Enhanced family history-based algorithms increase the identification of individuals meeting criteria for genetic testing of hereditary cancer syndromes but would not reduce disparities on their own.

OBJECTIVE: This study aimed to 1) investigate algorithm enhancements for identifying patients eligible for genetic testing of hereditary cancer syndromes using family history data from electronic health records (EHRs); and 2) assess their impact on relative differences across sex, race, ethnicity, and language preference. MATERIALS AND METHODS: The study used EHR data from a tertiary academic medical center. A baseline rule-base algorithm, relying on structured family history data (structured data; SD), was enhanced using a natural language processing (NLP) component and a relaxed criteria algorithm (partial match [PM]). The identification rates and differences were analyzed considering sex, race, ethnicity, and language preference. RESULTS: Among 120,007 patients aged 25-60, detection rate differences were found across all groups using the SD (all P < 0.001). Both enhancements increased identification rates; NLP led to a 1.9 % increase and the relaxed criteria algorithm (PM) led to an 18.5 % increase (both P < 0.001). Combining SD with NLP and PM yielded a 20.4 % increase (P < 0.001). Similar increases were observed within subgroups. Relative differences persisted across most categories for the enhanced algorithms, with disproportionately higher identification of patients who are White, Female, non-Hispanic, and whose preferred language is English. CONCLUSION: Algorithm enhancements increased identification rates for patients eligible for genetic testing of hereditary cancer syndromes, regardless of sex, race, ethnicity, and language preference. However, differences in identification rates persisted, emphasizing the need for additional strategies to reduce disparities such as addressing underlying biases in EHR family health information and selectively applying algorithm enhancements for disadvantaged populations. Systematic assessment of differences in algorithm performance across population subgroups should be incorporated into algorithm development processes.

Artificial Intelligence and Cancer Control: Toward Prioritizing Justice, Equity, Diversity, and Inclusion (JEDI) in Emerging Decision Support Technologies.

PURPOSE FOR REVIEW: This perspective piece has two goals: first, to describe issues related to artificial intelligence-based applications for cancer control as they may impact health inequities or disparities; and second, to report on a review of systematic reviews and meta-analyses of artificial intelligence-based tools for cancer control to ascertain the extent to which discussions of justice, equity, diversity, inclusion, or health disparities manifest in syntheses of the field's best evidence. RECENT FINDINGS: We found that, while a significant proportion of existing syntheses of research on AI-based tools in cancer control use formal bias assessment tools, the fairness or equitability of models is not yet systematically analyzable across studies. Issues related to real-world use of AI-based tools for cancer control, such as workflow considerations, measures of usability and acceptance, or tool architecture, are more visible in the literature, but still addressed only in a minority of reviews. Artificial intelligence is poised to bring significant benefits to a wide range of applications in cancer control, but more thorough and standardized evaluations and reporting of model fairness are required to build the evidence base for AI-based tool design for cancer and to ensure that these emerging technologies promote equitable healthcare.

A method for structuring complex clinical knowledge and its representational formalisms to support composite knowledge interoperability in healthcare.

