Maintenance Note

On Friday, August 31, 2018 at 12:00 pm Eastern Time, JMIR will be completing a server migration to improve site stability and user experience. We expect to be back online Friday, August 31, 2018 at 5:00 pm Eastern Time. Should any problems arise our technical team will be using the weekend to resolve them, and users will be able to access our site by Sunday, September 2, 2018 at 1:00pm Eastern Time.

Who will be affected?


Journal Description

JMIR Medical Informatics (JMI, ISSN 2291-9694) (Editor-in-chief: Christian Lovis MD MPH FACMI) is a top-rated, tier A journal which focuses on clinical informatics, big data in health and health care, decision support for health professionals, electronic health records, ehealth infrastructures and implementation. It has a focus on applied, translational research, with a broad readership including clinicians, CIOs, engineers, industry and health informatics professionals.

Published by JMIR Publications, publisher of the Journal of Medical Internet Research (JMIR), the leading eHealth/mHealth journal (Impact Factor 2017: 4.671), JMIR Med Inform has a slightly different scope (emphasizing more on applications for clinicians and health professionals rather than consumers/citizens, which is the focus of JMIR), publishes even faster, and also allows papers which are more technical or more formative than what would be published in the Journal of Medical Internet Research.

JMIR Medical Informatics journal features a rapid and thorough peer-review process, professional copyediting, professional production of PDF, XHTML, and XML proofs (ready for deposit in PubMed Central/PubMed). The site is optimized for mobile and iPad use.

JMIR Medical Informatics adheres to the same quality standards as JMIR and all articles published here are also cross-listed in the Table of Contents of JMIR, the worlds' leading medical journal in health sciences / health services research and health informatics (


Recent Articles:

  • Three-dimensional portable document format (montage). Source: The Authors / Rawpixel; Copyright: Rawpixel; URL:; License: Licensed by the authors.

    Three-Dimensional Portable Document Format (3D PDF) in Clinical Communication and Biomedical Sciences: Systematic Review of Applications, Tools, and Protocols


    Background: The Portable Document Format (PDF) is the standard file format for the communication of biomedical information via the internet and for electronic scholarly publishing. Although PDF allows for the embedding of three-dimensional (3D) objects and although this technology has great potential for the communication of such data, it is not broadly used by the scientific community or by clinicians. Objective: The objective of this review was to provide an overview of existing publications that apply 3D PDF technology and the protocols and tools for the creation of model files and 3D PDFs for scholarly purposes to demonstrate the possibilities and the ways to use this technology. Methods: A systematic literature review was performed using PubMed and Google Scholar. Articles searched for were in English, peer-reviewed with biomedical reference, published since 2005 in a journal or presented at a conference or scientific meeting. Ineligible articles were removed after screening. The found literature was categorized into articles that (1) applied 3D PDF for visualization, (2) showed ways to use 3D PDF, and (3) provided tools or protocols for the creation of 3D PDFs or necessary models. Finally, the latter category was analyzed in detail to provide an overview of the state of the art. Results: The search retrieved a total of 902 items. Screening identified 200 in-scope publications, 13 covering the use of 3D PDF for medical purposes. Only one article described a clinical routine use case; all others were pure research articles. The disciplines that were covered beside medicine were many. In most cases, either animal or human anatomies were visualized. A method, protocol, software, library, or other tool for the creation of 3D PDFs or model files was described in 19 articles. Most of these tools required advanced programming skills and/or the installation of further software packages. Only one software application presented an all-in-one solution with a graphical user interface. Conclusions: The use of 3D PDF for visualization purposes in clinical communication and in biomedical publications is still not in common use, although both the necessary technique and suitable tools are available, and there are many arguments in favor of this technique. The potential of 3D PDF usage should be disseminated in the clinical and biomedical community. Furthermore, easy-to-use, standalone, and free-of-charge software tools for the creation of 3D PDFs should be developed.

  • Map of key stakeholders of the NERP step 1 with whom qualitative in-depth interviews were conducted (montage). Source: The Authors /; Copyright: JMIR Publications; URL:; License: Creative Commons Attribution (CC-BY).

