Theory of radiologist interaction with instant messaging decision support tools: A sequential-explanatory study.

Radiology specific clinical decision support systems (CDSS) and artificial intelligence are poorly integrated into the radiologist workflow. Current research and development efforts of radiology CDSS focus on 4 main interventions, based around exam centric time points-after image acquisition, intra-report support, post-report analysis, and radiology workflow adjacent. We review the literature surrounding CDSS tools in these time points, requirements for CDSS workflow augmentation, and technologies that support clinician to computer workflow augmentation. We develop a theory of radiologist-decision tool interaction using a sequential explanatory study design. The study consists of 2 phases, the first a quantitative survey and the second a qualitative interview study. The phase 1 survey identifies differences between average users and radiologist users in software interventions using the User Acceptance of Information Technology: Toward a Unified View (UTAUT) framework. Phase 2 semi-structured interviews provide narratives on why these differences are found. To build this theory, we propose a novel solution called Radibot-a conversational agent capable of engaging clinicians with CDSS as an assistant using existing instant messaging systems supporting hospital communications. This work contributes an understanding of how radiologist-users differ from the average user and can be utilized by software developers to increase satisfaction of CDSS tools within radiology.

Generalizability of Human Activity Recognition Machine Learning Models from non-Parkinson's to Parkinson's Disease Patients.

Recent evidence shows that high-intensity exercises reduce tremors and stiffness in Parkinson's disease (PD). However, there is insufficient evidence on the types of exercises; in effect, high-intensity may be a personalized measure. Recent progress in automated Human Activity Recognition using machine learning (ML) models shows potential for better monitoring of PD patients. However, ML models must be calibrated to ignore tremors and accurately identify activity and its intensity. We report findings from a study where we trained ML models using data from medically validated triple synchronous sensors connected to 8 non-PD subjects performing 32 exercises. We then tested the models to identify exercises performed by 8 PD patients at different stages of the disease. Our analysis shows that better data preprocessing before modeling can provide some model generalizability. However, it is extremely challenging, as the models work with high accuracy on one group (Healthy or PD patients) (F1=0.88-0.94) but not on both groups.Clinical relevance-Patients with Parkinson's and other motor-generative diseases can now accurately measure physical activity with machine learning approaches. Clinicians, caregivers, and apps can make accurate, personalized exercise recommendations to augment medications that reduce tremors and stiffness.

Mapping the delineation of practice to the AMIA foundational domains for applied health informatics.

OBJECTIVE: This article reports on the alignment between the foundational domains and the delineation of practice (DoP) for health informatics, both developed by the American Medical Informatics Association (AMIA). Whereas the foundational domains guide graduate-level curriculum development and accreditation assessment, providing an educational pathway to the minimum competencies needed as a health informatician, the DoP defines the domains, tasks, knowledge, and skills that a professional needs to competently perform in the discipline of health informatics. The purpose of this article is to determine whether the foundational domains need modification to better reflect applied practice. MATERIALS AND METHODS: Using an iterative process and through individual and collective approaches, the foundational domains and the DoP statements were analyzed for alignment and eventual harmonization. Tables and Sankey plot diagrams were used to detail and illustrate the resulting alignment. RESULTS: We were able to map all the individual DoP knowledge statements and tasks to the AMIA foundational domains, but the statements within a single DoP domain did not all map to the same foundational domain. Even though the AMIA foundational domains and DoP domains are not in perfect alignment, the DoP provides good examples of specific health informatics competencies for most of the foundational domains. There are, however, limited DoP knowledge statements and tasks mapping to foundational domain 6-Social and Behavioral Aspects of Health. DISCUSSION: Both the foundational domains and the DoP were developed independently, several years apart, and for different purposes. The mapping analyses reveal similarities and differences between the practice experience and the curricular needs of health informaticians. CONCLUSIONS: The overall alignment of both domains may be explained by the fact that both describe the current and/or future health informatics professional. One can think of the foundational domains as representing the broad foci for educational programs for health informaticians and, hence, they are appropriately the focus of organizations that accredit these programs.

An answer recommendation framework for an online cancer community forum.

