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

Abstract

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.

Introduction

In 2006 medical expenditure in Japan amounted to 33.1 trillion yen (US$331 billion). More than half of all expenditure is for people over 65 years old. The population is aging rapidly and already at least 20% of the people are over 65 years of age. In order to cope with this situation, various policy measures have been taken, including requiring patients to bear more of their own medical costs. Japan has a well-organized universal public health insurance system. However, the percentage of costs reimbursed by public health insurance has been falling, due to increasing medical expenditures and deficits in the medical insurance budget. The elderly have thus been forced to pay more of their own medical costs.

Another measure to reduce medical expenditure is to focus on the prevention of diseases: the healthier people become, the fewer medical costs are required. One example is to enhance consciousness towards health and efforts to prevent illness. To this end, the government has begun initiatives such as the ‘Health Japan 21 Project.’ Recent campaigns against metabolic syndrome are another example, as this condition is thought to increase the risk of hypertension or hyperlipidaemia. The campaigns against metabolic syndrome include recommending regular physical exercises, and monitoring of diet and nutrition. Prevention of illness through health maintenance is an important measure to reduce medical expenditure.

The present study investigated the utilization of Information Technology (IT) to maintain the health of residents. We examined e-health systems which monitored the health of elderly people at home by transmitting data such as blood pressure, ECG and blood oxygen, to a remote medical site via a telecommunications network.¹ At present, more than 100 local governments in Japan employ such systems, using a total of more than 12,000 devices. Relatively simple equipment is used to record an elderly person's condition and the data are then used for diagnosis and consultation. Reports sent by the medical institution also help users to enhance their daily health consciousness and maintain good health. These positive effects have been identified through field surveys.^1–3

The present study was based on data from Nishi-aizu Town in Fukushima Prefecture. E-health in Nishi-aizu Town is quite simple, and senior people at home are connected to medical institutions through telecommunications networks. Their vital signs data are transmitted to the Town's health centre and nurses there provide advice to them based on the data. In our previous studies,^4–6 we focused on the savings resulting from use of the e-health system, and found that it reduced annual medical expenditure on lifestyle-related diseases by about 20%. In the present study paper, we analysed how and why e-health reduces medical expenditure.

Methods

The database was the same as that of our previous research which examined the reduction of medical expenditure. We identified users and non-users of e-health, and checked their receipts of National Health Insurance for five years. A receipt mainly includes the subject's medical expenditure and days needed for treatment of each disease. 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. In order to compare the two groups without bias, we added their characteristics such as gender, age and having chronic diseases as explanatory variables, with days for treatment as the dependent variable. That is, the days for treatment were affected not only by use of e-health but also by other user characteristics, and the effect due to the latter could be identified and subtracted from the total effect. Their data are summarised in Table 1.

Table 1

Summary statistics

	n	Mean	SD	Min	Max	P value
Medical expenditure on lifestyle-related illness*	2040	6836	10266	0	76573
Days spent for treatment (lifestyle-related illness)	2040	6.1	8.4	0	85
Sex (male 0, female 1)	2040	0.56	0.50	0	1
Age (years)	2040	69.98	8.99	43	94
Education^†	2040	1.58	0.81	0	4
Employment (employed 1, unemployed 0)	2040	0.53	0.50	0	1
Number of family members living together	2040	2.6	1.9	0	9	<0.01 (User > Non-user)
Income*	1820	42.53	59.98	10	300
Having chronic disease (yes 1, no 0)	1665	0.51	0.50	0	1	<0.05 (User > Non-user)

*Variable based on the medical score; 1 point is worth 10 yen

^†Variable scored from 1 = junior education to 4 = university

Based on the data we obtained, we tested four hypotheses as follows:

The days spent for treatment of users of e-health is shorter than those of non-users;

The longer they use the system, the shorter their days for treatment become;

The group of longer usage of the system has the larger elasticity; in other words, for users who utilize e-health for more years, an additional one year use decreases the days for treatment more than for those who use e-health for fewer years;

The effect of the reduction of days for treatment is higher, especially for people with chronic diseases.

Results

The analysis showed that the days for treatment for lifestyle-related illnesses of women were longer than those for men by 1.5 days per year. The regression coefficients showed that the days for treatment increased by 0.2 days per year, when subjects become one year older.

Lower income residents had longer days for treatment than higher income residents.

The days for treatment of residents who had chronic diseases were longer than those without them by 3.6 days per year.

The days for treatment of e-health users were shorter than those of non-users by 1.6 days per year (see Table 2).

Table 2

Result of estimation (Hypothesis 1)

Variables	Coefficient	SE	P value
Sex	1.4926	0.4007	<0.01
Age	0.1816	0.0243	<0.01
Education	−0.0619	0.2591
Employment	0.5945	0.4265
Number of family living together	0.2887	0.1072	<0.01
Income	−0.0128	0.0035	<0.01
Chronic disease	3.6024	0.3987	<0.01
User dummy	−1.6004	0.4055	<0.01
Constant	−8.4460	1.9786	<0.01
Adjusted R-squared	0.1165
Number of observations	1545

The days spent for treatment of lifestyle-related illnesses could be reduced by 0.14 days per year, if they extended their use of e-health by an additional one year (see Table 3).

