PheNominal: an EHR-integrated web application for structured deep phenotyping at the point of care.

BACKGROUND: Clinical phenotype information greatly facilitates genetic diagnostic interpretations pipelines in disease. While post-hoc extraction using natural language processing on unstructured clinical notes continues to improve, there is a need to improve point-of-care collection of patient phenotypes. Therefore, we developed "PheNominal", a point-of-care web application, embedded within Epic electronic health record (EHR) workflows, to permit capture of standardized phenotype data. METHODS: Using bi-directional web services available within commercial EHRs, we developed a lightweight web application that allows users to rapidly browse and identify relevant terms from the Human Phenotype Ontology (HPO). Selected terms are saved discretely within the patient's EHR, permitting reuse both in clinical notes as well as in downstream diagnostic and research pipelines. RESULTS: In the 16 months since implementation, PheNominal was used to capture discrete phenotype data for over 1500 individuals and 11,000 HPO terms during clinic and inpatient encounters for a genetic diagnostic consultation service within a quaternary-care pediatric academic medical center. An average of 7 HPO terms were captured per patient. Compared to a manual workflow, the average time to enter terms for a patient was reduced from 15 to 5 min per patient, and there were fewer annotation errors. CONCLUSIONS: Modern EHRs support integration of external applications using application programming interfaces. We describe a practical application of these interfaces to facilitate deep phenotype capture in a discrete, structured format within a busy clinical workflow. Future versions will include a vendor-agnostic implementation using FHIR. We describe pilot efforts to integrate structured phenotyping through controlled dictionaries into diagnostic and research pipelines, reducing manual effort for phenotype documentation and reducing errors in data entry.

Structured override reasons for drug-drug interaction alerts in electronic health records.

OBJECTIVE: The study sought to determine availability and use of structured override reasons for drug-drug interaction (DDI) alerts in electronic health records. MATERIALS AND METHODS: We collected data on DDI alerts and override reasons from 10 clinical sites across the United States using a variety of electronic health records. We used a multistage iterative card sort method to categorize the override reasons from all sites and identified best practices. RESULTS: Our methodology established 177 unique override reasons across the 10 sites. The number of coded override reasons at each site ranged from 3 to 100. Many sites offered override reasons not relevant to DDIs. Twelve categories of override reasons were identified. Three categories accounted for 78% of all overrides: "will monitor or take precautions," "not clinically significant," and "benefit outweighs risk." DISCUSSION: We found wide variability in override reasons between sites and many opportunities to improve alerts. Some override reasons were irrelevant to DDIs. Many override reasons attested to a future action (eg, decreasing a dose or ordering monitoring tests), which requires an additional step after the alert is overridden, unless the alert is made actionable. Some override reasons deferred to another party, although override reasons often are not visible to other users. Many override reasons stated that the alert was inaccurate, suggesting that specificity of alerts could be improved. CONCLUSIONS: Organizations should improve the options available to providers who choose to override DDI alerts. DDI alerting systems should be actionable and alerts should be tailored to the patient and drug pairs.

Design and Implementation of a Visual Analytics Electronic Antibiogram within an Electronic Health Record System at a Tertiary Pediatric Hospital.

BACKGROUND: Hospitals use antibiograms to guide optimal empiric antibiotic therapy, reduce inappropriate antibiotic usage, and identify areas requiring intervention by antimicrobial stewardship programs. Creating a hospital antibiogram is a time-consuming manual process that is typically performed annually. OBJECTIVE: We aimed to apply visual analytics software to electronic health record (EHR) data to build an automated, electronic antibiogram ("e-antibiogram") that adheres to national guidelines and contains filters for patient characteristics, thereby providing access to detailed, clinically relevant, and up-to-date antibiotic susceptibility data. METHODS: We used visual analytics software to develop a secure, EHR-linked, condition- and patient-specific e-antibiogram that supplies susceptibility maps for organisms and antibiotics in a comprehensive report that is updated on a monthly basis. Antimicrobial susceptibility data were grouped into nine clinical scenarios according to the specimen source, hospital unit, and infection type. We implemented the e-antibiogram within the EHR system at Children's Hospital of Philadelphia, a tertiary pediatric hospital and analyzed e-antibiogram access sessions from March 2016 to March 2017. RESULTS: The e-antibiogram was implemented in the EHR with over 6,000 inpatient, 4,500 outpatient, and 3,900 emergency department isolates. The e-antibiogram provides access to rolling 12-month pathogen and susceptibility data that is updated on a monthly basis. E-antibiogram access sessions increased from an average of 261 sessions per month during the first 3 months of the study to 345 sessions per month during the final 3 months. CONCLUSION: An e-antibiogram that was built and is updated using EHR data and adheres to national guidelines is a feasible replacement for an annual, static, manually compiled antibiogram. Future research will examine the impact of the e-antibiogram on antibiotic prescribing patterns.

Safety of Automatic End Dates for Antimicrobial Orders to Facilitate Stewardship.

Following implementation of automatic end dates for antimicrobial orders to facilitate antimicrobial stewardship at a large, academic children's hospital, no differences were observed in patient mortality, length of stay, or readmission rates, even among patients with documented bacteremia. Infect Control Hosp Epidemiol 2016;37:974-978.

Optimization of drug-drug interaction alert rules in a pediatric hospital's electronic health record system using a visual analytics dashboard.