Health community forums are a kind of online platform to discuss various matters related to management of illness. People are increasingly searching for answers online, particularly when they are diagnosed with cancer like life-threatening diseases. People seek suggestions or advice through these platforms to make decisions during their treatments. However, locating the correct information or similar people is often a great challenge for them. In this scenario, this paper proposes an answer recommendation system in an online breast cancer community forum that provide guidance and valuable references to users while making decisions. The answer is the summary of already discussed topic in the forum, so that they do not need to go through all the answer posts which spans over multiple pages or initiate a thread once again. There are three phases for the answer recommendation system, including query similarity model to retrieve the past similar query, query-answer pair generation and answer recommendation. Query similarity model is employed by a Siamese network with Bi-LSTM architecture which could achieve an F1-score of 85.5%. Also, the paper shows the efficacy of transfer learning technique to generalize the model well in our breast cancer query-query pair data set. The query-answer pairs are generated by an extractive summarization technique that is based on an optimization algorithm. The effectiveness of the generated summary is evaluated based on a manually generated summary, and the result shows a ROUGE-1 score of 49%.

The Kidney Transplant Self-Management Scale: Instrument Development and Psychometric Testing.

This study reports the development and psychometric testing of the Kidney Transplant Self-Management Scale (KT-SMS). The instrument development phase included the following: (a) conceptual definition, item generation, and framework; (b) face validity assessment; and (c) content validity assessment. The psychometric testing phase included the following: (a) construct validity testing; (b) internal consistency reliability testing; (c) convergent validity testing; and (d) predictive power of the KT-SMS using a cross-sectional sample of kidney transplant recipients (N = 153). Factor analysis results supported the 16-item KT-SMS as multidimensional with five domains (medication adherence, cardiovascular risk reduction, protecting kidney, ownership, and skin cancer prevention). Internal consistency reliability for the total scale and five subscales was adequate. Convergent validity was supported as the intercorrelations of the KT-SMS total score with the five subscales were significant. The KT-SMS total score and five subscales were significantly correlated with self-efficacy for managing chronic disease, patient activation, and health-related quality of life.

PhenoDEF: a corpus for annotating sentences with information of phenotype definitions in biomedical literature.

BACKGROUND: Adverse events induced by drug-drug interactions are a major concern in the United States. Current research is moving toward using electronic health record (EHR) data, including for adverse drug events discovery. One of the first steps in EHR-based studies is to define a phenotype for establishing a cohort of patients. However, phenotype definitions are not readily available for all phenotypes. One of the first steps of developing automated text mining tools is building a corpus. Therefore, this study aimed to develop annotation guidelines and a gold standard corpus to facilitate building future automated approaches for mining phenotype definitions contained in the literature. Furthermore, our aim is to improve the understanding of how these published phenotype definitions are presented in the literature and how we annotate them for future text mining tasks. RESULTS: Two annotators manually annotated the corpus on a sentence-level for the presence of evidence for phenotype definitions. Three major categories (inclusion, intermediate, and exclusion) with a total of ten dimensions were proposed characterizing major contextual patterns and cues for presenting phenotype definitions in published literature. The developed annotation guidelines were used to annotate the corpus that contained 3971 sentences: 1923 out of 3971 (48.4%) for the inclusion category, 1851 out of 3971 (46.6%) for the intermediate category, and 2273 out of 3971 (57.2%) for exclusion category. The highest number of annotated sentences was 1449 out of 3971 (36.5%) for the "Biomedical & Procedure" dimension. The lowest number of annotated sentences was 49 out of 3971 (1.2%) for "The use of NLP". The overall percent inter-annotator agreement was 97.8%. Percent and Kappa statistics also showed high inter-annotator agreement across all dimensions. CONCLUSIONS: The corpus and annotation guidelines can serve as a foundational informatics approach for annotating and mining phenotype definitions in literature, and can be used later for text mining applications.

Comparison of Ease of Use and Comfort in Fitness Trackers for Participants Impaired by Parkinson's Disease: An exploratory study.

Parkinson's disease (PD) is an incurable, fatal neurodegenerative disease, and only available treatment is to minimize symptoms. Anecdotal evidence suggests whole body workout can help to reduce PD severity; however, it is challenging to quantify its effect on PD. The increased availability of fitness trackers can help in quantifying the effect of whole-body workout on PD. Before using any over the counter fitness tracker, we must study the ease of use of the fitness trackers in PD patients. We interviewed 32 PD patients with six over the counter fitness trackers and determined their perceptions and attitude towards the fitness trackers. Although none of the fitness trackers received perfect scores for ease of use or comfort due to the presence of tremors, two trackers performed significantly better than the others. Further study is warranted to understand the potential for fitness trackers to be used by PD patients.