INTRODUCTION: The use and interoperability of clinical knowledge starts with the quality of the formalism utilized to express medical expertise. However, a crucial challenge is that existing formalisms are often suboptimal, lacking the fidelity to represent complex knowledge thoroughly and concisely. Often this leads to difficulties when seeking to unambiguously capture, share, and implement the knowledge for care improvement in clinical information systems used by providers and patients. OBJECTIVES: To provide a systematic method to address some of the complexities of knowledge composition and interoperability related to standards-based representational formalisms of medical knowledge. METHODS: Several cross-industry (Healthcare, Linguistics, System Engineering, Standards Development, and Knowledge Engineering) frameworks were synthesized into a proposed reference knowledge framework. The framework utilizes IEEE 42010, the MetaObject Facility, the Semantic Triangle, an Ontology Framework, and the Domain and Comprehensibility Appropriateness criteria. The steps taken were: 1) identify foundational cross-industry frameworks, 2) select architecture description method, 3) define life cycle viewpoints, 4) define representation and knowledge viewpoints, 5) define relationships between neighboring viewpoints, and 6) establish characteristic definitions of the relationships between components. System engineering principles applied included separation of concerns, cohesion, and loose coupling. RESULTS: A "Multilayer Metamodel for Representation and Knowledge" (M*R/K) reference framework was defined. It provides a standard vocabulary for organizing and articulating medical knowledge curation perspectives, concepts, and relationships across the artifacts created during the life cycle of language creation, authoring medical knowledge, and knowledge implementation in clinical information systems such as electronic health records (EHR). CONCLUSION: M*R/K provides a systematic means to address some of the complexities of knowledge composition and interoperability related to medical knowledge representations used in diverse standards. The framework may be used to guide the development, assessment, and coordinated use of knowledge representation formalisms. M*R/K could promote the alignment and aggregated use of distinct domain-specific languages in composite knowledge artifacts such as clinical practice guidelines (CPGs).

Considerations for using predictive models that include race as an input variable: The case study of lung cancer screening.

Indiscriminate use of predictive models incorporating race can reinforce biases present in source data and lead to an exacerbation of health disparities. In some countries, such as the United States, there is therefore a push to remove race from prediction models; however, there are still many prediction models that use race as an input. Biomedical informaticists who are given the responsibility of using these predictive models in healthcare environments are likely to be faced with questions like how to deal with race covariates in these models. Thus, there is a need for a pragmatic framework to help model users think through how to include race in their chosen model so as to avoid inadvertently exacerbating disparities. In this paper, we use the case study of lung cancer screening to propose a simple framework to guide how model users can approach the use (or non-use) of race inputs in the predictive models they are tasked with leveraging in electronic health records and clinical workflows.

Predicting pharmacotherapeutic outcomes for type 2 diabetes: An evaluation of three approaches to leveraging electronic health record data from multiple sources.

Electronic health record (EHR) data are increasingly used to develop prediction models to support clinical care, including the care of patients with common chronic conditions. A key challenge for individual healthcare systems in developing such models is that they may not be able to achieve the desired degree of robustness using only their own data. A potential solution-combining data from multiple sources-faces barriers such as the need for data normalization and concerns about sharing patient information across institutions. To address these challenges, we evaluated three alternative approaches to using EHR data from multiple healthcare systems in predicting the outcome of pharmacotherapy for type 2 diabetes mellitus(T2DM). Two of the three approaches, named Selecting Better (SB) and Weighted Average(WA), allowed the data to remain within institutional boundaries by using pre-built prediction models; the third, named Combining Data (CD), aggregated raw patient data into a single dataset. The prediction performance and prediction coverage of the resulting models were compared to single-institution models to help judge the relative value of adding external data and to determine the best method to generate optimal models for clinical decision support. The results showed that models using WA and CD achieved higher prediction performance than single-institution models for common treatment patterns. CD outperformed the other two approaches in prediction coverage, which we defined as the number of treatment patterns predicted with an Area Under Curve of 0.70 or more. We concluded that 1) WA is an effective option for improving prediction performance for common treatment patterns when data cannot be shared across institutional boundaries and 2) CD is the most effective approach when such sharing is possible, especially for increasing the range of treatment patterns that can be predicted to support clinical decision making.

Evaluation in Life Cycle of Information Technology (ELICIT) framework: Supporting the innovation life cycle from business case assessment to summative evaluation.

