Tracking medication information across medical records

A patient's electronic medical record can consist of a large number of reports, especially for an elderly patient or for one affected by a chronic disease. It can thus be cumbersome for a physician to go through all of the reports to understand the patient's complete medical history. This paper describes work in progress towards tracking medications and their dosages through the course of a patient's medical history. 923 reports associated with 11 patients were obtained from a university hospital. Drug names were identified using a dictionary look-up approach. Dosages corresponding to these drugs were determined using regular expressions. The state of a drug (ON, OFF), which determines whether or not the drug was being taken, was identified using a support vector machine with features based on expert knowledge. Results were promising: prec. approximately recall approximately 87%. The output is a timeline display of the drugs which the patient has been taking.

[1]  Peter J. Haug,et al.  Bmc Medical Informatics and Decision Making Automation of a Problem List Using Natural Language Processing , 2005 .

[2]  Kevin B. Johnson,et al.  Personal health records: evaluation of functionality and utility. , 2002, Journal of the American Medical Informatics Association : JAMIA.

[3]  Krzysztof J. Cios,et al.  Uniqueness of medical data mining , 2002, Artif. Intell. Medicine.

[4]  Anil K. Jain,et al.  Feature Selection: Evaluation, Application, and Small Sample Performance , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  M. D. Kalmeijer,et al.  Implementation of a computerized physician medication order entry system at the Academic Medical Centre in Amsterdam , 2004, Pharmacy World and Science.

[6]  Anne F. Kittler,et al.  A cost-benefit analysis of electronic medical records in primary care. , 2003, The American journal of medicine.

[7]  Carol Friedman,et al.  Extracting Information on Pneumonia in Infants Using Natural Language Processing of Radiology Reports , 2003, BioNLP@ACL.

[8]  Claire Gardent,et al.  Improving Machine Learning Approaches to Coreference Resolution , 2002, ACL.

[9]  Trisha Greenhalgh,et al.  How to read a paper: The Medline database , 1997, BMJ.

[10]  B L Humphreys,et al.  The UMLS project: making the conceptual connection between users and the information they need. , 1993, Bulletin of the Medical Library Association.

[11]  David W. Bates,et al.  Position Paper: A Proposal for Electronic Medical Records in U.S. Primary Care , 2003, J. Am. Medical Informatics Assoc..

[12]  I. Jolliffe.  Principal Component Analysis , 2002 .

[13]  D A Evans,et al.  Automating concept identification in the electronic medical record: an experiment in extracting dosage information. , 1996, Proceedings : a conference of the American Medical Informatics Association. AMIA Fall Symposium.

[14]  Roger W Bush,et al.  Reducing waste in US health care systems. , 2007, JAMA.

[15]  William W. Stead,et al.  Viewpoint Paper: Achievable Steps Toward Building a National Health Information Infrastructure in the United States , 2004, J. Am. Medical Informatics Assoc..

[16]  Carol Friedman,et al.  Facilitating Cancer Research using Natural Language Processing of Pathology Reports , 2004, MedInfo.