Deep learning based analysis of sentiment dynamics in online cancer community forums: An experience.

Online health communities (OHC) provide various opportunities for patients with chronic or life-threatening illnesses, especially for cancer patients and survivors. A better understanding of the sentiment dynamics of patients in OHCs can help in the precise formulation of the needs during their treatment. The current study investigated the sentiment dynamics in patients' narratives in a Breast Cancer community group (Breastcancer.org) to identify the changes in emotions, thoughts, stress, and coping mechanisms while undergoing treatment options, particularly chemotherapy, radiation, and surgery. Sentiment dynamics of users' posts was performed using a deep learning model. A sentiment change analysis was performed to measure change in the satisfaction level of the users. The deep learning model BiLSTM with sentiment embedding features provided a better F1-score of 91.9%. Sentiment dynamics can assess the difference in satisfaction level the users acquire by interacting with other users in the forum. A comparison of the proposed model with existing models revealed the effectiveness of this methodology.

Measuring and Visualizing Chlamydia and Gonorrhea Inequality: An Informatics Approach Using Geographical Information Systems.

BACKGROUND: Health inequality measurements are vital in understanding disease patterns in identifying high-risk patients and implementing effective intervention programs to treat and manage sexually transmitted diseases. OBJECTIVES: To measure and identify inequalities among chlamydia and gonorrhea rates using Gini coefficient measurements and spatial visualization mapping from geographical information systems. Additionally, we seek to examine trends of disease rate distribution longitudinally over a ten-year period for an urbanized county. METHODS: Chlamydia and gonorrhea data from January 2005 to December 2014 were collected from the Indiana Network for Patient Care, a health information exchange system that gathers patient data from electronic health records. The Gini coefficient was used to calculate the magnitude of inequality in disease rates. Spatial visualization mapping and decile categorization of disease rates were conducted to identify locations where high and low rates of disease persisted and to visualize differences in inequality. A multiple comparisons ANOVA test was conducted to determine if Gini coefficient values were statistically different between townships and time periods during the study. RESULTS: Our analyses show that chlamydia and gonorrhea rates are not evenly distributed. Inequalities in disease rates existed for different areas of the county with higher disease rates occurring near the center of the county. Inequality in gonorrhea rates were higher than chlamydia rates. Disease rates were statistically different when geographical locations or townships were compared to each other (p < 0.0001) but not for different years or time periods (p = 0.5152). CONCLUSION: The ability to use Gini coefficients combined with spatial visualization techniques presented a valuable opportunity to analyze information from health information systems in investigating health inequalities. Knowledge from this study can benefit and improve health quality, delivery of services, and intervention programs while managing healthcare costs.

Identification of Patients in Need of Advanced Care for Depression Using Data Extracted From a Statewide Health Information Exchange: A Machine Learning Approach.

BACKGROUND: As the most commonly occurring form of mental illness worldwide, depression poses significant health and economic burdens to both the individual and community. Different types of depression pose different levels of risk. Individuals who suffer from mild forms of depression may recover without any assistance or be effectively managed by primary care or family practitioners. However, other forms of depression are far more severe and require advanced care by certified mental health providers. However, identifying cases of depression that require advanced care may be challenging to primary care providers and health care team members whose skill sets run broad rather than deep. OBJECTIVE: This study aimed to leverage a comprehensive range of patient-level diagnostic, behavioral, and demographic data, as well as past visit history data from a statewide health information exchange to build decision models capable of predicting the need of advanced care for depression across patients presenting at Eskenazi Health, the public safety net health system for Marion County, Indianapolis, Indiana. METHODS: Patient-level diagnostic, behavioral, demographic, and past visit history data extracted from structured datasets were merged with outcome variables extracted from unstructured free-text datasets and were used to train random forest decision models that predicted the need of advanced care for depression across (1) the overall patient population and (2) various subsets of patients at higher risk for depression-related adverse events; patients with a past diagnosis of depression; patients with a Charlson comorbidity index of ≥1; patients with a Charlson comorbidity index of ≥2; and all unique patients identified across the 3 above-mentioned high-risk groups. RESULTS: The overall patient population consisted of 84,317 adult (aged ≥18 years) patients. A total of 6992 (8.29%) of these patients were in need of advanced care for depression. Decision models for high-risk patient groups yielded area under the curve (AUC) scores between 86.31% and 94.43%. The decision model for the overall patient population yielded a comparatively lower AUC score of 78.87%. The variance of optimal sensitivity and specificity for all decision models, as identified using Youden J Index, is as follows: sensitivity=68.79% to 83.91% and specificity=76.03% to 92.18%. CONCLUSIONS: This study demonstrates the ability to automate screening for patients in need of advanced care for depression across (1) an overall patient population or (2) various high-risk patient groups using structured datasets covering acute and chronic conditions, patient demographics, behaviors, and past visit history. Furthermore, these results show considerable potential to enable preventative care and can be easily integrated into existing clinical workflows to improve access to wraparound health care services.

