Physician Experiences with Perceived Pressure to Order Diagnostic Imaging Services

The overuse of diagnostic imaging (DI) services, which is estimated to be 30% in Canada, can expose patients to unnecessary radiation, and strain human and financial resources. This study explored the DI ordering practices of physicians in Canada through semi-structured interviews to gain a deeper understanding of the factors contributing to the overuse of DI services. The majority of participants (n=11; 91%) described feeling pressured by patients to order DI services in circumstances that were unwarranted. The results are followed by a discussion about ways technology (such as a decision support system) could aid in facilitating a dialogue between physicians and patients about when and when not to order DI.

A methodology for mining clinical data: experiences from TRANSFoRm project

Data mining of electronic health records (eHRs) allows us to identify patterns of patient data that characterize diseases and their progress and learn best practices for treatment and diagnosis. Clinical Prediction Rules (CPRs) are a form of clinical evidence that quantifies the contribution of different clinical data to a particular clinical outcome and help clinicians to decide the diagnosis, prognosis or therapeutic conduct for any given patient. The TRANSFoRm diagnostic support system (DSS) is based on the construction of an ontological repository of CPRs for diagnosis prediction in which clinical evidence is expressed using a unified vocabulary. This paper explains the proposed methodology for constructing this CPR repository, addressing algorithms and quality measures for filtering relevant rules. Some preliminary application results are also presented.

Multidisciplinary Modelling of Symptoms and Signs with Archetypes and SNOMED-CT for Clinical Decision Support

Clinical Decision Support Systems (CDSS) help to improve health care and reduce costs. However, the lack of knowledge management and modelling hampers their maintenance and reuse. Current EHR standards and terminologies can allow the semantic representation of the data and knowledge of CDSS systems boosting their interoperability, reuse and maintenance. This paper presents the modelling process of respiratory conditions' symptoms and signs by a multidisciplinary team of clinicians and information architects with the help of openEHR, SNOMED and clinical information modelling tools for a CDSS. The information model of the CDSS was defined by means of an archetype and the knowledge model was implemented by means of an SNOMED-CT based ontology.

Development and preliminary validation of a dynamic, patient-tailored method to detect abnormal laboratory test results

Laboratory test results in primary care are flagged as 'abnormal' when they fall outside a population-based Reference Interval (RI), typically generating many alerts with a low specificity. In order to decrease alert frequency while retaining clinical relevance, we developed a method to assess dynamic, patient-tailored RIs based on mixed-effects linear regression models. Potassium test results from primary care were used as proof-of-concept test bed. Clinical relevance was assessed via a survey administered to general practitioners (GPs). Overall, the dynamic, patient-tailored method and the combination of both methods flagged 20% and 36% fewer values as abnormal than the population-based method. Nineteen out of 43 invited GPs (44%) completed the survey. The population-based method yielded a better sensitivity than the patient-tailored and the combined methods (0.51 vs 0.41 and 0.38, respectively) but a lower PPV (0.66 vs 0.67 and 0.76, respectively). We conclude that a combination of population-based and patient-tailored RIs can improve the detection of abnormal laboratory results. We suggest that lab values outside both RIs be flagged with high priority in clinical practice.

Ontology-Driven Semantic Search for Brazilian Portuguese Clinical Notes

The emerging penetration of Health IT in Latin America (especially in Brazil) has exacerbated the ever-increasing amount of Electronic Health Record (EHR) clinical free text documents.This imposes a workflow efficiency challenge on clinicians who need to synthesize such documents during the typically time-constrained patient care. We propose an ontology-driven semantic search framework that effectively supports clinicians' information synthesis at the point of care.

Follow-up Recommendation Detection on Radiology Reports with Incidental Pulmonary Nodules

The management of follow-up recommendations is fundamental for the appropriate care of patients with incidental pulmonary findings. The lack of communication of these important findings can result in important actionable information being lost in healthcare provider electronic documents. This study aims to analyze follow-up recommendations in radiology reports containing pulmonary incidental findings by using Natural Language Processing and Regular Expressions. Our evaluation highlights the different follow-up recommendation rates for oncology and non-oncology patient cohorts. The results reveal the need for a context-sensitive approach to tracking different patient cohorts in an enterprise-wide assessment.

