 (14) 0/1 a n represents the number of orders that fit the therapeutic interchange criteria out of the selected 400 formulary medication orders in each study period. b p = 0.24 for comparison with period 1. c p < 0.0001 for comparison with period 2. d p < 0.0001 for comparison with period 1. e p = 0.058 for comparison with period 2. View Large Table 2 Inappropriate Conversions and Discharge Medications During 3 Study Periods Type of Inappropriate Order Fraction (%) Inappropriate Ordersa Period 1 (n = 31) Period 2 (n = 43) Period 3 (n = 42) Inappropriate conversion 16/31 (52) 16/43 (37)b 1/42 (2)c,d  Incorrect dosage 15/16 (94) 16/16 (100) 1/1 (100)  Incorrect formulation 1/16 (6) 0/16 0/1 Inappropriate discharge medication 20/31 (65) 7/43 (16) 1/42 (2)  Duplication of therapy 1/20 (5) 0/7d 0/1d,e  Formulary drug continued 12/20 (60) 6/7 (86) 1/1 (100)  Other reason 7/20 (35) 1/7 (14) 0/1 Type of Inappropriate Order Fraction (%) Inappropriate Ordersa Period 1 (n = 31) Period 2 (n = 43) Period 3 (n = 42) Inappropriate conversion 16/31 (52) 16/43 (37)b 1/42 (2)c,d  Incorrect dosage 15/16 (94) 16/16 (100) 1/1 (100)  Incorrect formulation 1/16 (6) 0/16 0/1 Inappropriate discharge medication 20/31 (65) 7/43 (16) 1/42 (2)  Duplication of therapy 1/20 (5) 0/7d 0/1d,e  Formulary drug continued 12/20 (60) 6/7 (86) 1/1 (100)  Other reason 7/20 (35) 1/7 (14) 0/1 a n represents the number of orders that fit the therapeutic interchange criteria out of the selected 400 formulary medication orders in each study period. b p = 0.24 for comparison with period 1. c p < 0.0001 for comparison with period 2. d p < 0.0001 for comparison with period 1. e p = 0.058 for comparison with period 2. View Large Table 3 Costs for Drug Classes Involved in Evaluated Therapeutic Interchangesa Variable Period 1 Period 2 Period 3 Cost per dose ($)  ACEI 0.05 0.03 0.03  ARB 0.59 0.37 0.13  HMG-CoA RI 0.23 0.26 0.25 Estimated annual cost ($)  ACEI 410 246 246  ARB 2,745 1,721 605  HMG-CoA RI 6,108 6,904 6,639   Total 9,263 8,872 7,490 Variable Period 1 Period 2 Period 3 Cost per dose ($)  ACEI 0.05 0.03 0.03  ARB 0.59 0.37 0.13  HMG-CoA RI 0.23 0.26 0.25 Estimated annual cost ($)  ACEI 410 246 246  ARB 2,745 1,721 605  HMG-CoA RI 6,108 6,904 6,639   Total 9,263 8,872 7,490 a ACEI = angiotensin-converting enzyme inhibitor, ARB = angiotensin II receptor blocker, HMG-CoA RI = 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. View Large Table 3 Costs for Drug Classes Involved in Evaluated Therapeutic Interchangesa Variable Period 1 Period 2 Period 3 Cost per dose ($)  ACEI 0.05 0.03 0.03  ARB 0.59 0.37 0.13  HMG-CoA RI 0.23 0.26 0.25 Estimated annual cost ($)  ACEI 410 246 246  ARB 2,745 1,721 605  HMG-CoA RI 6,108 6,904 6,639   Total 9,263 8,872 7,490 Variable Period 1 Period 2 Period 3 Cost per dose ($)  ACEI 0.05 0.03 0.03  ARB 0.59 0.37 0.13  HMG-CoA RI 0.23 0.26 0.25 Estimated annual cost ($)  ACEI 410 246 246  ARB 2,745 1,721 605  HMG-CoA RI 6,108 6,904 6,639   Total 9,263 8,872 7,490 a ACEI = angiotensin-converting enzyme inhibitor, ARB = angiotensin II receptor blocker, HMG-CoA RI = 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. View Large Discussion Due to the growing number of medications that are chemically related to their respective prototype drugs and the increasing need to reduce healthcare costs, the role of TI has been expanding.4 As a result, prescribers often need to switch medications to hospital formulary alternatives when patients are admitted to the hospital, which could be a laborious, time-consuming, and error-prone process without proper TI programs in place.7 Additional challenges encountered with TI programs include unique indications for nonformulary agents, renal dosing, and administration and/or formulation contraints. Ensuring correct conversions to formulary medications on admission and appropriate discharge medications during transitions of care is critical to avoid potential harm from inappropriate TI. The study results suggest that formulary adherence, cost reduction, and medication error minimization are achieved by expanding the use of prescriber EMR alerts and pharmacist prescribing authorities in TI programs. Furthermore, use of provider EMR alerts can be expanded to other quality-improvement initiatives such as the management of drug shortages resulting from interruptions or permanent discontinuances in manufacturing. There were several limitations to the study. First, as it was a retrospective single-center study, the findings may not be generalizable to other institutions and were subject to inherent biases due to the retrospective study design. Second, our selection of the 3-month time frame of November–January for all 3 study periods may have resulted in an inaccurate representation of nonformulary medication orders at UCSFMC overall. Third, the study was limited to the evaluation of 3 classes of medications that constitute only a subset of the medications targeted in TI programs at UCSFMC. Fourth, drug acquisition costs may vary by hospital, and the costs evaluated in our study are not necessarily applicable to other institutions. Fifth, the data on estimated annual cost savings only accounted for direct medication costs. The most prominent advantages of TI are difficult to quantify due to the indirect costs associated with use of nonformulary medications, such as human labor associated with inventory management and stocking of medications of variable strengths and with limited shelf space available; additionally, the time taken by ordering physicians and verifying pharmacists to approve and dispense nonformulary medications is likely to be prolonged, which directly affects the efficiency of the medication processing workflow and delays patient receipt of medications. Lastly, the 3 classes of medications analyzed in the study have relatively low costs due to the availability of alternative generic medications. Future studies are warranted to evaluate all aspects of TI effectiveness and safety with other medication classes. There are several factors that can be taken into consideration to maximize the benefits of TI programs. To take advantage of savings from reduced direct medication costs, additional classes of medications can be included in TI programs, such as dipeptidyl peptidase-4 inhibitors, long-acting β-agonist or corticosteroid inhalers, and prostaglandin ophthalmic solutions. To promote the proper use of TI programs, it is imperative that pharmacy departments establish necessary orientation and training regarding the policies and procedures related to TI protocols for both newly hired and existing pharmacists. Furthermore, a copy of the TI protocols, including recommended dosage conversion ratios, should be distributed to all involved employees and prescribers and be readily available to all pharmacists involved in order-entry processes. Additionally, documentation of TI should be strongly encouraged to ensure accurate and timely communication among providers and minimize medication errors. Lastly, not only pharmacists but all other involved clinicians, including physicians, nurse practitioners, and physician assistants, should receive adequate education on the TI program via memoranda, inservice sessions, and posters displayed around the hospital. Conclusion Percentages of formulary medications in 3 medication classes, considered separately and together, increased with the implementation of TI alerts for prescribers and the addition of TI-related pharmacist prescribing authority. Over the same study periods, percentages of inappropriate dosing conversions and inappropriate discharge medications decreased. Disclosures The authors have declared no potential conflicts of interest. References 1 Principles of a sound drug formulary system. In: Hawkins B , ed. Best practices for hospital and health-system pharmacy: positions and guidance documents of ASHP . Bethesda, MD: American Society of Health-System Pharmacists; 2006 : 110 – 3 . 2 American Society of Health-System Pharmacists. ASHP statement on the pharmacy and therapeutics committee and the formulary system. Am J Health-Syst Pharm. 2008 ; 65 : 2384 – 6 . 3 Gray T , Bertch K, Galt K, et al. Guidelines for therapeutic interchange. Pharmacotherapy. 2004 ; 25 : 1666 – 80 . Crossref Search ADS 4 Holmes DR , Becker JA, Granger CB, et al. ACCF/AHA 2011 health policy statement on therapeutic interchange and substitution: a report of the American College of Cardiology Foundation Clinical Quality Committee. J Am Coll Cardiol. 2011 ; 58 : 1287 – 307 . Crossref Search ADS PubMed 5 Hess G , Sanders KN, Hill J, et al. Therapeutic dose assessment of patient switching from atorvastatin to simvastatin. Am J Manag Care. 2007 ; 13 : 80 – 5 . 6 Teich JM , Merchia PR, Schmiz JL, et al. Effects of computerized physician order entry on prescribing practices. Arch Intern Med. 2000 ; 160 : 2741 – 7 . Crossref Search ADS PubMed 7 Walk SU , Bertsche T, Kaltschmidt J, et al. Rule-based standardised switching of drugs at the interface between primary and tertiary care. Eur J Clin Pharmacol. 2008 ; 64 : 319 – 27 . Crossref Search ADS PubMed 8 Pruszydlo MG , Walk-Fritz SU, Hoppe-Tichy T, et al. Development and evaluation of a computerised clinical decision support system for switching drugs at the interface between primary and tertiary care. BMC Med Inform Dec Mak. 2012 ; 12 : 137 . Crossref Search ADS Copyright © 2018, American Society of Health-System Pharmacists, Inc. All rights reserved.