Alignment of American Association of Colleges of Nursing Graduate-Level Nursing Informatics Competencies With American Medical Informatics Association Health Informatics Core Competencies.

This study yielded a map of the alignment of American Association of Colleges of Nursing Graduate-Level Nursing Informatics Competencies with American Medical Informatics Association Health Informatics Core Competencies in an effort to understand graduate-level accreditation and certification opportunities in nursing informatics. Nursing Informatics Program Directors from the American Medical Informatics Association and a health informatics expert independently mapped the American Association of Colleges of Nursing competencies to the American Medical Informatics Association Health Informatics knowledge, skills, and attitudes. The Nursing Informatics Program Directors' map connected an average of 4.0 American Medical Informatics Association Core Competencies per American Association of Colleges of Nursing competency, whereas the health informatics expert's map connected an average of 5.0 American Medical Informatics Association Core Competencies per American Association of Colleges of Nursing competency. Agreement across the two maps ranged from 14% to 60% per American Association of Colleges of Nursing competency, revealing alignment between the two groups' competencies according to knowledge, skills, and attitudes. These findings suggest that graduates of master's degree programs in nursing, especially those specializing in nursing informatics, will likely be prepared to sit for the proposed Advanced Health Informatics Certification in addition to the American Nurses Credentialing Center bachelor's-level Informatics Nursing Certification. This preliminary map sets the stage for further in-depth mapping of nursing informatics curricula with American Medical Informatics Association Core Competencies and will enable interprofessional conversations around nursing informatics specialty program accreditation, nursing workforce preparation, and nursing informatics advanced certification. Nursing informaticists should examine their need for credentials as key contributors who will address critical health informatics needs.

Using Facebook in Recruiting Kidney Transplant Recipients for a REDCap Study.

This web-based study recruited kidney transplant recipients from Facebook using three recruiting methods over a 5-week period. Participants completed 125 survey items via REDCap (Research Electronic Data Capture) survey. Facebook recruitment generated 153 eligible participants who completed surveys. The average survey response time was 15.07 min (SD = 6.12; range: 4-43), with a low missing item rate (<5%). Facebook's standard ads were most effective for recruiting subjects (n = 78, 51%), followed by three targeted Facebook kidney transplant support groups (n = 52, 34%) and a pay-to-promote study page (n = 12, 7.8%). The average cost paid for each valid survey was US$2.19 through standard Facebook ads and US$2.92 from the study page. The cost for online survey completion is economically feasible even for those with limited funds. Issues related to online surveys including extreme survey response times and participant misrepresentation were reported in this study.

Impact of document consolidation on healthcare providers' perceived workload and information reconciliation tasks: a mixed methods study.

Background: Information reconciliation is a common yet complex and often time-consuming task performed by healthcare providers. While electronic health record systems can receive "outside information" about a patient in electronic documents, rarely does the computer automate reconciling information about a patient across all documents. Materials and Methods: Using a mixed methods design, we evaluated an information system designed to reconcile information across multiple electronic documents containing health records for a patient received from a health information exchange (HIE) network. Nine healthcare providers participated in scenario-based sessions in which they manually consolidated information across multiple documents. Accuracy of consolidation was measured along with the time spent completing 3 different reconciliation scenarios with and without support from the information system. Participants also attended an interview about their experience. Perceived workload was evaluated quantitatively using the NASA-TLX tool. Qualitative analysis focused on providers' impression of the system and the challenges faced when reconciling information in practice. Results: While 5 providers made mistakes when trying to manually reconcile information across multiple documents, no participants made a mistake when the system supported their work. Overall perceived workload decreased significantly for scenarios supported by the system (37.2% in referrals, 18.4% in medications, and 31.5% in problems scenarios, P < 0.001). Information reconciliation time was reduced significantly when the system supported provider tasks (58.8% in referrals, 38.1% in medications, and 65.1% in problem scenarios). Conclusion: Automating retrieval and reconciliation of information across multiple electronic documents shows promise for reducing healthcare providers' task complexity and workload.

