Linear equations with random variables by Toshiro Tango, Statistics in Medicine 2005; 24:3213-3222.
Explore the source record for details and available documents.
Adolescent↗
Biomedical subjects
Etsuji Okamoto
Publications and source records attributed to Etsuji Okamoto.
Analysis of administrative sanctions and criminal prosecutions of doctors in Japan.
OBJECTIVES: To illustrate how administrative sanctions (AS) and criminal prosecution (CP) differ with regard to application with doctors' misconducts. METHODS: A total 465 doctors who were punished twice by AS and CP were analyzed using the proportional distribution method (PDM) to break down into the charge-specific months of suspension or imprisonment. RESULTS: Overall, the Minister of Health, Labor & Welfare (MHLW) sanctioned doctors by suspending their licenses for twice the number of months that the court ordered for imprisonment. Charge-specific analysis of months (suspension or prison terms) revealed a different pattern of judgment. The MHLW judged obscenity more unethical, allocating a larger share of the total months of suspension to punish this misconduct, but judged bribery less unethical allocating a smaller share of the total months of suspension to punish this than the court. For traditional crimes like swindling, murder and psychostimulant abuse, both judgments followed similar patterns allocating the same share of months for punishment of such acts. DISCUSSION: CP and AS were shown to have different patterns in their judgments of doctors' crimes or misconducts reflecting the different purposes they pursue: justice by CP and ethics by AS. (186 words).
Criminal Law↗
[Encryption technique for linkable anonymizing].
Linkage of different records such as health insurance claims or medical records for the purpose of cohort studies or cancer registration usually requires matching with personal names and other personally identifiable data. The present study was conducted to examine the possibility of performing such privacy-sensitive procedures in a "linkable anonymizing" manner using encryption. While bidirectional communication entails encryption and deciphering, necessitating both senders and receivers sharing a common secret "key", record linkage entails only encryption and not deciphering because researchers do not need to know the identity of the linked person. This unidirectional nature relieves researchers from the historical problem of "key sharing" and enables data holders such as municipal governments and insurers to encrypt personal names in a relatively easy manner. The author demonstrates an encryption technique using readily available spread-sheet software, Microsoft Excel in a step-by-step fashion. Encoding Chinese characters into the numeric JIS codes and replacing the codes with a randomly assigned case-sensitive alphabet, all names of Japanese nationals will be encrypted into gibberish strings of alphabet, which can not be deciphered without the secret key. Data holders are able to release personal data without sacrificing privacy, even when accidental leakage occurs and researchers are still able to link records of the same name because encrypted texts, although gibberish, are unique to each name. Such a technical assurance of privacy protection is expected to satisfy the Privacy Protection Act or the Ethical Guidelines for Epidemiological Research and enhance public health research. Traditional encryption techniques, however, cannot be applied to cancer or stroke registration, because the registrar receives reports from numerous unspecified senders. The new public key encryption technique will enable disease registry in a linkable anonymizing manner. However various technical problems such as complexity, difficulties in registrar inquiries and risk of code-breaking make the encryption technique unsuitable for disease registry in the foreseeable future.
Computer Security↗
Estimation of disease-specific costs in health insurance claims: a comparison of three methods.
OBJECTIVE: To compare the accuracy and validity of three different methods (Proportional Disease Magnitude method [PDM] with two different magnitude estimations: arithmetic means with correction by the authors; Proportional Allotment Estimator [PAE] by Tango; Maximum Likelihood Estimator [MLE] also by Tango) for estimating disease-specific costs in health insurance claims. METHODS: Application of the three methods to a computer-generated simulation dataset whose disease-specific costs were known and to actual outpatient claims whose disease-specific costs were unknown. OUTCOME MEASURES: For simulation data, the accuracy was assessed by correlation between known disease-specific costs and estimated disease-specific costs by the three methods. For actual claims, concurrent validity was assessed by inter-method correlations between pairs of the two methods. RESULTS: All three methods showed good agreement and accuracy with the simulation data but marked disagreement when they applied to actual claims. MLE yielded an aggregate total of disease-specific costs exceeding the actual total by 21.3% and showed negative disease-specific costs in 18 out of 154 categories. Inter-method correlations showed that PDM with PAE and MLE correlated most strongly (R2 = 0.9022) while the least correlation was observed for PDM with arithmetic means and MLE (R2 = 0.6861). CONCLUSION: MLE is not usable for claims analysis but PDM yielded good estimates with two different methods of magnitude estimation using actual claims.
