Identifying fall-related injuries: Text mining the electronic medical record

Abstract

Unintentional injury due to falls is a serious and expensive health problem among the elderly. This is especially true in the Veterans Health Administration (VHA) ambulatory care setting, where nearly 40% of the male patients are 65 or older and at risk for falls. Health service researchers and clinicians can utilize VHA administrative data to identify and explore the frequency and nature of fallrelated injuries (FRI) to aid in the implementation of clinical and prevention programs. Here we define administrative data as structured (coded) values that are generated as a result clinical services provided to veterans and stored in databases. However, the limitations of administrative data do not always allow for conclusive decision making, especially in areas where coding may be incomplete. This study utilizes data and text mining techniques to investigate if unstructured text-based information included in the electronic medical record can validate and enhance those records in the administrative data that should have been coded as fallrelated injuries. The challenges highlighted by this study include data extraction and preparation from administrative sources and the full electronic medical records, de-indentifying the data (to assure HIPAA compliance), conducting chart reviews to construct a ''gold standard'' dataset, and performing both supervised and unsupervised text mining techniques in comparison with traditional medical chart review.

Data Mining& Health Care: Techniques of Application

“…According to Monica Chiarini Tremblay et al (2009)it can be easily comprehended and applied. For this, around hundred papers from three leading journals spread over last ten years have been considered.…”

“…In a country of over 1.1 billion people, the healthcare system will have to innovate to double the utilization of its existing resources so as to reach a stage available in developing countries (Aqueel Ahmed et al 2012). According to Monica Chiarini Tremblay et al (2009) "telemedicine is one such innovative technology", and "if used effectively can double utilization of scarce human resources" (Leigh Turner et al 2007). If telemedicine models are integrated with the healthcare model, such models may become viable.…”

Aircraft interior failure pattern recognition utilizing text mining and neural networks

“…There are many interesting fields of research such as detection of similarities between patent documents and scientific publications (Magerman et al 2010); examining mobile learning trends (Hung and Zhang 2011); discovering a multi-functional metal-binding glycoprotein that exhibits many biological functions of interest to many researchers from the fields of clinical medicine, dentistry, pharmacology, veterinary medicine, nutrition and milk science (Shimazaki and Kushida 2010); identifying fall-related injuries in electronic medical record (Tremblay et al 2009); mining business policy texts for discovering process models (Li et al 2010); discovering knowledge by opinion mining from noisy text data (Dey and Haque 2009); tracking what people are saying, finding influencers, and using many social network analytic tools to analyze the underlying social networks embedded within the blogosphere (Macskassy 2011) and (Huang et al 2011) and with emails via clustering and pattern discovery (Manco et al 2008); identifying the anomaly cases for knowledge discovery from the warranty and service data in the automotive domain (Rajpathak et al 2011); discovering frequent musical patterns (motifs) that is a relevant problem in musicology (Jiménez et al 2011). In Biology, text mining has new challenges as can be seen in Dai et al (2010); a good example of text mining on language recognition can be seen in Al-Jumaily et al (2011), where Arabic, the most widely spoken language in the Arab World is identified on the web.…”

Negation scope and spelling variation for text-mining of Danish electronic patient records

“…Previously unstructured data has been used for a range of purposes such as diagnosis detection (e.g. Meyste, 2006;Suzuki, 2008;Liao, 2010), decision support (Tremblay, 2009), and temporal investigation of adverse drug reactions (Eriksson, to appear 2014). Structured EPR data will primarily contain diagnoses relevant to the current hospitalization, whereas free text will contain additional information about adverse drug reactions and the general health status of the patient.…”