Best Practice Advisory Research Articles

83. Impact of a Urine Culture Best Practice Advisory on Collection of Urine Cultures and Subsequent Antibiotic Therapy

BackgroundIn the acute care setting, urinary tract infections (UTIs) may be over diagnosed in up to 40% of cases. In most scenarios, asymptomatic bacteriuria (ASB) is not an indication for antibiotic therapy; inappropriate therapy is associated with a higher incidence of antibiotic-resistant bacteria and adverse drug reactions. Limiting inappropriate collection of urine cultures may decrease unnecessary treatment of ASB. The objective of this study is to assess the impact of a urine culture best practice advisory (BPA) on collection of unnecessary urine cultures. MethodsThis retrospective, observational, single-center study included adult inpatients with an order for urinalysis/urine culture. Those who were pregnant, had a concomitant infection other than UTI and/or were taking antimicrobials for a non-UTI indication, and were undergoing urological procedures were excluded. Duplicate urine culture collections and/or admissions were excluded. Incorporation of a BPA into computerized provider order entry, allowing providers to assess need and document indication for urine culture collection, was implemented on July 2019. The following clinical outcomes were assessed: number of unnecessary urine cultures collected, number of antibiotic treatments, and antibiotic-associated adverse reactions. ResultsTwo hundred met criteria for inclusion; 96 in the pre-BPA group (Aug – Oct 2018) and 104 in the post-BPA group (Aug – Oct 2019). Seventy-four (37%) were male and the mean age was 64 and 70 years (p=0.249), respectively. The Charlson Comorbidity Index (CCI) was similar between groups (4 vs. 5, p=0.162) and majority were admitted to a general medical ward (94.5%). Seventy patients (72.9%) in the pre-BPA group and 47 (51.6%) in the post-BPA group had inappropriately ordered urinalysis/urine cultures (OR 0.40; 95% CI 0.22-0.73; p=0.003). Of these patients, 15 (21.4%) and 9 (19.1%) from the pre- and post-BPA groups, respectively, were treated (p=0.077). Among those treated, only two adverse drug reactions were reported. ConclusionImplementation of a BPA significantly reduced the number of inappropriate urinalysis/urine culture orders. There was a trend towards decreased antibiotic use for ASB. Future studies are warranted to assess sustainability of these results. Disclosures All Authors: No reported disclosures

Abstract 11386: Self-Service Analysis of Best Practice Alert Acknowledgement Comments as a Method for Evaluation Reveals Opportunities for Optimization of Clinical Decision Support for Patients with Prolonged QT Syndrome

Introduction: The main goal of Clinical Decision Support (CDS) is to provide timely and useful information for providers to deliver high-quality and safe care. Issues, like alert fatigue and provider burnout are commonly cited as unintended consequences of the CDS systems. Herein, we describe leveraging electronic health record (EHR) self-service reporting tools to gain insights on a CDS rule designed to avoid providers from ordering QT-prolonging medication in patients with history of QT-prolongation. Methods: Our EHR (Epic) SlicerDicer reporting tool was used to identify the frequency of QT-prolongation Best Practice Advisory (BPA) triggerings from 1/1/2020 to 12/31/2020 across Mayo Clinic sites. Additionally, associated patient encounter and BPA-specific data were queried. Text from acknowledgement comments to the BPA were reviewed and central themes and comments were categorized. Epic Clarity Data Warehouse was queried to identify the BPA-triggering medications and frequencies. Comparisons of percent action taken on the BPA were compared internally to Mayo Clinic’s BPA benchmark, and with national benchmarks using Epic BestPractice Pulse dashboard. Results: A total of 17,538 BPAs of interest were triggered across the Mayo Clinic sites in 2020, of which 3,696 (20.1%) were acted upon through either monitoring or medication discontinuation. This is below Mayo Clinic’s median of 29.2% for inpatient BPAs and 45.6% BPAs for outpatient, but above the national median of 10% (IQR 6.1-13.1%) for inpatient and 10% (IQR 4.3-19.5%) for outpatient BPAs actioned upon. A total of 30 different medications triggered the BPA, with the most common being ondansetron (41.5%), amiodarone (11.3%), and haloperidol (10.1%). User comments were left on 2,382 orders (13.5%) that acknowledged the BPA, but provided argument to prescribe drug. Three most common themes identified from comments on overridden BPAs: patient previously tolerated medication (19.0%), QT interval not prolonged on recent ECG (14.2%), and patient on hospice care (5.2%). Conclusion: Analysis of free text comments of CDS can be and should be performed to reveal opportunities for further improvement of CDS. Providers should be encouraged to provide comments when encountering CDS interventions.

