Management of unruptured intracranial aneurysm in Japan: a Markovian decision analysis with utility measurements based on the Glasgow Outcome Scale.

The purpose of this study was to analyze the management of individual patients with unruptured intracranial aneurysms (UN-ANs) using a decision-analytic approach. Transition probabilities among Glasgow Outcome Scale (GOS) categories were estimated from the published literature and data from patients who had been treated at Kitasato University Hospital. Utilities were obtained from 140 health providers based principally on the GOS. Baseline analysis for a healthy 40-year-old man with an anterior UN-AN less than 10 mm in diameter showed that the quality-adjusted life expectancies for preventive operation and follow-up were 15.34 and 14.66 years, respectively. For a follow-up strategy to be preferred, the annual rupture rate had to be as low as 0.9%. These results were sustained through extensive sensitivity analysis. The results support preventive operation for UN-ANs, and identify problems that can be clarified with a well-designed stratified clinical trial.

Utility and treatment decisions: 15 clinical cases in Japan.

The purpose of this study was to investigate the applicability of utility concepts to treatment decision making in different clinical situations. In each of the 15 independent cases studied, subjective utility values were obtained directly from the patients before the treatment decisions were made. The optimal decision based on the utility values was then calculated. The actual treatment decisions were made without any input or influence by this study. The data were obtained at two large public hospitals and two privately owned multi-service hospitals in Kobe, Japan. Active participation by patients in treatment decision making seems particularly useful when two or more treatments are approximately equivalent in terms of outcome. A careful assessment of an individual patient's utilities can help the patient choose a treatment option meeting individual goals. Patient autonomy must be strengthened by giving the patients realistic and unbiased information of the alternatives available and the possible outcomes.

[Imipenem/cilastatin sodium and other beta-lactams for respiratory tract infections: clinical benefit and treatment days for cure].

Therapeutic efficacy and the treatment days for cure of imipenem/cilastatin sodium (IPM/CS) in treatment of pulmonary infections were prospectively determined in comparison with those of beta-lactams other than carbapenems mainly ceftazidime (CAZ) or sulbactam/cefoperazone (SBT/CPZ). The overall response rate was 84.9% (62/73) in the IPM/CS group and 74.7% (56/75) in the beta-lactam group, the difference not being significant. In the subjects having underlying respiratory diseases, the response rate was 91.1% (41/45) and 73.9% (34/46) in the IPM/CS and beta-lactam groups, respectively. In patients with infections secondary to chronic respiratory disease, the rate was 91.2% (31/34) in the former group and 66.7% (24/36) in the latter group, respectively. The differences were significant for both stratified analyses. The treatment days for cure judged by the attending physician were 12.9 +/- 0.6 days in the IPM/CS group, and 14.5 +/- 0.7 days in the beta-lactam group. The difference was not, however, significant. In patients with mild to moderate infections, the treatment days for cure was 12.0 +/- 0.6 days (n = 64) in the IPM/CS group and 14.3 +/- 0.7 days (n = 70) in the beta-lactam group. In patients with underlying respiratory disease, the treatment days for cure were 11.8 +/- 0.7 days (n = 45) and 14.7 +/- 0.9 days (n = 46) in the IPM/CS and beta-lactam groups, respectively. In patients with infections secondary to chronic respiratory disease, the days were 11.1 +/- 0.7 days (n = 34) and 14.7 +/- 1.1 days (n = 36), respectively. Thus, IPM/CS therapy significantly reduced the number of treatment days until cure. There was, however, no significant difference between the two therapy groups in treatment of the patients with severe infections, those without underlying respiratory disease, or those with pneumonia and/or lung abscess. The treatment days for cure were also assessed by the members of review committee taking into consideration of body temperature, leukocyte count, and C-reactive protein. As the result, it was 6.9 +/- 0.5 days in the IPM/ CS and 10.3 +/- 0.7 days in the beta-lactam groups; respectively, and the difference was significant. Time (days) until cure was also compared between the two groups using survival time analysis, confirming a more rapid response in the IPM/CS group. Although IPM/CS therapy was associated with a shorter response time as assessed by both the attending physicians and the review committee, there were considerable differences between the results of these judgements. Thus, the duration of treatment with injectable antibiotics requires reevaluation in the future. No significant differences were observed between the groups with respect to parameters indicating side effects and laboratory abnormalities. There were no severe symptoms or laboratory findings, and symptoms and changes in laboratory values, if any resolved during the course of therapy or after the withdrawal of treatment. In conclusion, IPM/CS seems to be very useful as first-line therapy for respiratory tract infections and for shortening the duration of treatment.

