Measuring the effect of telecare on medical expenditures without bias using the propensity score matching method.

OBJECTIVE: This article examines the effect of telecare on medical expenditures for chronic diseases using survey data from Nishi-aizu Town, Fukushima Prefecture, Japan. The study uses the propensity score matching (PSM) method, a rigorous analytical method that overcomes sample selection bias, a common problem when using survey data. SUBJECTS AND METHODS: One hundred ninety-nine users (treatment) of telecare and 209 nonusers (control) were selected from residents, and their medical expenditures were obtained from the National Health Insurance scheme for comparison. Individual characteristics of the two groups, including age, sex, income, and health conditions, were compared, and variables that contained biases were specified by a t test. After calculation of their propensity scores and elimination of biases, the effect of telecare on medical expenditures was estimated. To obtain robust results, four different matching methods were applied: caliper matching, single nearest-neighbor matching, Epanechnikov kernel matching, and biweight kernel matching. RESULTS: No independent variable showed significant differences between the two groups after matching, indicating that selection biases were successfully eliminated using PSM. Using PSM, we saw a decrease in medical expenditures in Japanese yen of 25,538-39,936 (USD 319.23-499.20) per year per user and a decrease in the number of treatment days of 2.6-4.0 days. In comparison, our previous analyses using the same data underestimated the effects of telecare. PSM provides greater effects by reducing bias. CONCLUSIONS: Using PSM to compare subjects in two groups with similar characteristics except for their use or nonuse of telecare, we demonstrated that the treatment group has lower medical expenditures for chronic diseases than the control group. Proper matching is important in evaluating the impact of telecare interventions. Limitations of PSM include its requirement for a large number of samples and the limited ability to explain why and how telemedicine produces these effects. Other empirical methods are required to identify the mechanism of how telemedicine works.

Does telecare reduce the number of treatment days?

This paper analyzes the effect of telecare in reducing the number of days required for the treatment of telecare users. Statistical analysis was conducted on panel data about 400 individuals from 2002 to 2006 in Nishi-aizu Town, Fukushima Prefecture, Japan. Three estimation models were used, namely the Instrument Variable, Instrument Variable Panel, and system GMM (Generalized Method of Moments) methods, both to demonstrate causality and to ensure the robustness of the results. We attempted to reduce sample selection bias in this framework by controlling variables with bias. A negative correlation between outpatient days and telecare use was observed with the former two methods, while the system GMM proved causality for the effect of telecare in reducing the number of outpatient days. Telecare use reduced outpatient days in telecare users by 2.0 days per year. This result is consistent with those of other telecare implementations in the US and the UK.

Effect of e-health on medical expenditures of outpatients with lifestyle-related diseases.

We analyzed the effect of e-health on medical expenditures in Nishi-aizu Town, Fukushima Prefecture, Japan, using panel data of medical expenditures for about 400 residents from 2002 to 2006. The Nishi-aizu Town system was introduced in 1994 and is still successfully operating as one of the longest running implementations of e-health in Japan. The town office maintains a register of receipts for medical expenditures paid by the National Health Insurance system and provides data on e-health users, allowing users and nonusers of e-health and their respective costs to be distinguished. Here, we focus on patients with lifestyle-related diseases such as high blood pressure, diabetes, stroke, heart failure, etc. This article postulates that e-health reduces medical expenditures via two mechanisms, decreasing travel expenses and preventing symptoms from worsening. The former implies that e-health monitoring allows patients at home to visit medical institutions less frequently, and the latter that the symptoms experienced by e-health users are less severe than those experienced by nonusers. We termed these the travel cost effect and opportunity cost effect, respectively. Chronic conditions tend not to occur singly, and many patients have more than one; for example, patients with high blood pressure or diabetes also likely have heart disease at the same time. This multiplicity of conditions hampers cost analysis. Among methodological issues, a number of recent empirical health analyses have focused on the endogenous problem of explanatory variables. Here, we solved this problem using the generalized method moments (GMM) system, which allows treatment of not only the endogenous problem of explanatory variables but also the dynamic relationship among variables, which arise due to the chronic time-lagged effect of lifestyle-related diseases on patients. We also examined a second important methodological problem related to reverse correlation between the medical expenditures of an outpatient and e-health and took sampling biases into consideration. We concluded that this control of endogeneity through system GMM confirms that the relationship between the medical expenditures of an outpatient and e-health shows causation rather than simple correlation and that e-health use, duration of e-health use, and frequency of e-health use can reduce outpatient medical expenditures for lifestyle-related diseases.

Empirical analysis of the reduction of medical expenditures by e-health.