Impact of Automated Best Practice Advisories on Provider Response to CYP2C19 Genotyping Results for Patients on Clopidogrel

Objective The study objective was to examine provider acceptance and genotyping responses to a best practice advisory (BPA) concerning clopidogrel and CYP2C19 intermediate and poor metabolizers within the context of a new pharmacogenomics program at a Midwestern health system. Other secondary objectives analyzed included appropriate BPA firing, the distribution of alleles in study population, indications for clopidogrel use, and impact of indication on therapy change. Methods: In this study, the progress of this program was assessed by quantifying how providers respond to BPAs generated in the electronic medical record (EMR), in the context of a single representative gene–drug–outcome relationship. Patient data was pulled via reports yielding patients with genotyped information in the EMR and cross-referenced with a report evaluating BPA firing occurrences. Results: By capturing antiplatelet therapy changes in response to CYP2C19 genotyping results, 37 patients were found that had 73 BPAs fire. Nine of those patients had alternative antiplatelet therapy ordered. Of these, 6 alternative antiplatelet therapies were ordered from the BPA. Conclusion: Providers utilized BPAs, but responded differently based on individual knowledge of genotypes and indications. Information obtained from this study can be used for provider education and as reference for future design and wording of BPAs.

Improving palliative care integration for patients with stage IV lung cancer at UConn Health by initiating a best practice advisory (BPA) in the electronic medical record (EMR).

331 Background: The American Society of Clinical Oncology (ASCO) practice updates on the integration of palliative care into standard oncology practice provides a key recommendation that patients with advanced cancer should receive palliative care services. Specifically, ASCO recommends that all patients with stage IV cancer receive a referral to an interdisciplinary palliative care team early in their course of disease and within 8 weeks of diagnosis. At UConn Health, it has been previously documented that only 28% of patients with stage IV cancer receive a referral to the palliative care team. To improve the quality of cancer care, a BPA with standardized criteria for palliative care referral was developed and implemented for patients with stage IV lung cancer. In this pilot study, the goal was to get 80-90% of patients a referral to palliative care team within 8 weeks of stage IV lung cancer diagnosis. Methods: The Model for Improvement using Plan-Do-Study-Act Cycles was completed: 1) created an EMR report to identify patients with stage IV lung cancer; 2) completed a retrospective review of patients with stage IV lung cancer referred to palliative care 6 months prior to BPA implementation; 3) created and implemented BPA; 4) educated providers about palliative care referral and BPA; 5) retrospectively reviewed referrals 6 months following BPA implementation; and 6) evaluated potential barriers. The palliative care EMR BPA was developed in collaboration with information technology specialists. The BPA alert populates the EMR when the provider opens a patient chart or visit encounter for all patients with 1) diagnosis of lung cancer; 2) stage IV disease; and 3) does not have a current order for palliative care referral. Results: Prior to BPA implementation (January 1, 2020- July 31, 2020), 8 of 28 patients (32%) with stage IV lung cancer were referred to palliative care service. The BPA became active on 9/15/2020. Within the six months following BPA implementation, 16 patients were newly diagnosed with stage IV lung cancer. Of these 16 patients, 81% of them had a referral to palliative care made within 8 weeks of clinical staging. Several barriers were identified with current process, including lack of staging tool use by all providers; lack of documentation of all data required for staging tool to automatically calculate stage, and inability to track patients who declined palliative care appointment. Conclusions: Incorporating a BPA reminding providers to consider a palliative care referral improved referrals of patients with newly diagnosed stage IV lung cancer to the palliative care clinic within 8 weeks of diagnosis, improving compliance with ASCO’s practice guidance on integration of palliative care. Next steps are to address barriers and expand the use of palliative care referral BPA to all patients with stage IV cancer.

262 An Emergency Department-Based Response to the Hepatitis A Outbreak in Michigan: Improving Hepatitis A Vaccination Rates in High-Risk Populations

A non-food-borne hepatitis A outbreak occurred in Michigan between August 2016 and September 2019, resulting in 920 cases, 738 hospitalizations, and 30 deaths. To support the Michigan Department of Health and Human Services’ efforts to increase hepatitis A vaccination rates among high-risk individuals, our multicenter health system implemented an electronic medical record (EMR)-based vaccination intervention across its nine emergency departments (ED). The primary objective of this retrospective cohort and survey analysis was to quantitatively determine whether this intervention was successful in increasing vaccination rates. The secondary objective was to qualitatively assess the attitudes towards, and barriers to use of, the computerized vaccine reminder system. All patients 18 years or older who arrived to any of the nine EDs between August 2018 and January 2020 were screened using an electronic nursing questionnaire embedded in the EMR (Epic). If a patient was determined to be high-risk based on the questionnaire (homeless, incarceration history, illicit drug use, liver disease, or a man who has sex with men), an electronic best practice advisory (BPA) would trigger and give the patient’s physician the option to order the hepatitis A vaccine. If consented, patients would receive a one-time dose of the hepatitis A vaccine in the ED. We also administered a survey to physicians and nurses to evaluate perceptions and barriers to use of the EMR intervention. During the pre-intervention period from August 2016 to July 2018, 885,344 patients visited the EDs. 49 vaccines were ordered (5.5 per 100,000 patients) and 34 were administered (3.8 per 100,000 patients). During the intervention period from August 2018 to January 2020, 774,034 patients visited the EDs and 574,865 (74.3%) were screened. Of those screened, 11,016 patients were found to be high-risk and triggered the BPA. Among this group of patients, 1,929 vaccines were ordered (249 per 100,000 patients) and 883 were administered (114 per 100,000 patients). We also found that during the intervention period, an additional 565 vaccines were ordered and 322 vaccines were administered without a BPA prompt. Nurses consistently screened 70-80% of patients per month. Physicians were initially more compliant with the BPA’s use (301 vaccines in September 2018), but compliance declined over time (67 vaccines in January 2020) (Graph 1). Surveys revealed that two major barriers to consistent BPA use by physicians was lack of time and the perception that vaccinations are low-priority in the ED. EMR screening tools and BPAs can be utilized in the ED as an effective strategy to vaccinate high-risk individuals. This may be translatable to outbreaks of other vaccine-preventable illnesses like influenza, measles, or SARS-CoV-2. Providing recurrent education about the importance of public health initiatives and eligibility criteria for vaccine administration are needed to sustain compliance. It is essential to frequently audit and provide feedback to physicians on their compliance, and address their concerns about barriers to use.