    Implementing a National Electronic Referral Program: Qualitative Study


    Background: Electronic referrals or e-referrals can be defined as the electronic transmission of patient data and clinical requests between health service providers. National electronic referral systems have proved challenging to implement due to problems of fit between the technical systems proposed and the existing sociotechnical systems. In seeming contradiction to a sociotechnical approach, the Irish Health Service Executive initiated an incremental implementation of a National Electronic Referral Programme (NERP), with step 1 including only the technical capability for general practitioners to submit electronic referral requests to hospital outpatient departments. The technology component of the program was specified, but any changes required to embed that technology in the existing sociotechnical system were not specified. Objective: This study aimed to theoretically frame the lessons learned from the NERP step 1 on the design and implementation of a national health information technology program. Methods: A case study design was employed, using qualitative interviews with key stakeholders of the NERP step 1 (N=41). A theory-driven thematic analysis of the interview data was conducted, using Barker et al’s Framework for Going to Full Scale. Results: The NERP step 1 was broadly welcomed by key stakeholders as the first step in the implementation of electronic referrals—delivering improvements in the speed, completeness of demographic information, and legibility and traceability of referral requests. National leadership and digitalized health records in general practice were critical enabling factors. Inhibiting factors included policy uncertainty about the future organizational structures within which electronic referrals would be implemented; the need to establish a central referral office consistent with these organizational structures; outstanding interoperability issues between the electronic referral solution and hospital patient administration systems; and an anticipated need to develop specialist referral templates for some specialties. A lack of specification of the sociotechnical elements of the NERP step 1 inhibited the necessary testing and refinement of the change package used to implement the program. Conclusions: The key strengths of the NERP step 1 are patient safety benefits. The NERP was progressed beyond the pilot stage despite limited resources and outstanding interoperability issues. In addition, a new electronic health unit in Ireland (eHealth Ireland) gained credibility in delivering national health information technology programs. Limitations of the program are its poor integration in the wider policy and quality improvement agenda of the Health Service Executive. The lack of specification of the sociotechnical elements of the program created challenges in communicating the program scope to key stakeholders and restricted the ability of program managers and implementers to test and refine the change package. This study concludes that while the sociotechnical elements of a national health information technology program do not need to be specified in tandem with technical elements, they do need to be specified early in the implementation process so that the change package used to implement the program can be tested and refined.

  • Source: StockSnap; Copyright: William Stitt; URL:; License: Public Domain (CC0).

    Uncovering a Role for Electronic Personal Health Records in Reducing Disparities in Sexually Transmitted Infection Rates Among Students at a Predominantly...


    Background: Black youth continue to bear an overwhelming proportion of the United States sexually transmitted infection (STI) burden, including HIV. Several studies on web-based and mobile health (mHealth) STI interventions have focused on characterizing strategies to improve HIV-related prevention and treatment interventions, risk communication, and stigma among men who have sex with men (MSM), people who use substances, and adolescent populations. The Electronic Sexual Health Information Notification and Education (eSHINE) Study was an exploratory mixed-methods study among students at a historically black university exploring perceptions on facilitating STI testing conversations with partners using electronic personal health records (PHRs). Objective: The purpose of this paper is to use eSHINE Study results to describe perceived impacts of electronic PHRs on facilitating STI testing discussions between sexual partners. Methods: Semistructured focus groups and individual in-depth interviews were conducted on a heterogeneous sample of students (n=35) between May and July 2014. Qualitative phase findings guided development of an online survey instrument for quantitative phase data collection. Online surveys were conducted using a convenience sample of students (n=354) between January and May 2015. Online survey items collected demographic information, sexual behaviors, beliefs and practices surrounding STI testing communication between partners, and beliefs about the impact of electronic PHR access on facilitating these discussions with partners. Chi-square analysis was performed to assess gender differences across quantitative measures. A Wilcoxon signed rank sum test was used to test the null hypothesis that electronic PHRs are believed to have no effect on the timing of dyadic STI health communication. Results: Participants described multiple individual and dyadic-level factors that inhibit initiating discussions about STI testing and test results with partners. Electronic PHRs were believed to improve ability to initiate conversations and confidence in STI screening information shared by partners. Among online survey participants, men were more likely to believe electronic PHRs make it easier to facilitate STI talks with potential partners (59.9% vs 51.9%; χ2=3.93, P=.05). The Wilcoxon signed-rank test results indicate significant increases in perceived discussion timing before sex with electronic PHR access (61.0% vs 40.4%; P<.001). Conclusions: Findings suggest that electronic PHR access in STI screening settings among similar populations of Black youth may improve both motivation and personal agency for initiating dyadic STI health communication. Results from this study will likely inform novel interventions that use access to electronic PHRs to stimulate important health-related discussions between sexual partners. Moving forward requires studying strategies for implementing interventions that leverage electronic PHRs to create new sexual health communication channels with providers, peers, and family among black youth.

  • Source: Freepik; Copyright: pressfoto; URL:; License: Licensed by JMIR.

    Task-Data Taxonomy for Health Data Visualizations: Web-Based Survey With Experts and Older Adults


    Background: Increasingly, eHealth involves health data visualizations to enable users to better understand their health situation. Selecting efficient and ergonomic visualizations requires knowledge about the task that the user wants to carry out and the type of data to be displayed. Taxonomies of abstract tasks and data types bundle this knowledge in a general manner. Task-data taxonomies exist for visualization tasks and data. They also exist for eHealth tasks. However, there is currently no joint task taxonomy available for health data visualizations incorporating the perspective of the prospective users. One of the most prominent prospective user groups of eHealth are older adults, but their perspective is rarely considered when constructing tasks lists. Objective: The aim of this study was to construct a task-data taxonomy for health data visualizations based on the opinion of older adults as prospective users of eHealth systems. eHealth experts served as a control group against the bias of lacking background knowledge. The resulting taxonomy would then be used as an orientation in system requirement analysis and empirical evaluation and to facilitate a common understanding and language in eHealth data visualization. Methods: Answers from 98 participants (51 older adults and 47 eHealth experts) given in an online survey were quantitatively analyzed, compared between groups, and synthesized into a task-data taxonomy for health data visualizations. Results: Consultation, diagnosis, mentoring, and monitoring were confirmed as relevant abstract tasks in eHealth. Experts and older adults disagreed on the importance of mentoring (χ24=14.1, P=.002) and monitoring (χ24=22.1, P<.001). The answers to the open questions validated the findings from the closed questions and added therapy, communication, cooperation, and quality management to the aforementioned tasks. Here, group differences in normalized code counts were identified for “monitoring” between the expert group (mean 0.18, SD 0.23) and the group of older adults (mean 0.08, SD 0.15; t96=2431, P=.02). Time-dependent data was most relevant across all eHealth tasks. Finally, visualization tasks and data types were assigned to eHealth tasks by both experimental groups. Conclusions: We empirically developed a task-data taxonomy for health data visualizations with prospective users. This provides a general framework for theoretical concession and for the prioritization of user-centered system design and evaluation. At the same time, the functionality dimension of the taxonomy for telemedicine—chosen as the basis for the construction of present taxonomy—was confirmed.

