The Rede NUTES telemedicine question submission system provides second opinions to remote healthcare professionals in the northeastern state of Pernaumbuco, Brazil. Submitted questions to the telemedicine system by general practitioners and nurses were analyzed using big data exploration techniques to summarize topic, trends and lexical features.
Philipp Bruland, Ulrich Kathöfer, Maximilian Treder, Nicole Eter, Martin Dugas
1254 - 1254
Reading centers provide centralized high-quality diagnostics in ophthalmic clinical trials. Since ophthalmic images are captured in electronic format at peripheral clinics, an integrated workflow for image transfer and creation of structured reports is needed, including quality assurance. The image portal and the study database are separate components. We assessed whether this integration is feasible with trial-related IT standards and built a prototype system as a proof-of-concept. CDISC ODM and OAuth authentication were used to integrate the image portal with x4T-EDC, facilitating automatic data transfer and single sign-on.
Pedro J. Caraballo, Joseph A. Sutton, Ann M. Moyer, David Blair, Lois C. Hines, Padma S. Rao, Mark F. Adams, Sahana Murthy, Tina Garza, Mary E. Karow, Harmanjit Singh, Jyothsna Giri, Donald B. Gabrielson, Jennifer St. Sauver, Suzette J. Bielinski, Mark A. Parkulo
1255 - 1255
Clinical use of pharmacogenomic (PGx) knowledge at the bedside is new and complex. Our program has implemented multiple PGx-CDS interventions in different clinical settings and in multiple commercial EHRs. Herein, we discuss lessons learned and propose general technical guidelines related to PGx implementation.
High fidelity simulation-based teaching has played an important role in medical education, especially in anesthesiology and emergency. But there is not any currently validated scoring system or prediction model for high fidelity simulation. We will develop a validated prediction model to enhance the efficiency and validation of clinical training with high fidelity simulation.
This paper introduces the characteristics and complexity of traditional Chinese medicine (TCM) data, considers that modern big data processing technology has brought new opportunities for the research of TCM, and gives some ideas and methods to apply big data technology in TCM.
This study describes the process of identifying use cases for next generation electronic nursing records in a high-level view. Literature review, clinical observation, and a Delphi survey were employed. Eight use cases were identified with importance and convergence scores.
Tony Cornford, Valentina Lichtner, Jane Dickson, Ralph Hibberd, Ela Klecun, Will Venters, Bryony Dean Franklin
1259 - 1259
Medicines' supply and use is incresingly reliant on digital means and information. This poster presents exploratory research over five episodes of digitalisation of medicines across the supply network. We ‘follow the drug’ through this emerging field, providing an initial map of this new territory.
Automatic encoding of diagnosis and procedures can increase the interoperability and efficacy of the clinical cooperation. The concept, rule-based and machine learning classification methods for automatic code generation can easily reach their limit due to the handcrafted rules and a limited coverage of the vocabulary in a concept library. As the first step to apply deep learning methods in automatic encoding in the clinical domain, a suitable semantic representation should be generated. In this work, we will focus on the embedding mechanism and dimensional reduction method for text representation, which mitigate the sparseness of the data input in the clinical domain. Different methods such as word embedding and random projection will be evaluated based on logs of query-document matching.
Samah J. Fodeh, Dezon Finch, Lina Bouayad, Stephen Luther, Robert D. Kerns, Cynthia Brandt
1261 - 1261
Pain is a significant public health problem, affecting an estimated 100 million Americans. Evidence has highlighted that patients with chronic pain often suffer from deficits in pain care quality (PCQ). Efforts to improve PCQ hinge on the identification of reliable PCQ indicators such as pain assessment. In this study, we developed a classifier that leverages narratives in clinical notes to derive indicators of pain assessment for patients with chronic pain.
Serious adverse events (AE) or reactions (AR) may occur in clinical trials and require particularly regulated reporting. Manual management is inefficient and ineffective. Based on a description of regulations, we have developed a data model with class-, state-, use-case- and activity diagrams, which can be used for automatic code generation of an assisting software tool for AE / AR data management.
Yuan Gao, Yu Tian, Shengqiang Chi, Yao Lu, Xinhang Li, Tianshu Zhou, Jing-song Li
1263 - 1263
This paper presents a data-driven method to study the relationship of survival and clinical information of patients. The machine learning models were established to study the survival situation at the time of interest based on survival analysis. The way to determine the time of interest is an innovation of this paper. The distribution of survival time is considered, namely the three quartiles, as well as the traditional analysis experience is taken into consideration.
Data entry remains the slowest link in the value chain inhibiting growth of Electronic Medical Records and attendant benefits towards Meaningful use. We designed templates for user forms customized by specialty. Here, we demonstrate the functionality of our software and provide instructions on how these can be adopted by other developers.
Clinical trials generate gold standard medical evidence, but are often criticized for the lack of population representativeness. We performed a comparative meta-analysis of drug trials that focus on older adults (>= 65 years old) and adults (18–64 years old). The major finding is that a higher percentage of geriatric drug trials were terminated or withdrawn than that of adult drug trials.
openEHR is a widely used EHR specification. Given its technology-independent nature, different approaches for implementing openEHR data repositories exist. Public openEHR datasets are needed to conduct benchmark analyses over different implementations. To address their current unavailability, we propose a method for generating openEHR test datasets that can be publicly shared and used.
