Shivani Batra

Applying data mining techniques to standardized electronic health records for decision support

Data Mining is very popular in todays era because it provides access to the information that is hidden from a normal human being eye. The hidden information that is made available through data mining can benefit the person involved by providing an efficient decision support. Today data mining can be applied to various areas such as marketing, banking, aerospace and healthcare. It is identified that providing decision support in healthcare domain can help in saving human life. Although providing decision support through data mining in healthcare is very beneficial but it lacks standardization. A comparison of benefits gained from applying data mining techniques to standardized and non-standardized EHRs is provided. This paper aims to indentify the need of applying data mining techniques to standardized electronic healthcare records. It interrogates various issues that need to be resolved for providing an efficient standardized decision support system.

Mining Standardized Semantic Interoperable Electronic Healthcare Records

There is vast potential for data mining applications in healthcare. Generally, these applications can be grouped as the evaluation of treatment effectiveness; management of healthcare; customer relationship management; and detection of fraud and abuse. To build the knowledge that is universally true, data has to be collected from all over the world. Collecting such large amount of data to prepare a single database that can be used to apply data mining techniques requires many challenges to be faced by the researcher. Till date no standard is adopted universally that imposes some guidelines on storing the data in a particular format. We propose to use a generic database to collect standardized EHR data that is available in different formats and at different geographical regions. This paper proposes a framework for applying data mining techniques to healthcare database stored on the basis of row model. We also try to incorporate protective measures in the architecture for privacy of the user, try to secure the data collected from various sources and also use of authentication mechanism at interface level.

Suitability of Data Models for Electronic Health Records Database

With the advancement in technology, data is also growing exponentially. Storing this BIG DATA in an efficient manner is the key for any successful project. Work done in this paper is dedicated towards presenting the possible efficient ways available to store Electronic Health Records (EHRs). The main hurdles in storing EHRs are sparseness and volatility which relational model is incapable to handle. The other models present for storing EHRs are Entity Attribute Value (EAV), Dynamic Tables, Optimized Entity Attribute Value (OEAV) and Optimized Column Oriented Model (OCOM). Authors have provided a comparative study which will help the administrator to choose the best model among the models specified above. Authors have also discussed about the different scenarios (standardized and non-standardized EHRs) in which a combination of these models can be used. Authors have simulated EAV, Dynamic tables, OEAV and OCOM models to provide comparison results of time taken for executing basic operations (queries) and memory consumed by different models.

Enhancing Access to Standardized Clinical Application for Mobile Interfaces

As Electronic Health Records (EHRs) become more prevalent in health care, research is needed to understand the efficacy within clinical contexts for a standard based health application. The current research explores ‘Opereffa’ to be used for handheld moveable devices. Opereffa stands for openEHR REFerence Framework and Application. It is a project for creating an open source clinical application, which will be driven by the Clinical Review Board of openEHR [2]. It is based on openEHR standard which combines structure of archetypes and terminology codes. This is the first effort for its exploration on mobile devices. The aim is to generate an application programming interface for Android based mobile for its testing on a sample set of archetypes. Later, we will extend this research to other mobile operating systems. The study has been done for increasing the usability and reach ability of EHRs. It enhances data sharing through mobile for standardized EHRs (through use of archetypes).

Healthsurance – Mobile App for Standardized Electronic Health Records Database

With the increasing popularity of Electronic Health Records (EHRs), there arises a need to understand its importance in terms of clinical contexts for a standard based health application. Standards for semantic interoperability propose the use of archetypes for building a health application. A usual practice followed for storing of EHRs is through graphical user interfaces. Generally, user interface is static corresponding to the underlying medical concept, often made manually and are prone to errors. However, evolution in knowledge demands for dynamically generated user interfaces to reduce time, minimize cost and enhance reliability. Current research implements mobile app for standardized Electronic Health Records Database termed as HEALTHSURANCE. The application maintains its dynamic behavior through creation of graphical user interfaces at runtime by gaining knowledge from the artefacts (known as archetypes) available from standard clinical repositories (such as Clinical Knowledge Manager). This provides easy and hassle-free user operability without any need of mobile developer. A standardized format and content helps to uplift the credibility of data and maintains a uniform and specific set of constraints used to evaluate the user’s health. A generic centralized database is chosen for data storage to support evolution in clinical knowledge and to handle heterogeneity of EHRs data. Implementing mobile app based on archetype paradigm avoids reimplementation of systems, migrating databases and allows the creation of future-proof systems.