OBJECTIVE: Our objective was to develop an evaluation framework for electronic health record (EHR)-integrated innovations to support evaluation activities at each of four information technology (IT) life cycle phases: planning, development, implementation, and operation. METHODS: The evaluation framework was developed based on a review of existing evaluation frameworks from health informatics and other domains (human factors engineering, software engineering, and social sciences); expert consensus; and real-world testing in multiple EHR-integrated innovation studies. RESULTS: The resulting Evaluation in Life Cycle of IT (ELICIT) framework covers four IT life cycle phases and three measure levels (society, user, and IT). The ELICIT framework recommends 12 evaluation steps: (1) business case assessment; (2) stakeholder requirements gathering; (3) technical requirements gathering; (4) technical acceptability assessment; (5) user acceptability assessment; (6) social acceptability assessment; (7) social implementation assessment; (8) initial user satisfaction assessment; (9) technical implementation assessment; (10) technical portability assessment; (11) long-term user satisfaction assessment; and (12) social outcomes assessment. DISCUSSION: Effective evaluation requires a shared understanding and collaboration across disciplines throughout the entire IT life cycle. In contrast with previous evaluation frameworks, the ELICIT framework focuses on all phases of the IT life cycle across the society, user, and IT levels. Institutions seeking to establish evaluation programs for EHR-integrated innovations could use our framework to create such shared understanding and justify the need to invest in evaluation. CONCLUSION: As health care undergoes a digital transformation, it will be critical for EHR-integrated innovations to be systematically evaluated. The ELICIT framework can facilitate these evaluations.

Decision fatigue in low-value prostate cancer screening.

BACKGROUND: Low-value prostate-specific antigen (PSA) testing is common yet contributes substantial waste and downstream patient harm. Decision fatigue may represent an actionable target to reduce low-value urologic care. The objective of this study was to determine whether low-value PSA testing patterns by outpatient clinicians are consistent with decision fatigue. METHODS: Outpatient appointments for adult men without prostate cancer were identified at a large academic health system from 2011 through 2018. The authors assessed the association of appointment time with the likelihood of PSA testing, stratified by patient age and appropriateness of testing based on clinical guidelines. Appointments included those scheduled between 8:00 am and 4:59 pm, with noon omitted. Urologists were examined separately from other clinicians. RESULTS: In 1,581,826 outpatient appointments identified, the median patient age was 54 years (interquartile range, 37-66 years), 1,256,152 participants (79.4%) were White, and 133,693 (8.5%) had family history of prostate cancer. PSA testing would have been appropriate in 36.8% of appointments. Clinicians ordered testing in 3.6% of appropriate appointments and in 1.8% of low-value appointments. Appropriate testing was most likely at 8:00 am (reference group). PSA testing declined through 11:00 am (odds ratio [OR], 0.57; 95% CI, 0.50-0.64) and remained depressed through 4:00 pm (P < .001). Low-value testing was overall less likely (P < .001) and followed a similar trend, declining steadily from 8:00 am (OR, 0.48; 95% CI, 0.42-0.56) through 4:00 pm (P < .001; OR, 0.23; 95% CI, 0.18-0.30). Testing patterns in urologists were noticeably different. CONCLUSIONS: Among most clinicians, outpatient PSA testing behaviors appear to be consistent with decision fatigue. These findings establish decision fatigue as a promising, actionable target for reducing wasteful and low-value practices in routine urologic care. LAY SUMMARY: Decision fatigue causes poorer choices to be made with repetitive decision making. This study used medical records to investigate whether decision fatigue influenced clinicians' likelihood of ordering a low-value screening test (prostate-specific antigen [PSA]) for prostate cancer. In more than 1.5 million outpatient appointments by adult men without prostate cancer, the chances of both appropriate and low-value PSA testing declined as the clinic day progressed, with a larger decline for appropriate testing. Testing patterns in urologists were different from those reported by other clinicians. The authors conclude that outpatient PSA testing behaviors appear to be consistent with decision fatigue among most clinicians, and interventions may reduce wasteful testing and downstream patient harms.

Improvement Initiative: End-Tidal Carbon Monoxide Measurement in Newborns Receiving Phototherapy.