Enhancing Patient Safety Event Reporting by K-nearest Neighbor Classifier

Data quality was placed as a major reason for the low utility of patient safety event reporting systems. A pressing need in improving data quality has advanced recent research focus in data entry associated with human factors. The debate on structured data entry or unstructured data entry reveals not only a trade-off problem among data accuracy, completeness, and timeliness, but also a technical gap on text mining. The present study suggested a text classification method, k-nearest neighbor (KNN), for predicting subject categories as in our proposed reporting system. Our results demonstrated the feasibility of KNN classifier used for text classification and indicated the advantage of such an application to raise data quality and clinical decision support in reporting patient safety events.

Clinical data warehousing for evidence based decision making

Large volumes of heterogeneous health data silos pose a big challenge when exploring for information to allow for evidence based decision making and ensuring quality outcomes. In this paper, we present a proof of concept for adopting data warehousing technology to aggregate and analyse disparate health data in order to understand the impact various lifestyle factors on obesity. We present a practical model for data warehousing with detailed explanation which can be adopted similarly for studying various other health issues.

Detecting Learning and Reasoning Patterns in a CDSS for Dementia Investigation

Reasoning conducted in clinical practice is manifested through different and often combined reasoning and learning strategies, adjusted to the characteristics of the available information, the medical professional's experience and skills, and the available tools, such as clinical practice guidelines. This research outlines a design model for supporting the commonly used strategies. This design model was implemented into a clinical decision-support system (CDSS), in addition to a method for detecting reasoning strategies applied when using the CDSS. This method was applied in a case study, with preliminary results presented in this paper and will be further implemented in future studies.

An Intelligent Ecosystem for Providing Support in Prehospital Trauma Care in Cuenca, Ecuador

According to facts given by the World Health Organization, one in ten deaths worldwide is due to an external cause of injury. In the field of pre-hospital trauma care, adequate and timely treatment in the golden period can impact the survival of a patient. The aim of this paper is to show the design of a complete ecosystem proposed to support the evaluation and treatment of trauma victims, using standard tools and vocabulary such as OpenEHR, as well as mobile systems and expert systems to support decision-making. Preliminary results of the developed applications are presented, as well as trauma-related data from the city of Cuenca, Ecuador.

Extraction Of Adverse Events From Clinical Documents To Support Decision Making Using Semantic Preprocessing

Clinical documentation is usually stored in unstructured format in electronic health records (EHR). Processing the information is inconvenient and time consuming and should be enhanced by computer systems. In this paper, a rule-based method is introduced that identifies adverse events documented in the EHR that occurred during treatment. For this purpose, clinical documents are transformed into a semantic structure from which adverse events are extracted. The method is evaluated in a user study with neurosurgeons. In comparison to a bag of word classification using support vector machines, our approach achieved comparably good results of 65% recall and 78% precision. In conclusion, the rule-based method generates promising results that can support physicians' decision making. Because of the structured format the data can be reused for other purposes as well.

Knowledge-based immunosuppressive therapy for kidney transplant patients--from theoretical model to clinical integration

Immunosuppressive therapy is a risky necessity after a patient received a kidney transplant. To reduce risks, a knowledge-based system was developed that determines the right dosage of the immunosuppresive agent Tacrolimus. A theoretical model, to classify medication blood levels as well as medication adaptions, was created using data from almost 500 patients, and over 13.000 examinations. This model was then translated into an Arden Syntax knowledge base, and integrated directly into the hospital information system of the Vienna General Hospital. In this paper we give an overview of the construction and integration of such a system.

Participatory design for drug-drug interaction alerts

The utilization of decision support systems, in the point of care, to alert drug-drug interactions has been shown to improve quality of care. Still, the use of these systems has not been as expected, it is believed, because of the difficulties in their knowledge databases; errors in the generation of the alerts and the lack of a suitable design. This study expands on the development of alerts using participatory design techniques based on user centered design process. This work was undertaken in three stages (inquiry, participatory design and usability testing) it showed that the use of these techniques improves satisfaction, effectiveness and efficiency in an alert system for drug-drug interactions, a fact that was evident in specific situations such as the decrease of errors to meet the specified task, the time, the workload optimization and users overall satisfaction in the system.

