Eunja Chung

A personalized mobile patient guide system for a patient-centered smart hospital: Lessons learned from a usability test and satisfaction survey in a tertiary university hospital

OBJECTIVE The present study focused on the design, implementation, and evaluation of a personalized mobile patient guide system that utilizes smart phones, indoor navigation technology and a hospital information system (HIS) to address the difficulties that outpatients face in finding hospital facilities, recognizing their daily treatment schedule, and accessing personalized medical and administrative information. MATERIALS AND METHODS The present study was conducted in a fully digitized tertiary university hospital in South Korea. We developed a real-time location-based outpatient guide system that consists of Bluetooth access points (APs) for indoor navigation, an Android-based guide application, a guide server, and interfaces with the HIS. A total of 33 subjects and 43 outpatients participated in the usability test (UT) and the satisfaction survey, respectively. RESULTS We confirmed that the indoor navigation feature can be applied to outpatient departments with precision using a position error test. The participants in the UT completed each scenario with an average success rate of 67.4%. According to the results, we addressed the problems and made improvements to the user interface by providing users with context-based guidance information. The satisfaction rating of the system was high, with an average score of 4.0 out of 5.0, showing its utility as a patient-centered hospital service. CONCLUSION The innovative mobile patient guide system for outpatients is feasible and can be successfully implemented to provide personalized information with high satisfaction. Additionally, the issues identified and lessons learned from our experiences regarding task scheduling, indoor navigation, and usability should be considered when developing the system.

Implementation of a Next-Generation Electronic Nursing Records System Based on Detailed Clinical Models and Integration of Clinical Practice Guidelines

Objectives The purpose of this paper is to describe the components of a next-generation electronic nursing records system ensuring full semantic interoperability and integrating evidence into the nursing records system. Methods A next-generation electronic nursing records system based on detailed clinical models and clinical practice guidelines was developed at Seoul National University Bundang Hospital in 2013. This system has two components, a terminology server and a nursing documentation system. Results The terminology server manages nursing narratives generated from entity-attribute-value triplets of detailed clinical models using a natural language generation system. The nursing documentation system provides nurses with a set of nursing narratives arranged around the recommendations extracted from clinical practice guidelines. Conclusions An electronic nursing records system based on detailed clinical models and clinical practice guidelines was successfully implemented in a hospital in Korea. The next-generation electronic nursing records system can support nursing practice and nursing documentation, which in turn will improve data quality.

Predictive Bayesian Network Model Using Electronic Patient Records for Prevention of Hospital-Acquired Pressure Ulcers

PURPOSE The study was designed to determine the discriminating ability of a Bayesian network (BN) for predicting risk for pressure ulcers. METHODS Analysis was done using a retrospective cohort, nursing records representing 21,114 hospital days, 3,348 patients at risk for ulcers, admitted to the intensive care unit of a tertiary teaching hospital between January 2004 and January 2007. A BN model and two logistic regression (LR) versions, model-I and -II, were compared, varying the nature, number and quality of input variables. Classification competence and case coverage of the models were tested and compared using a threefold cross validation method. RESULTS Average incidence of ulcers was 6.12%. Of the two LR models, model-I demonstrated better indexes of statistical model fits. The BN model had a sensitivity of 81.95%, specificity of 75.63%, positive and negative predictive values of 35.62% and 96.22% respectively. The area under the receiver operating characteristic (AUROC) was 85.01% implying moderate to good overall performance, which was similar to LR model-I. However, regarding case coverage, the BN model was 100% compared to 15.88% of LR. CONCLUSION Discriminating ability of the BN model was found to be acceptable and case coverage proved to be excellent for clinical use.

Development and User Research of a Smart Bedside Station System toward Patient-Centered Healthcare System

User experience design that reflects real-world application and aims to support suitable service solutions has arisen as one of the current issues in the medical informatics research domain. The Smart Bedside Station (SBS) is a screen that is installed on the bedside for the personal use and provides a variety of convenient services for the patients. Recently, bedside terminal systems have been increasingly adopted in hospitals due to the rapid growth of advanced technology in healthcare at the point of care. We designed user experience (UX) research to derive users’ unmet needs and major functions that are frequently used in the field. To develop the SBS service, a service design methodology, the Double Diamond Design Process Model, was undertaken. The problems or directions of the complex clinical workflow of the hospital, the requirements of stakeholders, and environmental factors were identified through the study. The SBS system services provided to patients were linked to the hospital’s main services or to related electronic medical record (EMR) data. Seven key services were derived from the results of the study. The primary services were as follows: Bedside Check In and Out, Bedside Room Service, Bedside Scheduler, Ready for Rounds, My Medical Chart, Featured Healthcare Content, and Bedside Community. This research developed a patient-centered SBS system with improved UX using service design methodology applied to complex and technical medical services, providing insights to improve the current healthcare system.

Integration of Evidence into a Detailed Clinical Model-based Electronic Nursing Record System.