OBJECTIVES: To determine, as part of our Utah Newborn Nursery Bilirubin Management Program, whether end-tidal carbon monoxide concentration (ETCOc) measurements in all newborns in our nursery receiving phototherapy were associated with outcomes related to the management of hyperbilirubinemia, including time (hours after birth) when phototherapy was initiated, total duration of phototherapy during the nursery stay, repeat phototherapy treatments, and hospital readmission for phototherapy. STUDY DESIGN: We performed a planned interim analysis of a component of our program in which we measured ETCOc noninvasively using CoSense on each newborn in our nursery receiving phototherapy and recorded specific outcomes related to phototherapy management. RESULTS: Of 1856 newborns admitted to our nursery in a 6-month period in 2020, 170 (9.8%) were treated with phototherapy. An ETCOc reading was successfully obtained in 145 of 151 attempts (96%). Higher ETCOc values were associated with earlier institution of phototherapy and longer duration of phototherapy. For every 1-ppm increase in ETCOc, phototherapy was started 9 hours earlier (95% CI, 3.3-14.8; P = .002) and was administered for an additional 9.3 hours (95% CI, 4.1-14.6; P < .001). Three newborns were readmitted to the hospital for intensive phototherapy; while in the nursery, all 3 had an elevated ETCOc (2.2, 2.6, and 2.9 ppm). CONCLUSIONS: Our findings provide answers to questions raised in the 2004 American Academy of Pediatrics bilirubin guidelines. In our neonatal nursery, measuring ETCOc in all phototherapy recipients was feasible and safe, and the results were associated with multiple aspects of phototherapy management. Higher ETCOc values predicted earlier and longer phototherapy courses.

Comparing models of delivery for cancer genetics services among patients receiving primary care who meet criteria for genetic evaluation in two healthcare systems: BRIDGE randomized controlled trial.

BACKGROUND: Advances in genetics and sequencing technologies are enabling the identification of more individuals with inherited cancer susceptibility who could benefit from tailored screening and prevention recommendations. While cancer family history information is used in primary care settings to identify unaffected patients who could benefit from a cancer genetics evaluation, this information is underutilized. System-level population health management strategies are needed to assist health care systems in identifying patients who may benefit from genetic services. In addition, because of the limited number of trained genetics specialists and increasing patient volume, the development of innovative and sustainable approaches to delivering cancer genetic services is essential. METHODS: We are conducting a randomized controlled trial, entitled Broadening the Reach, Impact, and Delivery of Genetic Services (BRIDGE), to address these needs. The trial is comparing uptake of genetic counseling, uptake of genetic testing, and patient adherence to management recommendations for automated, patient-directed versus enhanced standard of care cancer genetics services delivery models. An algorithm-based system that utilizes structured cancer family history data available in the electronic health record (EHR) is used to identify unaffected patients who receive primary care at the study sites and meet current guidelines for cancer genetic testing. We are enrolling eligible patients at two healthcare systems (University of Utah Health and New York University Langone Health) through outreach to a randomly selected sample of 2780 eligible patients in the two sites, with 1:1 randomization to the genetic services delivery arms within sites. Study outcomes are assessed through genetics clinic records, EHR, and two follow-up questionnaires at 4 weeks and 12 months after last genetic counseling contactpre-test genetic counseling. DISCUSSION: BRIDGE is being conducted in two healthcare systems with different clinical structures and patient populations. Innovative aspects of the trial include a randomized comparison of a chatbot-based genetic services delivery model to standard of care, as well as identification of at-risk individuals through a sustainable EHR-based system. The findings from the BRIDGE trial will advance the state of the science in identification of unaffected patients with inherited cancer susceptibility and delivery of genetic services to those patients. TRIAL REGISTRATION: BRIDGE is registered as NCT03985852 . The trial was registered on June 6, 2019 at clinicaltrials.gov .

Designing, Conducting, and Reporting Clinical Decision Support Studies: Recommendations and Call to Action.

