The financial burden of healthcare expenditures has been increasing and it is becoming more and more important for local governments and insurers to conduct health expenditure analyses.
We conducted an ecological study to define factors, which are associated with specific disease medical costs per insured person for inpatients and outpatients using claim data of the national health insurance of 53 municipalities in Tokyo from May 2006.
A model was developed using the proportion of elderly seniors over 75years, gender, supply of medical service, income and educational level as possible factors for predicting healthcare costs. The targeted diseases were 4 lifestyle related diseases (diabetes, hypertension, ischemic heart disease, cerebrovascular disease), 5 cancers (stomach, colon, rectum, liver, lung) and 2 mental disorders (dementia, mood disorder). We selected one principal diagnosis for each claim.
Our analyses suggest that the factors related to medical costs of specific disease are different by disease. The proportion of patients over 75 years was most often related to medical costs for inpatients and outpatients. Furthermore gender or educational levels were seldom related to medical costs. Disease specific medical costs for outpatients were related significantly to the proportion of patients over 75 years, male gender and the high school entrance rate. Multiple regression analyses suggest that the proportion of patients over 75 years was related to 5 diseases for inpatients and 8 diseases for outpatients. Coefficients of determination after adjustment of the degree of freedom were 0.165-0.326 for inpatients and 0.141-0.434 for outpatients.
The results of this study suggest that predicting factors for medical costs are different for every disease and clinical setting and this should be taken into account when developing models for economic analyses.
View full abstract