Dempsey,, Jillian;Gillis,, Christine;Sibicky,, Stephanie;Matta,, Lina;MacRae,, Calum;Kirshenbaum,, James;Faxon,, David;Churchill,, William

doi: 10.2146/ajhp170099pmid: 29976830

Abstract Purpose The utility of a transitions-of-care (TOC) pharmacist intervention focused on improving the quality and safety of the medication process for high-risk cardiovascular patients was evaluated. Methods A quality-improvement initiative was developed for patients with heart failure or acute coronary syndrome followed longitudinally at a hospital˴s outpatient cardiovascular clinic. The TOC pharmacist intervention occurred either before a patient˴s outpatient cardiovascular clinic appointment or during a hospitalization. The major outcome analyzed was the number of unplanned hospital readmissions within 30 days. Additional endpoints evaluated included the time to healthcare utilization, number of medication discrepancies identified, percentage of therapeutic recommendations accepted by a provider, number of medication access issues resolved, patient cost savings, patient satisfaction, and mean time spent on an intervention by the pharmacist per patient encounter. Results A total of 118 patients received the TOC pharmacist intervention. A total of 516 medication discrepancies were identified and corrected, with 55.6% of discrepancies involving cardiovascular medications. A total of 244 recommendations for therapeutic optimization were provided, with an 81% provider acceptance rate and a 100% patient satisfaction rate. Fifty-five patients were provided with medication cost savings, and medication-access issues were resolved for 8 patients. A TOC pharmacist spent means of 98 and 73 minutes on patient education and coordination of care during inpatient and ambulatory encounters, respectively. The 30-day hospital readmission rate for patients with heart failure was reduced by 20%. Conclusion A TOC pharmacist intervention improved the quality and safety of care across both inpatient and ambulatory settings for high-risk cardiovascular patients at our institution. acute coronary syndrome, discharge planning, heart failure, medication access, medication reconciliation, transitions of care Transitions of care (TOCs) are best defined as actions designed to ensure the coordination and continuity of healthcare as patients transfer between different locations or different levels of care within the same location.1,2 Nearly two thirds of postdischarge adverse drug events (ADEs) are medication related, with 29% posing a serious or life-threatening risk, often resulting in preventable emergency department (ED) visits and unplanned hospital admissions.3 Many themes and pitfalls in the transitional care process have been identified and include a paucity of communication among healthcare providers, both within and between the inpatient and ambulatory patient care settings, and scant patient and caregiver education and preparedness for discharge, as well as a lack of provider assessment of patients˴ health literacy, medication adherence, financial capabilities, and overall access to medications.4,–7 Under the Affordable Care Act, enacted in 2010, hospitals face monetary penalties for increased readmissions for selected groups of high-risk patients, including those with heart failure (HF), a history of acute myocardial infarction, pneumonia, and chronic obstructive pulmonary disease.3,8 These penalties are expected to increase through the Hospital Readmissions Reduction Program under the Centers for Medicare and Medicaid Services (CMS) in years to come. Another CMS-based initiative, known as the Hospital Value-Based Purchasing program, rewards acute care hospitals with incentive payments for providing quality care.3,8 Efforts to avoid imminent fiscal penalties and embrace opportunities for incentive reimbursements for quality care have paved the way for a new focus on expansion of advanced hospital-based TOC services. The majority of TOC literature published to date has focused on transitioning high-risk patients from the hospital to the home setting. High-risk populations may be broadly defined or more narrowly identified based on a patient˴s concomitant diseases, number of home medications, age, socioeconomic status, or frequency of readmission. The development of TOC programs continues to gain momentum on a national level, as many studies have proven the value of multidisciplinary interventions by health professionals, including pharmacists, at or shortly after discharge, on the amelioration of postdischarge ADEs.3,9 As medication experts, pharmacists play a unique role in the TOC process by performing medication-related activities and providing various services to facilitate medication education and adherence. Limited studies have solely evaluated a TOC pharmacist˴s intervention within a cardiovascular patient population. Furthermore, the impact of pharmacists on 30-day hospital readmissions and healthcare use remains uncertain. This article describes the outcomes of Project PURPOSE (Pharmacist Utilized Resources to ImProve CardiOvaScular CarE), a pharmacy resident–run TOC pilot service focused on a high-risk cardiovascular patient population. The goal of the service was to improve the quality and safety of cardiovascular patient care across both inpatient and ambulatory settings at a tertiary academic medical center. Methods Project design. We developed a single-center, 6-month, quality-improvement pilot initiative at Brigham and Women˴s Hospital in Boston, Massachusetts. Because the project was a quality-improvement initiative, approval was not required by the hospital˴s institutional review board. The pilot service was implemented in September 2015 and ended in February 2016. Two TOC postgraduate year 2 pharmacy residents and 1 pharmacy practice academic faculty member were dedicated to the service. Residents were available Monday through Friday during afternoon hours only to staff the TOC service. Before the launch of the project, pharmacists were not largely involved in institution-specific TOC initiatives. The TOC pharmacy service was developed in collaboration with our institution˴s cardiology department and aimed to identify high-risk patients in either the hospital˴s outpatient cardiovascular clinic or an inpatient cardiology unit. Once identified, patients were followed longitudinally by a TOC pharmacy resident for the 6-month duration of the pilot service. In the outpatient cardiovascular clinic, a TOC pharmacist was scheduled to meet with patients 30–45 minutes before their cardiology appointment. In the inpatient cardiology unit, a TOC pharmacist met with patients within 24 hours of admission and again before discharge. Postdischarge phone calls were completed 48–72 hours following discharge, and postclinic phone calls were completed on a case-by-case basis as warranted for additional follow-up. Patient selection. Project PURPOSE˴s targeted patient population was defined as (1) patients age 65 years or older with a history of HF and at least 1 admission to our hospital within the past 12 months and (2) patients age 65 years or older with a history of acute coronary syndrome (ACS). Providers were also able to refer patients to the TOC pharmacy service who may not have met the inclusion criteria but who the provider felt would benefit from the intervention. All patients enrolled in the pilot service had to have been followed longitudinally by a cardiologist within our institution˴s outpatient cardiovascular clinic. Patients were excluded from receiving the TOC pharmacist intervention if they did not speak English, were followed by the integrated care management group, had received a heart transplant or ventricular assist device, had cognitive impairment requiring caregiver assistance, resided in a long-term care or assisted-living facility, planned to be discharged to a rehabilitation or skilled nursing facility, or were actively abusing alcohol or other substances. In addition, patients were excluded if they received their outpatient cardiovascular care outside our institution. Intervention. During the outpatient and inpatient visits, the following activities were completed as time permitted: performance of gold-standard medication history and reconciliation, distribution of adherence tools (e.g., pillboxes, medication grids), triage and resolution of medication access issues, review of Medicare prescription insurance plan, identification of cost-saving opportunities, and recommendations for drug therapy optimization after conducting a full pharmacotherapy review of each patient˴s medication regimen. Additional TOC services completed during inpatient encounters included discharge medication review and education and medication bedside delivery. TOC coaching, as defined by Coleman and colleagues,10 was incorporated in all patient encounters. The role of the transitions coach is to encourage patients to assume a more active role in their disease management and care, assert their preferences, and foresee needs as they relate to their care management. The Patient Activation Assessment (PAA) tool was used to monitor patient progress in the following arenas as they relate to the 4 pillars of patient self-management: medication management, use of a patient-centered personal health record, timely care and follow-up, and identification and response to red flags indicative of disease progression. Using this tool, patient progress in each area was scored on a scale of 0–10, with 0 signifying low patient activation and 10 signifying high patient activation.11,12 Outcomes. Metrics for the TOC pharmacist intervention were captured through the residents˴ documentation in the patient˴s electronic health record (EHR), and data were collected and analyzed using Research Electronic Data Capture (REDCap) (Vanderbilt University, Nashville, TN). Outcomes included 30-day hospital readmissions and time to healthcare utilization (e.g., ED and urgent care visits). Data for these outcomes were collected from September 2015 through March 2016. A readmission was defined as a second admission to our institution, regardless of readmission diagnosis, within 30 days after the initial discharge. Baseline readmission rates for the HF patient population were extracted from our institution˴s internal data, calculated by dividing the total number of patients with HF readmitted to the hospital within 30 days of discharge by the total number of HF discharges. Additional data collected and analyzed included change in PAA score, number of identified medication discrepancies, number of medication access issues resolved, patient cost savings procured through prescription coupon distribution, medication substitutions, and Medicare prescription insurance plan reviews. The number of clinical recommendations provided by the pharmacist and accepted by providers, in addition to the mean time spent by a TOC pharmacist on each patient encounter, was also analyzed. In addition, patient satisfaction was assessed in the outpatient clinic via distribution of a postpharmacist visit survey to patients. Statistical analysis. Due to the descriptive nature of this analysis, all outcomes were expressed as