By enabling more efficient and effective medical decision making, computer-based clinical decision support (CDS) could unlock widespread benefits from the significant investment in electronic health record (EHR) systems in the United States. Evidence from high-quality CDS studies is needed to enable and support this vision of CDS-facilitated care optimization, but limited guidance is available in the literature for designing and reporting CDS studies. To address this research gap, this article provides recommendations for designing, conducting, and reporting CDS studies to: 1) ensure that EHR data to inform the CDS are available; 2) choose decision rules that are consistent with local care processes; 3) target the right users and workflows; 4) make the CDS easy to access and use; 5) minimize the burden placed on users; 6) incorporate CDS success factors identified in the literature, in particular the automatic provision of CDS as a part of clinician workflow; 7) ensure that the CDS rules are adequately tested; 8) select meaningful evaluation measures; 9) use as rigorous a study design as is feasible; 10) think about how to deploy the CDS beyond the original host organization; 11) report the study in context; 12) help the audience understand why the intervention succeeded or failed; and 13) consider the financial implications. If adopted, these recommendations should help advance the vision of more efficient, effective care facilitated by useful and widely available CDS.

Patient Interactions With an Automated Conversational Agent Delivering Pretest Genetics Education: Descriptive Study.

BACKGROUND: Cancer genetic testing to assess an individual's cancer risk and to enable genomics-informed cancer treatment has grown exponentially in the past decade. Because of this continued growth and a shortage of health care workers, there is a need for automated strategies that provide high-quality genetics services to patients to reduce the clinical demand for genetics providers. Conversational agents have shown promise in managing mental health, pain, and other chronic conditions and are increasingly being used in cancer genetic services. However, research on how patients interact with these agents to satisfy their information needs is limited. OBJECTIVE: Our primary aim is to assess user interactions with a conversational agent for pretest genetics education. METHODS: We conducted a feasibility study of user interactions with a conversational agent who delivers pretest genetics education to primary care patients without cancer who are eligible for cancer genetic evaluation. The conversational agent provided scripted content similar to that delivered in a pretest genetic counseling visit for cancer genetic testing. Outside of a core set of information delivered to all patients, users were able to navigate within the chat to request additional content in their areas of interest. An artificial intelligence-based preprogrammed library was also established to allow users to ask open-ended questions to the conversational agent. Transcripts of the interactions were recorded. Here, we describe the information selected, time spent to complete the chat, and use of the open-ended question feature. Descriptive statistics were used for quantitative measures, and thematic analyses were used for qualitative responses. RESULTS: We invited 103 patients to participate, of which 88.3% (91/103) were offered access to the conversational agent, 39% (36/91) started the chat, and 32% (30/91) completed the chat. Most users who completed the chat indicated that they wanted to continue with genetic testing (21/30, 70%), few were unsure (9/30, 30%), and no patient declined to move forward with testing. Those who decided to test spent an average of 10 (SD 2.57) minutes on the chat, selected an average of 1.87 (SD 1.2) additional pieces of information, and generally did not ask open-ended questions. Those who were unsure spent 4 more minutes on average (mean 14.1, SD 7.41; P=.03) on the chat, selected an average of 3.67 (SD 2.9) additional pieces of information, and asked at least one open-ended question. CONCLUSIONS: The pretest chat provided enough information for most patients to decide on cancer genetic testing, as indicated by the small number of open-ended questions. A subset of participants were still unsure about receiving genetic testing and may require additional education or interpersonal support before making a testing decision. Conversational agents have the potential to become a scalable alternative for pretest genetics education, reducing the clinical demand on genetics providers.

A systematic review of theoretical constructs in CDS literature.

BACKGROUND: Studies that examine the adoption of clinical decision support (CDS) by healthcare providers have generally lacked a theoretical underpinning. The Unified Theory of Acceptance and Use of Technology (UTAUT) model may provide such a theory-based explanation; however, it is unknown if the model can be applied to the CDS literature. OBJECTIVE: Our overall goal was to develop a taxonomy based on UTAUT constructs that could reliably characterize CDS interventions. METHODS: We used a two-step process: (1) identified randomized controlled trials meeting comparative effectiveness criteria, e.g., evaluating the impact of CDS interventions with and without specific features or implementation strategies; (2) iteratively developed and validated a taxonomy for characterizing differential CDS features or implementation strategies using three raters. RESULTS: Twenty-five studies with 48 comparison arms were identified. We applied three constructs from the UTAUT model and added motivational control to characterize CDS interventions. Inter-rater reliability was as follows for model constructs: performance expectancy (κ = 0.79), effort expectancy (κ = 0.85), social influence (κ = 0.71), and motivational control (κ = 0.87). CONCLUSION: We found that constructs from the UTAUT model and motivational control can reliably characterize features and associated implementation strategies. Our next step is to examine the quantitative relationships between constructs and CDS adoption.