This paper aims to examine reduction of medical expenditures by utilizing the system of Nishi-aizu Town, Fukushima Prefecture. The town office has been implementing it since 1994 and keeps receipts on medical expenditures of its approximately 4,000 residents paid by National Health Insurance for 5 years from 2002 to 2006. We select (1) users; and (2) non-users of the e-health system, and by comparing their medical expenditures, we examine: (i) difference in medical expenditures between two groups; and (ii) negative correlation between medical expenditures and the length of usage of the e-health system. We find that total medical expenditures of users are larger than those of non-users, whereas by restricting to lifestyle-related illnesses such as high blood pressure, cerebral infarction, strokes, and diabetes, medical expenditures of users are found to be smaller than those of non-users. The results we obtained here provide the rigorous economic foundation of the e-health system.

Relation between telecare implementation and number of treatment days in a Japanese town.

We analysed data on the medical expenditure of 199 telecare users in Nishi-aizu Town, Fukushima Prefecture, which has one of the oldest ongoing telecare implementations in Japan. As controls, 450 out of 3528 non-users residents covered by National Health Insurance were randomly selected in the same age and sex ratios as the telecare users. An analysis by the Generalized Method of Moments (GMM) was conducted in order to examine causality, i.e. that telecare use reduces the number of treatment days. To reduce sample selection bias, the presence of chronic diseases, age and education were added as control variables in the estimation. The results show that the treatment days of those who had chronic diseases were greater than those who did not have chronic diseases by 8.7 days per year (P < 0.10), and they were increased by 5.6 days (P < 0.01) according to their age. Finally, telecare use decreased treatment days by 3.1 days (P < 0.05).

Empirical analysis of the long-term effects of telecare use in Nishi-aizu Town, Fukushima Prefecture, Japan.

This paper examined the long-term effects of the use of telecare (e-Health) on the residents of Nishi-aizu Town, Fukushima, Japan, between 2002 and 2010. We compared medical expenditure and days of treatment between telecare users (treatment group) and non-users (control group) based on receipt data obtained from the National Health Insurance, which is operated by the government. In previous studies, we used receipt data obtained for the years 2002 to 2006; this study expands the analysis period four more years with respect to respondents who were included in previous analyses. Ninety users and 118 non-users were included in both analyses. Using rigorous statistical methods, including system generalized method of moments (GMM), this paper demonstrates that telecare users require fewer days of treatment and lower medical expenditure than non-users with respect to the chronic diseases of stroke, hypertension, heart failure, and diabetes. To date, there have been no publications examining the long-term economic effects of the use of telemedicine, so the current study presents a new facet to the research in this field.

An empirical approach to estimating the effect of e-health on medical expenditure.

We studied the e-health system used in Nishi-aizu Town in Fukushima Prefecture in Japan. The system allows elderly people at home to transmit vital signs data to the Town's health centre, where nurses provide advice based on the data. Our hypothesis was that the e-health system in Nishi-aizu Town reduced the need to visit clinics. We attempted to prove this by a regression analysis, in which days for treatment were compared between users and non-users of e-health. The results showed that days for treatment of e-health users were shorter than those of non-users by 1.6 days per year. Thus the total reduction in expenditure as a result of fewer hospital visits (emergency and elective) was about 16,000 yen per year. In a previous study, we proved that in Nishi-aizu Town the medical expenditure of e-health users was smaller than those of non-users by 15,688 yen. The results of the present study therefore coincide with those of the previous one and show that the reduction of medical expenditure is principally caused by the reduction of days for treatment.

An empirical analysis of the reduction in medical expenditure by e-health users.

We examined the medical expenditure of elderly residents who used an e-health system in Nishiaizu town. Since 1994, health-related data, such as blood pressure, ECG and blood oxygen, have been transmitted to a remote medical institution via a telecommunications network. We selected 412 users from the list of registered e-health users in the town. We also selected 450 residents who were not e-health users. We sent them questionnaires and obtained 199 valid responses from e-health users and 209 from non-users. Then we examined the receipts of these 408 people, which were kept in paper form in the town office. Based on a multiple regression analysis, we found that users of the e-health system had lower medical expenditure for lifestyle-related illness than non-users. The medical expenditure of e-health users was lower than that of non-users by 15,302 yen (US $133) per year. This amount was approximately 21% of the average annual medical expenditure of the residents. The results also showed that: long-time users of e-health had lower medical expenditure on lifestyle-related illness; long-time users of e-health had lower medical expenditure than those who used it for a shorter time; e-health had more effect on people with diseases than those without.