  • Source: Flickr; Copyright: US Department of Agriculture; URL:; License: Creative Commons Attribution (CC-BY).

    Adverse Drug Event Reporting From Clinical Care: Mixed-Methods Analysis for a Minimum Required Dataset


    Background: Patients commonly transition between health care settings, requiring care providers to transfer medication utilization information. Yet, information sharing about adverse drug events (ADEs) remains nonstandardized. Objective: The objective of our study was to describe a minimum required dataset for clinicians to document and communicate ADEs to support clinical decision making and improve patient safety. Methods: We used mixed-methods analysis to design a minimum required dataset for ADE documentation and communication. First, we completed a systematic review of the existing ADE reporting systems. After synthesizing reporting concepts and data fields, we conducted fieldwork to inform the design of a preliminary reporting form. We presented this information to clinician end-user groups to establish a recommended dataset. Finally, we pilot-tested and refined the dataset in a paper-based format. Results: We evaluated a total of 1782 unique data fields identified in our systematic review that describe the reporter, patient, ADE, and suspect and concomitant drugs. Of these, clinicians requested that 26 data fields be integrated into the dataset. Avoiding the need to report information already available electronically, reliance on prospective rather than retrospective causality assessments, and omitting fields deemed irrelevant to clinical care were key considerations. Conclusions: By attending to the information needs of clinicians, we developed a standardized dataset for adverse drug event reporting. This dataset can be used to support communication between care providers and integrated into electronic systems to improve patient safety. If anonymized, these standardized data may be used for enhanced pharmacovigilance and research activities.

  • Source: Wikimedia Commons; Copyright: Kgbo; URL:; License: Creative Commons Attribution + ShareAlike (CC-BY-SA).

    Nurses’ Experience With Health Information Technology: Longitudinal Qualitative Study


    Background: Nurses are the largest group of health information technology (HIT) users. As such, nurses’ adaptations are critical for HIT implementation success. However, longitudinal approaches to understanding nurses’ perceptions of HIT remain underexplored. Previous studies of nurses’ perceptions demonstrate that the progress and timing for acceptance of and adaptation to HIT varies. Objective: This study aimed to explore nurses’ experience regarding implementation of HIT over time. Methods: A phenomenological approach was used for this longitudinal qualitative study to explore nurses’ perceptions of HIT implementation over time, focusing on three time points (rounds) at 3, 9, and 18 months after implementation of electronic health records and bar code medication administration. The purposive sample was comprised of clinical nurses who worked on a medical-surgical unit in an academic center. Results: Major findings were categorized into 7 main themes with 54 subthemes. Nurses reported personal-level and organizational-level factors that facilitated HIT adaptation. We also generated network graphs to illustrate the occurrence of themes. Thematic interconnectivity differed due to nurses’ concerns and satisfaction at different time points. Equipment and workflow were the most frequent themes across all three rounds. Nurses were the most dissatisfied approximately 9 months after HIT implementation. Eighteen months after HIT implementation, nurses’ perceptions appeared more balanced. Conclusions: It is recommended that organizations invest in equipment (ie, wireless barcode scanners), refine policies to reflect nursing practice, and improve systems to focus on patient safety. Future research is necessary to confirm patterns of nurses’ adaptation to HIT in other samples.

  • Preterm infant in the NICU. Source: Flickr; Copyright: The Hudson Family; URL:; License: Creative Commons Attribution (CC-BY).

    The Impact of Implementation of a Clinically Integrated Problem-Based Neonatal Electronic Health Record on Documentation Metrics, Provider Satisfaction, and...


