Screen shot 2017 03 13 at 11 28

Seminarier Vitalis 2018

100 mtf

Kursen Patientsäkra IT-system i vården och nya regelverk arrangeras på Vitalis av MTF (Svensk Förening för Medicinsk Teknik och Fysik) Kursen är nu fullbokad och anmälan stängd.

100 mspre

Bortom videobesöket – Hologram och AI

Måndag 15:00 - 18:00 The Theatre
Preconference, Vitalis

Dagens teknik möjliggör mycket mer än ett videosamtal. För att fånga kraften i digitaliseringen krävs att vi tillsammans vågar tänka om och tänka nytt. Allt fler hälso- och sjukvårdsorganisationer runt om i världen har börjat undersöka Mixed Reallity (MR) genom användandet av Microsoft HoloLens. Maskinlärning och Artificiell Intelligens (AI) blir allt vanligare tack vare åtkomsten till beräkningskapacitet och AI-plattformar i publika molntjänster som Microsoft Azure. Och vad händer när dessa trender börjar konvergera? Vad är potentialen med MR+AI?
Det är hög tid att förstå hur MR+AI kommer bidra till digitaliseringen av svensk hälso- och sjukvård. Det är dags att blicka bortom videobesöket om Sverige ska bli världsbäst på eHälsa.Microsoft’s Globala Chief Medical Officer Simon Kos delar med sig av det som händer runt om i världen samt Microsoft vision för MR+AI.Tillsammans med ledande personer från Prototyp, Sigma och Tieto diskuterar och visar vi på vi hur Mixed Reallity och AI kommer förändra den svenska vården på kort och lång sik
Detta är en Vitalis Preconference arrangerad av Microsoft. Separat anmälan krävs. Klicka på Mer info för länk.

100 bild mingel

ehälsopub

Måndag 18:00 - 20:00 Twentyfourseven - Lobbyn Gothia Towers
Preconference, Vitalis


Välkommen till Public House!-          Vitalis
och VGR bjuder in till ehälsopub

Kom och ladda upp inför Vitalis redan på måndagkvällen och
träffa kollegor i branschen!
Vi behöver fler arenor där vi kan mötas över gränserna och
diskutera framtidens vård och omsorg under neutral flagg. Därför har Vitalis
och VGR tagit ett gemensamt initiativ till en enkel afterwork där vårdgivare,
myndigheter, leverantörer och andra kan träffas under enkla former.
Vi utlovar trevligt sällskap, intressanta diskussioner och
after work priser i baren!
Drop in, ingen föranmälan.
Tid: 23 april, kl
18-20
 Plats: Gothia Towers lobbybar Twentyfourseven  

Varmt välkomna önskar Vitalis-teamet och VGR genom Anette
Falkenroth!

We introduce 3000PA, a clinical document corpus composed of 3,000 EPRs from three different clinical sites, which is intended to serve as a national reference language resource for German clinical NLP. We outline its design principles, its role within a currently evolving new medical informatics infrastructure in Germany, first results from a medication annotation campaign and the evaluation of a medication information extraction prototype using a subset from 3000PA.

The emergence and utilization of mobile phone technology have been increased in a variety of health interventions. Firm conclusion about whether text-messaging program impact on smoking cessation remains inconclusive. Therefore, our study objective was systematically to review literatures, and evaluate the acceptability and feasibility of a smoking cessation text-messaging program for pregnant women. PubMed, EMBASE, and the bibliographies of articles were searched for studies published between January 1, 1990, and August 1, 2017, which reported on the smoking cessation text-messaging program for pregnant women. We selected 12 out of 225 unique abstracts for full-text reviewing using our selection criteria and 12 out of these 10 studies, comprising 3,601 pregnant smokers, met all of our inclusion criteria. Findings from this study show that a text-messaging program was acceptable and feasible for pregnant smokers and provided motivational, educational support for quitting smoking. This smoking quitting intervention also provided service at a cheap cost and potential impact on pregnant women smoking cessation. It will be a public health priority in near future but more research is needed with long follow-up period

Health Records (PHRs) are an opportunity to help patients in taking an active role in their healthcare. PHR components such as medication information are health care management tools. The aim of this paper is to describe a tool for the automated purchase of medication through a PHR. The medication module, which is part of the PHR, designed with user centered design (UCD) techniques, it is composed by different interfaces by which medication is shown, renewed, purchased and delivered to patients. PHR medication module, represents an opportunity to provide patients with a tool to visualize their medication list. This might potentially represent the first stage in the development of a medication reconciliation tool for patients.