Iterative heuristic design of temporal graphic displays with clinical domain experts.

Conventional electronic health record information displays are not optimized for efficient information processing. Graphical displays that integrate patient information can improve information processing, especially in data-rich environments such as critical care. We propose an adaptable and reusable approach to patient information display with modular graphical components (widgets). We had two study objectives. First, reduce numerous widget prototype alternatives to preferred designs. Second, derive widget design feature recommendations. Using iterative human-centered design methods, we interviewed experts to hone design features of widgets displaying frequently measured data elements, e.g., heart rate, for acute care patient monitoring and real-time clinical decision-making. Participant responses to design queries were coded to calculate feature-set agreement, average prototype score, and prototype agreement. Two iterative interview cycles covering 64 design queries and 86 prototypes were needed to reach consensus on six feature sets. Interviewers agreed that line graphs with a smoothed or averaged trendline, 24-h timeframe, and gradient coloring for urgency were useful and informative features. Moreover, users agreed that widgets should include key functions: (1) adjustable reference ranges, (2) expandable timeframes, and (3) access to details on demand. Participants stated graphical widgets would be used to identify correlating patterns and compare abnormal measures across related data elements at a specific time. Combining theoretical principles and validated design methods was an effective and reproducible approach to designing widgets for healthcare displays. The findings suggest our widget design features and recommendations match critical care clinician expectations for graphical information display of continuous and frequently updated patient data.

Learning hidden patterns from patient multivariate time series data using convolutional neural networks: A case study of healthcare cost prediction.

OBJECTIVE: To develop an effective and scalable individual-level patient cost prediction method by automatically learning hidden temporal patterns from multivariate time series data in patient insurance claims using a convolutional neural network (CNN) architecture. METHODS: We used three years of medical and pharmacy claims data from 2013 to 2016 from a healthcare insurer, where data from the first two years were used to build the model to predict costs in the third year. The data consisted of the multivariate time series of cost, visit and medical features that were shaped as images of patients' health status (i.e., matrices with time windows on one dimension and the medical, visit and cost features on the other dimension). Patients' multivariate time series images were given to a CNN method with a proposed architecture. After hyper-parameter tuning, the proposed architecture consisted of three building blocks of convolution and pooling layers with an LReLU activation function and a customized kernel size at each layer for healthcare data. The proposed CNN learned temporal patterns became inputs to a fully connected layer. We benchmarked the proposed method against three other methods: (1) a spike temporal pattern detection method, as the most accurate method for healthcare cost prediction described to date in the literature; (2) a symbolic temporal pattern detection method, as the most common approach for leveraging healthcare temporal data; and (3) the most commonly used CNN architectures for image pattern detection (i.e., AlexNet, VGGNet and ResNet) (via transfer learning). Moreover, we assessed the contribution of each type of data (i.e., cost, visit and medical). Finally, we externally validated the proposed method against a separate cohort of patients. All prediction performances were measured in terms of mean absolute percentage error (MAPE). RESULTS: The proposed CNN configuration outperformed the spike temporal pattern detection and symbolic temporal pattern detection methods with a MAPE of 1.67 versus 2.02 and 3.66, respectively (p < 0.01). The proposed CNN outperformed ResNet, AlexNet and VGGNet with MAPEs of 4.59, 4.85 and 5.06, respectively (p < 0.01). Removing medical, visit and cost features resulted in MAPEs of 1.98, 1.91 and 2.04, respectively (p < 0.01). CONCLUSIONS: Feature learning through the proposed CNN configuration significantly improved individual-level healthcare cost prediction. The proposed CNN was able to outperform temporal pattern detection methods that look for a pre-defined set of pattern shapes, since it is capable of extracting a variable number of patterns with various shapes. Temporal patterns learned from medical, visit and cost data made significant contributions to the prediction performance. Hyper-parameter tuning showed that considering three-month data patterns has the highest prediction accuracy. Our results showed that patients' images extracted from multivariate time series data are different from regular images, and hence require unique designs of CNN architectures. The proposed method for converting multivariate time series data of patients into images and tuning them for convolutional learning could be applied in many other healthcare applications with multivariate time series data.