S1227 Improving Hepatocellular Carcinoma Screening in Patients With Cirrhosis

Introduction: Hepatocellular Carcinoma (HCC) is currently the 4th leading cause of death worldwide and the fastest growing cause of cancer-related deaths in the United States. This increase of incidence in HCC is largely due to the prevalence of Hepatitis B, Hepatitis C, alcoholic liver disease, as well as the surge in obesity-related fatty liver disease. The prognosis of patients diagnosed with HCC remains poor with a 5-year survival rate of 18%, largely due to late detection. Current screening guidelines recommend abdominal ultrasound every 6 months 1/- AFP, and bi-annual screening has been shown to reduce mortality by 37%. Dynamic contrast enhanced CT and MRI are alternatives with higher sensitivity but have never been studied looking at mortality. The goal of this study is to determine how the healthcare system is meeting the needs of these patients and find ways to provide better care for this population. Methods: Using tools within the Electronic Medical Record (EMR), we were able to determine that the number of patients in our community with diagnoses of cirrhosis and being up to date on HCC screening was 189/1123 (16.8%). Working with IT, we developed a Best Practice Advisory (BPA) to ”flag” these patients who have not undergone HCC screening within the past 6 months when seen by outpatient providers. When accepted, the order for abdominal ultrasound was automatically sent and associated with the corresponding diagnosis. Results: The project was initiated in January of 2020 but due to the COVID pandemic, the BPA did not “go live” until October 14th, 2020. At the 6-month mark, the data was analyzed and while the screening rates only increased to 18.2% (221/1213), the incidence of ultrasound usage increased 123% with no change in usage of CT or MRI. Most common causes for the BPA being bypassed were “Not addressed this visit” and “Follows with Gastroenterology.” Conclusion: As the burden of liver disease increases with the obesity epidemic and incidence of NASH, mortality from liver disease is expected to increase. How can we use EMR to continue to improve screening rates in these patients and identify areas of deficiency? We will continue to monitor efficacy of the BPA implemented, with future areas of focus being revamped patient education on hospital discharge, improved follow up, and identification of different socioeconomic barriers to patient care..

S1224 Computable Phenotypes for NAFLD and NASH Identify Patients with Significant Comorbidities yet Most Remain Undiagnosed

Introduction: Non-alcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease, yet remains underdiagnosed in real-world practice. Several clinical tools exist to identify possible NAFLD and nonalcoholic steatohepatitis (NASH). Incorporating these into computable phenotypes may help identify patients using electronic health records (EHR), improve clinical care and long-term outcomes. Our objective was to create and evaluate computable phenotypes for NAFLD and NASH using previously validated scores. Methods: We conducted a retrospective study of adults using the Duke University Health System (DUHS) between 2014 and 2019. Patients (pts) with a body mass index < 25 kg/m2 and liver disease diagnosis other than NAFLD or NASH were excluded. The first cohort (NAFLD1) included pts with alanine aminotransferase (ALT) > 40 for men and > 31 for women on 2 occasions six months apart within 2 years. The second NAFLD cohort (NAFLD2) included patients with an ICD9/10 code for NAFLD (K76.0). The first cohort of NASH (NASH1) pts included those that met NAFLD1 criteria plus a FIBROSIS-4 score > 2.67. The second NASH (NASH2) cohort included pts with a NASH diagnosis (K75.81). Chi squared and post hoc Z tests were used to compare cohorts. Results: Out of 359,643 pts who met inclusion criteria, 33,971 pts met criteria for NAFLD1, and 12,544 had a NAFLD diagnosis (NAFLD2). 7,888 met criteria for NASH1 and 2,648 had a NASH diagnosis (NASH2). For both NAFLD and NASH, clinical comorbidities (Table) were significantly more prevalent in NAFLD1 and NASH1 cohorts. White pts were more likely to be diagnosed with NASH than those identified by FIBROSIS-4 score alone (82% vs 69%, P < 0.005). Fewer black pts were diagnosed with NASH (7.8% vs. 24.8%, P < 0.005). Fewer NAFLD2 or NASH2 pts were treated with statins. All cause hospitalizations were high regardless of diagnosis (Figure 1). Conclusion: Very few pts were diagnosed with NAFLD (3.5%) or NASH (0.7%) in this 5 years period. The number of pts meeting criteria by computable phenotype was higher suggesting significant underdiagnosis, particularly among black pts. Our results suggest that pts with NAFLD and NASH phenotypes but no diagnosis are more likely to have downstream comorbidities. Metabolic syndrome features like hypertension and hyperlipidemia should trigger suspicion for NAFLD and NASH in overweight pts. Further studies are needed to understand the reasons behind the underdiagnosis and the increased downstream comorbidities.Figure 1.: Best Practice Advisory (BPA) alerting provider of the need for screening ultrasoundTable 1.: Comparison of Demographics, Comorbidities, and Medications of Different Phenotypes of NAFLD and NASH. T2DM = Type 2 Diabetes Mellitus. CAD = Coronary artery disease. PAD = peripheral artery disease. HFrEF = Heart failure with reduced ejection fraciton. HFpEF = Heart failure with preserved ejection fraction. ACEi = ACE inhibitor. ARB = angiotensin receptor blocker. BB = beta blocker.

