Automatic summarization of risk factors preceding disease progression an insight-driven healthcare service case study on using medical records of diabetic patients

Abstract

In this study we consider the problem of how to derive insight from medical records to define and improve healthcare services. As noted in many guidelines, risk factors are important to determining the care plan of chronic disease patients, e.g., pre-diabetic or diabetic patients who have started on hemoglobin A1c (HbA1c) control medications. Whereas the traditional management of chronic disease relies on a predetermined set of risk factors, without regard to patient-specific status, literature and recently released guidelines have suggested a less-prescriptive approach that allows flexibility in disease management plans to account for patient-centric information shown in medical records. However, methods of systematically summarizing medical records into risk factors have not been evaluated to support such a patient-centric focus in healthcare services. In this study, we evaluated automatic methods that can identify risk factors important for classifying Diabetic patients at risk of worsen disease progression. In particular, we used the prescription of cardiovascular disease (CVD) medication as the indicator of CVD co-morbidity development in Diabetic patients. We evaluated the summaries obtained with different sources of health information on the risk stratification task and examined the quality of the generated summaries using various proposed intrinsic metrics. In addition, we evaluated to what extent we can reduce the whole medical records into a small set of risk factors. The evaluation illustrates the potential of risk factor summarization and hints on how it can be used to enable practitioners in care planning and to support complex followup services at both the point of care and the extended care settings.

Patient Information Summarization in Clinical Settings: Scoping Review

“…Moreover, the queries are formulated in English; therefore, results not in English and containing non-English terms might be missed if their abstract was not translated or if they do not appear in the citation list or references in the retrieved articles. Finally, there were potentially relevant records [ 149 , 180 , 181 ], where the full text could not be read and analyzed as it was not available at the time of writing the manuscript.…”

Service design in the healthcare space with a special focus on non-clinical service departments: A synthesis and future directions

“…Yet another study spoke about utilising evidence based and experience-based models so as to create a system wherein one can predict the risks associated with the progression of diseases and thus creating a healthcare environment that is insight based and evidence based, which makes the diagnosis accurate. 16 Another very interesting study pointed out the effectiveness of designing for safety, the study was conducted for the NHS using journey mapping and the process was focussed on understanding which interface framework were the most successful. 17…”

A probabilistic topic model for clinical risk stratification from electronic health records

“…With the widely adoption of electronic health record (EHR) in healthcare organizations, more advanced machine learning and data mining algorithms were introduced into risk stratification [6,13,19,20]. EHR typically contains a diverse set of information types, including patient demographics, symptoms, vital signs, laboratory tests and treatments, etc., which provides a comprehensive source for risk stratification.…”