Semantic Interoperability in Electronic Health Record Databases: Standards, Architecture and e-Health Systems

Information systems have been deployed in different clinics and hospitals to preserve patient data. In order to promote the exchange of data among systems (and organizations), standards are being adopted for data exchange. Further, the clinics and hospitals aim to manage a patient’s life-time history of records. A piece of the individual patient’s medical record can be captured, stored, queried, and shared over a network through enrichment in information technology. Thus, electronic health records (EHRs) are being standardized for incorporating semantic interoperability. In addition, a generic storage structure is required to capture distinguished data requirements of various organizations. The generic structure must be capable of dealing with sparseness and frequent evolution behavior of EHRs. A subsequent step requires that healthcare professionals and patients get to use the EHRs, with the help of technological developments, such as workflow toolkits and new (easy) query languages. The goal is to present an overview of different approaches in understanding some current and challenging concepts in e-health informatics. Successful handling of these challenges will lead to improved quality in healthcare by reducing medical errors, decreasing costs, and enhancing patient care. The report is focused on the following objectives: (1) understanding the role of EHRs Databases; (2) understanding the need for standardization to enhance quality; (3) establishing interoperability in maintaining EHRs; (4) explicating a framework for standardization and interoperability (the openEHR architecture); (5) exploring various data models for managing EHRs; and (6) understanding the difficulties in querying data in EHR and e-health systems.

Entity Attribute Value Style Modeling Approach for Archetype Based Data

Entity Attribute Value (EAV) storage model is extensively used to manage healthcare data in existing systems, however it lacks search efficiency. This study examines an entity attribute value style modeling approach for standardized Electronic Health Records (EHRs) database. It sustains qualities of EAV (i.e., handling sparseness and frequent schema evolution) and provides better performance for queries in comparison to EAV. It is termed as the Two Dimensional Entity Attribute Value (2D EAV) model. Support for ad-hoc queries is provided through a user interface for better user-interaction. 2D EAV focuses on how to handle template-centric queries as well as other health query scenarios. 2D EAV is analyzed (in terms of minimum non-null density) to make a judgment about the adoption of 2D EAV over n-ary storage model of RDBMS. The primary aim of current research is to handle sparseness, frequent schema evolution, and efficient query support altogether for standardized EHRs. 2D EAV will benefit data administrators to handle standardized heterogeneous data that demands high search efficiency. It will also benefit both skilled and semi-skilled database users (such as, doctors, nurses, and patients) by providing a global semantic interoperable mechanism of data retrieval.

Managing Large-Scale Standardized Electronic Health Records

Electronic health records (EHRs) contain data about a person’s health history. Increasingly, EHRs have the characteristics of big data in terms of their volume, velocity, and variety (the 3 “V”s). Volume is a major concern for EHRs especially due to the presence of huge amount of null data, i.e., for storing sparse data that leads to storage wastage. Reducing storage wastage due to sparse values requires amendments to the storage mechanism that stores only non-null data, and also allows faster data retrieval and supports multidimensional heterogeneous data. Another area of concern regarding EHRs data is standardization. Standardization can aid in semantic interoperability that resolves the discrepancies in interpretation of health records among different medical organizations or persons involved. Various proposals have been made at the logical layer of relational database management system for managing large-scale standardized records in terms of data volume, velocity, and variety. Every proposed modification to logical layer has its pros and cons. In this chapter, we will discuss various aspects of the solutions proposed for managing standardized EHRs, and the approaches to adopt these standards. After efficient management of EHR data, analytics can be applied to minimize the overall cost of healthcare.

Moving from Relational Data Storage to Decentralized Structured Storage System

The utmost requirement of any successful application in today’s environment is to extract the desired piece of information from its Big Data with a very high speed. When Big Data is managed via traditional approach of relational model, accessing speed is compromised. Moreover, relational data model is not flexible enough to handle big data use cases that contains a mixture of structured, semi-structured, and unstructured data. Thus, there is a requirement for organizing data beyond relational model in a manner which facilitates high availability of any type of data instantly. Current research is a step towards moving relational data storage (PostgreSQL) to decentralized structured storage system (Cassandra), for achieving high availability demand of users for any type of data (structured and unstructured) with zero fault tolerance. For reducing the migration cost, the research focuses on reducing the storage requirement by efficiently compressing the source database before moving it to Cassandra.

Usability and information retrieval issues for electronic healthcare record databases

Standardized Electronic Healthcare Records (EHRs) data have a complex structure. They are based on dual model approach where domain knowledge is saved as an archetype [1, 2]. It is technically very difficult to generate query language for medical expert to retrieve the data. It gives rise to issue of retrieval and usability of electronic healthcare records databases. To improve the retrieval process, Graphical User Interface (GUI) must be simplified. Other issues related with electronic healthcare record systems are mobility, reachability and optimal use of advancements in information technology. This paper highlights the problems and issues of electronic healthcare record systems. It investigates the available graphical user interfaces for electronic healthcare record databases and compares their usability. Finally, it presents the comparison of a graphical user interface for a standardized Electronic healthcare records application for various handheld mobile devices.