An EMR should support the informational needs of nursing practice. However, a multidimensional measure of the actual use of an EMR in hospitals ranked at different adoption stages revealed significant results that should be addressed to enable nurses to bring their full contribution to their patients and to the healthcare team.
Satish M. Mahajan, Amey S. Mahajan, Sahand Negahban
245 - 249
Many researchers are working toward the goal of data-driven care by predicting the risk of 30-day readmissions for patients with heart failure. Most published predictive models have used only patient level data from either single-center studies or secondary data analysis of randomized control trials. This study describes a hierarchical model that captures regional differences in addition to patient-level data from 1778 unique patients across 31 geographically distributed hospitals from one health system. The model was developed using Bayesian techniques operating on a large set of predictors. It provided Area Under Curve (AUC) of 0.64 for the validation cohort. We confirmed that the regional differences indeed exist in the observed data and verified that our model was able to capture the regional variances in predicting the risk of 30-day readmission for patients in our cohort.
Satish M. Mahajan, Amey S. Mahajan, Robert King, Sahand Negahban
250 - 255
Decades-long research efforts have shown that Heart Failure (HF) is the most expensive diagnosis for hospitalizations and the most frequent diagnosis for 30-day readmissions. If risk stratification for readmission of HF patients could be carried out at the time of discharge from the index hospitalization, corresponding appropriate post-discharge interventions could be arranged to avoid potential readmission. We, therefore, sought to explore and compare two newer machine learning methods of risk prediction using 56 predictors from electronic health records data of 1778 unique HF patients from 31 hospitals across the United States. We used two approaches boosted trees and spike-and-slab regression for analysis and found that boosted trees provided better predictive results (AUC: 0.719) as compared to spike-and-slab regression (AUC: 0.621) in our dataset.
Alexandra González Aguña, Virginia Díaz Teruel, Adrián Satamaría Pérez, Jorge Luis Gómez González, Ma Lourdes Jiménez Rodríguez, José Ma Santamaría García, Sylvia Claudine Ramírez Sánchez, Niurka Vialart Vidal, Daniel Flavio Condor Camara
256 - 260
The diagnosis of care problems is a complex process that involves many variables and inferences. This competence begins to acquire during the student stage, but matures later. The qualified professional continues to settle and perfect this judgment ability. Expert Systems are technologies that can help in making decisions such as diagnosis. The objective of this study is to build an Expert System in order to help in diagnosis of care problems by means of taxonomic triangulation technique. The deductive method follows three phases that result in the representation of expert knowledge through an associative network, the construction of a verified and validated prototype and, finally, the design of an app through a document of requirements specification in IEEE standard.
Value is defined as outcomes/cost of care . Although nurses contribute significantly within the interdisciplinary care team, we struggle to measure the value of nursing care or the added value of each nurse caring for a patient. This presentation reveals findings of the Nursing Value Workgroup (Workgroup), a multi-year component of the Big Data and Nursing Knowledge Initiative . Panelists will discuss: Construction of a common model to provide a basis for developing nursing business intelligence and analytics; development and publication of definitions and metrics for nursing value; creation of user stories to measure nurse sensitive phenomena; and establishment of a data warehouse to facilitate research. Initial findings from a multi-hospital study on pain management and value of care will be presented.
This presentation includes action proposals with impact to growing older populations and shows how a plan for developing elderly service system was created as one part of a bigger welfare plan at one hospital district in Finland. The central part of preparing a plan was the public administration anonymous Big Data concerning ageing and functional capacity. The aim was to establish a customer-oriented plan based on the need for population service.
The study demonstrated an application of machine learning techniques in building a depression prediction model. We used the NSHAP II data (3,377 subjects and 261 variables) and built the models using a logistic regression with and without L1 regularization. Depression prediction rates ranged 58.33% to 90.48% and 83.33% to 90.44% in the model with and without L1 regularization, respectively. The moderate to high prediction rates imply that the machine learning algorithms built the prediction models successfully.
A dissertation project at the Witten/Herdecke University  is investigating which (nursing sensitive) patient characteristics are suitable for predicting a higher or lower degree of nursing workload.
For this research project four predictive modelling methods were selected. In a first step, SUPPORT VECTOR MACHINE, RANDOM FOREST, and GRADIENT BOOSTING were used to identify potential predictors from the nursing sensitive patient characteristics. The results were compared via FEATURE IMPORTANCE. To predict nursing workload the predictors identified in step 1 were modelled using MULTINOMIAL LOGISTIC REGRESSION. First results from the data mining process will be presented. A prognostic determination of nursing workload can be used not only as a basis for human resource planning in hospital, but also to respond to health policy issues.
We were challenged to design an obeservation support system. To improve evidence-based observations, we anylysed nursing records from electronic medical records (EMR) and patient experiences (from blogs) regarding pain symptoms using text mining methods. As a result, it was found that the view point of pain differed between the patient and nurse. It is reccomended that an observation advice message should be inplemented into EMR to improve nurses' observations on pain management.
Electronic health record (EHR) systems have been used widely in research. However, most of the EHRs are highly dimensional and it is challenging to analyze such large data set. Bioinformatics is an interdisciplinary science with a focus on data management and interpretation for complex biological phenomena. We investigated biomarkers of nutrition from 3001 patients. Multivariate-adjusted hazard ratios of mortality were calculated according to both albumin and sodium levels. We explore the association of aging predicted by all-cause mortality in future.
All the doctoral dissertations and master's theses of “nursing” in CNKI database were retrieved to explore the hot spots on nursing research field in China, we processed, visualized the data above, and analyzed those data by using co-word analysis informetrics method. Then, the figure for co-occurrence analysis of high-frequency keywords in dissertation on nursing research were plotted, which represented nine topics and could help health staff to understand hot topics in nursing research field.
José Ma Santamaría García, Ma Lourdes Jiménez Rodríguez, Jorge Luis Gómez González, Marta Fernández Batalla, Alexandra González Aguña, Sara Herrero Jaén, Adriana Cercas Duque, Sylvia Claudine Ramírez Sánchez, Estela Melguizo Herrera, Daniel Flavio Condor Camara
271 - 271
The axioms of care postulate energy and time as underlying entities that allow us to study the dimensions of the person as care: the sorge. Logic allows to formalize the deductive study of sorge knowledge in order to determine what are the dimensions of care. These dimensions will be the guiding framework in the analysis of big data contained in health information systems.