means or medians and standard deviations. Sample size calculation to determine statistical significance of 30-day readmission rates for patients with HF was performed using chi-square analysis. This project was designed with 80% power to detect a 20% reduction in 30-day hospital readmission rates for HF patients. The a priori level of significance was 0.05. Results Patient selection. A total of 118 patients were enrolled in Project PURPOSE and received the TOC pharmacist intervention during the 6-month pilot initiative. A total of 2,550 patients met the inclusion criteria and were eligible to receive the intervention; however, given resident availability restrictions, only 118 patients actually received the intervention. Ninety-five percent of patients who received the intervention were age 65 years or older; the remaining 5% of patients were younger than 65 years but were referred to the TOC pharmacy service for medication-related issues. A greater percentage of patients with at least 1 admission to our institution in the preceding year had a history of HF versus ACS (60% versus 19%). Sixteen percent of patients had a history of both HF and ACS. Patients˴ mean ± S.D. age was 72.9 ± 9.8 years, and more than half were male (64%) and Caucasian (81%). The 5 most common comorbidities in descending order of prevalence were hypertension (73%), hyperlipidemia (51%), arrhythmia (45%), coronary artery disease (40%), and diabetes (38%). Approximately 60% of patients were Medicare beneficiaries, 15% received both state and federal funding through Medicaid, and the remaining 30% were insured by private commercial payers. In addition, 60% of patients had a low understanding of their medication regimen, and only 6% were determined to have a high understanding of their medications by a TOC pharmacist. Patients˴ baseline demographics are summarized in Table 1. Table 1 Baseline Demographics of High-Risk Cardiovascular Patients (n = 118) Variablea Value Mean ± S.D. age, yr 72.9 ± 9.8 Male, no. (%) patients 76 (64) Ethnicity, no. (%) patients  Caucasian 96 (81)  African American 18 (15)  Other 4 (4) High-risk criteria, no. (%) patients  Age ≥ 65 yr 112 (95)  HF plus more than 1 HF-related admission in past 12 mo 71 (60)  ACS (NSTEMI/STEMI/unstable angina) 22 (19)  HF plus more than 1 HF-related admission in past 12 mo plus ACS 19 (16)  Concern about medication-related issues 6 (5) Medical history, no. (%) patients  Hypertension 86 (73)  Hyperlipidemia 60 (51)  Arrhythmia 53 (45)  Coronary artery disease 47 (40)  Diabetes 45 (38)  End-stage renal disease 43 (36)  Stroke/transient ischemic attack 20 (17) Prescription insurance type, no. (%) patients  Medicare 70 (59)  Medicaid 18 (15)  Commercial/private 35 (30) Degree of patient medication understanding, no. (%) patientsb  Low 70 (59)  Intermediate 41 (35)  High 7 (6) Variablea Value Mean ± S.D. age, yr 72.9 ± 9.8 Male, no. (%) patients 76 (64) Ethnicity, no. (%) patients  Caucasian 96 (81)  African American 18 (15)  Other 4 (4) High-risk criteria, no. (%) patients  Age ≥ 65 yr 112 (95)  HF plus more than 1 HF-related admission in past 12 mo 71 (60)  ACS (NSTEMI/STEMI/unstable angina) 22 (19)  HF plus more than 1 HF-related admission in past 12 mo plus ACS 19 (16)  Concern about medication-related issues 6 (5) Medical history, no. (%) patients  Hypertension 86 (73)  Hyperlipidemia 60 (51)  Arrhythmia 53 (45)  Coronary artery disease 47 (40)  Diabetes 45 (38)  End-stage renal disease 43 (36)  Stroke/transient ischemic attack 20 (17) Prescription insurance type, no. (%) patients  Medicare 70 (59)  Medicaid 18 (15)  Commercial/private 35 (30) Degree of patient medication understanding, no. (%) patientsb  Low 70 (59)  Intermediate 41 (35)  High 7 (6) aHF = heart failure, ACS = acute coronary syndrome, NSTEMI = non-ST-segment elevation myocardial infarction, STEMI = ST-segment elevation myocardial infarction. bLow = inconsistent or incomplete patient understanding of the majority of medication indications, doses, strengths, and frequencies, intermediate = patient is able to identify medications by name or indication but not both and has little understanding of dose (e.g., ˵I take the blue blood pressure pill once a day˶), high = patient understands the majority of medication indications, doses, strengths, and frequencies. View Large Table 1 Baseline Demographics of High-Risk Cardiovascular Patients (n = 118) Variablea Value Mean ± S.D. age, yr 72.9 ± 9.8 Male, no. (%) patients 76 (64) Ethnicity, no. (%) patients  Caucasian 96 (81)  African American 18 (15)  Other 4 (4) High-risk criteria, no. (%) patients  Age ≥ 65 yr 112 (95)  HF plus more than 1 HF-related admission in past 12 mo 71 (60)  ACS (NSTEMI/STEMI/unstable angina) 22 (19)  HF plus more than 1 HF-related admission in past 12 mo plus ACS 19 (16)  Concern about medication-related issues 6 (5) Medical history, no. (%) patients  Hypertension 86 (73)  Hyperlipidemia 60 (51)  Arrhythmia 53 (45)  Coronary artery disease 47 (40)  Diabetes 45 (38)  End-stage renal disease 43 (36)  Stroke/transient ischemic attack 20 (17) Prescription insurance type, no. (%) patients  Medicare 70 (59)  Medicaid 18 (15)  Commercial/private 35 (30) Degree of patient medication understanding, no. (%) patientsb  Low 70 (59)  Intermediate 41 (35)  High 7 (6) Variablea Value Mean ± S.D. age, yr 72.9 ± 9.8 Male, no. (%) patients 76 (64) Ethnicity, no. (%) patients  Caucasian 96 (81)  African American 18 (15)  Other 4 (4) High-risk criteria, no. (%) patients  Age ≥ 65 yr 112 (95)  HF plus more than 1 HF-related admission in past 12 mo 71 (60)  ACS (NSTEMI/STEMI/unstable angina) 22 (19)  HF plus more than 1 HF-related admission in past 12 mo plus ACS 19 (16)  Concern about medication-related issues 6 (5) Medical history, no. (%) patients  Hypertension 86 (73)  Hyperlipidemia 60 (51)  Arrhythmia 53 (45)  Coronary artery disease 47 (40)  Diabetes 45 (38)  End-stage renal disease 43 (36)  Stroke/transient ischemic attack 20 (17) Prescription insurance type, no. (%) patients  Medicare 70 (59)  Medicaid 18 (15)  Commercial/private 35 (30) Degree of patient medication understanding, no. (%) patientsb  Low 70 (59)  Intermediate 41 (35)  High 7 (6) aHF = heart failure, ACS = acute coronary syndrome, NSTEMI = non-ST-segment elevation myocardial infarction, STEMI = ST-segment elevation myocardial infarction. bLow = inconsistent or incomplete patient understanding of the majority of medication indications, doses, strengths, and frequencies, intermediate = patient is able to identify medications by name or indication but not both and has little understanding of dose (e.g., ˵I take the blue blood pressure pill once a day˶), high = patient understands the majority of medication indications, doses, strengths, and frequencies. View Large Patient encounters. A total of 176 encounters incorporating the TOC pharmacist intervention were completed, including 89 outpatient cardiovascular encounters, 60 inpatient encounters with admission and discharge visits reported as separate entities, and 27 postdischarge and postclinic follow-up phone call encounters. Thirty-seven percent (n = 22) of the inpatient encounters included only an admission visit, either because a TOC pharmacist was not available to see the patient at discharge or the patient˴s goals of care changed (e.g., palliative care, hospice placement) during hospitalization. Sixty-three percent (n = 24) of patients discharged from the hospital were reached via telephone within 72 hours of discharge for medication review and follow-up. Hospital readmission and healthcare utilization. Baseline hospital readmission rates were determined to be 20% and 9% for patients with a history of HF and ACS, respectively. Seven of the 31 HF patients who received the TOC pharmacist intervention in the inpatient setting were readmitted within 30 days, and none of the 4 ACS patients seen in the inpatient setting were readmitted in the 30 days after their intervention. Thus, the 30-day hospital readmission rate for patients with HF or ACS was reduced by 20% (p = 0.596) through the TOC pharmacy service. Notably, 2 of the 7 patients who were readmitted were referred to the service by providers. The overall 30-day rate of healthcare use, defined as an urgent care or ED visit to our institution, was 4% (n = 5). Medication discrepancies. Medication discrepancies were determined by a process previously developed and described by Pippins et al.13 A total of 516 medication discrepancies occurred during outpatient encounters (383 [74.2%]) and inpatient encounters or postdischarge phone calls (133 [25.8%]) (Table 2). Most inpatient medication discrepancies (88.0%, n = 117) were identified during admission; 9.0% (n = 12) were identified at discharge. A mean ± S.D. of 4 ± 2.76 medication discrepancies were identified per patient in the overall population, with each patient reporting a mean ± S.D. of 14 ± 5.81 home medications. Table 2 Medication Discrepancies Identified for High-Risk Cardiovascular Patients Variablea Total No. (%) Discrepancies Ambulatory (n = 383) Inpatient (n = 130) Postdischarge Phone Call (n = 3) Category of medication discrepancy  Cardiovascular agent 127 (33.2) 45 (34.6) 2 (66.7)  Antiplatelet 17 (4.4) 6 (4.6) 1 (33.3)  Anticoagulant 13 (3.4) 4 (3.1) 0  Insulin 6 (1.6) 3 (2.3) 0  Oral antihyperglycemic 10 (2.6) 5 (3.8) 0  Opioid 8 (2.1) 2 (1.5) 0  Benzodiazepine 6 (1.6) 4 (3.1) 0  Other 196 (51.2) 61 (46.9) 0 Patient level of contribution  Discontinuation due to medication ADE/intolerance 11 (2.9) 1 (0.8) 0  Chose not to fill prescription at pharmacy 2 (0.5) 5 (3.8) 0  Did not take medication due to cost 2 (0.5) 7 (5.4) 0  Intentional nonadherence 36 (9.4) 21 (16.2) 0  Nonintentional nonadherence 8 (2.1) 7 (5.4) 3 (100.0)  Performance deficit 3 (0.8) 2 (1.5) 0 System level of contribution  Conflicting information between sources 5 (1.3) 4 (3.1) 0  Discharge instructions incomplete 10 (2.6) 7 (5.4) 1 (33.3)  Medication omission 146 (38.1) 27 (20.8) 2 (66.7)  Medication duplication 6 (1.5) 6 (4.6) 1 (33.3)  Additional medication 83 (21.7) 37 (28.5) 0  Incorrect medication dose 92 (24.0) 32 (24.6) 0  Incorrect medication route 16 (4.2) 4 (3.1) 0  Incorrect medication frequency 59 (15.4) 18 (13.8) 0  Incorrect medication formulation 14 (3.7) 8 (6.2) 0  Patient cognitive impairment not recognized/dexterity issue 3 (0.7) 0 0  No caregiver assistance 2 (0.5) 3 (2.3) 3 (100.0)  PAML error 0… 66 (50.8) 0  Failure to reconcile 0… 15 (11.5) 0 Discrepancy resolution  Discussed potential benefits of adhering or harm of nonadherence to medication regimen with patient 300 (78.3) 8 (6.2) 3 (100.0)  Encouraged patient to call PCP and/or specialist 9 (2.3) 4 (3.1) 0  Provided resource information to facilitate adherence 15 (3.9) 15 (11.5) 3 (100.0)  Notified responding clinician in inpatient setting or provider in ambulatory setting 383 (100.0) 130 (100.0) 3 (100.0)  Adjusted medication regimen based on TOC pharmacist recommendation 22 (5.9) 16 (12.6) 0 Variablea Total No. (%) Discrepancies Ambulatory (n = 383) Inpatient (n = 130) Postdischarge Phone Call (n = 3) Category of medication discrepancy  Cardiovascular agent 127 (33.2) 45 (34.6) 2 (66.7)  Antiplatelet 17 (4.4) 6 (4.6) 1 (33.3)  Anticoagulant 13 (3.4) 4 (3.1) 0  Insulin 6 (1.6) 3 (2.3) 0  Oral antihyperglycemic 10 (2.6) 5 (3.8) 0  Opioid 8 (2.1) 2 (1.5) 0  Benzodiazepine 6 (1.6) 4 (3.1) 0  Other 196 (51.2) 61 (46.9) 0 Patient level of contribution  Discontinuation due to medication ADE/intolerance 11 (2.9) 1 (0.8) 0  Chose not to fill prescription at pharmacy 2 (0.5) 5 (3.8) 0  Did not take medication due to cost 2 (0.5) 7 (5.4) 0  Intentional nonadherence 36 (9.4) 21 (16.2) 0  Nonintentional nonadherence 8 (2.1) 7 (5.4) 3 (100.0)  Performance deficit 3 (0.8) 2 (1.5) 0 System level of contribution  Conflicting information between sources 5 (1.3) 4 (3.1) 0  Discharge instructions incomplete 10 (2.6) 7 (5.4) 1 (33.3)  Medication omission 146 (38.1) 27 (20.8) 2 (66.7)  Medication duplication 6 (1.5) 6 (4.6) 1 (33.3)  Additional medication 83 (21.7) 37 (28.5) 0  Incorrect medication dose 92 (24.0) 32 (24.6) 0  Incorrect medication route 16 (4.2) 4 (3.1) 0  Incorrect medication frequency 59 (15.4) 18 (13.8) 0  Incorrect medication formulation 14 (3.7) 8 (6.2) 0  Patient cognitive impairment not recognized/dexterity issue 3 (0.7) 0 0  No caregiver assistance 2 (0.5) 3 (2.3) 3 (100.0)  PAML error 0… 66 (50.8) 0  Failure to reconcile 0… 15 (11.5) 0 Discrepancy resolution  Discussed potential benefits of adhering or harm of nonadherence to medication regimen with patient 300 (78.3) 8 (6.2) 3 (100.0)  Encouraged patient to call PCP and/or specialist 9 (2.3) 4 (3.1) 0  Provided resource information to facilitate adherence 15 (3.9) 15 (11.5) 3 (100.0)  Notified responding clinician in inpatient setting or provider in ambulatory setting 383 (100.0) 130 (100.0) 3 (100.0)  Adjusted medication regimen based on TOC pharmacist recommendation 22 (5.9) 16 (12.6) 0 aADE = adverse drug effect, PAML = preadmission medication list, PCP = primary care provider, TOC = transitions of care. bLow = inconsistent or incomplete patient understanding of the majority of medication indications, doses, strengths, and frequencies, intermediate = patient is able to identify medications by name or indication but not both and has little understanding of dose (e.g., ˵I take the blue blood pressure pill once a day˶), high = patient understands the majority of medication indications, doses, strengths, and frequencies. View Large Table 2 Medication Discrepancies Identified for High-Risk Cardiovascular Patients Variablea Total No. (%) Discrepancies Ambulatory (n = 383) Inpatient (n = 130) Postdischarge Phone Call (n = 3) Category of medication discrepancy  Cardiovascular agent 127 (33.2) 45 (34.6) 2 (66.7)  Antiplatelet 17 (4.4) 6 (4.6) 1 (33.3)  Anticoagulant 13 (3.4) 4 (3.1) 0  Insulin 6 (1.6) 3 (2.3) 0  Oral antihyperglycemic 10 (2.6) 5 (3.8) 0  Opioid 8 (2.1) 2 (1.5) 0  Benzodiazepine 6 (1.6) 4 (3.1) 0  Other 196 (51.2) 61 (46.9) 0 Patient level of contribution  Discontinuation due to medication ADE/intolerance 11 (2.9) 1 (0.8) 0  Chose not to fill prescription at pharmacy 2 (0.5) 5 (3.8) 0  Did not take medication due to cost 2 (0.5) 7 (5.4) 0  Intentional nonadherence 36 (9.4) 21 (16.2) 0  Nonintentional nonadherence 8 (2.1) 7 (5.4) 3 (100.0)  Performance deficit 3 (0.8) 2 (1.5) 0 System level of contribution  Conflicting information between sources 5 (1.3) 4 (3.1) 0  Discharge instructions incomplete 10 (2.6) 7 (5.4) 1 (33.3)  Medication omission 146 (38.1) 27 (20.8) 2 (66.7)  Medication duplication 6 (1.5) 6 (4.6) 1 (33.3)  Additional medication 83 (21.7) 37 (28.5) 0  Incorrect medication dose 92 (24.0) 32 (24.6) 0  Incorrect medication route 16 (4.2) 4 (3.1) 0  Incorrect medication frequency 59 (15.4) 18 (13.8) 0  Incorrect medication formulation 14 (3.7) 8 (6.2) 0  Patient cognitive impairment not recognized/dexterity issue 3 (0.7) 0 0  No caregiver assistance 2 (0.5) 3 (2.3) 3 (100.0)  PAML error 0… 66 (50.8) 0  Failure to reconcile 0… 15 (11.5) 0 Discrepancy resolution  Discussed potential benefits of adhering or harm of nonadherence to medication regimen with patient 300 (78.3) 8 (6.2) 3 (100.0)  Encouraged patient to call PCP and/or specialist 9 (2.3) 4 (3.1) 0  Provided resource information to facilitate adherence 15 (3.9) 15 (11.5) 3 (100.0)  Notified responding clinician in inpatient setting or provider in ambulatory setting 383 (100.0) 130 (100.0) 3 (100.0)  Adjusted medication regimen based on TOC pharmacist recommendation 22 (5.9) 16 (12.6) 0 Variablea Total No. (%) Discrepancies Ambulatory (n = 383) Inpatient (n = 130) Postdischarge Phone Call (n = 3) Category of medication discrepancy  Cardiovascular agent 127 (33.2) 45 (34.6) 2 (66.7)  Antiplatelet 17 (4.4) 6 (4.6) 1 (33.3)  Anticoagulant 13 (3.4) 4 (3.1) 0  Insulin 6 (1.6) 3 (2.3) 0  Oral antihyperglycemic 10 (2.6) 5 (3.8) 0  Opioid 8 (2.1) 2 (1.5) 0  Benzodiazepine 6 (1.6) 4 (3.1) 0  Other 196 (51.2) 61 (46.9) 0 Patient level of contribution  Discontinuation due to medication ADE/intolerance 11 (2.9) 1 (0.8) 0  Chose not to fill prescription at pharmacy 2 (0.5) 5 (3.8) 0  Did not take medication due to cost 2 (0.5) 7 (5.4) 0  Intentional nonadherence 36 (9.4) 21 (16.2) 0  Nonintentional nonadherence 8 (2.1) 7 (5.4) 3 (100.0)  Performance deficit 3 (0.8) 2 (1.5) 0 System level of contribution  Conflicting information between sources 5 (1.3) 4 (3.1) 0  Discharge instructions incomplete 10 (2.6) 7 (5.4) 1 (33.3)  Medication omission 146 (38.1) 27 (20.8) 2 (66.7)  Medication duplication 6 (1.5) 6 (4.6) 1 (33.3)  Additional medication 83 (21.7) 37 (28.5) 0  Incorrect medication dose 92 (24.0) 32 (24.6) 0  Incorrect medication route 16 (4.2) 4 (3.1) 0  Incorrect medication frequency 59 (15.4) 18 (13.8) 0  Incorrect medication formulation 14 (3.7) 8 (6.2) 0  Patient cognitive impairment not recognized/dexterity issue 3 (0.7) 0 0  No caregiver assistance 2 (0.5) 3 (2.3) 3 (100.0)  PAML error 0… 66 (50.8) 0  Failure to reconcile 0… 15 (11.5) 0 Discrepancy resolution  Discussed potential benefits of adhering or harm of nonadherence to medication regimen with patient 300 (78.3) 8 (6.2) 3 (100.0)  Encouraged patient to call PCP and/or specialist 9 (2.3) 4 (3.1) 0  Provided resource information to facilitate adherence 15 (3.9) 15 (11.5) 3 (100.0)  Notified responding clinician in inpatient setting or provider in ambulatory setting 383 (100.0) 130 (100.0) 3 (100.0)  Adjusted medication regimen based on TOC pharmacist recommendation 22 (5.9) 16 (12.6) 0 aADE = adverse drug effect, PAML = preadmission medication list, PCP = primary care provider, TOC = transitions of care. bLow = inconsistent or incomplete patient understanding of the majority of medication indications, doses, strengths, and frequencies, intermediate = patient is able to identify medications by name or indication but not both and has little understanding of dose (e.g., ˵I take the blue blood pressure pill once a day˶), high = patient understands the majority of medication indications, doses, strengths, and frequencies. View Large Approximately 55.6% of all discrepancies involved cardiovascular medications, with various levels of patient and system contributors. Relative to the total number of discrepancies in both the inpatient and outpatient settings, the rate of intentional nonadherence (e.g., patient chose not to take medication as prescribed) was greater than the rate of nonintentional nonadherence (e.g., patient did not understand how to take the medication) (11.1% versus 3.1%, respectively). Over 50% of inpatient medication discrepancies were categorized as preadmission medication list errors. Preadmission medication list errors were defined as medications that were incorrectly listed, added, or omitted by the nurse or provider from a patient˴s EHR home medication list during medication reconciliation on hospital admission. A total of 11.5% of inpatient discrepancies were categorized as failure to reconcile, meaning that medications were listed on a patient˴s preadmission medication list but, without explanation or reason, were not translated as inpatient orders. Pharmacist recommendations and time utilization. A total of 244 recommendations for therapeutic optimization, 54 in the inpatient setting and 190 in the outpatient cardiovascular setting, were provided by a pharmacist through the TOC pharmacy service, with 198 of those recommendations (81.2%) accepted and implemented by a provider. The majority of therapeutic recommendations included initiating therapeutic drug and laboratory monitoring (37.2%), medication initiation (25.7%), titration of current medication (19.7%), and discontinuation of current medication (16.9%). In the inpatient setting, a TOC pharmacist spent a mean ± S.D. of 55 ± 25.6 minutes counseling or educating each patient and 42.5 ± 21.8 minutes coordinating patient care and follow-up on medication-related issues. In the outpatient clinic, a TOC pharmacist spent a mean ± S.D. of 41.2 ± 13.1 minutes on counseling and education and 32.2 ± 13.1 minutes on coordination of care and follow-up. It is worth noting that a pharmacist˴s visit in the outpatient clinic was often abbreviated due to delays in patient arrival and unpredictable changes in provider scheduling. Figure 1 provides additional information regarding pharmacists˴ time spent on patient interventions. Figure 1 View largeDownload slide Average time spent by pharmacist per patient on education and coordination of care and follow-up. Figure 1 View largeDownload slide Average time spent by pharmacist per patient on education and coordination of care and follow-up. Patient cost savings. A total of 55 patients were afforded cost-saving opportunities. Patient cost savings were provided through several different avenues including drug coupon distribution for patients with commercial or private prescription insurance plans, medication discontinuation, generic substitution, Medicare tier exception approval, and Medicare Part D plan review during the Medicare open-enrollment period. After a pharmacist encounter, at least 1 medication was discontinued for 29 patients, generic medication substitution occurred for 3 patients (annual savings of $50–$1,850), 11 patients received drug coupons (annual savings of $90–$1,500), and a tier exception request was processed for 1 patient who was unable to afford a medication copayment (annual savings of $372). Medicare Part D reviews were completed for 11 patients during the 2-month Medicare open-enrollment period. The mean saving per patient per year was $460 (range, $50–$1,800). A TOC pharmacist recommended that 17 patients switch to the preferred pharmacy associated with their current Medicare plan, and 54 patients were referred to a group within the state Department of Elder Affairs for insurance plan review and drug assistance program screening. As an additional means to increase medication access for patients, prior authorizations were triaged before discharge and medications were delivered to patients˴ bedside through our institution˴s outpatient pharmacy medication delivery program. A total of 7 prior authorizations were successfully processed through the TOC pharmacy service, and 19 prescriptions were filled through the medication bedside delivery program. Patient activation and satisfaction. The mean overall baseline PAA