Multi-agent Architecture for the Multi-Skill Tasks Modeling at the Pediatric Emergency Department

Patient journey in the Pediatric Emergency Department is a highly complex process. Current approaches for modeling are insufficient because they either focus only on the single ancillary units, or therefore do not consider the entire treatment process of the patients, or they do not account for the dynamics of the patient journey modeling. Therefore, we propose an agent based approach in which patients and emergency department human resources are represented as autonomous agents who are able to react flexible to changes and disturbances through pro-activeness and reactiveness. The main aim of this paper is to present the overall design of the proposed multi-agent system, emphasizing its architecture and the behavior of each agent of the model. Besides, we describe inter-agent communication based on the agent interaction protocol to ensure cooperation between agents when they perform the coordination of tasks for the users. This work is integrated into the ANR HOST project (ANR-11-TecSan-010).

Filling the gap between guideline development and formalization process - a requirement analysis

Clinical guidelines are made to aid diagnosis, management and treatment of patients. Authoring, publishing, updating and maintaining clinical guidelines are time-consuming, labour-intensive and complex. Unfortunately, it is time-consuming to search and retrieve patient-specific recommendations in free-text documents. Literature is rich with methods proposed to support encoding of clinical guidelines into computer-interpretable formats (CIG). However, there is a lack of studies covering the actual guideline development and authoring. So, the objective of this research is to explore gap between tools and methods for authoring guideline content and for designing and implementing computer-interpretable guidelines. Towards this objective, we have performed a user requirements analysis to arrive at a set of design recommendations. The resulting functionality framework can be used to design and develop authoring tools for the entire life-cycle of the computerized clinical guideline.

GA-ANFIS Expert System Prototype for Prediction of Dermatological Diseases

This paper presents novel GA-ANFIS expert system prototype for dermatological disease detection by using dermatological features and diagnoses collected in real conditions. Nine dermatological features are used as inputs to classifiers that are based on Adaptive Neuro-Fuzzy Inference Systems (ANFIS) for the first level of fuzzy model optimization. After that, they are used as inputs in Genetic Algorithm (GA) for the second level of fuzzy model optimization within GA-ANFIS system. GA-ANFIS system performs optimization in two steps. Modelling and validation of the novel GA-ANFIS system approach is performed in MATLAB environment by using validation set of data. Some conclusions concerning the impacts of features on the detection of dermatological diseases were obtained through analysis of the GA-ANFIS. We compared GA-ANFIS and ANFIS results. The results confirmed that the proposed GA-ANFIS model achieved accuracy rates which are higher than the ones we got by ANFIS model.

Clinical Decision Support Knowledge Management: Strategies for Success

Clinical Decision Support Systems have been shown to increase quality of care, patient safety, improve adherence to guidelines for prevention and treatment, and avoid medication errors. Such systems depend mainly on two types of content; the clinical information related to patients and the medical knowledge related to the specialty that informs the system rules and alerts. At King Faisal Specialist Hospital and Research Center, Saudi Arabia, the Health Information Technology Affairs worked on identifying best strategies and recommendations for successful CDSS knowledge management. A review of literature was conducted to identify main areas of challenges and factors of success. A qualitative survey was used over six months' duration to collect opinions, experiences and suggestions from both IT and healthcare professionals. Recommendations were categorized into ten main topics that should be addressed during the development and implementation of CDSS knowledge management tools in the hospital.

An Ecosystem of Intelligent ICT Tools for Speech-Language Therapy Based on a Formal Knowledge Model

The language and communication constitute the development mainstays of several intellectual and cognitive skills in humans. However, there are millions of people around the world who suffer from several disabilities and disorders related with language and communication, while most of the countries present a lack of corresponding services related with health care and rehabilitation. On these grounds, we are working to develop an ecosystem of intelligent ICT tools to support speech and language pathologists, doctors, students, patients and their relatives. This ecosystem has several layers and components, integrating Electronic Health Records management, standardized vocabularies, a knowledge database, an ontology of concepts from the speech-language domain, and an expert system. We discuss the advantages of such an approach through experiments carried out in several institutions assisting children with a wide spectrum of disabilities.

