An exploratory method of data retrieval from the electronic medical record for the evaluation of quality in healthcare

Health care organizations are under increasing demands to demonstrate the level of quality in the services they provide from federal, state, regulatory agencies and third party payers. Most healthcare organizations are ill equipped to handle these demands due to the costs in time and manpower needed to retrieve data from health care databases.; The medical record is considered a 'gold standard' of clinical databases in health care due to the breadth and depth of clinically relevant data. Data from the electronic medical record (EMR) are often required to demonstrate quality of health care services, although these data are difficult to retrieve without incurring the costs related to the traditional method of manual data retrieval.; Natural language processing (NLP) presents a framework for understanding and retrieving narrative data from the medical record. Based on concepts of semantics and syntax of language, this framework provides a means for organizing data retrieval using an automated software technology.; The study method incorporated principles of NLP to guide the automated retrieval of quality of care data from electronic medical records. The purpose of the study was to evaluate the efficiency and effectiveness of the study NLP method, as well as test the reliability and validity of the retrieved data compared to manual retrieval of the same data.; The study NLP method used a qualitative software program (Nudist) to retrieve and code narrative quality of care data from the EMR. The study NLP method performed with acceptable reliability (.73) when the narrative variables were documented in a consistent format in the EMR. The format of the EMR and consistent narrative format of the documented variables were key factors related to accurate data retrieval using the study NLP system. Although the study NLP method did not perform as well as manual data retrieval in terms of efficiency and accuracy, subsequent trials with the study NLP method data may demonstrate levels of efficiency and reliability to rival manual data retrieval methods. This study demonstrated the potential usefulness of NLP in the retrieval of quality of care data from the EMR.