Outcomes of clinical decision support for outpatient management of Clostridioides difficile infection.

To determine the impact of clinical decision support on guideline-concordant Clostridioides difficile infection (CDI) treatment. Quasi-experimental study in >50 ambulatory clinics. Primary, specialty, and urgent-care clinics. Adult patients were eligible for inclusion if they were diagnosed with and treated for a first episode of symptomatic CDI at an ambulatory clinic between November 1, 2019, and November 30, 2020. An outpatient best practice advisory (BPA) was implemented to notify prescribers that "vancomycin or fidaxomicin are preferred over metronidazole for C.difficile infection" when metronidazole was prescribed to a patient with CDI. In total, 189 patients were included in the study: 92 before the BPA and 97 after the BPA. Their median age was 59 years; 31% were male; 75% were white; 30% had CDI-related comorbidities; 35% had healthcare exposure; 65% had antibiotic exposure; 44% had gastric acid suppression therapy within 90 days of CDI diagnosis. The BPA was accepted 23 of 26 times and was used to optimize the therapy of 16 patients in 6 months. Guideline-concordant therapy increased after implementation of the BPA (72% vs 91%; P = .001). Vancomycin prescribing increased and metronidazole prescribing decreased after the BPA. There was no difference in clinical response or unplanned encounter within 14 days after treatment initiation. Fewer patients after the BPA had CDI recurrence within 14-56 days of the initial episode (27% vs 7%; P < .001). Clinical decision support increased prescribing of guideline-concordant CDI therapy in the outpatient setting. A targeted BPA is an effective stewardship intervention and may be especially useful in settings with limited antimicrobial stewardship resources.

A quality improvement initiative to improve primary care referral rates for penicillin allergy delabeling

BackgroundOf the US population, 10% reports a penicillin allergy but more than 90% can ultimately tolerate penicillin. Confirmation of these allergies in the pediatric population may improve future health outcomes and decrease costs. Referring patients for confirmatory testing is the first step in clarifying penicillin allergies. ObjectiveTo increase the number of referrals of patients with listed penicillin allergies from the University of California, San Diego academic general pediatrics clinics to Rady Children's Hospital allergy clinics using an educational session and a best practice advisory (BPA) in the electronic medical record. MethodsAn educational session with attendings and 3 plan-do-study-act (PDSA) cycles were completed using a BPA alert that triggered for all patients with a documented penicillin-class drug allergy to draw attention and facilitate referral. The BPA was modified at each PDSA cycle based on physician input. ResultsAt baseline, 1.9% of referrals to the allergy clinic were for penicillin-class drug allergies. After an attending physician educational session, the percentage increased to 13.7%. The BPA was implemented with further increase to 27.8% of all allergy referrals in the course of 3 PDSA cycles. Not all patients with penicillin-class drug allergies were referred, and the reasons were documented when the physicians dismissed the BPA. Physicians did not refer 35% of the time because of time constraints, as opposed to patient or parent disinterest, which was 8% of the time. ConclusionReferrals to the allergist for confirmatory testing in patients with listed penicillin allergies increased by more than 10 fold. This study illustrates successful tools to support delabeling.

Implementation of Electronic Health Record Tools for Prevention of Venous Thromboembolism in Hospitalized Children with Cancer and Blood Disorders