[Sensitivity and specificity in AIDS screening system].

The interest in the optimal combination of the diagnostic tests is growing in terms of reasoning rationale and cost-effectiveness of the diagnostic system. As an example of the two-stage test system, we investigated the sensitivity, specificity, and posterior probabilities of the HIV screening system with incomplete information of the conditional independency between the enzyme immunoassay and the Western Blot. The theoretical analysis of the probabilistic reasoning of the HIV screening system could provide us with the upper and lower limits of the predictive positive value, given the prior probability of HIV infection. Consequently, we could show that the lower and the upper boundaries of the predictive positive value are, respectively, 4.07% and 100% for the low risk group (prevalence of HIV: 0.01%), while 99.86% and 100% for the high risk group (prevalence of HIV: 62%). These analyses imply that clinicians must be very careful in the interpretation of the positive test result in the HIV screening system, depending on how the risk of HIV infection of the patient is estimated in advance. In addition to the probabilistic analyses of the test parameters, the approach by use of the utility theory is favorable for the issue of optimizing the diagnostic test sequences. It remains for future investigation.

A computational model of approximate Bayesian inference for associating clinical algorithms with decision analyses.

The lack of rationale or explanation is a major deficiency of clinical algorithms. To address this issue, the authors present a computational model for associating decision analyses with clinical algorithms. Automata theory is used to model categorical reasoning with approximate Bayesian inference based on probability intervals. This approximation reduces the number of computations to linear-order instead of the exponential-order combinations of clinical findings in exact Bayes. The linkage of decision analyses and clinical algorithms by means of this model exploits a new concept of "regular" clinical algorithms and their equivalency in theory and provides valuable perspectives in practice for developers of clinical algorithms.

Education on medical informatics integrated in the campus information network system at Shimane Medical University.

An integrated campus information network system at Shimane Medical University has been developed to organize medical information generated from each section and provide information services useful for education, research, and clinical practice. This report outlines: the education-research system in connection with a campus information network system, the MUMPS programming self-directed learning software, and the curriculum of education on medical informatics.

The frequencies of disease names with the natural language used in the hospital information system.

The statistical behavior of disease names referred by physicians with the natural language in a large hospital information system is little known despite the theoretical and practical interest. To address this issue, we reviewed and investigated the usage-frequencies of 18,274 disease names, 10,288 for outpatient care and 7986 for inpatient care, referred from October 1983 to June 1992 with the notation of the natural language in Japanese by the use of the registration-retrieval system of disease names at Fukui Medical School, Japan. Consequently, we found that the investigated distributions did not conform to the Poisson distribution, but conformed well to the Polya-Eggenberger distribution in both cases of outpatient and inpatient care. It implies that the disease names with the natural language are possibly referred by physicians with some interrelations.

Obesity among school children in a province of southern Thailand and its association with socioeconomic status.

The association of nutrition status of children aged 7-12 years (n=663) with socioeconomic factors in a province of southern Thailand in 1995 was investigated. Three type of schools were surveyed: a school with a higher educational standard (elite school) in the municipality of the province, a school with many children from low-income families (low-income school) in the same municipality, and five ordinary schools in rural areas of the province (district schools). The proportions of obese children were 22.1%, 5.8% and 2.7%, respectively for the three type of schools, when obesity was defined as weight to height of over 120% of the median of children in Bangkok. The risk ratios and 95% confidence intervals for obesity in the elite and the low-income schools were 5.0 (3.5-7.2) and 1.9 (0.8-4.8), respectively, taking the district schools as a reference. Our research suggested that the high prevalence of obesity among elite-school children could be related to the comparatively high socioeconomic status of the children's families. It also shows that the children in the province studied were as a whole considerably leaner than children in the big cities of Thailand. These results imply a need for appropriate interventions which cannot only prevent obesity, but also improve the malnutrition of school children in the rural provinces of southern Thailand.