Electronic health record (EHR) systems are necessary for the sharing of medical information between care delivery organizations (CDOs). Although many standardized data can be shared, it is still difficult to share non-standardized medical records. Moreover, EHR systems do not have detailed control over which documents are disclosed. We have been operating a document-based EMR system, DACS, in which all medical records are stored in the PDF format. In the present study, we describe the development of our document-based EHR system. With this system, all of the documents stored in the DACS can be opened and the user can select the target CDOs and the documents to be disclosed. An access control list (ACL) file was designed based on the HL7 CDA header, which allows for the management of the documents that are to be disclosed. The PDF file and the corresponding ACL file are delivered to a Japanese standard repository, SS-MIX2. The doctor of the target CDO can access the documents according to the information contained within the ACL file. From March 2017 to September 2017, three hundred and eight documents of 20 patients were disclosed to 3 medical facilities using this document-based EMR system.

Agile Ontology for Complex Chronic Patient Personalized Decision Support

Francisco José Núñez Benjumea
Tisdag 08:00 - 09:30 HALL - posters
MIE, Poster, Poster session #1

Complex Chronic Patient management needs to complement the recommendations defined in clinical guidelines from recommendations based on treatments performed on a representative set of patients, identifying conflicts between the recommendations of different guidelines designed for handling isolated chronic diseases. It also requires its extension to specific protocols in areas not described with sufficient detail in clinical guidelines in terms of safety and quality. This problem is overcome by designing and deploying an Ontology-Based Personalized Clinical Decision Support Tool which helps to improve individualized decisions and reduces variability in clinical practice.
 

Introduction
Nonvalvular Atrial Fibrillation, is estimated to affect 5.8 million people in the US. This study compared the accuracy of structured ICD9 vs. electronic health record (EHR) data using NLP, to identify NVAF cases and the CHA2DS2-VASc and HAS-BLED Scores.
Methods
The retrospective cohort study included patients 18 to 90 with a diagnosis of NVAF. An electronic model for structured data using ICD-9 criteria and for unstructured data using a NLP to a semi-supervised machine learning technique SNOMED CT algorithm, a high throughput phenotyping system that rapidly assigns ontology terms to text in patient records, was applied to identify the NVAF population and their CHA2DS2-VASc and HAS-BLED Scores. We used a cohort of 63,296,120 patients to determine the national rates of NVAF, CHA2DS2VASc and Stroke and Death Rates.
Results
Out of the 96,681 patients identified in the AllScripts EHR data, 2.8% (2722 cases) were identified with NVAF by the Structured+NLP method as opposed to 1.9% for Structured alone (1849 cases) with a difference of 873 cases (32.1%, p<0.001). The area under the curve for the structured plus NLP was 0.914 and for structured data alone was 0.864 (p<0.001) .
Conclusion
The Structured+NLP biosurveillance method had a higher sensitivity in comparison to Structured data alone, allowing for an increased number of true positive cases to be identified. This program when implemented is predicted to prevent 153,623 strokes and save 9,202 lives at a savings of over 15 billion dollars in year one to the country.

The AID-stroke project developed a clinical decision support engine based on ontological definition of concepts and rules for acute stroke patient care. The project implemented clinical decision support rules focused on providing best evidence and assessment during the emergency patient care. The scope of decision support rules includes patient inclusion criteria for TeleStroke service and fibrinolysis, transferring patient for cerebral angiography, intervention or neurosurgery. This research has been partly funded by the ITEMAS Platform (PT13/0006/0036) and by the AID Ictus project (DTS14/0143) by Carlos III National Health Institute, both co-funded by FEDER from European Regional Development Funds (European Union).

Medical researchers are interested in how potentially interconnected diseases progress over time. A range of methods can be used to extract disease trajectories from the temporal data in electronic health records (EHRs). Our research attempts to reproduce cardiovascular disease (CVD) trajectories published in Nature Communications 2014, where gout was found to be a central diagnosis. The process, using techniques from process mining helped to identify potential refinements to the method. We describe how approaches from process mining can help to advance the study of disease trajectories from EHRs.

Semantic interoperability is crucial in the frame of federated Electronic Health Records (EHRs) to share understandable and reusable data among the different healthcare information systems. The use of specific and integrated coding systems is recommended by Governments all over Europe, in particular concerning the registration of clinical data in the Patient Summary (PS). To support General Practitioners (GPs), we propose a Coding Support System (CSS) provided as a Web Service that aims, on one hand, to validate and, on the other, to suggest the most appropriate coding of clinical diagnoses in the PS digitized by the GP in his EHR tool, to avoid errors and misspecifications. The system leverages on a Knowledge Base composed of an ICD-9-CM ontology, built for the project, integrated with different terminological resources, and based on standardized coding rules. Preliminary tests for the validation of the Problem List coding show promising results and new tests are being carried out on other elements of the PS.