▪IntroductionVenous thromboembolism (VTE) is the 2nd most common hospital acquired condition at our pediatric hospital. Children with cancer and blood disorders have acquired prothrombotic risk factors. As part of a quality improvement (QI) initiative to decrease VTE in hospitalized children with cancer and blood disorders, we implemented VTE risk assessment, prevention, and documentation tools within the Electronic Health Record (EHR). The baseline documented screening prior to the tool was 0% and sequential compression device (SCD) and prophylactic anticoagulation were rarely prescribed. The primary outcome was percentage of patients with documented screening.MethodsA VTE risk assessment tool, including the Braden Q activity score (score <3 is defined as chair or bedridden) and clinical risk factors, was developed based on published literature, clinical practice guidelines, local experience, and recommendations from the Children's Hospitals' Solutions for Patient Safety. Patients were categorized as LOW, MODERATE, or HIGH risk for VTE and risk-based recommendations for VTE prevention were provided. MODERATE risk was defined as Braden Q activity score <3 plus 0-1 other risk factors; SCD were recommended if not contraindicated. HIGH risk was defined as central venous catheter (CVC) in place and/or Braden Q activity score <3 plus 2 or more other risk factors; SCD and prophylactic anticoagulation were recommended if no contraindications. Patients are eligible for risk assessment if they are 10-17.99 years and admitted to the hematology/oncology unit at Rady Children's Hospital San Diego for >24 hours.ResultsBetween March 1 to May 31, 2017, 57 unique patients had a total of 107 admissions, and VTE risk assessment was completed for 102 admissions (95.3%), see Table 1. Members of the VTE Prevention Taskforce reviewed the accuracy of the risk assessments. Four LOW risk patients were reclassified as HIGH risk and 1 MODERATE risk patient was reclassified as HIGH risk. One HIGH risk patients with new diagnosis of VTE had therapeutic anticoagulation ordered and SCD placed on the unaffected lower extremity. Three MODERATE risk patients had SCD placed. Two patients who were initially classified as LOW risk during hospitalization were subsequently diagnosed with CVC related upper extremity deep vein thrombosis (DVT) after hospital discharge. New risk factors at the time of DVT diagnosis were obesity in one patient and recent asparaginase treatment in the other.Tools which positively impacted adherence with the risk assessment are a Best Practice Advisory (BPA) which fired for medical providers when an eligible patient was admitted, a BPA which fired once the risk category was assigned, a VTE Registry which included all eligible patients, and a VTE Quality Measure Screen which the VTE Prevention Champion used to identify patients who had not been screened and those who were identified at HIGH and MODERATE risk. Medical decision making was documented with a VTE risk assessment note located in the Admission Navigator.ConclusionsWe effectively implemented VTE risk assessment, prevention, and documentation tools in the EHR. Additional QI cycles are required to determine the optimal VTE risk assessment, to improve accuracy of screening, and to increase adherence with recommended interventions to decrease VTE. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Universal Screening for Hepatitis C Virus in the ED Using a Best Practice Advisory.

IntroductionIn 2019 the United States Preventive Services Task Force (USPSTF) released draft guidelines recommending universal hepatitis C virus (HCV) screening for individuals aged 18–79. We aimed to assess the efficacy of an emergency department-based HCV screening program, by comparing screening practices before and after its implementation.MethodsWe performed a retrospective cohort analysis of two temporally matched, 11-month study periods, corresponding to before and after the implementation of a best practice advisory (BPA). Patients were screened for anti-HCV antibody (Ab), and positive results were followed by HCV viral load (VL) testing. The primary implementation outcome was ED testing volume (number of tests performed/month). The primary screening outcomes were the seroprevalence of anti-HCV Ab and HCV VL. We describe data with simple descriptive statistics.ResultsThe median age of patients was similar between periods (pre: 50 years [interquartile range [IQR] 34–62], post: 47 years [IQR 33–59]). Patients screened were more likely to be males in the pre-BPA period (Male, pre: 60%, post: 49%). During the pre-BPA study period, a total of 69,604 patients were seen in the ED, and 218 unique patients were screened for HCV (mean 19.8 tests/month). During the post-BPA study period, a total of 68,225 patients were seen in the ED, and 14,981 unique patients were screened for HCV (mean 1361.9 tests/month). Anti-HCV Ab seroprevalence was 23% (51/218) and 9% (1340/14,981) in the pre-BPA and post-BPA periods, respectively. In the pre-BPA period, six patients with a positive anti-HCV Ab level had follow-up VL testing (detectable in three). In the post-BPA period, reflex VL testing was performed in most patients (91%, 1225/1,340), and there were 563 patients with detectable VLs, indicating active infection.ConclusionOur study shows that using a universal BPA-driven screening protocol can dramatically increase the number of patients screened for HCV and increase the number of new HCV diagnoses.

Evaluation of a Pharmacist-Led Diabetes Collaborative Drug Therapy Management Service

Objective: To evaluate the pharmacist-led diabetes collaborative drug therapy management services in a family medicine and internal medicine clinic. Design: Mixed methods of evaluation based on the Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) framework. Results: Reach: 71.3% of patients who were independently consulted (n = 184/258) and 1.6% (n = 11/680) of patients who triggered a best practice advisory (BPA) were enrolled. Effectiveness: 27.7% of patients (n = 54/195) enrolled were lost to follow-up. Adoption: 55% of eligible providers (n = 77/140) have placed a consult. Implementation: Providers independently choose to refer patients and are also prompted to place consults by a BPA that triggers for patients with an HbA1c ≥ 9%. Common reasons providers did not place a consult include: alignment with workflow, patient refusal, and patients followed by other services. Regarding patient perceptions, patients valued the service. Patients reported increased accountability with disease state maintenance and increased self-efficacy. Regarding how to improve the service, patients wanted more information on expectations before engaging with the pharmacist. Patients suggested to replicate this service for pain, cancer, and blood pressure management. Maintenance: 96.7% of providers (n = 30/31) reported they were very likely/likely to place a consult in the future and 60% of providers (n = 18/30) reported they were very likely/likely to place a consult when prompted by the BPA. Conclusion: These results can be utilized to make improvements to the pharmacist-led diabetes collaborative drug therapy management service to ensure sustainability. This study also provides lessons learned and strategies for future adoption, implementation, and maintenance of similar services for other disease states.

