Ausgabe 00/2016

The aim of the Infomed project is to implement a shared electronic patient record for all the healthcare professionals and patients in the canton du Valais. Just before opening Infomed to the patients we ask the participating physicians about their platform’s satisfaction. We also try to get their opinion about the access by the patients to Infomed and the included medical records.

The rapid adoption of mobile applications for wellness and health tracking has resulted in vast amounts of patient-generated data. However, these data are often underutilized in the traditional patient care. In this paper, we explore how to use these patient-generated data to improve patient care. Based on a review of healthcare model and recommendations, we propose and compare four models with increasing integration with electronic health records. We also compared the freedom of choice of apps, as well as content validity and expected effectiveness. In the first model, patients have the full range of app choice, and full control over their data, in particular for sharing with healthcare providers. In the second model, patients use a selection of apps to export their data to a repository, which can be accessed by their providers (without integration into the EHR). In the third model, interoperability between the apps and the EHR allows full integration, but restricts app choice. Finally, the last model adds the notion of cost-effectiveness to the previous model. Although the EHR-integrated models limit app choice for patients, the app content is medically validated and patient-generated data is more easily accessed to improve patient care. However, these integrated models require decision support algorithms to avoid overwhelming the healthcare providers with data, and may not necessarily imply better quality patient care.

Introduction Venous thromboembolism (VTE) as a hospital-acquired condition (HAC) – i.e. not ‘present on admission’ (POA) – is a potentially preventable complication. A decrease of HAC VTE events indicates success of efforts to prevent VTE in hospitalized patients. However, so far, costly chart reviews were needed to identify patients with HAC VTE. We investigated whether electronic health record data such as medication orders and their temporal relations allow for differentiating between HAC and POA. Therefore, we modeled a tree and two random forests and evaluated the automated classification of HAC VTE. Methods All inpatients with a length of stay of ≥24 hours (h), discharged from the Brigham and Women’s Hospital, a large tertiary care hospital in Boston, MA, between January 2009 and April 2014 were searched for ICD-9 diagnosis codes of acute venous thrombosis or pulmonary embolism. Patients were included who had VTE in the admitting diagnosis field – defined as POA VTE – or in one of up to 50 discharge diagnoses. Of those, only patients who received heparin, dalteparin, enoxaparin, alteplase, rivaroxaban or fondaparinux were considered, and the time from admission to the first order was calculated for each drug. Additionally included predictors: dose information, demographics (age, gender, race, language), length of stay, admission service, discharge service, transfer destination of the patient after discharge, and whether the patient was alive or died during the hospitalization or within 30 days after discharge. A single tree and two random forests (each with 5,000 trees) were generated to analyze the predictors and to assess the predictive power of the chosen approach. Since medication orders are electronically available in real time, such prospective predictors may have implications for clinical decision support – therefore, prospective predictors (i.e. demographics, admission service, time to order a drug, route and dose information for each drug) were separately analyzed in the first random forest. Half of the data served as calibration set, half as validation set. Statistical computing was performed using the software R version 3.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Results A total of 5,374 patient stays featured a VTE diagnosis with a defined drug order. If VTE was POA (n=1,262; 23.5%), the median time to order one of the aforementioned drugs was 2.5h (IQR 1.3-5.0h). Among HAC VTE cases without an admitting diagnosis of VTE (n=4,112; 76.5%), the median time to order the drug was 4.2h (IQR 1.7-18.2h). Unsurprisingly, a single tree – after cross-validation and pruning – identified the time from admission to the ordering of intravenous (IV) heparin as the most significant predictor (Fig. 1). This tree’s validation resulted in an accuracy of 78.8% and a positive predictive value (PPV) of 83.3% for the classification of HAC VTE. The first validated random forest used predictors which are available in real time: the forest had an accuracy of 79.7% and a PPV of 85.3% for the classification of HAC VTE. The second validated random forest considered all variables and resulted in an accuracy of 81.7% and a PPV of 87.8% (variables’ importance is shown in Fig. 2). Discussion We modeled a tree and two random forests using structured data predictors to differentiate between HAC and POA VTE. Our validated tree (Fig. 1), considering the first order for IV heparin and the length of stay, could immediately be implemented as a first step to identifying HAC VTE patients. However, the random forests performed better, even when exclusively prospective predictors were used – and such real time models may have implications for clinical decision support tools. In conclusion, our random forests could help to evaluate interventions to improve thromboprophylaxis regimens for inpatients, where costly chart reviews are needed to differentiate between POA VTE and potentially preventable complications.

Background and Introduction Last year’s introduction of ResearchKit, an open source toolkit for iOS facilitating the creation of smartphone research apps, has sparked renewed interest in smartphone-driven biomedical research. In addition to the initial five research apps, about a dozen more ResearchKit-powered apps are now available to iOS-using participants in the United States. In April 2016, ResearchStack – the Android counterpart to ResearchKit – has been released, enabling researchers to finally include participants using the most popular mobile operating system. The field now has powerful informatics tools at its disposal, but it still needs to prove that the approach of collecting patient data for biomedical research via smartphones is useful and sustainable. Methods The C Tracker study is an apps-based trial, assessing hepatitis C patients’ activity levels over time. The app distributes surveys to study participants on a 2-weekly basis and returns activity data, such as steps taken and time spent exercising, along with survey answers. Users are identified by a random number, all data is de-identified and encrypted before being sent over the internet. The well-known i2b2 research backend serves as data storage. To provide value to participants, the app also contains a dashboard showing their recent activity, resources informing about hepatitis C and its treatment and other tidbits, such as a map of the US, showing participant origin. We are bringing C Tracker to Switzerland, extending its target population from anonymous “in the wild” recruitment to patients already enrolled in the Swiss Hepatitis C Cohort Study (SCCS). The data delivery toolchain, available open source under the name “C3-PRO” and using the upcoming Fast Healthcare Interoperability Resources (FHIR) standard, is extended with a separate backend system storing participant identity data, linking the app’s user identifier to participants’ SCCS study identifier. Circumnavigating the cloudy waters of electronic consent in Switzerland, we collect paper-based consent from participants during their annual clinic visit, at least initially. We are also adapting our toolchain to ResearchStack and hope to port the complete app to Android in a timely manner. Results & Discussion At this early stage in the project, we have identified steps in the original approach in need of adaptation to Switzerland. Most importantly, we have built an “identity manager”, allowing us to collect paper based consent from patients, recording the consent electronically and provide participants with a link to “unlock” the app, allowing access to the research study part of the app as a fully consented user. While this adds another system that research coordinators need to use, its use is straightforward, only requiring entry of five data items. The link to the app can either be established immediately via QR code or by emailing a link to the participant that will open the app. We are in the process of finalizing the server components and the app and hope to enroll our first participants in the near future. All our tools will be made available open source.