Table 3

Result of estimation (Hypothesis 2)

Variables	Coefficient	SE	P value
Sex	1.6113	0.4003	<0.01
Age	0.1881	0.0248	<0.01
Education	−0.0511	0.2602
Employment	0.6429	0.4289
Number of family living together	0.2697	0.1081	<0.05
Income	−0.0128	0.0035	<0.01
Chronic disease	3.5651	0.4000	<0.01
Years of using e-health	−0.1354	0.0554	<0.05
Constant	−9.4203	2.0099	<0.01
Adjusted R-squared	0.1095
Number of observations	1545

The elasticity of reduction of days spent for treatment with respect to the length of utilizing the system becomes larger, as shown in Table 4. This implies that elasticity increases according to years of its use, and the more years one uses the system, the larger the reduction in days spent for treatment becomes.

Table 4

Elasticity (Hypothesis 3)

Years of e-health use	Elasticity (days spent on treatment)
Non-user	0
Less than 2	−0.0006
2–4	−0.0690
4–6	−0.0993
6–8	−0.1413
8–10	−0.1816
10–12	−0.1784
Over 12	−0.2503

The comparison of the elasticity of medical expenditure and days spent for treatment, showed that the latter had larger elasticity. This implies that the effect of reduction of days spent for treatment might be larger than that of medical expenditure.

Finally, the e-health system had a larger effect on people with chronic diseases. The difference in days for treatment between user and non-user groups was found to be 2.74 days (see Table 5).

Table 5

Result of estimation (Hypothesis 4)

	Without chronic disease			With chronic disease
Variables	Coefficient	SE	P value	Coefficient	SE	P value
Sex	1.2603	0.4595	<0.01	1.8182	0.6503	<0.01
Age	0.2141	0.0260	<0.01	0.1405	0.0418	<0.01
Education	−0.2698	0.3092		−0.0647	0.4053
Employment	0.6798	0.4797		0.5156	0.7015
Number of family living together	0.2002	0.1198		0.4364	0.1772	<0.05
Income	−0.0021	0.0039		−0.0237	0.0057	<0.01
User dummy	−0.5405	0.4633		−2.7403	0.6598	<0.01
Constant	−11.3001	2.1434	<0.01	−1.2197	3.3308
Adjusted R-squared	0.1095			0.0663
Number of observations	755			790

Discussion

As a result of the use of e-health, the days for treatment were reduced by 1.6 days per year compared to non-users. The average cost of one day of treatment or one diagnosis is 10,000 yen. Thus the total reduction in expenditure as a result of fewer hospital visits (emergency and elective) was about 16,000 yen per year. In the previous study, we proved that the medical expenditure of e-health users was smaller than those of non-users by 15,688 yen in Nishi-aizu Town.⁵ The results of the present study, therefore, coincide with those of the previous one. In other words, the reduction of medical expenditure is principally caused by the reduction of days for treatment. As medical expenditure is highly correlated with the days for treatment, the results seem to be reasonable. The results of the present study reveal the mechanism for the reduction of medical expenditures by users of e-health.

Increasing medical costs is a common phenomenon all over the world. There are two measures to cope with this: utilization of IT in medical work and prevention of illness (or maintenance of health). An e-health system can assist with both. However, some local governments in Japan decided to terminate their e-health services, because they doubted whether e-health was really effective or not. The results we have obtained provide a rigorous foundation for the evidence supporting the use of e-health.

Footnotes

Acknowledgements

We are grateful for financial support from the Ministry of Welfare and Labour and the Ministry of Education, Science and the Japan Society for the Promotion of Science. We also thank Mr H Yamaguchi, Mr K Takahashi and other officials of the Nishi-aizu Town Office who supported our field studies.

References

Tsuji

. The telehomecare/ehealth system in Japan. Business Briefing: Global Healthcare 2002:72–6

Tsuji

, Suzuki

, Taoka

. An empirical analysis of a telehealth system in terms of cost-sharing. J Telemed Telecare 2003;9(Suppl. 1):41–3

Tsuji

, Iizuka

, Taoka

, Teshima

. The willingness of Japanese citizens to pay for e-Health systems. Journal on Information Technology in Healthcare 2003;4:103–10

Akematsu

, Tsuji

, Taoka

. An empirical analysis of reduction of medical expenditures by the eHealth system: further results. HealthCom 2008. Proceedings of IEEE 10th International Conference on e-Health Networking, Applications and Services , 2008:140–7

Akematsu

, Tsuji

. An empirical analysis of the reduction in medical expenditure by e-health users. J Telemed Telecare 2009;15:109–11

Tsuji

, Akematsu

, Taoka

. How much can eHealth system save medical expenditures. Journal of eHealth Technology and Application 2007;5:92–9