Increasing Naloxone Prescribing in the Emergency Department Through Education and Electronic Medical Record Work-Aids

Emergency department (ED) visits for opioid overdose continue to rise. Evidence-based harm reduction strategies for opioid use disorder (OUD), such as providing home naloxone, can save lives, but ED implementation remains challenging. The researchers aimed to increase prescribing of naloxone to ED patients with OUD and opioid overdose by employing a model for improvement methodology, a multidisciplinary team, and high-reliability interventions. Monthly naloxone prescribing rates among discharged ED patients with opioid overdose and OUD-related diagnoses were tracked over time. Interventions included focused ED staff education on OUD and naloxone, and creation of electronic medical record (EMR)-based work-aids, including a naloxone Best Practice Advisory (BPA) and order set. Autoregressive interrupted time series was used to model the impact of these interventions on naloxone prescribing rates. The impact of education on ED staff confidence and perceived barriers to prescribing naloxone was measured using a published survey instrument. After adjusting for education events and temporal trends, ED naloxone BPA and order set implementation was associated with a significant immediate 21.1% increase in naloxone prescribing rates, which was sustained for one year. This corresponded to increased average monthly prescribing rates from 1.5% before any intervention to 28.7% afterward. ED staff education had no measurable impact on prescribing rates but was associated with increased nursing perceived importance and increased provider confidence in prescribing naloxone. A significant increase in naloxone prescribing rates was achieved after implementation of high-reliability EMR work-aids and staff education. Similar interventions may be key to wider ED staff engagement in harm reduction for patients with OUD.

Universal HCV Screening of Baby Boomers is Feasible, but It is Difficult

Our objective was to assess the impact of mass mailing and the inclusion of Best Practice Advisory (BPA) "Pop-Up" tool in the electronic medical record (EMR) on HCV screening rates. Between June 2015 and March 2020, two interventions were developed for primary care physicians (PCP). An educational letter along with a blood requisition form, signed on behalf of the PCPs, was sent to patients. We also developed a BPA "Pop-Up" screening tool to alert PCPs to order HCV screening tests on patients with no previous screening. Data were collected and analyzed prospectively. When we started the screening program in June 2015, 33,736 baby boomers were eligible for screening, and the hospital system added an additional 26,027 baby boomers between June 2015 and March 2020. Of the 89 primary care providers employed by the hospital, 75 agreed to participate at different time periods. We screened 23,291 (43.5%) of 53,526 eligible patients during study period. Of these, 399 (1.7%) had HCV antibody, but HCV RNA was positive in only 195 (1%). HCV antibody positivity rates were higher in men, blacks, and in 1951-1960 birth cohorts. Spontaneous clearance rates appeared to be lower in men (OR 0.59, 95% CI 0.39-0.90, P= 0.015) and in blacks (OR 0.31, 95% CI 0.20-0.50, P < 0.001). Although a formal screening program increased screening rates for HCV among baby boomers, about 50% of baby boomers remained unscreened. In this community screening program, we found that men and blacks are less likely to have spontaneous HCV clearance.

Urinary tract infection stewardship: A urinary antibiogram and electronic medical record alert nudging narrower-spectrum antibiotics for urinary tract infections.

An antimicrobial stewardship intervention consisting of a urinary antibiogram and an electronic health record best-practice advisory promoted narrower-spectrum antibiotics for uncomplicated urinary tract infections in hospitalized patients. Over 20 months, the intervention significantly reduced ceftriaxone orders by 48% (P < .001) and increased cefazolin use 19 times from baseline (P < .001).

Abstract 14379: Implementation of Pharmacist-led Antithrombotic Medication Management for Elective Endoscopic Procedures