Objectives The purpose of this study was to test the feasibility of an electronic nursing record system for perinatal care that is based on detailed clinical models and clinical practice guidelines in perinatal care. Methods This study was carried out in five phases: 1) generating nursing statements using detailed clinical models; 2) identifying the relevant evidence; 3) linking nursing statements with the evidence; 4) developing a prototype electronic nursing record system based on detailed clinical models and clinical practice guidelines; and 5) evaluating the prototype system. Results We first generated 799 nursing statements describing nursing assessments, diagnoses, interventions, and outcomes using entities, attributes, and value sets of detailed clinical models for perinatal care which we developed in a previous study. We then extracted 506 recommendations from nine clinical practice guidelines and created sets of nursing statements to be used for nursing documentation by grouping nursing statements according to these recommendations. Finally, we developed and evaluated a prototype electronic nursing record system that can provide nurses with recommendations for nursing practice and sets of nursing statements based on the recommendations for guiding nursing documentation. Conclusions The prototype system was found to be sufficiently complete, relevant, useful, and applicable in terms of content, and easy to use and useful in terms of system user interface. This study has revealed the feasibility of developing such an ENR system.

Exploring use of a clinical data repository containing international classification for nursing practice-based nursing practice data.

To determine the usefulness of a clinical data repository for nursing, we conducted two studies (1) investigating the gaps between required nursing care time based on patient classification and actual nursing care time based on nurse staffing level and (2) exploring the practice variations of nurses by comparing nursing interventions documented to prevent and treat pressure ulcers. We reviewed the nursing records of 124 416 patients discharged from 2005 to 2007 to identify the gaps in nursing care time. We also reviewed records of 41 891 patients discharged in 2007 to identify those who had pressure ulcers or were at risk of pressure ulcers and analyzed the nursing interventions documented to prevent and treat pressure ulcers. The pediatric and geriatric units showed relatively high staffing needs and the trends of understaffing over time. For pressure ulcer care, nursing interventions vary by nursing unit. Position change was the most common nursing intervention documented except in the maternity unit, followed by ulcer wound care, use of devices, and nutritional assessment. This study showed that data in a clinical data repository can provide nurse managers and nurses with valuable information about nurse staffing and patient care.

Novel Approach to Inpatient Fall Risk Prediction and its Cross-site Validation using Time-Variant Data (Preprint)

Background Electronic medical records (EMRs) contain a considerable amount of information about patients. The rapid adoption of EMRs and the integration of nursing data into clinical repositories have made large quantities of clinical data available for both clinical practice and research. Objective In this study, we aimed to investigate whether readily available longitudinal EMR data including nursing records could be utilized to compute the risk of inpatient falls and to assess their accuracy compared with existing fall risk assessment tools. Methods We used 2 study cohorts from 2 tertiary hospitals, located near Seoul, South Korea, with different EMR systems. The modeling cohort included 14,307 admissions (122,179 hospital days), and the validation cohort comprised 21,172 admissions (175,592 hospital days) from each of 6 nursing units. A probabilistic Bayesian network model was used, and patient data were divided into windows with a length of 24 hours. In addition, data on existing fall risk assessment tools, nursing processes, Korean Patient Classification System groups, and medications and administration data were used as model parameters. Model evaluation metrics were averaged using 10-fold cross-validation. Results The initial model showed an error rate of 11.7% and a spherical payoff of 0.91 with a c-statistic of 0.96, which represent far superior performance compared with that for the existing fall risk assessment tool (c-statistic=0.69). The cross-site validation revealed an error rate of 4.87% and a spherical payoff of 0.96 with a c-statistic of 0.99 compared with a c-statistic of 0.65 for the existing fall risk assessment tool. The calibration curves for the model displayed more reliable results than those for the fall risk assessment tools alone. In addition, nursing intervention data showed potential contributions to reducing the variance in the fall rate as did the risk factors of individual patients. Conclusions A risk prediction model that considers longitudinal EMR data including nursing interventions can improve the ability to identify individual patients likely to fall.

Use of electronic critical care flow sheet data to predict unplanned extubation in ICUs

This study utilized critical care flow sheet data to develop prediction models for unplanned extubation. A total of 5180 patients with 5412 cases of endotracheal tube extubation treated in a tertiary care teaching hospital were evaluated. A total of 60 extubation cases were classified as unplanned, and 5352 as planned. Features documented in the critical care flow sheet for the 24 h prior to extubation were grouped into those with recording frequencies ≤3 and >3. The nearest values to the extubation were identified for all features. For features recorded >3 times, the maximum, minimum, mean, and recording frequencies were calculated. Univariate analyses were performed to select features for inclusion in multivariate analyses. Three multivariate logistic regression models were compared. Model 1 contained only the nearest value, Model 2 added a recording frequency, and Model 3 replaced the nearest value with the maximum, minimum, or mean that had the highest effect size for each feature recorded >3 times. Univariate analyses showed that 18 features differed significantly between the unplanned extubation and control groups. These included vital signs (e.g., pulse and respiration rates, body temperature), ventilator parameters (e.g., minute volume, peak pressure), and consciousness indicators (e.g., Glasgow coma scale score, Richmond agitation sedation scale score, motor power). On all three multivariate analyses, the Glasgow coma scale score, pulse rate, and peak pressure were statistically significant. The frequency of patient positioning (Model 2) and the minimum respiration rate (Model 3) were also significant. Area under the curve, sensitivity, and positive and negative predictive values improved slightly from Model 1 to Model 2 and from Model 2 to Model 3. This study found that minute volume, peak pressure, and motor power are significant risk factors for unplanned extubation that have not been previously reported. Recording frequency, which reflects how often nursing activities were provided, was also a useful predictor. The indicators identified in this study may help to predict and prevent unplanned extubation in clinical settings.