    Background: A goal of effective electronic health record provider documentation platforms is to provide an efficient, concise, and comprehensive notation system that will effectively reflect the clinical course, including the diagnoses, treatments, and interventions. Objective: The aim is to fully redesign and standardize the provider documentation process, seeking improvement in documentation based on ongoing All Patient Refined Diagnosis Related Group–based coding records, while maintaining noninferiority comparing provider satisfaction to our existing documentation process. We estimated the fiscal impact of improved documentation based on changes in expected hospital payments. Methods: Employing a multidisciplinary collaborative approach, we created an integrated clinical platform that captures data entry from the obstetrical suite, delivery room, neonatal intensive care unit (NICU) nursing and respiratory therapy staff. It provided the sole source for hospital provider documentation in the form of a history and physical exam, daily progress notes, and discharge summary. Health maintenance information, follow-up appointments, and running contemporaneous updated hospital course information have selected shared entry and common viewing by the NICU team. The interventions were to (1) improve provider awareness of appropriate documentation through a provider education handout and follow-up group discussion and (2) fully redesign and standardize the provider documentation process building from the native Epic-based software. The measures were (1) hospital coding department review of all NICU admissions and 3M All Patient Refined Diagnosis Related Group–based calculations of severity of illness, risk of mortality, and case mix index scores; (2) balancing measure: provider time utilization case study and survey; and (3) average expected hospital payment based on acuity-based clinical logic algorithm and payer mix. Results: We compared preintervention (October 2015-October 2016) to postintervention (November 2016-May 2017) time periods and saw: (1) significant improvement in All Patient Refined Diagnosis Related Group–derived severity of illness, risk of mortality, and case mix index (monthly average severity of illness scores increased by 11.1%, P=.008; monthly average risk of mortality scores increased by 13.5%, P=.007; and monthly average case mix index scores increased by 7.7%, P=.009); (2) time study showed increased time to complete history and physical and progress notes and decreased time to complete discharge summary (history and physical exam: time allocation increased by 47%, P=.05; progress note: time allocation increased by 91%, P<.001; discharge summary: time allocation decreased by 41%, P=.03); (3) survey of all providers: overall there was positive provider perception of the new documentation process based on a survey of the provider group; (4) significantly increased hospital average expected payments: comparing the preintervention and postintervention study periods, there was a US $14,020 per month per patient increase in average expected payment for hospital charges (P<.001). There was no difference in payer mix during this time period. Conclusions: A problem-based NICU documentation electronic health record more effectively improves documentation without dissatisfaction by the participating providers and improves hospital estimations of All Patient Refined Diagnosis Related Group–based revenue.

  • Source: Rawpixel; Copyright: Rawpixel; URL:; License: Public Domain (CC0).

    Validation of a Natural Language Processing Algorithm for Detecting Infectious Disease Symptoms in Primary Care Electronic Medical Records in Singapore


    Background: Free-text clinical records provide a source of information that complements traditional disease surveillance. To electronically harness these records, they need to be transformed into codified fields by natural language processing algorithms. Objective: The aim of this study was to develop, train, and validate Clinical History Extractor for Syndromic Surveillance (CHESS), an natural language processing algorithm to extract clinical information from free-text primary care records. Methods: CHESS is a keyword-based natural language processing algorithm to extract 48 signs and symptoms suggesting respiratory infections, gastrointestinal infections, constitutional, as well as other signs and symptoms potentially associated with infectious diseases. The algorithm also captured the assertion status (affirmed, negated, or suspected) and symptom duration. Electronic medical records from the National Healthcare Group Polyclinics, a major public sector primary care provider in Singapore, were randomly extracted and manually reviewed by 2 human reviewers, with a third reviewer as the adjudicator. The algorithm was evaluated based on 1680 notes against the human-coded result as the reference standard, with half of the data used for training and the other half for validation. Results: The symptoms most commonly present within the 1680 clinical records at the episode level were those typically present in respiratory infections such as cough (744/7703, 9.66%), sore throat (591/7703, 7.67%), rhinorrhea (552/7703, 7.17%), and fever (928/7703, 12.04%). At the episode level, CHESS had an overall performance of 96.7% precision and 97.6% recall on the training dataset and 96.0% precision and 93.1% recall on the validation dataset. Symptoms suggesting respiratory and gastrointestinal infections were all detected with more than 90% precision and recall. CHESS correctly assigned the assertion status in 97.3%, 97.9%, and 89.8% of affirmed, negated, and suspected signs and symptoms, respectively (97.6% overall accuracy). Symptom episode duration was correctly identified in 81.2% of records with known duration status. Conclusions: We have developed an natural language processing algorithm dubbed CHESS that achieves good performance in extracting signs and symptoms from primary care free-text clinical records. In addition to the presence of symptoms, our algorithm can also accurately distinguish affirmed, negated, and suspected assertion statuses and extract symptom durations.

  • Source: Freepik; Copyright: peoplecreations; URL:; License: Licensed by JMIR.