Introduction: Mismanagement of antithrombotic medications often leads to cancelled elective procedures. Pharmacist-led anticoagulation management prior to elective procedures is not well studied. Methods: We implemented a best practice advisory (BPA) that offered referral to a pharmacist-led medication management service prior to elective outpatient gastrointestinal endoscopies. Eligible patients were taking warfarin, a direct oral anticoagulant, and/or a P2Y12 inhibitor for management of atrial fibrillation (AF), venous thromboembolism (VTE), coronary artery disease (CAD), peripheral arterial disease (PAD), and/or mechanical heart valve. Patients referred to the antithrombotic service were compared to those managed by the ordering provider. Outcomes assessed included documentation of a medication management plan, guideline-appropriateness of drug cessation prior to endoscopy, and guideline-appropriate rates of enoxaparin bridging for high risk warfarin patients. Results: Seventy-four percent of patients (448/645) were referred to the antithrombotic service. These patients were more likely to have had VTE (33.0% vs. 18.2%), less likely to have CAD (31.9% vs. 40.9%; p=0.009 for indication category), and more likely to be prescribed warfarin (32.8% vs 20.9%, p&lt;0.001). After adjusting for baseline differences, patients referred to the antithrombotic service were more likely to have had pre-procedure medication plan documentation (98.7% vs 56.1%, p &lt;0.001), and pre-procedure antithrombotic management was more likely to follow guidelines (97.5% vs. 91.5%, p=0.024). Guideline-appropriate use of bridging enoxaparin for warfarin patients was not different (10/16 [62.5%] vs. 2/6 [33.3%], p=0.221). Conclusion: Pre-procedural antithrombotic medication management by a pharmacist-led team for elective endoscopies was associated with more guideline-adherent management and better documentation of the medication plans.

The Impact of a New Best Practice Advisory on the Management of Diabetic Ketoacidosis.

According to the Center for Disease Control and Prevention, diabetic ketoacidosis (DKA) hospitalization rates have been steadily increasing. Due to the increasing incidence and the economic impact associated with its morbidity and treatment, effective management is key. We aimed to streamline the management of DKA in our intensive care units (ICU) by implementing a Best-Practice Advisory (BPA) that notifies providers when DKA has resolved. A BPA was implemented on 9/15/2018. We conducted a retrospective review of patients admitted to the ICU with DKA a year before and after 9/15/2018. Adults (≥18 age) meeting DKA criteria on admission and treated with continuous insulin infusion (CII) were included. Pre-intervention group included patients admitted before BPA implementation and post-intervention group included patients admitted after. Summary and univariate analyses were performed. A total of 282 patients were included; 162 (57%) pre-intervention and 120 (43%) post-intervention. Mean (±SD) age of the patients was 44 (±17) years. There was no significant difference in baseline characteristics such as age, sex, race, BMI, HbA1c, initial blood glucose, anion gap or bicarbonate concentration between both the groups (p>0.05). Mean (±SD) total time on CII in hours was significantly lower in the post-intervention group {14.8 (±7.7) vs. 17.5 (±14.3) p=0.041, 95% CI: 0.11-5.3}. The incidence of hypoglycemia was lower in the post-intervention group {n=4 (3%) vs. 17 (10%), p=0.024}. There was no significant difference in hypokalemia, mortality, LOS or ICU stay between both the groups (p>0.05). The BPA introduced in our DKA management algorithm successfully reduced the total time on insulin and the incidence of hypoglycemia.

The Impact of a Best-Practice Advisory on Inpatient Use of Piperacillin-Tazobactam

Background: Antibiotic “time outs” have been identified as a way to decrease inappropriate use of antibiotics in hospitals.1 The University of Vermont Medical Center created a best-practice advisory (BPA) to alert clinicians to review piperacillin-tazobactam prescriptions after 72 hours (Fig. 1). Data examining the use of a BPA as a method to prompt clinicians to perform an antibiotic “time out” are limited. Objective: The purpose of our retrospective study was to evaluate the effectiveness of the BPA on the rate of piperacillin-tazobactam prescribing as measured by defined daily dose per 1,000 patient days (DDD). Methods: The BPA was integrated into the electronic health record and designed to activate once piperacillin-tazobactam has been prescribed for ≥72 hours. Under approval of the University of Vermont’s Institutional Review Board, administered data for piperacillin-tazobactam and 3 control antibiotics (cefazolin, ceftriaxone, and meropenem) were collected for 1 year prior to and 1 year following the launch of the BPA. Administered data were converted to DDD, and an interrupted time-series analysis was performed to evaluate for changes in antibiotic use. Results: The data included 7,094 patients in the preintervention group and 6,661 patients in the postintervention group. The BPA fired 1,478 times. The prescribing rate of piperacillin-tazobactam 1 year prior to the BPA was 32.34 DDD and decreased every month both before (−1.22 DDD) and after (−0.27 DDD) the BPA initiation, with no significant difference in prescribing trends (P = .10). Meropenem prescribing in the BPA era increased each month compared to the pre-BPA period (1.16 DDD; P = 0.02), whereas cefazolin use (P = .93) and ceftriaxone (P = .09) use did not significantly change. Conclusions: The data show that piperacillin-tazobactam utilization at our institution is decreasing. Considering that this trend started prior to the launch of the BPA and that rate of decline remained unchanged post-BPA, we conclude that the BPA did not further impact our piperacillin-tazobactam consumption. It is possible that other factors influencing prescribing account for the observed decline, including an institution-wide educational campaign regarding the appropriate use of broad-spectrum antibiotics that was initiated in the months prior to the BPA. The reason for the significant rise in meropenem post-BPA is unclear. This may be unrelated to the BPA; however, it requires further investigation.1. Core elements of hospital antibiotic stewardship programs. Centers for Disease Control and Prevention website. https://www.cdc.gov/antibioticuse/healthcare/implementation/core-elements.html. Updated July 19, 2019. Accessed October 6, 2019.Funding: NoneDisclosures: None