Novel Approach to Cluster Patient-Generated Data Into Actionable Topics: Case Study of a Web-Based Breast Cancer Forum.

BACKGROUND: The increasing use of social media and mHealth apps has generated new opportunities for health care consumers to share information about their health and well-being. Information shared through social media contains not only medical information but also valuable information about how the survivors manage disease and recovery in the context of daily life. OBJECTIVE: The objective of this study was to determine the feasibility of acquiring and modeling the topics of a major online breast cancer support forum. Breast cancer patient support forums were selected to discover the hidden, less obvious aspects of disease management and recovery. METHODS: First, manual topic categorization was performed using qualitative content analysis (QCA) of each individual forum board. Second, we requested permission from the Breastcancer.org Community for a more in-depth analysis of the postings. Topic modeling was then performed using open source software Machine Learning Language Toolkit, followed by multiple linear regression (MLR) analysis to detect highly correlated topics among the different website forums. RESULTS: QCA of the forums resulted in 20 categories of user discussion. The final topic model organized >4 million postings into 30 manageable topics. Using qualitative analysis of the topic models and statistical analysis, we grouped these 30 topics into 4 distinct clusters with similarity scores of ≥0.80; these clusters were labeled Symptoms & Diagnosis, Treatment, Financial, and Family & Friends. A clinician review confirmed the clinical significance of the topic clusters, allowing for future detection of actionable items within social media postings. To identify the most significant topics across individual forums, MLR demonstrated that 6 topics-based on the Akaike information criterion values ranging from -642.75 to -412.32-were statistically significant. CONCLUSIONS: The developed method provides an insight into the areas of interest and concern, including those not ascertainable in the clinic. Such topics included support from lay and professional caregivers and late side effects of therapy that consumers discuss in social media and may be of interest to clinicians. The developed methods and results indicate the potential of social media to inform the clinical workflow with regards to the impact of recovery on daily life.

AMIA Board White Paper: AMIA 2017 core competencies for applied health informatics education at the master's degree level.

This White Paper presents the foundational domains with examples of key aspects of competencies (knowledge, skills, and attitudes) that are intended for curriculum development and accreditation quality assessment for graduate (master's level) education in applied health informatics. Through a deliberative process, the AMIA Accreditation Committee refined the work of a task force of the Health Informatics Accreditation Council, establishing 10 foundational domains with accompanying example statements of knowledge, skills, and attitudes that are components of competencies by which graduates from applied health informatics programs can be assessed for competence at the time of graduation. The AMIA Accreditation Committee developed the domains for application across all the subdisciplines represented by AMIA, ranging from translational bioinformatics to clinical and public health informatics, spanning the spectrum from molecular to population levels of health and biomedicine. This document will be periodically updated, as part of the responsibility of the AMIA Accreditation Committee, through continued study, education, and surveys of market trends.

Toward Timely Data for Cancer Research: Assessment and Reengineering of the Cancer Reporting Process.

BACKGROUND: Cancer registries systematically collect cancer-related data to support cancer surveillance activities. However, cancer data are often unavailable for months to years after diagnosis, limiting its utility. OBJECTIVE: The objective of this study was to identify the barriers to rapid cancer reporting and identify ways to shorten the turnaround time. METHODS: Certified cancer registrars reporting to the Indiana State Department of Health cancer registry participated in a semistructured interview. Registrars were asked to describe the reporting process, estimate the duration of each step, and identify any barriers that may impact the reporting speed. Qualitative data analysis was performed with the intent of generating recommendations for workflow redesign. The existing and redesigned workflows were simulated for comparison. RESULTS: Barriers to rapid reporting included access to medical records from multiple facilities and the waiting period from diagnosis to treatment. The redesigned workflow focused on facilitating data sharing between registrars and applying a more efficient queuing technique while registrars await the delivery of treatment. The simulation results demonstrated that our recommendations to reduce the waiting period and share information could potentially improve the average reporting speed by 87 days. CONCLUSIONS: Knowing the time elapsing at each step within the reporting process helps in prioritizing the needs and estimating the impact of future interventions. Where some previous studies focused on automating some of the cancer reporting activities, we anticipate much shorter reporting by leveraging health information technologies to target this waiting period.

Review of Factors, Methods, and Outcome Definition in Designing Opioid Abuse Predictive Models.