The Case for Conceptual and Computable Cross-Fertilization Between Audit and Feedback and Clinical Decision Support

Many patients do not receive care consistent with best practice. Health informatics interventions often attempt to address this problem by comparing care provided to patients (e.g., from electronic health record data) to quality standards (e.g., described in clinical guidelines) and feeding this information back to clinicians. Traditionally these interventions are delivered at the patient-level as computerized clinical decision support (CDS) or at the population level as audit and feedback (A&F). Both CDS and A&F can improve care for patients but are variably effective; the challenge is to understand how the efficacy can be maximized. Although CDS and A&F are traditionally considered separate approaches, we argue that the systems share common mechanisms, and efficacy may be improved by cross-fertilizing relevant features and concepts. We draw on the Health Informatics and Implementation Science literature to argue that common mechanisms include functions typically associated with the other system, in addition to other features that may prove fruitful for further research.

Improving Clinical Decisions on T2DM Patients Integrating Clinical, Administrative and Environmental Data

This work describes an integrated informatics system developed to collect and display clinically relevant data that can inform physicians and researchers about Type 2 Diabetes Mellitus (T2DM) patient clinical pathways and therapy adherence. The software we developed takes data coming from the electronic medical record (EMR) of the IRCCS Fondazione Maugeri (FSM) hospital of Pavia, Italy, and combines the data with administrative, pharmacy drugs (purchased from the local healthcare agency (ASL) of the Pavia area), and open environmental data of the same region. By using different use cases, we explain the importance of gathering and displaying the data types through a single informatics tool: the use of the tool as a calculator of risk factors and indicators to improve current detection of T2DM, a generator of clinical pathways and patients' behaviors from the point of view of the hospital care management, and a decision support tool for follow-up visits. The results of the performed data analysis report how the use of the dashboard displays meaningful clinical decisions in treating complex chronic diseases and might improve health outcomes.

Fusing Heterogeneous Data for Alzheimer's Disease Classification

In multi-view learning, multimodal representations of a real world object or situation are integrated to learn its overall picture. Feature sets from distinct data sources carry different, yet complementary, information which, if analysed together, usually yield better insights and more accurate results. Neuro-degenerative disorders such as dementia are characterized by changes in multiple biomarkers. This work combines the features from neuroimaging and cerebrospinal fluid studies to distinguish Alzheimer's disease patients from healthy subjects. We apply statistical data fusion techniques on 101 subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. We examine whether fusion of biomarkers helps to improve diagnostic accuracy and how the methods compare against each other for this problem. Our results indicate that multimodal data fusion improves classification accuracy.

A Scalable Architecture for Rule Engine Based Clinical Decision Support Systems

Clinical Decision Support systems (CDSS) have reached a fair level of sophistication and have emerged as the popular system of choice for their aid in clinical decision making. These decision support systems are based on rule engines navigate through a repertoire of clinical rules and multitudes of facts to assist a clinical expert to decide on the set of actuations in response to a medical situation. In this paper, we present the design of a scalable architecture for a rule engine based clinical decision system.

Extraction of Vital Signs from Clinical Notes

Assessment of vital signs is an essential part of surveillance of critically ill patients to detect condition changes and clinical deterioration. While most modern electronic medical records allow for vitals to be recorded in a structured format, the frequency and quality of what is electronically stored may differ from how often these measures are actually recorded. We created a tool that extracts blood pressure, heart rate, temperature, respiratory rate, blood oxygen saturation, and pain level from nursing and other clinical notes recorded in the course of inpatient care to supplement structured vital sign data.

Rule-based Cervical Spine Defect Classification Using Medical Narratives

Classifying the defects occurring at the cervical spine provides the basis for surgical treatment planning and therapy recommendation. This process requires evidence from patient records. Further, the degree of a defect needs to be encoded in a standardized from to facilitate data exchange and multimodal interoperability. In this paper, a concept for automatic defect classification based on information extracted from textual data of patient records is presented. In a retrospective study, the classifier is applied to clinical documents and the classification results are evaluated.