Clostridioides difficile Testing Stewardship for Laxative Use Is Effective and Safe When Combined With Expert Clinical Input

Background: In January 2019, our large academic medical center implemented hard stops for ordering Clostridiodes difficile nucleic acid amplification testing (NAAT), and required a discussion with an infectious diseases physician if the order was placed in a clinical scenario not consistent with the 2017 IDSA/SHEA C. difficile infection (CDI) testing guidelines. Recently, some groups have expressed concerns that requiring the discontinuation of laxatives may delay the diagnosis of CDI and result in serious adverse outcomes. Methods:C. difficile testing stewardship interventions were performed at 2 hospitals within the same university health system to reduce inappropriate testing. In January 2019, a best practice advisory (BPA) was implemented to alert providers ordering C. difficile NAAT if patients had received laxatives within 24 hours, requiring a discussion with the ID physician to override the hard stop. We reviewed clinical outcomes of patients who had a BPA alert due to laxative use within the past 24 hours April 23 to October 23, 2019. Results: During the study period, there were 235 patients with a BPA because of laxative use within the past 24 hours. Moreover, 55 (23.4%) continued to experience diarrhea after the discontinuation of laxatives and were retested for CDI within 7 days. Only 8 tests returned positive, suggesting that, at most, 3.4% of cases had delayed diagnoses because of the hard stop. This finding is supported by the increase in the percentage of tests positive from 11.6% observed overall to 14.6% (8 of 55) after this intervention. There were no severe CDI cases (ICU admission, colectomy, or death) among patients who had delayed testing due to laxative use. Conclusions: In the setting of laxative use, C. difficile testing stewardship interventions with C. difficile NAAT using a hard-stop BPA are effective in reducing unnecessary testing and safe if they are used in combination with a real-time expert input of the risk of clinical disease.Funding: NoneDisclosures: None

THE IMPACT OF A NEW BEST PRACTICE ADVISORY ON THE MANAGEMENT OF DIABETIC KETOACIDOSIS

SESSION TITLE: Critical Care Posters SESSION TYPE: Original Investigation Posters PRESENTED ON: October 18-21, 2020 PURPOSE: According to the Centers for Disease Control and Prevention, overall age-adjusted diabetic ketoacidosis (DKA) hospitalization rates have been steadily increasing from 2009 to 2014 at an average annual rate of 6.3%. Further, the direct and indirect annual cost of DKA hospitalizations is 2.4 billion US dollars. Due to the increasing incidence of DKA and the economic impact associated with its morbidity and treatment, effective management is key. Herein, we aimed to streamline the management of DKA in our intensive care units (ICU) by implementing a Best-Practice Advisory (BPA) that notifies care providers when DKA has resolved. METHODS: A BPA was implemented in our ICUs on 9/15/2018 to notify clinical providers when DKA resolves. DKA resolution was defined as a blood glucose (BG)≤250mg/dl, anion gap≤17 and bicarbonate concentration≥18 mEq/L. We conducted a retrospective review of patients admitted to the ICU with DKA a year before and after 9/15/2018. Adult (≥18 age) patients were included in the study if they met DKA criteria on admission and were treated with continuous insulin infusion (CII). Patients admitted prior to BPA implementation were in the pre-intervention group while patients admitted after BPA implementation were in the post-intervention group. Our primary outcome was total time on CII. Secondary outcomes included the incidence of hypoglycemia (<80mg/dl), hypokalemia (<3.5 mEq/L), mortality, length of stay (LOS) and ICU stay. Summary and univariate analyses were performed. RESULTS: A total of 282 patients were included in the study. There were 162 (57%) patients in the pre-intervention group and 120 (43%) patients in the post-intervention group; 143 patients (51%) were females. The mean (±SD) age was 44 (±17) years. There was no significant difference in baseline characteristics such as age, sex, race, BMI, HbA1c, initial BG, initial anion gap or initial bicarbonate concentration between both groups (p>0.05). The mean (±SD) total time on CII in hours was significantly lower in the post-intervention group {14.8 (±7.7) vs 17.5 (±14.3) p=0.041, 95% CI: 0.11-5.3}. Lastly, the incidence of hypoglycemia was lower in the post-intervention group {n=4 (3%) vs 17 (10%), p=0.024}. There was no significant difference in hypokalemia, mortality, LOS or ICU stay between both groups (p>0.05). CONCLUSIONS: The BPA introduced in our DKA management algorithm successfully reduced total time on insulin and the incidence of hypoglycemia. CLINICAL IMPLICATIONS: Incorporating clinical decision support tools, such as Best-Practice Advisory, in the management of Diabetic Ketoacidosis improves key clinical outcomes. DISCLOSURES: No relevant relationships by Ahmad Al-Shyoukh, source=Web Response No relevant relationships by Diala Alawneh, source=Web Response no disclosure on file for Majdi Hamarshi; No relevant relationships by Moustafa Younis, source=Web Response