Objective: Several opioid risk assessment tools are available to prescribers to evaluate opioid analgesic abuse among chronic patients. The objectives of this study are to 1) identify variables available in the literature to predict opioid abuse; 2) explore and compare methods (population, database, and analysis) used to develop statistical models that predict opioid abuse; and 3) understand how outcomes were defined in each statistical model predicting opioid abuse. Design: The OVID database was searched for this study. The search was limited to articles written in English and published from January 1990 to April 2016. This search generated 1,409 articles. Only seven studies and nine models met our inclusion-exclusion criteria. Results: We found nine models and identified 75 distinct variables. Three studies used administrative claims data, and four studies used electronic health record data. The majority, four out of seven articles (six out of nine models), were primarily dependent on the presence or absence of opioid abuse or dependence (ICD-9 diagnosis code) to define opioid abuse. However, two articles used a predefined list of opioid-related aberrant behaviors. Conclusions: We identified variables used to predict opioid abuse from electronic health records and administrative data. Medication variables are the recurrent variables in the articles reviewed (33 variables). Age and gender are the most consistent demographic variables in predicting opioid abuse. Overall, there is similarity in the sampling method and inclusion/exclusion criteria (age, number of prescriptions, follow-up period, and data analysis methods). Intuitive research to utilize unstructured data may increase opioid abuse models' accuracy.

Dietary intake of isoflavones and coumestrol and the risk of prostate cancer in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial.

Experimental studies have revealed that phytoestrogens may modulate the risk of certain sites of cancer due to their structural similarity to 17ß-estradiol. The present study investigates whether intake of these compounds may influence prostate cancer risk in human populations. During a median follow up of 11.5 years, 2,598 cases of prostate cancer (including 287 advanced cases) have been identified among 27,004 men in the intervention arm of the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. Dietary intake of phytoestrogens (excluding lignans) was assessed with a food frequency questionnaire. Cox proportional hazards regression analysis was performed to estimate hazard ratios (HRs) and 95% confidence intervals (CI) for dietary isoflavones and coumestrol in relation to prostate cancer risk. After adjustment for confounders, an increased risk of advanced prostate cancer [HR (95% CI) for quintile (Q) 5 vs. Q1] was found for the dietary intake of total isoflavones [1.91 (1.25-2.92)], genistein [1.51 (1.02-2.22), daidzein [1.80 (1.18-2.75) and glycitein [1.67 (1.15-2.43)] (p-trend for all associations ≤0.05). For example, HR (95% CI) for comparing the Q2, Q3, Q4 and Q5 with Q1 of daidzein intake was 1.45 (0.93-2.25), 1.65 (1.07-2.54), 1.73 (1.13-2.66) and 1.80 (1.18-2.75), respectively (p-trend: 0.013). No statistically significant associations were observed between the intake of total isoflavones and individual phytoestrogens and non-advanced and total prostate cancer after adjustment for confounders. This study revealed that dietary intake of isoflavones was associated with an elevated risk of advanced prostate cancer.

Reconciling disparate information in continuity of care documents: Piloting a system to consolidate structured clinical documents.

BACKGROUND: Due to the nature of information generation in health care, clinical documents contain duplicate and sometimes conflicting information. Recent implementation of Health Information Exchange (HIE) mechanisms in which clinical summary documents are exchanged among disparate health care organizations can proliferate duplicate and conflicting information. MATERIALS AND METHODS: To reduce information overload, a system to automatically consolidate information across multiple clinical summary documents was developed for an HIE network. The system receives any number of Continuity of Care Documents (CCDs) and outputs a single, consolidated record. To test the system, a randomly sampled corpus of 522 CCDs representing 50 unique patients was extracted from a large HIE network. The automated methods were compared to manual consolidation of information for three key sections of the CCD: problems, allergies, and medications. RESULTS: Manual consolidation of 11,631 entries was completed in approximately 150h. The same data were automatically consolidated in 3.3min. The system successfully consolidated 99.1% of problems, 87.0% of allergies, and 91.7% of medications. Almost all of the inaccuracies were caused by issues involving the use of standardized terminologies within the documents to represent individual information entries. CONCLUSION: This study represents a novel, tested tool for de-duplication and consolidation of CDA documents, which is a major step toward improving information access and the interoperability among information systems. While more work is necessary, automated systems like the one evaluated in this study will be necessary to meet the informatics needs of providers and health systems in the future.