In the last years, a variety of theoretical frameworks have been proposed to support decision-makers while assessing eHealth initiatives. This study, by means of a systematic literature review, investigates the actual application of the theoretical frameworks when assessing eHealth initiatives. Results show that empirical studies do not rely on theoretical frameworks and prefer tailor-made approaches, thus reducing the comparability (and robustness) of current assessments.

Most of HIT interventions are developed for chronic patients. The objective of this study was to determine the readiness of chronic patients attended clinics of educational hospitals in Khorramabad, Iran for using HIT.A sample of 410 chronic patients, who attended clinics of educational hospitals in Khorramabad, Iran (in 2016), was selected by random sampling in this descriptive-analytical study. Data were collected using the valid and reliable PRE-HIT questionnaire and analyzed using SPSS 20 using descriptive-analytical statistics. The results showed that 24.4% (n=100) of the participants had experience of using computers. The readiness of chronic patients having experience with computers was at a medium level (mean=2.77, SD=0.59).The successful deployment of information technology in patients requires the attention to topics such as the difference in their level of readiness for using information technology and ease of use.

The landscape of databases available for vaccine-benefit risk studies is complex due to the heterogeneity of health systems and different approaches to data recording in computerised medical records across Europe. Potential databases need to be systematically identified and profiled in order to ensure good representation of data in the region. We report results from a two round survey used to profile potential data sources for four types of vaccine studies and the challenges faced during the process. The four types of study and the number of European data bases able to support research are: (1) Vaccine coverage (n=31), (2) Effectiveness (n=32) and (3) Safety (n-39) and (4) Burden of vaccine preventable disease (n=29), Ontological approaches that improve metadata reuse and interoperability facilitate participation of databases across the whole spectrum of vaccine research.

Bringing the Socio-markers into Health Surveillance: A Predictive Model for Pediatric Asthma Patients at Risk of Hospital Revisiting

Eun Kyong SHIN
Tisdag 08:00 - 09:30 HALL - posters
MIE, Poster, Poster session #1

Building a national index of Portuguese ICU clinical data

Ricardo Santos
Tisdag 08:00 - 09:30 F3
MIE, Short communication, Standards and Interoperability

Clinical Decision Support Systems (CDSS) were conceived to alert of potential errors and dangers, and more recently evolved into more sophisticated systems as Electronic Health Records (EHR) were developed and implemented. Storing and displaying patient information is the foundation that makes possible the improvement of clinical decision-making as large amounts of clinical and laboratory data can be processed. The quantity of data in the Intensive Care Units (ICU) can challenge the ability of clinicians to process them. CDSS systems can improve medical performance by improving patient outcomes. There is also a need to incorporate items like safety, functionality, technical performance, economics and organizational implications into the health technology assessments. The quality of health systems depends on the right decisions being made, which depends on information been accurate, embracing, unmistakable and available at the right time and place, to the right person.
Some of the ICU data are from the bedside monitor devices attached to each critical care patient such as vital signs, waveforms, trends and alarms. Our team intends to build a Portuguese ICU database index as a first step to map current state of use and reuse of clinical and administrative ICU data. Our main concern is underutilization of unrecorded data that should be developed into information to support clinicians deliver better care. There is an urgent need to create and establish partnership, adjust synergy with well defined roles and tasks between health informatics development centres and ICUs, uniting efforts, capabilities, knowledge towards scientific improvement and better clinical

Building a XML data warehouse based on HL7 CDA documents

Fabrizio Pecoraro
Tisdag 08:00 - 09:30 HALL - posters
MIE, Poster, Poster session #1

Aim of this paper is to define a method for developing a data warehouse based on HL7 Clinical Document Architecture (CDA). The procedures adopted to extract information from source documents and transforms them to be loaded in the destination data warehouse are centered on a conceptual framework, that maps the primitives of the data warehouse dimensional model with the HL7 CDA classes. The method as well as the data warehouse have been tested using a web repository of Continuity of Care Documents to assess the quality of care of patients with diabetes.

The aim of our study was to adopt the explanatory data-mining approach to identify characteristics of emergency department (ED) visits for ED management. To that end, we adopted a novel behaviour-based decision tree (DT) induction method that considers the information of medical diagnoses and individual patients’ information in order to analyse characteristics of patients’ visits to an ED. We interpreted the results based on the clinical value, and communicability of the DT, represented as a behaviour-based DT profile in order to increase the explanatory power of patients’ behaviours in the ED and provide clinical decision-making.