    Health Information Technology in Healthcare Quality and Patient Safety: Literature Review


    Background: The area of healthcare quality and patient safety is starting to use health information technology to prevent reportable events, identify them before they become issues, and act on events that are thought to be unavoidable. As healthcare organizations begin to explore the use of health information technology in this realm, it is often unclear where fiscal and human efforts should be focused. Objective: The purpose of this study was to provide a foundation for understanding where to focus health information technology fiscal and human resources as well as expectations for the use of health information technology in healthcare quality and patient safety. Methods: A literature review was conducted to identify peer-reviewed publications reporting on the actual use of health information technology in healthcare quality and patient safety. Inductive thematic analysis with open coding was used to categorize a total of 41 studies. Three pre-set categories were used: prevention, identification, and action. Three additional categories were formed through coding: challenges, outcomes, and location. Results: This study identifies five main categories across seven study settings. A majority of the studies used health IT for identification and prevention of healthcare quality and patient safety issues. In this realm, alerts, clinical decision support, and customized health IT solutions were most often implemented. Implementation, interface design, and culture were most often noted as challenges. Conclusions: This study provides valuable information as organizations determine where they stand to get the most “bang for their buck” relative to health IT for quality and patient safety. Knowing what implementations are being effectivity used by other organizations helps with fiscal and human resource planning as well as managing expectations relative to cost, scope, and outcomes. The findings from this scan of the literature suggest that having organizational champion leaders that can shepherd implementation, impact culture, and bridge knowledge with developers would be a valuable resource allocation to consider.

  • Source: Flickr; Copyright: aaron gilson; URL:; License: Creative Commons Attribution + NoDerivatives (CC-BY-ND).

    Development and Validation of a Functional Behavioural Assessment Ontology to Support Behavioural Health Interventions


    Background: In the cognitive-behavioral approach, Functional Behavioural Assessment is one of the most effective methods to identify the variables that determine a problem behavior. In this context, the use of modern technologies can encourage the collection and sharing of behavioral patterns, effective intervention strategies, and statistical evidence about antecedents and consequences of clusters of problem behaviors, encouraging the designing of function-based interventions. Objective: The paper describes the development and validation process used to design a specific Functional Behavioural Assessment Ontology (FBA-Ontology). The FBA-Ontology is a semantic representation of the variables that intervene in a behavioral observation process, facilitating the systematic collection of behavioral data, the consequential planning of treatment strategies and, indirectly, the scientific advancement in this field of study. Methods: The ontology has been developed deducing concepts and relationships of the ontology from a gold standard and then performing a machine-based validation and a human-based assessment to validate the Functional Behavioural Assessment Ontology. These validation and verification processes were aimed to verify how much the ontology is conceptually well founded and semantically and syntactically correct. Results: The Pellet reasoner checked the logical consistency and the integrity of classes and properties defined in the ontology, not detecting any violation of constraints in the ontology definition. To assess whether the ontology definition is coherent with the knowledge domain, human evaluation of the ontology was performed asking 84 people to fill in a questionnaire composed by 13 questions assessing concepts, relations between concepts, and concepts’ attributes. The response rate for the survey was 29/84 (34.52%). The domain experts confirmed that the concepts, the attributes, and the relationships between concepts defined in the FBA-Ontology are valid and well represent the Functional Behavioural Assessment process. Conclusions: The new ontology developed could be a useful tool to design new evidence-based systems in the Behavioral Interventions practices, encouraging the link with other Linked Open Data datasets and repositories to provide users with new models of eHealth focused on the management of problem behaviors. Therefore, new research is needed to develop and implement innovative strategies to improve the poor reproducibility and translatability of basic research findings in the field of behavioral assessment.

  • Source: Flickr; Copyright: Tunstall; URL:; License: Creative Commons Attribution (CC-BY).

    Perspectives of Nurses Toward Telehealth Efficacy and Quality of Health Care: Pilot Study


    Background: Telehealth nursing, or the delivery, management, and coordination of nursing care services provided via telecommunications technology, is one of the methods of delivering health care to patients in the United States. It is important to assess the service quality of the involved health professionals as well as the telehealth nursing process. The focus of this study is the innovative model of telehealth care delivery by nurses for managing patients with chronic disease while they are living in their own residence. Objective: The primary objective of this pilot study was to examine whether telehealth technology impacts the perceived level of internal service quality delivered by nurses within a telehealth organization. To address this research goal, the notion of telehealth nursing service quality (TNSQ) is empirically tested and validated with a survey instrument. Methods: Data were collected from nurses belonging to a home care agency based on interview questions inquiring about facilitators and inhibitors to TNSQ. A survey to measure TNSQ based on the SERVQUAL instrument was completed by adjusting descriptions of the original instrument to suit the context. Follow-up interviews were conducted to validate questions on the revised instrument. Results: The findings of this survey research were positive, based on mean differences between expectations and perceptions of TNSQ. This indicates satisfaction with TNSQ and shows that the quality of the service is higher than what the respondents expect. The Wilcoxon signed-rank test using the P value for the test, which is .35, did not show a statistically significant change between the median differences of perception and expectation. The total number of respondents was 13. Results indicate that overall perceived service quality is a positive value (0.05332). This means the perceptions of the level of service are slightly higher than what they expect, indicating there is satisfaction with TNSQ. Conclusions: The responses to the interview questions and data gathered from the survey showed overall satisfaction with TNSQ. The SERVQUAL instrument was a good framework to assess TNSQ. In a nutshell, the study highlighted how the telehealth process provides daily monitoring of patient health, leading to the benefits of immediate feedback for patients, family, and caregivers as well as convenience of scheduling.

  • Source:; Copyright: Anoto AB; URL:; License: Creative Commons Attribution + Noncommercial + NoDerivatives (CC-BY-NC-ND).