Insurance Claim Reporting↗
[Algorithm for application of the "ethical guidelines for epidemiological research" and taxonomy of public health research].
BACKGROUND: "Ethical Guidelines for Epidemiological Research" took effect in July 2002, with a moral duty of all researchers to comply when conducting epidemiological studies although it is not legally binding. Public health research entails various forms of studies including not only epidemiological studies but also attention to psychological, societal and economic aspects, which are outside of the jurisdiction of the guidelines. Hence, confusion may arise among members of Japanese Society of Public Health as to whether the study they conduct falls within the definition of epidemiological research. The author discusses legal interpretations of the guidelines arising in the course of translation work as part of government-funded project, "Dissemination of the 'Ethical Guidelines for Epidemiological Research' via Internet (principal investigator: Toru Doi)" and argues that a case-method approach would be best suited to enhance understanding by researchers with diverse, non-legal backgrounds. METHODS: The author proposes an algorithm for classification of studies as to whether the guideline applies, and applies it to all original articles published in the Japanese Journal of Public Health (JJPH) in one year (March 2002 thru February 2003). The rationale for classification is discussed from the strict legal viewpoint in each case. RESULTS: Sixteen out of 46 original articles published in JJPH for one year were classified as epidemiological studies to which the guidelines apply. Those classified otherwise were psychological studies (10), epidemiological studies not targeting specific diseases and are exempt form the guidelines (3), purely methodological studies (4), economics studies (3), fact-finding or opinion surveys with no hypothesis testing (2), as well as studies authorized by law (4) or using unlinkable anonymous data only (4), all of which are exempt from the guidelines. Reference to ethical considerations in the methodology section as required by the instructions for authors was generally well performed in epidemiological studies although some shortcomings were noted. DISCUSSIONS: The guidelines affect approximately a third of original articles published in the JJPH. The ratio of applicable to non-applicable articles was roughly 1 to 2. This gives both authors and reviewers confusion as to whether articles are subject to the guidelines and would require ethical approval. This case-method approach using actual articles readily available to members should help unnecessary confusion and enhance appropriate application of the guidelines. (366 words).
Algorithms↗
[Estimation of disease-specific costs in a dataset of health insurance claims and its validation using simulation data].
PURPOSES: To estimate disease-specific costs in a dataset of health insurance claims with multiple diagnoses with known aggregate cost per claim and unknown disease-specific cost of each diagnosis using PDM (Proportional Disease Magnitude) method, validate its accuracy using simulation data with Monte Carlo method and improve its accuracy by developing an adjustment formula. METHODS: Developed simulation data with pre-assigned disease-specific costs, applied PDM method using arithmetic means of per-diem-per-disease cost as magnitude, validated its accuracy by observing the correlation between estimates by PDM method and known disease-specific costs and formulated an adjustment formula to improve accuracy. The reproducibility of the findings was assessed using Monte Carlo method by repeating the same procedures. RESULTS: The observed arithmetic means of per-diem-per-disease cost did not match well with actual values resulting in unsatisfactory accuracy. However, when the observed means were adjusted with a formula in which the observed mean is multiplied by (observed mean/overall mean) in the power of 2, PDM method yielded an accurate estimate of disease-specific cost. The accuracy was reproduced by Monte Carlo method with 0.9 or above R square value and slope of regression line in 76, 56 out of 100 iterations respectively. CONCLUSIONS: PDM method proved to be an objective, reproducible and accurate method for estimation of disease-specific costs of health insurance claims.
Insurance Claim Reporting↗