Hospitals promote education for healthcare safety. Kiken Yochi Training (KYT) is introduced in order to improve the ability to identify hazards. KYT sheets, which describe working conditions, are necessary during the practice of KYT. However, it is difficult to utilize incident information when medical staffs create these sheets. Moreover, the means by which the ability to identify hazards can be evaluated has not been established. This paper proposes a method for designing KYT sheets. It describes an analysis method of medication incidents when medical staffs create KYT sheets. This method enables the evaluation of the ability to identify hazards since the KYT sheets developed using this method contain answers. KYT was executed twice by 57 nurses with the KYT sheets designed according to the proposed method. Consequently, the average number of answers found increased in the second attempt. Therefore, it is confirmed that staffs improve their ability to identify hazards.

View full abstract

For hospitals, personnel allocation is a critical issue because the quality of healthcare depends on the competence of the personnel. Personnel allocation at hospitals must take into account the diversity of patients' needs, the complexity of the services delivered, and the variation in specialized skills proficiency among personnel. A methodology taking these factors into account has however not been developed yet. The purpose of this study is to develop a model method for personnel allocation that reflects the specific features of hospital jobs. For this purpose, we identified determining necessary competences, determining possessed competences, and deriving allocation patterns as the three core aspects of personnel allocation; and incorporated hospital-specific considerations to each of these aspects. Determining necessary competences involves identifying items to be evaluated and estimating the number of required personnel, based on a job description including hospital-specific features. Determining possessed competences involves evaluating personnel competences in terms of levels reflecting skill proficiency. Deriving personnel allocation patterns involves ensuring quality in situations where possessed competences do not satisfy the necessary competences. Finally, based on real world application, we conclude that our model allows for more adequate patterns of personnel allocation at hospitals than existing approaches.

View full abstract

Detection of abnormal items is discussed when both continuous and categorical variables are observed and parameters are estimated. Assuming the location model, two types of detection methods are considered. One is estimative method and the other is testing method. Some properties are derived concerning the conditional powers given the level of a categorical variable for a testing sample and the frequency distribution of a categorical variable based on the initial data as well. For comparing the methods, their expected error rates for normal items and detection powers are numerically evaluated when one dichotomous variable is included. Because expected error rates cannot be set equal to the nominal value, odds ratio of power relative to expected error rate is used for comparison. It is observed that estimative method based on the likelihood ratio test and testing method are superior to the other estimative methods in the sense that these two methods give larger values of odds ratios for the wide region of interest. Testing method is the best so far as the smaller probability between two levels of a dichotomous variable does not get much larger for abnormal items.

View full abstract

Process capability indices (PCI) are widely used in the activities of quality control. This paper discusses the improvement on the natural estimators for the basic PCI's, C_p, C_pL, C_pU and C_pK.
Under the squared error loss function, Nagata (1995) showed that the natural estimator for C_p can be uniformly improved on by the estimator which is multiplied by some constant, and gave the degrees of improvement numerically. He also suggested by numerical calculations that the natural estimator for C_pK cannot be improved on by the simple estimator which is multiplied by some constant.
In this paper it is shown that the natural estimator for C_pU and C_pL can be uniformly improved on by the estimators which are multiplied by some constant, and the degrees of improvement are given numerically. Furthermore, it is mathematically shown that the natural estimator for C_pK cannot be improved on by any simple estimators which are multiplied by some constant when the sample size is larger than or equal to 7.

View full abstract