    Privacy-Preserving Predictive Modeling: Harmonization of Contextual Embeddings From Different Sources


    Background: Data sharing has been a big challenge in biomedical informatics because of privacy concerns. Contextual embedding models have demonstrated a very strong representative capability to describe medical concepts (and their context), and they have shown promise as an alternative way to support deep-learning applications without the need to disclose original data. However, contextual embedding models acquired from individual hospitals cannot be directly combined because their embedding spaces are different, and naive pooling renders combined embeddings useless. Objective: The aim of this study was to present a novel approach to address these issues and to promote sharing representation without sharing data. Without sacrificing privacy, we also aimed to build a global model from representations learned from local private data and synchronize information from multiple sources. Methods: We propose a methodology that harmonizes different local contextual embeddings into a global model. We used Word2Vec to generate contextual embeddings from each source and Procrustes to fuse different vector models into one common space by using a list of corresponding pairs as anchor points. We performed prediction analysis with harmonized embeddings. Results: We used sequential medical events extracted from the Medical Information Mart for Intensive Care III database to evaluate the proposed methodology in predicting the next likely diagnosis of a new patient using either structured data or unstructured data. Under different experimental scenarios, we confirmed that the global model built from harmonized local models achieves a more accurate prediction than local models and global models built from naive pooling. Conclusions: Such aggregation of local models using our unique harmonization can serve as the proxy for a global model, combining information from a wide range of institutions and information sources. It allows information unique to a certain hospital to become available to other sites, increasing the fluidity of information flow in health care.

Citing this Article

Right click to copy or hit: ctrl+c (cmd+c on mac)

Latest Submissions Open for Peer-Review:

View All Open Peer Review Articles
  • Patient's Satisfaction with Dental Service Provided by Dental Students of Qassim University, KSA.

    Date Submitted: Aug 13, 2018

    Open Peer Review Period: Aug 15, 2018 - Oct 10, 2018

    Background: In the past, dental patient satisfaction has been measured by the dentist's technical competence and mechanical precision2. Patient's opinion played no part in this method of measuring qua...

    Background: In the past, dental patient satisfaction has been measured by the dentist's technical competence and mechanical precision2. Patient's opinion played no part in this method of measuring quality. Consequently, consideration of patient's satisfaction became an integral part of providing dental services3. Some patients want a dentist who listens to them, has a friendly caring attitude, discuses treatment options and procedures, and gives confidence7. Communication skills, a component of empathy, have been shown to be important in limiting patient's dissatisfaction12. The self-rated oral health status might be considered as the key factor of patient’s satisfaction and positive feedback. Objective: To achieve high levels of patient's satisfaction, it is crucial to find out which items patients consider important and how they operationalize these items 17. Finally, medical care cannot be of high quality, unless patients’ satisfaction is observed. Methods: A 22-item patient's satisfaction questionnaire was designed. The questionnaire was used to obtain data from patients who were then currently receiving care in the College of Dentistry, Qassim University, KSA after minimum of three visits to the dental chair. Two hundred and fifty copies of the questionnaire were distributed equally in both male and female sections over four months period (March-August 2017). Results: A total of 250 questionnaire papers were distributed of which 215 were returned for a response rate of 86%. The response rate in male section was (94%) higher than that in female section (78%) at College of Dentistry, Qassim University, KSA. Overall ranking of factors related to satisfaction was done by calculating the mean percentage of agreement regarding the different disciplines of satisfaction among the studied group. The mean percentage of agreement for the 3 disciplines was 82.5% denoting a high level of satisfaction. Conclusions: Results indicated that the majority of patients were satisfied with the provided dental service at College of Dentistry, Qassim University, KSA. Patient dentist interaction and Technical competency had the most significant effect on patient satisfaction levels in this study . On the other hand, Administrative efficiency (AE) and Clinical setup environment (CS) were the factor that had the most impact on levels of dissatisfaction. In this study, no significant differences were observed between the satisfaction score and Nationality of the patients. The only significant finding observed for the satisfaction score was gender .While there was a significant differences regarding satisfaction between both genders . Clinical Trial: The Research Ethics Committee and Research & Presentation Skills Supervision Committee at College of Dentistry, Qassim University, KSA approved this study (Code #:EA/203/2017) and (SRPSSC #: M-5003-17) . The waiver of the informed consent process was approved on the basis of the questionnaires being anonymous and self-administered and containing no identifiers.

  • Wearable Smart Blanket System Model for Monitoring the Vital Signs of Patients in Ambulance

    Date Submitted: Aug 4, 2018

    Open Peer Review Period: Aug 13, 2018 - Oct 8, 2018

    Abstract Background and objective: The timely and managed intervention reduces the consequences of disease and sudden death among the patients in emergency conditions. Monitoring the patients in emer...