score was 5.9 ± 2.3. The TOC pharmacist increased patients˴ mean ± S.D. activation score by 2.14 ± 1.21 points (p < 0.0001). In the ambulatory cardiovascular clinic, 50 patients were surveyed after the TOC pharmacist˴s visit to assess whether the visit was beneficial. All patients (100%) were completely satisfied with the visit and felt they benefited from the services offered through Project PURPOSE. Discussion Defining a target patient population. In the developmental stages of this project, we were faced with the difficult decision of whom the program should target. Some institutions have used predictive analytics or statistical algorithms as a means to identify the likelihood of future outcomes based on historical data. The challenge with this type of risk-model development is that it requires real-time analysis and additional staffing resources, as historical data are often not routinely collected.14 Predictive risk models are often multifaceted and, as they relate to TOC service development, can differentiate risk based on age, number of previous hospital admissions, chronic diseases, number of medications, and number of high-risk medications. Studies have aimed to derive and validate prediction models for potentially avoidable readmissions and therefore have not included specific TOC interventions.15,–18 A number of other studies targeted specific patient populations for TOC intervention.12,19,–24 Naylor et al.22 conducted 1 of the first trials that targeted patients age 65 years or older with 1 of 9 medical diagnoses (HF, angina, acute myocardial infarction, respiratory tract infection, coronary artery bypass graft, cardiac valve replacement, major small bowel procedure, large bowel procedure, or orthopedic procedures of lower extremities) for their nurse-driven comprehensive discharge planning and home follow-up initiative. In 2004, Naylor and colleagues25 developed a TOC program across 6 academic and community hospitals in Philadelphia that targeted patients age 65 years or older with HF. Coleman et al.12 designed a randomized controlled trial that targeted patients with 1 of 11 diagnoses (stroke, HF, coronary artery disease, cardiac arrhythmia, chronic obstructive pulmonary disease, diabetes mellitus, spinal stenosis, hip fracture, peripheral vascular disease, deep vein thrombosis, and pulmonary embolism) within a large integrated healthcare delivery system in Colorado. There is an abundance of patient-specific factors that can predict potentially avoidable readmissions and a number of patient populations to target for a TOC intervention. After reviewing internal retrospective readmission data and previously published literature highlighting TOC service development, the cardiology and pharmacy departments within our institution agreed on a target patient population that included both HF and ACS patients age 65 years or older. Support for such a target population was further enhanced by CMS˴s Hospital Readmissions Reduction Program initiative and incentives for reduced reimbursement penalties. TOC interventions and associated outcomes. Common themes have emerged among TOC interventions to prevent hospital readmissions. The majority of interventions commence early during a patient˴s hospitalization and continue for at least 1 week after discharge.26 Specific details of TOC interventions are not consistently described; however, 7 commonalities incorporated in several TOC intervention studies have been identified: (1) patient education incorporating the teach-back method to ensure patient understanding of content reviewed,7,12,25,27,–33 (2) postdischarge phone call follow-up review,7,12,25,27,–29,31,33,34 (3) early patient assessment after hospital admission,7,12,25,29,32,–34 (4) medication reconciliation,7,30,–32 (5) inclusion of caregivers,12,25,30,–33,35 (6) postdischarge home visits,12,25,27,29,30,33 and (7) postdischarge provider handoff.7,12,28,30 Various healthcare groups have directed TOC interventions that are often multidisciplinary in nature and include nurses, social workers, pharmacists, and physicians. There is varying involvement and assigned responsibilities from each discipline, though collectively the groups share the same goal—to keep patients out of the hospital, improve patient activation and empowerment, improve patient safety, and reduce ADEs. To our knowledge, few TOC interventional studies have either incorporated a clinical pharmacist into the TOC risk-reduction model or solely evaluated a TOC pharmacy resident˴s intervention on TOC-associated outcomes. Project RED demonstrated that a nurse discharge advocate and pharmacist collaborating to coordinate hospital discharge, educate patients, and reconcile medications led to fewer ED visits and rehospitalizations compared with usual care alone.7 The PILL-CVD study incorporated pharmacist-assisted medication reconciliation, inpatient pharmacist counseling, low-literacy adherence aids, and individualized phone call follow-up after discharge.36 Study investigators concluded that clinically important medication errors were very common, affecting 50.8% of patients during the first 30 days after hospital discharge. In both of the aforementioned studies, it was difficult to determine solely the pharmacist˴s impact on ED visits, hospital readmissions, and the reduction of medication errors.7,36 Conklin et al.37 derived a quality-improvement program encompassing admission, discharge, and postdischarge medication reconciliation via a pharmacist. Their TOC service incorporated 4 phases of medication reconciliation, involved only 1 pharmacist, and is 1 of very few TOC services that have truly followed patients across the continuum of care, including the inpatient, postdischarge, and outpatient settings. Another TOC intervention developed specifically for patients with HF included admission and discharge medication reconciliation in addition to patient counseling and education completed by pharmacists, pharmacy residents, and students.38 Lastly, Truong and Backes8 evaluated the impact of a resident pharmacist on 30-day HF readmission rates and HF core measure compliance rates. As part of this intervention, the pharmacy resident completed admission and discharge medication review, identified and resolved discrepancies, provided clinical recommendations to the inpatient pharmacy team based on daily monitoring of laboratory data, facilitated medication bedside delivery for nonadherent patients, provided education and counseling, and worked with case managers and social workers to secure appropriate medications for patients. Through this intervention, HF patients had a significantly lower mean 30-day all-cause readmission rate compared with HF patients who received usual care (12% versus 24%, respectively).8 This study most closely mirrored our quality-improvement initiative, though patients in this study only received the TOC intervention in the inpatient setting and were not followed across the continuum of care as they were with our resident-run service. In addition, with our service, patients were followed for more than 30 days until the pilot end date, and more extensive TOC interventions were used, including medication cost-saving opportunities, assessment of patient activation, and resolution of medication access issues. The success of Project PURPOSE can be attributed in part to service design, in that a TOC pharmacy resident was truly able to follow patients across the continuum of care for the 6-month duration of the pilot. The service was novel in that it captured high-risk patients in both the inpatient and outpatient settings. If a patient captured in the outpatient setting was admitted, the TOC pharmacy resident would see the patient again on admission and for a third encounter before discharge. If a patient enrolled in the service was discharged, a TOC pharmacy resident would see the patient again in our institution˴s outpatient cardiovascular clinic. The extent of activities completed and services offered as part of this TOC intervention surpasses previously published project and/or service interventions. In addition, TOC coaching was incorporated into all patient encounters, as the TOC pharmacy residents received training from the Coleman Care Transitions group. Although not significant, due to pharmacist availability restrictions resulting in an inability to meet project power, 30-day hospital readmission rates for patients with HF were reduced by 20%. Future randomized controlled trials are required to confirm a reduction in readmissions attributable to the intervention outlined above. Limitations and future research. Limitations of Project PURPOSE are important to recognize. This was a nonrandomized, single-site intervention that lacked a comparator group. Given that the TOC pharmacy service was a 6-month pilot led primarily by 2 TOC pharmacy residents devoted to the service only during weekday afternoons, in addition to having their time split between inpatient and ambulatory settings, only 118 of the 2,550 eligible patients received the TOC intervention. In addition, not all patients received the full intervention as intended, given resident availability restrictions, though the vast majority did. Patients were often discharged over the weekend, and the TOC pharmacist therefore missed their discharge intervention; however, follow-up via a postdischarge phone call with the patient was always attempted. Given implementation of a new EHR system during the pilot period, we were unable to capture readmissions and healthcare use outside of our institution; therefore, observed 30-day readmission and healthcare utilization rates may have been underestimated. On the other hand, patients referred to the service by a cardiologist due to medication-related issues were complicated, often having an extensive history of medication and dietary noncompliance with 5 or more hospital admissions within the past year. These complex patients constituted 40% of 30-day readmissions during the pilot period. Given the success of this quality-improvement pilot initiative, the pharmacy department is working with leadership to gain funding to implement this service full time. This project design could be implemented across the continuum of care for other high-risk patients (e.g., liver disease, renal disease, solid organ transplantation). Conclusion A TOC pharmacist intervention improved the quality and safety of care across both inpatient and outpatients settings for high-risk cardiovascular patients at our institution. Christine Gillis and Jillian Dempsey Christine M. Gillis, Pharm.D., BCPS, is a senior clinical pharmacist at Brigham and Women˴s Hospital (BWH) in Boston, MA. She received her doctor of pharmacy degree from MCPHS University Boston in 2014. Dr. Gillis completed an ASHP-accredited postgraduate year 1 residency in hospital pharmacy at BWH and a postgraduate year 2 residency focusing on transitions of care. She continues to contribute to the area of transitions of care through her work in general medicine and quality-improvement projects aimed at improving care transitions for patients taking high-risk medications. Jillian T. Dempsey, Pharm.D., BCPS, is a