In the field of medical information, the amount of text information generated is increasing day by day, and the realization of automatic processing by computers is an urgent issue. However, medical texts, such as medical records generated in hospitals and medical diaries posted by patients on social networking services, are difficult to handle through natural language processing because of the use of special styles and expressions.

　The Kanagawa Cancer Center keeps text data records of the cancer telephone counseling service that was previously provided. However, these text data are unsuitable for natural language processing because they are described using special expressions, like other medical texts. Therefore, in this study, we first conducted a qualitative analysis, and then used text mining to extract and visualize the main complaint of the consultant.

　First, based on the results of the qualitative analysis, the chief complaints “worry,” “anxiety,” “request,” “distrust and mistrust,” and “dissatisfaction” that appeared cross-sectionally were selected as the targets of text mining. Next, the object keywords of these chief complaints were extracted using extended Backus-Naur form and visualized using a graph line drawing tool.

　The extraction results were subjected to a performance evaluation using the F-measure. As a result, the F-measure for “worry,” “request,” “distrust and mistrust,” and “dissatisfaction” all exceeded 0.7, with the F-measure of “request” reaching approximately 0.8. In addition, this method was also confirmed to improve performance in comparative experiments with general text mining methods. On the contrary, the F-measure of “anxiety” was approximately 0.62 because the object keywords were frequently ambiguous. Dealing with ambiguity is a work for the future.

　These results demonstrate the effectiveness of our method, and the findings of this study may be useful for processing other medical texts.

　Effective utilization of nursing human resources requires practical quantification tools on a temporal scale. However, a simple and sustainable system for quantifying nursing on a temporal scale has not yet been developed. Therefore, we developed a system to temporally quantify the per-patient nursing burden. This system is simple in that it measures bedside stay time. Its characteristic feature is that it combines a radio frequency proximity sensor embedded in a beacon and an electronic tag with a patient authentication function. Regarding the proximity sensor, the following measurement accuracy issue arose: Because the sensor relies exclusively on radio frequency communication, radio frequency interference from nearby beacons was encountered. However, authentication using an electronic tag improved the measurement accuracy. Evaluation of the system suggested that the combined use of the electronic tag and radio frequency proximity ameliorated the radio frequency interference problem, facilitating simplified measurement of bedside stay time.

　Since efficient bed assignment is one of the most important issues of hospital administration, various studies within this issue analyze a clinical indicator using a large size database. This study aims to develop a method of matrix representation of a daily bed occupancy to improve the quality of data analysis regardless of computerization of medical information. This matrix is known as a hospitalization matrix （HM） that can be created using only the date of hospitalization and discharge. Its number of rows and columns indicates the number of staffed beds and observational days, and its component is a non-negative integer （zero means bed vacancy and otherwise means bed occupancy）. HM can easily calculate a periodic average value of a bed occupancy rate （＝sum of corresponding columns/｛number of rows * observational days｝）. Therefore, HM can calculate various types of BOR such as a moving average, and a floor difference. Moreover, HM can create a secondary dataset for data analysis using a reusable script with a low implementation cost.

　Task shifting/sharing plays an important role in promoting workstyle reform for physicians, and the smooth and reliable process from the delegate’s proxy input to the physician approval in the electronic medical record system holds the key to it. Approval is an absolute requirement for proxy input; thus, increased numbers of unapproved entries will hamper task shifting/sharing. Delegates’ unapproved proxy input is alerted on the electronic medical record system, but it is not effective. Therefore, improvements in the electronic medical record systems and their operational methods are necessary to promote workstyle reform. In this report, we introduce a new notification system for unapproved proxy input and show positive results in significantly reducing the number of unapproved entries and promoting task shifting/sharing in disease name input.