Healthcare cost prediction: Leveraging fine-grain temporal patterns.

OBJECTIVE: To design and assess a method to leverage individuals' temporal data for predicting their healthcare cost. To achieve this goal, we first used patients' temporal data in their fine-grain form as opposed to coarse-grain form. Second, we devised novel spike detection features to extract temporal patterns that improve the performance of cost prediction. Third, we evaluated the effectiveness of different types of temporal features based on cost information, visit information and medical information for the prediction task. MATERIALS AND METHODS: We used three years of medical and pharmacy claims data from 2013 to 2016 from a healthcare insurer, where the first two years were used to build the model to predict the costs in the third year. To prepare the data for modeling and prediction, the time series data of cost, visit and medical information were extracted in the form of fine-grain features (i.e., segmenting each time series into a sequence of consecutive windows and representing each window by various statistics such as sum). Then, temporal patterns of the time series were extracted and added to fine-grain features using a novel set of spike detection features (i.e., the fluctuation of data points). Gradient Boosting was applied on the final set of extracted features. Moreover, the contribution of each type of data (i.e., cost, visit and medical) was assessed. We benchmarked the proposed predictors against extant methods including those that used coarse-grain features which represent each time series with various statistics such as sum and the most recent portion of the values in the entire series. All prediction performances were measured in terms of Mean Absolute Percentage Error (MAPE). RESULTS: Gradient Boosting applied on fine-grain predictors outperformed coarse-grain predictors with a MAPE of 3.02 versus 8.14 (p < 0.01). Enhancing the fine-grain features with the temporal pattern extraction features (i.e., spike detection features) further improved the MAPE to 2.04 (p < 0.01). Removing cost, visit and medical status data resulted in MAPEs of 10.24, 2.22 and 2.07 respectively (p < 0.01 for the first two comparisons and p = 0.63 for the third comparison). CONCLUSIONS: Leveraging fine-grain temporal patterns for healthcare cost prediction significantly improves prediction performance. Enhancing fine-grain features with extraction of temporal cost and visit patterns significantly improved the performance. However, medical features did not have a significant effect on prediction performance. Gradient Boosting outperformed all other prediction models.

Clinical decision support models and frameworks: Seeking to address research issues underlying implementation successes and failures.

Computer-based clinical decision support (CDS) has been pursued for more than five decades. Despite notable accomplishments and successes, wide adoption and broad use of CDS in clinical practice has not been achieved. Many issues have been identified as being partially responsible for the relatively slow adoption and lack of impact, including deficiencies in leadership, recognition of purpose, understanding of human interaction and workflow implications of CDS, cognitive models of the role of CDS, and proprietary implementations with limited interoperability and sharing. To address limitations, many approaches have been proposed and evaluated, drawing on theoretical frameworks, as well as management, technical and other disciplines and experiences. It seems clear, because of the multiple perspectives involved, that no single model or framework is adequate to encompass these challenges. This Viewpoint paper seeks to review the various foci of CDS and to identify aspects in which theoretical models and frameworks for CDS have been explored or could be explored and where they might be expected to be most useful.

Single-reviewer electronic phenotyping validation in operational settings: Comparison of strategies and recommendations.