OBJECTIVE: To develop and evaluate an electronic dashboard of hospital-wide electronic health record medication alerts for an alert fatigue reduction quality improvement project. METHODS: We used visual analytics software to develop the dashboard. We collaborated with the hospital-wide Clinical Decision Support committee to perform three interventions successively deactivating clinically irrelevant drug-drug interaction (DDI) alert rules. We analyzed the impact of the interventions on care providers' and pharmacists' alert and override rates using an interrupted time series framework with piecewise regression. RESULTS: We evaluated 2 391 880 medication alerts between January 31, 2011 and January 26, 2014. For pharmacists, the median alert rate prior to the first DDI deactivation was 58.74 alerts/100 orders (IQR 54.98-60.48) and 25.11 alerts/100 orders (IQR 23.45-26.57) following the three interventions (p<0.001). For providers, baseline median alert rate prior to the first round of DDI deactivation was 19.73 alerts/100 orders (IQR 18.66-20.24) and 15.11 alerts/100 orders (IQR 14.44-15.49) following the three interventions (p<0.001). In a subgroup analysis, we observed a decrease in pharmacists' override rates for DDI alerts that were not modified in the system from a median of 93.06 overrides/100 alerts (IQR 91.96-94.33) to 85.68 overrides/100 alerts (IQR 84.29-87.15, p<0.001). The medication serious safety event rate decreased during the study period, and there were no serious safety events reported in association with the deactivated alert rules. CONCLUSIONS: An alert dashboard facilitated safe rapid-cycle reductions in alert burden that were temporally associated with lower pharmacist override rates in a subgroup of DDIs not directly affected by the interventions; meanwhile, the pharmacists' frequency of selecting the 'cancel' option increased. We hypothesize that reducing the alert burden enabled pharmacists to devote more attention to clinically relevant alerts.

Randomized Controlled Trial of RSS Reader Use and Resident Familiarity With Primary Literature.

BACKGROUND: Awareness of the primary literature is important for clinicians. Lack of time, poor access to information, and lack of personal initiative may be barriers for some trainees. Really Simple Syndication (RSS) readers aggregate web content, such as journal abstracts, in a single location for easy viewing. OBJECTIVE: We assessed whether use of an RSS reader would increase resident reading frequency, familiarity, and understanding of the primary literature. METHODS: We conducted a single-center, prospective, randomized, nonblinded, controlled trial of the effect of RSS reader use on knowledge of recent literature among pediatrics residents. Residents were randomly assigned to the RSS group (education in RSS use and receipt of the Pediatrics RSS feed) or a control group that followed standard reading practices. Outcome measures were differences on baseline and monthly surveys of reading frequency, familiarity with recent publications, and knowledge of recent articles (familiarity validation). RESULTS: Of 144 eligible residents, 79 (55%) were enrolled in the survey, with 81% (64 of 79) of participants completing all surveys. The RSS reader use was correlated with greater familiarity with selected articles, but not with improved understanding (as measured by ability to answer multiple-choice questions about content). Participants reported satisfaction with the RSS reader based on its ease of use, accessibility, and as an aid in supplementing reading. CONCLUSIONS: The RSS reader use was accepted by residents and associated with increased familiarity with the primary literature but not with increased understanding.

Core drug-drug interaction alerts for inclusion in pediatric electronic health records with computerized prescriber order entry.

OBJECTIVE: The study aims to develop a core set of pediatric drug-drug interaction (DDI) pairs for which electronic alerts should be presented to prescribers during the ordering process. METHODS: A clinical decision support working group composed of Children's Hospital Association (CHA) members was developed. CHA Pharmacists and Chief Medical Information Officers participated. RESULTS: Consensus was reached on a core set of 19 DDI pairs that should be presented to pediatric prescribers during the order process. CONCLUSIONS: We have provided a core list of 19 high value drug pairs for electronic drug-drug interaction alerts to be recommended for inclusion as high value alerts in prescriber order entry software used with a pediatric patient population. We believe this list represents the most important pediatric drug interactions for practical implementation within computerized prescriber order entry systems.

Improving surveillance for hyperammonemia in the newborn.

OBJECTIVE: Prompt ascertainment is crucial for the management of hyperammonemic infants. Because these patients are rare and recognition of hyperammonemia is often delayed, we designed and implemented an electronic medical record (EMR)-based tool to assist physicians in the detection of hyperammonemia. METHODS: We retrospectively evaluated the hospitalizations of prior hyperammonemic infants to identify codable elements that could trigger an EMR-based warning. An alert was designed and implemented and its utilization was prospectively analyzed. RESULTS: Blood gas studies were obtained universally and early in the retrospectively evaluated infants (x¯=26h before ammonia level). Prompting physicians to evaluate ammonia after ordering blood gas studies would have accelerated the initial ammonia order in 89% of retrospective cases. The alert has activated 184 times over the first six months of operation leading to 63 laboratory evaluations and detection of one hyperammonemic infant. CONCLUSION: Implementation of an EMR-based warning system can improve surveillance for hyperammonemia in a susceptible population.

Balancing entrepreneurship and business practices for e-collaboration: responsible information sharing in academic research.

Flexible, highly accessible collaboration tools can inherently conflict with controls placed on information sharing by offices charged with privacy protection, compliance, and maintenance of the general business environment. Our implementation of a commercial enterprise wiki within the academic research environment addresses concerns of all involved through the development of a robust user training program, a suite of software customizations that enhance security elements, a robust auditing program, allowance for inter-institutional wiki collaboration, and wiki-specific governance.

Use of an enterprise wiki as a research collaboration tool.

Biomedical research projects are highly collaborative endeavors with unique information management and communication needs. We describe the pilot use of an enterprise wiki solution to facilitate group communication, secure file sharing, and collaborative writing within a pediatric hospital and research center. We discuss the choice of software, examples of use, and initial user feedback. We conclude that a wiki is a low-cost and high-yield approach to enhance research collaboration.