We evaluate the effects of the 2006 DPC revision on the length of hospital stay for cataract surgeries (DPC category code: 020110) by the power transformation model (the Box-Cox transformation model [BC model] excluding the case in which the transformation parameter is zero). The BC model is widely used in various fields. The model is estimated by the maximum likelihood estimator (BC MLE) under the normality assumption of error terms. However, since the error terms cannot be normally distributed except the case where the transformation parameter is zero, it is not a proper estimator. In this paper, we first propose a new estimator and a test of the BC MLE for the power transformation model. We then analyze the length of hospital stay for cataract surgeries by the proposed methods. In the analysis, we utilize data of the patients who underwent cataract operations and insertion of a prosthetic lens on one eye only without secondary operations and treatments (DPC code is 020110xx97x0x0 after the 2006 revision). Data of 4,374 patients collected from 20 general hospitals before and after the 2006 DPC revision are used. The factors which may affect the length of hospital stay are analyzed by the proposed model. The variables found to affect the length of hospital stay are the gender, age and principal diseases H260 and H268. Even after eliminating the influence of patient characteristics, we find large differences among hospitals in average lengths of hospital stay. The estimate of the 2006-7 dummy is negative and significant at the 1% level. This means that the revision reduced the length of hospital stay.

Under the National Health Insurance (NHI) scheme in Japan, a municipality insurer can issue a short-term health insurance card (hereafter, short-term card) or an eligibility certification card (hereafter, certification card) to those who fall behind in their premium contributions for a year and longer without any justifiable reason, instead of a normal health insurance card (hereafter, normal card). However, short-term cards that entitle to usual 30% copayment are not often issued to the qualified beneficiaries. Certification card holders have to pay a full amount of health care costs. This policy arrangement may seriously restrict health care access of certification card holders, though few quantitative studies investigate this issue to date. This paper examines whether and in what way access to health care among short-term and certification card holders is limited relative to normal card holders by analyzing administrative claim data provided by a municipality insurer. The analysis has uncovered the following four findings:

First, when age and income are not controlled, short-term and certification card holders reduce their probability of visiting physicians respectively by 23-28% and 52-53%, compared to normal card holders.

Second, in comparison with those normal card holders who fall behind in their premium contributions, the probability of visiting physicians among short-term card holders is almost the same, and the probability among certification card holders is reduced by 27-28%.

Third, those short-term card holders who receive a large amount of medical care benefit in a previous year did not receive certification cards, suggesting that the municipality does not issue a certification card for those who have larger needs of care, even if they fall behind in their premium contributions.

Fourth, a low-income group among short-term card holders is likely to have their short-term cards replaced with certification cards; 1.0% decrease in income results in 0.6-0.7% increase int he probability of being delivered the certification card.

These results indicated that the decreased probability of visiting physicians is observed in the phase of delinquency in payment rather than in the phase of having short-term or certification cards. Furthermore, the association between the increased probability of issuing a certification card and the decreased income level of short-term card holders suggests that they actually face the liquidity constraints. The NHI premium rate for low-income groups should be carefully reviewed, as the liquidity constraint seems to cause arrears of premium payment.

This paper analyzes the one-month survival of patients with out-of-hospital cardiac arrest in Japan using the Utstein style data from April 2008 through March 2009, where the number of sample cases was 110,233. Based on the bivariate probit model, this paper derives the golden hour principle provided by Cara (1977), which describes the relationship between mortality and the time after out-of-hospital cardiac arrest, and identifies the effect of the presence of return of spontaneous circulation (ROSC). Also, the golden hour principle is drawn to each cause of cardiac arrest. By taking the presence of ROSC into account, the prediction of one-month survival improves well, compared to cases with no ROSC variable. Furthermore, the effect of defibrillation by citizens and the ambulance crew is evaluated with/without witnesses. It concludes that the effect of defibrillation by citizens is about twice than by the ambulance crew, since the timing of defibrillation by citizens is earlier than by the ambulance crew.