OBJECTIVE: Develop evidence-based recommendations for single-reviewer validation of electronic phenotyping results in operational settings. MATERIAL AND METHODS: We conducted a randomized controlled study to evaluate whether electronic phenotyping results should be used to support manual chart review during single-reviewer electronic phenotyping validation (N=3104). We evaluated the accuracy, duration and cost of manual chart review with and without the availability of electronic phenotyping results, including relevant patient-specific details. The cost of identification of an erroneous electronic phenotyping result was calculated based on the personnel time required for the initial chart review and subsequent adjudication of discrepancies between manual chart review results and electronic phenotype determinations. RESULTS: Providing electronic phenotyping results (vs not providing those results) was associated with improved overall accuracy of manual chart review (98.90% vs 92.46%, p<0.001), decreased review duration per test case (62.43 vs 76.78s, p<0.001), and insignificantly reduced estimated marginal costs of identification of an erroneous electronic phenotyping result ($48.54 vs $63.56, p=0.16). The agreement between chart review and electronic phenotyping results was higher when the phenotyping results were provided (Cohen's kappa 0.98 vs 0.88, p<0.001). As a result, while accuracy improved when initial electronic phenotyping results were correct (99.74% vs 92.67%, N=3049, p<0.001), there was a trend towards decreased accuracy when initial electronic phenotyping results were erroneous (56.67% vs 80.00%, N=55, p=0.07). Electronic phenotyping results provided the greatest benefit for the accurate identification of rare exclusion criteria. DISCUSSION: Single-reviewer chart review of electronic phenotyping can be conducted more accurately, quickly, and at lower cost when supported by electronic phenotyping results. However, human reviewers tend to agree with electronic phenotyping results even when those results are wrong. Thus, the value of providing electronic phenotyping results depends on the accuracy of the underlying electronic phenotyping algorithm. CONCLUSION: We recommend using a mix of phenotyping validation strategies, with the balance of strategies based on the anticipated electronic phenotyping error rate, the tolerance for missed electronic phenotyping errors, as well as the expertise, cost, and availability of personnel involved in chart review and discrepancy adjudication.

User-centered design of multi-gene sequencing panel reports for clinicians.

The objective of this study was to develop a high-fidelity prototype for delivering multi-gene sequencing panel (GS) reports to clinicians that simulates the user experience of a final application. The delivery and use of GS reports can occur within complex and high-paced healthcare environments. We employ a user-centered software design approach in a focus group setting in order to facilitate gathering rich contextual information from a diverse group of stakeholders potentially impacted by the delivery of GS reports relevant to two precision medicine programs at the University of Maryland Medical Center. Responses from focus group sessions were transcribed, coded and analyzed by two team members. Notification mechanisms and information resources preferred by participants from our first phase of focus groups were incorporated into scenarios and the design of a software prototype for delivering GS reports. The goal of our second phase of focus group, to gain input on the prototype software design, was accomplished through conducting task walkthroughs with GS reporting scenarios. Preferences for notification, content and consultation from genetics specialists appeared to depend upon familiarity with scenarios for ordering and delivering GS reports. Despite familiarity with some aspects of the scenarios we proposed, many of our participants agreed that they would likely seek consultation from a genetics specialist after viewing the test reports. In addition, participants offered design and content recommendations. Findings illustrated a need to support customized notification approaches, user-specific information, and access to genetics specialists with GS reports. These design principles can be incorporated into software applications that deliver GS reports. Our user-centered approach to conduct this assessment and the specific input we received from clinicians may also be relevant to others working on similar projects.

The genomic CDS sandbox: An assessment among domain experts.

Genomics is a promising tool that is becoming more widely available to improve the care and treatment of individuals. While there is much assertion, genomics will most certainly require the use of clinical decision support (CDS) to be fully realized in the routine clinical setting. The National Human Genome Research Institute (NHGRI) of the National Institutes of Health recently convened an in-person, multi-day meeting on this topic. It was widely recognized that there is a need to promote the innovation and development of resources for genomic CDS such as a CDS sandbox. The purpose of this study was to evaluate a proposed approach for such a genomic CDS sandbox among domain experts and potential users. Survey results indicate a significant interest and desire for a genomic CDS sandbox environment among domain experts. These results will be used to guide the development of a genomic CDS sandbox.