    Abstract Background and objective: The timely and managed intervention reduces the consequences of disease and sudden death among the patients in emergency conditions. Monitoring the patients in emergency conditions requires rapid and appropriate decisions to save their lives. The present study aimed at modeling the wearable smart blanket system for monitoring the patients in the emergency conditions of ambulance. Method: The present study was based on an applied and descriptive-developmental design. Firstly, the requirements and features of wearable smart blanket system were elicited and secondly a smart blanket system was modeled by using the UML charts and elicited requirements. Finally, the designed architecture was evaluated using ARID scenario-based method. Findings: The functional requirements of wearable smart blanket system with its data elements and physical-structural features of this system as well as non-functional requirements were elicited. Based on the requirements and data elements elicited from the questionnaire, class diagram, activity, use-case diagram, sequence, deployment, and component were drawn. Then, the ARID scenario-based evaluation method was used to indicate that the designed architecture can provide the expected scenarios from the proposed system by using the UML and the relationships between components, systems, and users from the structural and behavioral perspectives. Conclusion: Wearable smart blanket system collects the data related to medical signals by the sensors installed on the blanket and such data are processed by the smart system. Therefore, it can be concluded that the design of this system makes it possible to monitor and control patients in risky conditions with better quality and to integrate vital signs. The analyzing biological data makes it easy for doctors to take early diagnosis and interventions.

  • Transcription Errors of Blood Glucose Values and Insulin Errors in an Intensive Care Unit: Toward Electronic Medical Record-Glucometer Interoperability

    Date Submitted: Aug 8, 2018

    Open Peer Review Period: Aug 9, 2018 - Oct 4, 2018

    Background: Critically ill patients require constant point-of-care-testing (POCT) for blood glucose using glucometers to guide initiation and titration decisions of continuous insulin infusion. Transc...

    Background: Critically ill patients require constant point-of-care-testing (POCT) for blood glucose using glucometers to guide initiation and titration decisions of continuous insulin infusion. Transcribing blood glucose tests’ values from the glucometers into a paper log and the electronic medical record (EMR) is a very common yet error-prone practice in intensive care units (ICUs) given the lack of connectivity between glucometers and the EMR in many U.S. hospitals. Objective: This study examined (1) transcription errors of blood glucose values obtained by a glucometer and documented in the paper log and in the EMR vital signs flowsheet in a surgical trauma ICU, (2) insulin errors resulted from transcription errors of blood glucose values, (3) lack of documenting blood glucose values in the paper log and the EMR vital signs flowsheet, and (4) the average time for docking the glucometer. Methods: This secondary data analysis examined 5049 point of care blood glucose tests for transcription errors, insulin errors, lack of documenting blood glucose values in the paper log and the EMR, and the glucometer docking time in a surgical trauma ICU. Transcription errors include errors in the paper log and errors in the EMR vital signs flowsheet. The results of the 5049 blood glucose tests were obtained from RALS®-Plus V1.5.1, a bi-directional interface software for in-hospital glucometers that uploads meters data after docking the meters into the EMR lab flowsheet. Chart audit was used to obtain the transcribed values of blood glucose tests from the EMR vital signs flowsheet and the patient demographic and clinical-related information. The paper log was accessed to obtain the corresponding transcribed values of the blood glucose tests. Three nurse educators obtained the data. Results: The 5049 blood glucose tests were pertinent to 234 patients. There were 98 transcription errors in the paper log (2%, out of 4441 transcribed tests in the paper logs) that resulted in 8 insulin errors. Out of the 2985 transcribed values in the EMR vital signs flowsheet, there were 242 (8%) errors related to 63 patients (27% of the patients) and resulted in 24 insulin errors. The total number of transcription errors in both (paper log and vital signs flowsheet) was 43 (2%, out of 2138). Overall, there were 68 (29% out of 234 patients) unique patients involved in all types of errors and 30 unique insulin errors that affected 25 patients (11% of the patients). In the vital signs flowsheet, 41% of the tests (out of 5049 tests) were not transcribed. The average time from the POCT to the time meters were docked (readings were uploaded into the EMR lab flowsheet) was 8 hours with a median of 5.5 hours. Some of the readings took 56 hours (2.3 days) to be uploaded into the EMR lab flowsheet. Conclusions: Transcription errors of blood glucose values obtained by glucometers do exist and result in insulin errors. Given the high dependence on glucometers for POCT of blood glucose in ICUs, full EMR-glucometer interoperability is required for complete and accurate documentation of blood glucose values and elimination of transcription errors and the subsequent insulin-related errors in ICUs.

  • Tracking development with Baby CROINC: The interplay of parent experience, usage, and technology approach

    Date Submitted: Aug 7, 2018

    Open Peer Review Period: Aug 9, 2018 - Oct 4, 2018

    Background: To lower barriers to developmental screening, we designed Baby CROINC (CROwd INtelligence Curation), a digital platform to help parents track and assess their children’s development thro...

    Background: To lower barriers to developmental screening, we designed Baby CROINC (CROwd INtelligence Curation), a digital platform to help parents track and assess their children’s development through crowd wisdom. Objective: To understand users’ experiences using Baby CROINC in relation to users’ technological competence and attitudes, while considering the influence of their children’s presented developmental evaluations and parents’ actual use of the system. Methods: Mothers of 260 children (M age= 17.6 months, SD=13.7) used Baby CROINC for two weeks. They entered developmental milestones on their children’s developmental diary timeline and received statistical developmental percentile reports. Mothers then completed Usability and Technology Profile Questionnaires. Results: Mothers’ experiences of the Baby CROINC system usability were associated with their attitudes toward solving technological problems, mediated by frequency of engagement in Internet activities. Mothers with a proactive approach toward solving technology problems, engage in a wide range of Internet activities, and/or view the Internet as integral to their lives had a better experience with Baby CROINC than mothers who did not. The system’s perceived usability was not associated with the crowd-based child developmental percentiles or quantity of mothers’ usage of the system. Conclusions: Parent’s user experiences correlate with their technology competence and problem solving attitude but is not correlated with their child’s developmental status. Developmental screening platforms need to solve the tension between requiring active engagement and encouraging proactive parenting.