Recent studies have identified some genes related to cancer phenotypes, but lack of comprehensive clinical information. Here we present an integrated data mining method to find associations between genes and multiple clinical conditions of breast cancer by using gene expression data from Gene Expression Omnibus. The associated modules of genes and clinical conditions were built and some potential related genes were suggested.
Jing Huang, Xinyuan Zhang, Jingcheng Du, Rui Duan, Liu Yang, Jason H. Moore, Yong Chen, Cui Tao
1268 - 1268
US Food and Drug Administration (FDA) Adverse Event (AE) Reporting System (FAERS) is a major source of data for monitoring drug safety. However, there is not general procedure to systematically compare drugs group. We present a statistical method, which can effectively identify significant differences in AE rates among drugs and estimates the differences in age and gender distributions.
Gretchen Hultman, Reed McEwan, Serguei Pakhomov, Elizabeth Lindemann, Steven Skube, Genevieve B. Melton
1269 - 1269
NLP-PIER (Natural Language Processing – Patient Information Extraction for Research) is a self-service platform with a search engine for clinical researchers to perform natural language processing (NLP) queries using clinical notes. We conducted user-centered testing of NLP-PIER's usability to inform future design decisions. Quantitative and qualitative data were analyzed. Our findings will be used to improve the usability of NLP-PIER.
Tianyao Huo, Thomas J. George Jr., Yi Guo, Zhe He, Mattia Prosperi, François Modave, Jiang Bian
1270 - 1270
Patients with colorectal cancer (CRC) often face treatment delays and the exact reasons have not been well studied. This study is to explore clinical workflow patterns for CRC patients using electronic health records (EHR). In particular, we modeled the clinical workflow (provider-provider interactions) of a CRC patient's workup period as a social network, and identified clusters of workflow patterns based on network characteristics. Understanding of these patterns will help guide healthcare policy-making and practice.
Doyeop Kim, Sukhoon Lee, Tae Young Kim, Sanghyung Jin, JaeYeon Park, JeongGil Ko, Rae Woong Park, Dukyong Yoon
1271 - 1271
Bio-signals can be crucial evidence in detecting urgent clinical events. However, until now, access to this data was limited. We aim to construct and provide a new open bio-signal repository with data gathered from more than 40 intensive care unit (ICU) beds. For doing so, we completed the interfacing system with the patient monitors at the target beds and plan to expand this data set to more than 100 ICU beds. Once completed, we plan to publicly open the data to catalyze interesting clinical-event detection research.
We examined depression, impulse control disorder, and life style by degree of smartphone addiction. Chi-square tests and ANOVA were used to identify significant variables. CART was used to generate a decision making diagram of variables affecting smartphone addiction. The severe smartphone addiction group had rates of depression and impulse control disorder than the initial smartphone group.
Jae Kwon Kim, In Hye Yook, Mun Joo Choi, Jong Sik Lee, Yong Hyun Park, Ji Youl Lee, In Young Choi
1273 - 1273
This study objectives to investigate a range of Partin table and several machine learning methods for pathological stage prediction and assess them with respect to their predictive model performance based on Koreans data. The data was used SPCDB and gathered records from 944 patients treated with tertiary hospital. Partin table has low accuracy (65.68%) when applied on SPCDB dataset for comparison on patients with OCD NOCD conditions. SVM (75%) represents a promising alternative to Partin table from which pathology staging can benefit.
We have used deep neural networks (DNNs) to generate clinical opinions from general blood test results. DNNs have overfitting problem in general. We believe the complex structure of DNN and insufficient data to be the major reasons of overfitting in our case. In this paper, we apply dropout and batch normalization to avoid overfitting. Experimental results show the improvement in the performance of the DNNs.
Raphael W. Majeed, Tingyan Xu, Mark R. Stöhr, Rainer Röhrig
1275 - 1275
Since its release in 2004, the i2b2 data warehouse software has become a valuable tool for clinical researchers. Physicians can use its browser-based query frontend intuitively without additional training or reading through documentation. While the i2b2 software describes it's API as “REST”, it is neither stateless nor does it follow the common guidelines for RESTful APIs. Thus, interfacing other software with i2b2's custom RPC-style XML-API is a very cumbersome process. To overcome these issues, we developed a lightweight software abstraction layer “lightweight i2b2 façade” (li2b2-façade).
Raphael W. Majeed, Mark R. Stöhr, Volker S. Thiemann, Rainer Röhrig, Andreas Günther
1276 - 1276
Clinical Data Ware Houses are established sources for research and quality management. The open source data warehouse software i2b2 enjoys good reputation and wide-spread use in the international medical informatics community. We developed a novel infrastructure to allow queries to be distributed asynchronously between i2b2 data warehouses.
In healthcare, applying deep learning models to electronic health records (EHRs) has drawn considerable attention. This sequential nature of EHR data make them wellmatched for the power of Recurrent Neural Network (RNN). In this poster, we propose “Deep Diabetologist” – using RNNs for EHR sequential data modeling to provide personalized hypoglycemic medication prediction for diabetic patients. Our results demonstrate improved performance compared with a baseline classifier using logistic regression.