senior clinical pharmacist at BWH. She graduated with a doctor of pharmacy degree from MCPHS University Boston in 2014. Dr. Dempsey completed an ASHP-accredited postgraduate year 1 residency in pharmacy practice at BWH and remained at BWH to complete a postgraduate year 2 pharmacy residency specializing in transitions of care. Throughout her residency training, Dr. Dempsey collaborated with several multidisciplinary care teams on research and projects aimed at streamlining the care transitions process in an effort to decrease healthcare utilization and costs, reduce unnecessary hospitalizations, and improve the quality and safety of patient care. Disclosures The authors have declared no potential conflicts of interest. References 1 Coleman EA and Boult C. Improving the quality of transitional care for persons with complex care needs. J Am Geriatr Soc. 2003 ; 51 : 556 – 67 . Crossref Search ADS PubMed 2 Coleman EA . Falling through the cracks: challenges and opportunities for improving transitional care for persons with continuous complex care needs. J Am Geriatr Soc. 2003 ; 51 : 549 – 55 . Crossref Search ADS PubMed 3 Walker PC , Tucker JN and Mason NA. An advanced pharmacy practice experience in transitional care. Am J Pharm Educ. 2010 ; 74 ( 2 ): 1 – 6 . Crossref Search ADS PubMed 4 Greenwald JL , Denham CR and Jack BW. The hospital discharge: a review of a high risk care transition with highlights of a reengineered discharge process. J Patient Saf. 2007 ; 3 : 97 – 106 . Crossref Search ADS 5 Wachter RM . Hospitalists in the United States—mission accomplished or work in progress? N Engl J Med. 2004 ; 350 : 1935 – 6 . Crossref Search ADS PubMed 6 Moore C , Wisnivesky J, Williams S and McGinn T. Medical errors related to discontinuity of care from an inpatient to an outpatient setting. J Gen Intern Med. 2003 ; 18 : 646 – 51 . Crossref Search ADS PubMed 7 Jack BW , Chetty VK, Anthony D, et al. A reengineered hospital discharge program to decrease rehospitalization. Ann Intern Med. 2009 ; 150 : 178 – 87 . Crossref Search ADS PubMed 8 Truong JT and Backes AC. The impact of a continuum of care resident pharmacist on heart failure readmissions and discharge instructions at a community hospital. SAGE Open Med. 2015 ; 3 : 2050312115577986 . Crossref Search ADS PubMed 9 Naylor M and Keating SA. Transitional care. Am J Nurs. 2008 ; 108 ( suppl ): 58 – 63 . Crossref Search ADS PubMed 10 Coleman EA , Rosenbek SA and Roman SP. Disseminating evidence-based care into practice. Popul Health Manag. 2013 ; 16 : 227 – 34 . Crossref Search ADS PubMed 11 Coleman EA , Smith JD, Frank JC, et al. Preparing patients and caregivers to participate in care delivered across settings: the Care Transitions Intervention. J Am Geriatr Soc. 2004 ; 52 : 1817 – 25 . Crossref Search ADS PubMed 12 Coleman EA , Parry C, Chalmers S and Min SJ. The Care Transitions Intervention: results of a randomized controlled trial. Arch Intern Med. 2006 ; 166 : 1822 – 8 . Crossref Search ADS PubMed 13 Pippins JR , Gandhi TK, Hamann C, et al. Classifying and predicting errors of inpatient medication reconciliation. J Gen Intern Med. 2008 ; 23 : 1414 – 22 . Crossref Search ADS PubMed 14 Predictive Analytics in Practice . HBR SAS Institute Inc. Accessed 1 Jun 2016. 15 Hasan O , Meltzer DO, Shaykevich SA, et al. Hospital readmission in general medicine patients: a prediction model. J Gen Intern Med. 2009 ; 25 : 211 – 9 . Crossref Search ADS PubMed 16 Amarasingham R , Moore BJ, Tabak YP, et al. An automated model to identify heart failure patients at risk for 30-day readmission or death using electronic medical record data. Med Care. 2010 ; 48 : 981 – 8 . Crossref Search ADS PubMed 17 Allaudeen N , Vidyarthi A, Maselli J and Auerbach A. Redefining readmission risk factors for general medicine patients. J Hosp Med. 2011 ; 6 : 54 – 60 . Crossref Search ADS PubMed 18 Donze J , Aujesky D, Williams D and Schnipper JL. Potentially avoidable 30-day hospital readmissions in medical patients. JAMA Intern Med. 2013 ; 173 : 632 – 8 . Crossref Search ADS PubMed 19 Anderson C , Deepak BV, Amoateng-Adjepong Y and Zarich S. Benefits of comprehensive inpatient education and discharge planning combined with outpatient support in elderly patients with congestive heart failure. Congest Heart Fail. 2005 ; 11 : 315 – 21 . Crossref Search ADS PubMed 20 Andersen HE , Schultz-Larsen K, Kreiner S, et al. Can readmission after stroke be prevented? Results of a randomized clinical study: a postdischarge follow-up service for stroke survivors. Stroke. 2000 ; 31 : 1038 – 45 . Crossref Search ADS PubMed 21 Naylor M , Brooten D, Jones R, et al. Comprehensive discharge planning for the hospitalized elderly. A randomized clinical trial. Ann Intern Med. 1994 ; 120 : 999 – 1006 . Crossref Search ADS PubMed 22 Naylor MD , Brooten D, Campbell R, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: a randomized clinical trial. JAMA. 1999 ; 281 : 613 – 20 . Crossref Search ADS PubMed 23 Evans RL and Hendricks RD. Evaluating hospital discharge planning: a randomized clinical trial. Med Care. 1993 ; 31 : 358 – 70 . Crossref Search ADS PubMed 24 Phillips CO , Wright SM, Kern DE, et al. Comprehensive discharge planning with postdischarge support for older patients with congestive heart failure: a meta-analysis. JAMA. 2004 ; 291 : 1358 – 67 . Crossref Search ADS PubMed 25 Naylor MD , Brooten DA, Campbell RL, et al. Transitional care of older adults hospitalized with heart failure: a randomized controlled trial. J Am Geriatr Soc. 2004 ; 52 : 675 – 84 . Crossref Search ADS PubMed 26 Albert NM , Barnason S, Deswal A, et al. Transitions of care in heart failure: a scientific statement from the American Heart Association. Circ Heart Fail. 2015 ; 8 : 384 – 409 . Crossref Search ADS PubMed 27 Voss R , Gardner R, Baier R, et al. The Care Transitions Intervention: translating from efficacy to effectiveness. Arch Intern Med. 2011 ; 171 : 1232 – 7 . Crossref Search ADS PubMed 28 Harrison MB , Browne GB, Roberts J, et al. Quality of life of individuals with heart failure: a randomized trial of the effectiveness of two models of hospital-to-home transition. Med Care. 2002 ; 40 : 271 – 82 . Crossref Search ADS PubMed 29 Pugh LC , Tringali RA, Boehmer J, et al. Partners in care: a model of collaboration. Holist Nurs Pract. 1999 ; 13 : 61 – 5 . Crossref Search ADS PubMed 30 Saleh SS , Freire C, Morris-Dickinson G and Shannon T. An effectiveness and cost-benefit analysis of a hospital-based discharge transition program for elderly Medicare recipients. J Am Geriatr Soc. 2012 ; 60 : 1051 – 6 . Crossref Search ADS PubMed 31 Hansen LO , Greenwalk JL, Budnitz T, et al. Project BOOST: effective effort to reduce rehospitalization. J Hosp Med. 2013 ; 8 : 421 – 8 . Crossref Search ADS PubMed 32 Rutherford P , Nielsen GA, Taylor J, et al. How-to guide: improving transitions from the hospital to community settings to reduce avoidable rehospitalizations . Cambridge, MA: Institute for Healthcare Improvement; 2013 . 33 Williams MV , Li J, Hansen LO, et al. Project BOOST implementation: lessons learned. South Med J. 2014 ; 107 : 455 – 65 . Crossref Search ADS PubMed 34 Bridge Model Collaborative. The Bridge Model . www.transitionalcare.org/the-bridgemodel/ (accessed 2016 Jun 1). 35 Altfeld SJ , Shier GE, Rooney M, et al. Effects of an enhanced discharge planning intervention for hospitalized older adults: a randomized trial. Gerontologist. 2013 ; 53 : 430 – 40 . Crossref Search ADS PubMed 36 Kripalani S , Roumie CL, Dalal AK, et al. Effect of a pharmacist intervention on clinically important medication errors after hsopital discharge: a randomized controlled trial. Ann Intern Med. 2012 ; 157 : 1 – 10 . Crossref Search ADS PubMed 37 Conklin JR , Togami JC, Burnett A, et al. Care transitions service: a pharmacy-driven program for medication reconciliation through the continuum of care. Am J Health-Syst Pharm. 2014 ; 71 : 802 – 10 . Crossref Search ADS PubMed 38 Gunadi S , Upfield S, Pham ND, et al. Development of a collaborative transitions-of-care program for heart failure patients. Am J Health-Syst Pharm. 2015 ; 72 : 1147 – 52 . Crossref Search ADS PubMed Copyright © 2018, American Society of Health-System Pharmacists, Inc. All rights reserved.

Phan,, Anthony;Allen,, Bryan;Epps,, Kevin;Alikhil,, Maryam;Kamataris,, Katherine;Tucker,, Calvin

doi: 10.2146/ajhp170400pmid: 30139725

Abstract Purpose Evaluation of the clinical impact of a pharmacist led-penicillin allergy assessment initiative to enhance antibiotic selection is reported. Methods A retrospective analysis was conducted on patients with a self-reported penicillin allergy (SRPA) at a 529-bed community teaching hospital and compared clinical response rate before and after implementation of a penicillin allergy assessment initiative, consisting of pharmacy staff education and pocket card development. Patients admitted with SRPA who received antibiotics with gram-negative coverage for at least 48 hours were included. The primary outcome was the clinical response rate of penicillin-allergic patients determined preimplementation and postimplementation of the initiative and was based upon improvement in signs and symptoms of infection. Secondary outcomes included antibiotics used, antibiotic durations, length of stay, survival rate, antibiotic discontinuation rate, and Clostridium difficile infection rate. Results A total of 280 patients were reviewed. Clinical response rate improved after implementation of the initiative (p = 0.047). There were significant differences in the type of antibiotics prescribed between the preimplementation group and the postimplementation group: increased cephalosporin use (p < 0.001), decreased aztreonam use (p = 0.017), and lower fluoroquinolone use (p = 0.008). Median length of stay (p = 0.943), in-hospital mortality rate (p = 0.173), and C. difficile infection rate (p = 0.426) were similar before and after implementation of the initiative. Conclusion After implementation of an initiative to encourage the use of cephalosporins rather than aztreonam in patients with SRPA, the rate of clinical response and cephalosporin use increased and rates of exposure to aztreonam and fluoroquinolones decreased. antibiotics, antimicrobial stewardship, aztreonam, beta-lactam, drug allergy, infectious disease Penicillin allergy is one of the most commonly reported antibiotic allergies, with a reported frequency of 8–12% across the United States.1,–4 However, 85–90% of patients who report a penicillin allergy are found to have a negative penicillin skin test, and fewer than 1% of patients who report a penicillin allergy have a true immunoglobulin E-mediated allergic reaction. Self-reported penicillin allergy (SRPA) presents a challenge to treatment and impacts optimal antibiotic selection. Penicillin and other β-lactams are first-line treatment options for many infections requiring gram-negative coverage, but penicillin-allergic patients may be treated with alternative antibiotics.5,–8 The consequences of avoiding β-lactams include higher rates of treatment failure, risk of infections with resistant organisms, and increased costs.9,–11 Furthermore, the cross-reactivity between penicillins