  • Health care provider utilization of the VA transgender e-consultation program

    Date Submitted: Jul 25, 2018

    Open Peer Review Period: Jul 28, 2018 - Sep 22, 2018

    Background: In 2015, the Department of Veterans Affairs (VA) nationally implemented a transgender e-consultation (e-consult) program with expert clinical guidance for providers. This mixed-methods pro...

    Background: In 2015, the Department of Veterans Affairs (VA) nationally implemented a transgender e-consultation (e-consult) program with expert clinical guidance for providers. This mixed-methods project aimed to describe providers’ program experiences, reasons for not using the program, and how to improve program use. Objective: This mixed-methods project aimed to describe providers’ program experiences, reasons for not using the program, and how to improve program use. Methods: From January to May 2017, 15 urban and rural VA providers who submitted at least one e-consult in the last year participated in semi-structured interviews about their program experiences that were analyzed using content analysis. From November to December 2017, 53 providers with transgender patient encounters but did not utilize the program participated in a brief online survey of reasons for program non-use and facilitators encouraging use. Results: Qualitative analysis found providers learned of the program through email, colleagues, the Electronic Health Record (EHR) system, and participation in VA LGBT committees/educational trainings. Providers used the program for establishing a care plan, hormone therapy recommendations, sexual and reproductive health education, surgical treatment education, patient-provider communication guidance, and second opinions. Facilitators of use included understandable recommendations, ease of use through the EHR system, and being the only transgender resource for rural providers. Barriers to use included time constraints, communication-related problems with e-consult, impractical recommendations for under-resourced sites, and misunderstanding e-consult purpose. Improvement suggestions included adding concise/sectioned responses, expanding provider/patient program awareness, designating a follow-up contact person, and increasing provider education about transgender Veterans and related care. Quantitative analysis found common reasons for not using the program were: not knowing about it, not needing to use it yet, and getting help from a colleague outside of e-consult. Common suggestions to improve program use were: more information about where to find e-consult in the chart, guidance on talking with patients about the program, and e-mail announcements to improve provider awareness of the program. Post hoc exploratory analyses showed no differences between urban and rural providers. Conclusions: The VA transgender e-consult program is useful to providers, but there are several barriers to implementing recommendations, some of which are especially challenging for rural providers. Addressing noted barriers and enhancing the facilitators may improve program use and improve quality care for transgender Veterans.

  • Understanding of and Barriers to Electronic Health Records in a Culturally Diverse Pediatric Population

    Date Submitted: Jul 17, 2018

    Open Peer Review Period: Jul 21, 2018 - Sep 15, 2018

    Background: Electronic health records (EHR) have become standard in the healthcare setting. In an effort to improve health literacy, foster patient- provider communication, and ease the transition fro...

    Background: Electronic health records (EHR) have become standard in the healthcare setting. In an effort to improve health literacy, foster patient- provider communication, and ease the transition from adolescent to adult care, our institution created a policy that allows patients aged 13-17 to have EHR access. Literature review revealed predictable differences in portal registration between different ethnicities and socioeconomic statuses. Consequently a cross-sectional survey was developed to investigate barriers to EHR access in a sample of culturally diverse adolescents. Objective: Assess for barriers to EHR access in a culturally diverse adolescent population. Methods: A 42-item anonymous survey was completed by 97 adolescents age 13-18 attending general pediatrics clinics. Results were analyzed using descriptive statistics and t-tests. Results: Average participant age was 15.5 + 1.5 years with 59.8% male and 40.2% female. Participants were 44.3% Black, 41.2% Hispanic, 9.3% Caucasian, 3.1% Asian, and 2.1% other. There were statistically significant differences in perceptions of confidentiality in age (13-15 vs. 16-18; P=0.001) and insurance status (government vs. private, P=0.012), but not gender, ethnicity or parental education level. Older, privately insured girls had more confidence in the confidentiality of their relationship with their doctors. 93.8% of participants had heard the term "EHR," but only 54.6% were familiar with its function. 77.3% of patients primarily accessed the internet through phones and 50% of participants knew that patients under 18 could obtain care for mental health, substance abuse, sexual health and pregnancy. Conclusions: Research has identified gaps in EHR technology with regards to the pediatric patient population. Results of our survey show adolescents may have misconceptions regarding the doctor-patient relationship, their ability to obtain care, and the modalities present in an Electronic Health Record. As technology progresses, it is essential to have a deeper understanding of adolescents' perceptions of confidentiality, technology, and available resources in order to design an EHR system that encourages patient education and communication while limiting barriers to care.