and cephalosporins is associated with similarities in the side-chain group of the β-lactam ring.12 This finding suggests that many penicillin-allergic patients may still be safely treated with a cephalosporin with less structural similarity to penicillin and that alternative antibiotics such as aztreonam and fluoroquinolones may be avoided. Local institution susceptibility patterns showing increased resistance rates of Pseudomonas aeruginosa, with a susceptibility rate to aztreonam as low as 45%, prompted a pharmacist-led initiative at our institution. Reduction in aztreonam use was chosen as the target of this antimicrobial stewardship initiative, while focusing on improving empirical antibiotic therapy. This initiative was implemented to optimize antibiotic selection in penicillin-allergic patients and reduce the use of high-cost and less-effective antibiotics. Similar antimicrobial stewardship initiatives, which have included penicillin allergy assessment and pharmacist intervention, have demonstrated reductions in the use of alternative antibiotics such as aztreonam.12,–16 In fact, various penicillin allergy assessment initiatives have demonstrated a 36–76% decrease in aztreonam use, resulting in cost savings of $297 per patient in 1 study.15 Although studies have found that reduced aztreonam use provides economic benefits, further studies are needed to evaluate whether clinical outcomes are maintained or improved when reducing the use of alternative antibiotics. Methods This single-center, retrospective, observational study was reviewed and approved by the institutional review board at St. Vincent˴s HealthCare, where the study was conducted. The initiative, which had 2 main components, was implemented in May 2015. The first component of this initiative involved education of the pharmacy staff. Pharmacists were educated about the low risk of penicillin–cephalosporin cross-reactivity and instructed to clarify all SRPAs when receiving an order for aztreonam. After clarification of the allergy and review of a patient˴s tolerance of previous cephalosporin use, pharmacists were expected to contact the prescriber to switch the order from aztreonam to a cephalosporin when possible. The second part of the initiative involved the development of a penicillin-allergy guidance pocket card. This card was distributed hospital-wide and explained the low frequency of cross-reactivity and listed the β-lactams with similar side chains. View largeDownload slide Anthony Phan, Pharm.D., received his doctor of pharmacy degree from the University of Florida College of Pharmacy in 2016 and completed an ASHP-accredited postgraduate year 1 pharmacy residency at St. Vincent˴s HealthCare. His current areas of interest include infectious diseases, critical care, and pharmacy informatics. View largeDownload slide Anthony Phan, Pharm.D., received his doctor of pharmacy degree from the University of Florida College of Pharmacy in 2016 and completed an ASHP-accredited postgraduate year 1 pharmacy residency at St. Vincent˴s HealthCare. His current areas of interest include infectious diseases, critical care, and pharmacy informatics. This analysis included all patients 18 years of age or older with an SRPA who received antibiotics with gram-negative coverage for at least 48 hours. Data were gathered for periods before (December 2013 through December 2014) and after (June 2015 through June 2016) implementation of the initiative. Antibiotics with gram-negative coverage were examined, reducing aztreonam use should not affect the use of antibiotics not used for gram-negative infections. Patients with a history of allergy to cephalosporin, penicillin desensitization, antibiotics for prophylaxis or before admission, pregnancy, or incarceration were excluded. Patients were divided into groups based on whether they were treated before or after the initiative was implemented. The primary outcome was to compare the clinical response rate of patients with SRPA who were treated with antibiotics before and after implementation of the initiative. Clinical response rate, which was based on a modified version of the clinical stability definition from the 2007 Infectious Diseases Society of America community-acquired pneumonia guidelines,5 was defined as documentation of resolution or improvement in signs and symptoms of infection, survival at discharge, and criteria for clinical failure were not met. Clinical failure was defined as prolonged clinical instability of at least 48 hours occurring after 2 days of antimicrobial therapy. Clinical instability criteria included any 1 of the following: temperature of ≥37.8 °C, heart rate of ≥100 beats/min, respiratory rate of ≥24 breaths/min, systolic blood pressure of <90 mm Hg, inability to maintain oral intake, and altered mental status. Secondary outcomes included length of stay, survival rate, intensive care unit admission rate, aztreonam use, aztreonam drug cost, Clostridium difficile infection rate, antibiotic selected, rate of discontinuation due to an adverse reaction, and appropriate antibiotic use. Appropriate antibiotic use was defined as (1) the use of an administered antibiotic regimen that is active against the identified pathogen based on the results of in vitro antimicrobial susceptibility testing or (2) the patient had a clinical response to guideline-recommended antibiotics without the need for changing antibiotic therapy. Antibiotics with at least 1 dose administered were recorded. If an antibiotic was discontinued or switched during the patient˴s hospital stay, documentation was examined. If it was determined that the change in antibiotic regimen was due to dosage de-escalation, the antibiotic was defined as appropriate. All data analyses were performed using SAS OnDemand for Academics, version 9.4 (SAS Inc., Cary, NC). All patients were randomized using a random-number generator and evaluated until the sample size to obtain 80% power was achieved. Fisher˴s exact test or chi-square test was used to analyze nominal data, and Student˴s t test or Mann-Whitney U test was used to analyze continuous data. Based on an estimated 90% clinical response rate after implementation of the initiative, a total of 140 patients would be required in each group to detect a 12.5% absolute risk reduction with a power of 80% and an α of 0.05. Results After randomization and exclusions, the study included 280 patients, 140 each in the preimplementation and postimplementation groups. For the preimplementation group, randomization led to the review of 386 of 4,610 medical records. Of the 386 records reviewed, 140 remained after those for 246 patients were excluded—188 with less than 48 hours of antibiotic treatment, 30 with prophylactic antibiotic treatment, 25 with pregnancy, 2 with a reported cephalosporin allergy, and 1 with antibiotic treatment before admission. For the postimplementation group, randomization led to the review of 343 of 4,397 medical records. Of the 343 records reviewed, 140 remained after those for 203 patients were excluded—143 with less than 48 hours of antibiotic treatment, 47 with prophylactic antibiotic treatment, 10 with pregnancy, and 3 with a reported cephalosporin allergy. Baseline characteristics were similar between both groups except that there were significantly more patients with hypertension and diabetes mellitus in the postimplementation group (Table 1). The most common infections treated were respiratory and urinary tract infections. The clinical response rate improved from 83.6% of patients in the preimplementation group to 91.4% of patients in the postimplementation group (p = 0.0468). Overall, there were significant differences in the distribution of antibiotics prescribed between groups (p < 0.001). There was significantly less frequent aztreonam use, less frequent fluoroquinolone use, and more frequent cephalosporin use in the preimplementation group (Table 2). Secondary outcomes were similar between groups except that there were significantly more patients who discontinued antibiotics due to an adverse reaction in the preimplementation group (4.3% versus 0%, p = 0.03) (Table 2). The majority of adverse reactions in these patients were swelling and rash in patients receiving a fluoroquinolone; however, direct causation could not be determined. In addition, the hospital total acquisition cost of aztreonam decreased from $7,903 preimplementation to $1,092 postimplementation. Table 1 Baseline Characteristics of Study Patients Characteristic Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Mean ± S.D. age, yr 60.4 ± 20.6 65.0 ± 16.0 0.159 No. (%) women 103 (73.6) 99 (70.7) 0.807 Mean ± S.D. height, cm 166.0 ± 14.6 167.0 ± 10.9 0.323 Mean ± S.D. weight, kg 80.2 ± 28.8 81.8 ± 27.3 0.413 Comorbidities, no. (%)  Diabetes mellitus 41 (29.3) 57 (40.7) 0.045  Chronic kidney disease 6 (4.3) 14 (10.0) 0.063  Atrial fibrillation 14 (10.0) 16 (11.4) 0.699  Congestive heart failure 7 (5.0) 6 (4.3) 0.776  Coronary artery disease 22 (15.7) 31 (22.1) 0.170  Hypertension 72 (51.4) 101 (72.1) <0.001  Chronic obstructive pulmonary disease 39 (27.9) 41 (29.3) 0.791  Malignancy 29 (20.7) 22 (15.7) 0.357 Infection type, no. (%)  Pulmonary 57 (40.7) 61 (43.5) 0.400  Intraabdominal 26 (18.6) 18 (12.9) 0.190  Genitourinary 33 (23.6) 33 (23.6) 0.677  Sepsis, bacteremia 9 (6.4) 15 (10.7) 0.277  Skin or soft tissue 20 (14.3) 26 (18.6) 0.420 Characteristic Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Mean ± S.D. age, yr 60.4 ± 20.6 65.0 ± 16.0 0.159 No. (%) women 103 (73.6) 99 (70.7) 0.807 Mean ± S.D. height, cm 166.0 ± 14.6 167.0 ± 10.9 0.323 Mean ± S.D. weight, kg 80.2 ± 28.8 81.8 ± 27.3 0.413 Comorbidities, no. (%)  Diabetes mellitus 41 (29.3) 57 (40.7) 0.045  Chronic kidney disease 6 (4.3) 14 (10.0) 0.063  Atrial fibrillation 14 (10.0) 16 (11.4) 0.699  Congestive heart failure 7 (5.0) 6 (4.3) 0.776  Coronary artery disease 22 (15.7) 31 (22.1) 0.170  Hypertension 72 (51.4) 101 (72.1) <0.001  Chronic obstructive pulmonary disease 39 (27.9) 41 (29.3) 0.791  Malignancy 29 (20.7) 22 (15.7) 0.357 Infection type, no. (%)  Pulmonary 57 (40.7) 61 (43.5) 0.400  Intraabdominal 26 (18.6) 18 (12.9) 0.190  Genitourinary 33 (23.6) 33 (23.6) 0.677  Sepsis, bacteremia 9 (6.4) 15 (10.7) 0.277  Skin or soft tissue 20 (14.3) 26 (18.6) 0.420 View Large Table 1 Baseline Characteristics of Study Patients Characteristic Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Mean ± S.D. age, yr 60.4 ± 20.6 65.0 ± 16.0 0.159 No. (%) women 103 (73.6) 99 (70.7) 0.807 Mean ± S.D. height, cm 166.0 ± 14.6 167.0 ± 10.9 0.323 Mean ± S.D. weight, kg 80.2 ± 28.8 81.8 ± 27.3 0.413 Comorbidities, no. (%)  Diabetes mellitus 41 (29.3) 57 (40.7) 0.045  Chronic kidney disease 6 (4.3) 14 (10.0) 0.063  Atrial fibrillation 14 (10.0) 16 (11.4) 0.699  Congestive heart failure 7 (5.0) 6 (4.3) 0.776  Coronary artery disease 22 (15.7) 31 (22.1) 0.170  Hypertension 72 (51.4) 101 (72.1) <0.001  Chronic obstructive pulmonary disease 39 (27.9) 41 (29.3) 0.791  Malignancy 29 (20.7) 22 (15.7) 0.357 Infection type, no. (%)  Pulmonary 57 (40.7) 61 (43.5) 0.400  Intraabdominal 26 (18.6) 18 (12.9) 0.190  Genitourinary 33 (23.6) 33 (23.6) 0.677  Sepsis, bacteremia 9 (6.4) 15 (10.7) 0.277  Skin or soft tissue 20 (14.3) 26 (18.6) 0.420 Characteristic Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Mean ± S.D. age, yr 60.4 ± 20.6 65.0 ± 16.0 0.159 No. (%) women 103 (73.6) 99 (70.7) 0.807 Mean ± S.D. height, cm 166.0 ± 14.6 167.0 ± 10.9 0.323 Mean ± S.D. weight, kg 80.2 ± 28.8 81.8 ± 27.3 0.413 Comorbidities, no. (%)  Diabetes mellitus 41 (29.3) 57 (40.7) 0.045  Chronic kidney disease 6 (4.3) 14 (10.0) 0.063  Atrial fibrillation 14 (10.0) 16 (11.4) 0.699  Congestive heart failure 7 (5.0) 6 (4.3) 0.776  Coronary artery disease 22 (15.7) 31 (22.1) 0.170  Hypertension 72 (51.4) 101 (72.1) <0.001  Chronic obstructive pulmonary disease 39 (27.9) 41 (29.3) 0.791  Malignancy 29 (20.7) 22 (15.7) 0.357 Infection type, no. (%)  Pulmonary 57 (40.7) 61 (43.5) 0.400  Intraabdominal 26 (18.6) 18 (12.9) 0.190  Genitourinary 33 (23.6) 33 (23.6) 0.677  Sepsis, bacteremia 9 (6.4) 15 (10.7) 0.277  Skin or soft tissue 20 (14.3) 26 (18.6) 0.420 View Large Table 2 Outcomes Before and After Initiative Implementation Outcome Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Specific antibiotic type received, no. (%) pts  Aminoglycoside 7 (5.0) 2 (1.4) 0.090  Carbapenem 7 (5.0) 7 (5.0) >0.999  Cephalosporin 56 (40.0) 106 (75.7) <0.001  Fluoroquinolone 71 (50.7) 49 (35.0) 0.008  Macrolide 11 (7.9) 24 (17.1) 0.019  Monobactam 17 (12.1) 6 (4.3) 0.017 Mean ± S.D. length of stay (days) 7.2 ± 5.7 7.3 ± 6.7 0.943 No. (%) pts surviving 133 (95.0) 138 (98.6) 0.173 No. (%) pts admitted to intensive care unit 45 (32.1) 49 (35.0) 0.268 No. (%) pts with Clostridium difficile infection 9 (6.4) 6 (4.3) 0.426 No. (%) pts receiving appropriate antibiotics 94 (67.1) 108 (77.1) 0.062 Mean ± S.D. antibiotic duration (days) 4.3 ± 2.9 4.1 ± 3.2 0.466 No. (%) pts with antibiotic discontinued due to adverse reaction 6 (4.3) 0 0.030 Outcome Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Specific antibiotic type received, no. (%) pts  Aminoglycoside 7 (5.0) 2 (1.4) 0.090  Carbapenem 7 (5.0) 7 (5.0) >0.999  Cephalosporin 56 (40.0) 106 (75.7) <0.001  Fluoroquinolone 71 (50.7) 49 (35.0) 0.008  Macrolide 11 (7.9) 24 (17.1) 0.019  Monobactam 17 (12.1) 6 (4.3) 0.017 Mean ± S.D. length of stay (days) 7.2 ± 5.7 7.3 ± 6.7 0.943 No. (%) pts surviving 133 (95.0) 138 (98.6) 0.173 No. (%) pts admitted to intensive care unit 45 (32.1) 49 (35.0) 0.268 No. (%) pts with Clostridium difficile infection 9 (6.4) 6 (4.3) 0.426 No. (%) pts receiving appropriate antibiotics 94 (67.1) 108 (77.1) 0.062 Mean ± S.D. antibiotic duration (days) 4.3 ± 2.9 4.1 ± 3.2 0.466 No. (%) pts with antibiotic discontinued due to adverse reaction 6 (4.3) 0 0.030 View Large Table 2 Outcomes Before and After Initiative Implementation Outcome Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Specific antibiotic type received, no. (%) pts  Aminoglycoside 7 (5.0) 2 (1.4) 0.090  Carbapenem 7 (5.0) 7 (5.0) >0.999  Cephalosporin 56 (40.0) 106 (75.7) <0.001  Fluoroquinolone 71 (50.7) 49 (35.0) 0.008  Macrolide 11 (7.9) 24 (17.1) 0.019  Monobactam 17 (12.1) 6 (4.3) 0.017 Mean ± S.D. length of stay (days) 7.2 ± 5.7 7.3 ± 6.7 0.943 No. (%) pts surviving 133 (95.0) 138 (98.6) 0.173 No. (%) pts admitted to intensive care unit 45 (32.1) 49 (35.0) 0.268 No. (%) pts with Clostridium difficile infection 9 (6.4) 6 (4.3) 0.426 No. (%) pts receiving appropriate antibiotics 94 (67.1) 108 (77.1) 0.062 Mean ± S.D. antibiotic duration (days) 4.3 ± 2.9 4.1 ± 3.2 0.466 No. (%) pts with antibiotic discontinued due to adverse reaction 6 (4.3) 0 0.030 Outcome Before Initiative Implementation (n = 140) After Initiative Implementation (n = 140) p Specific antibiotic type received, no. (%) pts  Aminoglycoside 7 (5.0) 2 (1.4) 0.090  Carbapenem 7 (5.0) 7 (5.0) >0.999  Cephalosporin 56 (40.0) 106 (75.7) <0.001  Fluoroquinolone 71 (50.7) 49 (35.0) 0.008  Macrolide 11 (7.9) 24 (17.1) 0.019  Monobactam 17 (12.1) 6 (4.3) 0.017 Mean ± S.D. length of stay (days) 7.2 ± 5.7 7.3 ± 6.7 0.943 No. (%) pts surviving 133 (95.0) 138 (98.6) 0.173 No. (%) pts admitted to intensive care unit 45 (32.1) 49 (35.0) 0.268 No. (%) pts with Clostridium difficile infection 9 (6.4) 6 (4.3) 0.426 No. (%) pts receiving appropriate antibiotics 94 (67.1) 108 (77.1) 0.062 Mean ± S.D. antibiotic duration (days) 4.3 ± 2.9 4.1 ± 3.2 0.466 No. (%) pts with antibiotic discontinued due to adverse reaction 6 (4.3) 0 0.030 View Large Discussion Selecting optimal antibiotics in patients with an SRPA is challenging. Improper interpretation of this allergy and its associated cross-reactivity with cephalosporins may limit treatment options for many types of infections. Penicillin allergy assessment initiatives aim to identify candidates for alternative antibiotics such as cephalosporins with different side-chain structure, which may not have been previously considered attributable to SRPA. Due to the high level of resistance observed at our institution and the high cost associated with alternative antibiotics, this study intervention focused on improving empirical antibiotic therapy for patients with an SRPA. The goal of this antimicrobial stewardship initiative was to provide more-appropriate antibiotic therapy by identifying candidates for cephalosporins in penicillin-allergic patients and minimizing the use of broad-spectrum antibiotics to limit resistance rates and adverse reactions. With this initiative, a high number of SRPAs were found to be inaccurate, as many patients were able to tolerate cephalosporin. To the best of our knowledge, this is 1 of the first studies of SRPA to examine clinical response rate as a primary outcome with an adequately powered sample size. Our study findings demonstrate that the initiative was associated with a significant improvement in clinical response rate, with rates similar to previous findings.15 There was also a decrease in the use of aztreonam and other alternative antibiotics, which was consistent with previous findings.12,–16 Evidently, this antimicrobial stewardship initiative successfully facilitated a decreased use of the broad-spectrum antibiotic treatment by prescribers. Overall, this initiative can serve as an effective tool that can be expanded to other institutions and allow for reduced exposure to and adverse effects associated with less favorable alternative antibiotics while providing clinically effective and less costly care. One limitation of this study was the variable use of antibiotics selected for each type of infection, which may have confounded the results for the primary outcome. Further, there were other concurrent antimicrobial stewardship initiatives, including fluoroquinolone and carbapenem use restriction and antibiotic de-escalation, during the time frame of this study that may have affected antibiotic selection. In addition, this study was a retrospective chart review and relied on accurate and proper documentation in the electronic medical record. A final limitation was that patients˴ antibiotic use after discharge and readmission rates were not captured. Aztreonam was targeted as a high-cost antibiotic. Since the implementation of this initiative, pharmacists made interventions on aztreonam orders sooner, leading to a reduction of aztreonam use and increased use of less-expensive alternatives. This was associated with a decrease in aztreonam expenditure by $6,811 for the 140 patients studied after implementation of the initiative. Conclusion After implementation of an initiative to encourage the use of cephalosporins rather than aztreonam in patients with SRPA, the rate of clinical response and cephalosporin use increased and rates of exposure to aztreonam and fluoroquinolones decreased. Disclosures The authors have declared no potential conflicts of interest. References 1 Chang C , Mahmood MM, Teuber SS and Gershwin ME. Overview of penicillin allergy. Clin Rev Allergy Immunol. 2012 ; 43 : 84 – 97 . Crossref Search ADS PubMed 2 Solensky R and Khan DA. Drug allergy: an updated practice parameter. Ann Allergy Asthma Immunol. 2010 ; 105 : 259 – 73 . Crossref Search ADS PubMed 3 Macy E and Poon K-YT. Self-reported antibiotic allergy incidence and prevalence: age and sex effects. Am J Med. 2009 ; 122 : 778.e1 – 7 . Crossref Search ADS 4 Albin S and Agarwal S. Prevalence and characteristics of reported penicillin allergy in an urban outpatient adult population. Allergy Asthma Proc. 2014 ; 35 : 489 – 94 . Crossref Search ADS PubMed 5 Mandell LA , Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007 ; 44 ( Supplement 2 ): S27 – 72 . Crossref Search ADS PubMed 6 Kalil AC , Metersky ML, Klompas M, et al. Management of adults with hospital-acquired and ventilator-associated pneumonia: 2016 clinical practice guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016 ; 63 : e61 – 111 . Crossref Search ADS PubMed 7 Panel on Opportunistic Infections in HIV-Infected Adults and Adolescents. Guidelines for the prevention and treatment of opportunistic infections in HIV-infected adults and adolescents: recommendations from the Centers for Disease Control and Prevention, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America . http://aidsinfo.nih.gov/contentfiles/lvguidelines/adult_oi.pdf (accessed 2016 Jul 30). 8 Freifeld AG , Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis. 2011 ; 52 : e56 – 93 . Crossref Search ADS PubMed 9 Macy E and Contreras R. Health care use and serious infection prevalence associated with penicillin ˵allergy˶ in hospitalized patients: a cohort study. J Allergy Clin Immunol. 2014 ; 133 : 790 – 6 . Crossref Search ADS PubMed 10 Lee CE , Zembower TR, Fotis MA, et al. The incidence of antimicrobial allergies in hospitalized patients: implications regarding prescribing patterns and emerging bacterial resistance. Arch Intern Med. 2000 ; 160 : 2819 – 22 . Crossref Search ADS PubMed 11 Pichichero ME and Zagursky R. Penicillin and cephalosporin allergy. Ann Allergy Asthma Immunol. 2014 ; 112 : 404 – 12 . Crossref Search ADS PubMed 12 Swearingen SM , White C, Weident S, et al. A multidimensional antimicrobial stewardship intervention targeting aztreonam use in patients with a reported penicillin allergy. Int J Clin Pharm. 2016 ; 38 : 213 – 7 . Crossref Search ADS PubMed 13 King EA , Challa S, Curtin P, et al. Penicillin skin testing in hospitalized patients with β-lactam allergies. Effect on antibiotic selection and cost. Ann Allergy Asthma Immunol. 2016 ; 117 : 67 – 71 . Crossref Search ADS PubMed 14 Staicu ML , Brundige ML, Ramsey A, et al. Implementation of a penicillin allergy screening tool to optimize aztreonam use. Am J Health-Syst Pharm. 2016 ; 73 : 298 – 306 . Crossref Search ADS PubMed 15 Estep PM , Ferreira JA, Dupree LH, et al. Impact of an antimicrobial stewardship initiative to evaluate β-lactam allergy in patients ordered aztreonam. Am J Health-Syst Pharm. 2016 ; 73 ( suppl 1 ): S8 – 13 . Crossref Search ADS PubMed 16 Jiang F , Jain R, Pottinger PS, et al. Automated allergy and infectious disease pharmacy consult to limit the use of aztreonam in patients with reported beta-lactam allergy. J Allergy Clin Immunol. 2016 ; 137 ( suppl ): B196 . Crossref Search ADS Copyright © 2018, American Society of Health-System Pharmacists, Inc. All rights reserved.