Could Investment in Preventive Health Care Services Reduce Health Care Costs among Those Insured with Health Insurance Societies in Japan?

Abstract

This study examined the impact of expenditures for preventive health care services on health care costs among those insured with health insurance societies in Japan using cross-sectional and longitudinal designs. The subjects of the study were those insured with Japan's 1481 health insurance societies belonging to the National Federation of Health Insurance Societies in 2003 and 2007. Multiple regression analyses were conducted using the forced entry method. Case rates, number of service days, and health care costs were used as dependent variables, and preventive health care expenditures, average age, number of the insured, gender ratio, average monthly salary, and dependents ratio were used as independent variables. Expenditures for preventive health care services showed significant negative correlations with both the number of service days and health care costs for inpatient and outpatient services in 2003 and 2007. The results showed that expenditures for preventive health care services had a negative relationship with health care costs. Thus, these findings support the effects of investment in preventive health activities as promoted by health insurance societies to reduce health care costs. (Population Health Management 2014;17:42–47)

Introduction

As developed countries worldwide face aging populations,¹ they are encountering the common issue of increasing health care costs.^2,3 However, disease prevention is one effective strategy to both promote health and to reduce health care costs.⁴ Current health care reforms under way in the United States focus on improving primary care services and disease prevention.⁵

In Japan, there are various preventive health care services.⁶ As primary and secondary measures for the prevention of lifestyle-related diseases, the national policy for health promotion was developed in 1978, followed by the Active 80 Health Plan 14 in 1988.⁶ The People's Health Promotion Campaign for the 21st Century (Healthy Japan 21) was then initiated in 2000. Healthy Japan 21, which received legislative backing with The Health Promotion Act in 2002, has increased public awareness of the prevention of lifestyle-related diseases.⁶

Health insurance societies are obliged to facilitate preventive health care services to maintain and promote the health of the insured and their dependents, including activities for health management, the promotion of sports, health education and consumer awareness, and recreation facilities.⁷ Health management services include health checkups for the prevention of lifestyle-related diseases, comprehensive health examinations, and health counseling provided by nurses, registered dietitians, and health fitness programmers. The promotion of sports aims to maintain and improve the health of people and enhance their physical health and strength through exercise, sports, and walking, using facilities for health promotion. Activities for health education and consumer awareness include the publication of reports, distribution of leaflets, and organization of lectures to provide health-related education. Recreation facilities were opened to help people reduce stress, alleviate fatigue, and feel more energetic.

A lively debate exists over whether preventive health care services can in fact curb health care costs. Some reports state that effective prevention programs could substantially reduce health expenditures,^8
–10 while others argue that they are not cost-effective.^11
–13 There have been many economic evaluations of preventive health care programs worldwide.^14

–20 However, they usually evaluate costs related to preventive health care programs and rarely include health care costs.

The present study examines the relationship between expenditures for preventive health care services and health care costs using cross-sectional and longitudinal analyses, and examines the effects of preventive health care services to reduce health care costs based on data obtained from health insurance societies.

Methods

In 2003, the fixed insurance premium rate for those insured under employee health insurance plans increased from 20% to 30%. A 2008 amendment to the law included changes to preventive health services, incorporating the implementation of “specific health checkups and health care advice.”⁶ Thus, insurers now must determine the risk of metabolic syndrome based on patients' obesity, blood glucose, serum lipids, blood pressure, and smoking habits.²¹

This study examined the effects of preventive health services implemented in 2003 based on indices of health care expenditures for 2003 and 2007. In 2003, there were 1640 health insurance societies and the total number insured was 14,776,193; in 2007, there were 1521 health insurance societies and the total number insured was 15,855,286. Subjects were health insurance societies that were members of the National Federation of Health Insurance Societies and whose complete data were available in annual reports for 2003 and 2007. Of the 1640 health insurance societies whose complete data were included in the 2003 annual report, 1481 still existed in 2007, with complete data described in the 2007 annual report. This study involved these 1481 health insurance societies; their total number insured in 2003 was 14,043,626.

The following indices of health care costs were used in the statistical analysis as dependent variables: case rates, number of service days, and health care costs for inpatient, outpatient, and dental services in 2003 and 2007. The case rates are expressed as the number of health care claims per patient. The number of service days required for care per patient is calculated by dividing the total number of days for services by the number of health care claims. Out-of-pocket health expenses were excluded from health benefits. In both 2003 and 2007, 30% of the total health care costs were incurred by the insured (1 US dollar=80 yen).

Regarding the characteristics of health insurance societies that affected those dependent variables in 2003, the following independent variables were selected: expenditures for preventive health care services, average age, number of insured, gender ratio (number of insured males/number of insured females), average monthly salary, and dependent ratio (number of dependents/number of insured). First, independent variables in 2003 were compared, and paired t tests were used to address changes from 2003 to 2007. Second, for dependent variables, the following were calculated: minimum and maximum values, medians, interquartile ranges, mean values, and standard deviations. A multiple regression analysis was then conducted using the forced entry method, with the indices of health care costs in 2003 and 2007 as dependent variables and the characteristics of health insurance societies in 2003 as independent variables. Dependent variables were log-transformed (log (x+1)) because the distribution of these dependent variables was skewed to the right. In addition, expenditures for preventive health care costs, number of insured, gender ratio, and average monthly salary were log-transformed (log (x)) because they also were skewed. A standardized partial regression coefficient was used to quantify the effects of expenditures for preventive health care services, and a coefficient of determination was employed to establish the validity of the model. SPSS 11.5J (IBM, Armonk, NY) for Windows was used to conduct the statistical analyses, with a significance level of P=0.05.

Results

Table 1 shows the differences in independent variables among health insurance societies in 2003 and 2007. There were significant differences in expenditures for preventive health care services, average age (years), number of insured, average monthly salary, and dependent ratio (gender ratio was unavailable in 2007). Expenditures for preventive health care services increased moderately. Although the number of insured increased, the number of health insurance societies decreased. This suggests that larger societies survived during the study period and smaller ones did not. Furthermore, the average age increased and the dependent ratio decreased, which reflects a larger demographic trend. The reason for the increase in average monthly salary was the increase in average age (salaries increase with age in the Japanese salary system).

Table 1.

Descriptive Statistics of Independent Variables in 2003 and 2007

Variables	2003 (N=1481) Mean(SD)	2007 (N=1481) Mean(SD)	Difference Mean(SD)
Expenditures for preventive health care services ($)	253.15 (168.45)	279.25 (190.66)	26.10 (106.78)^***
Average age	39.86 (3.64)	41.42 (3.10)	1.56 (2.35)^***
Number of insured	9482.53 (19,730.79)	10,180.27 (22,130.33)	697.74 (5,249.15)^***
Gender ratio^#	4.64 (4.49)	unavailable	unavailable
Average monthly salary ($)	4647.17 (888.05)	4741.31 (989.10)	94.14 (324.38)^***
Dependent ratio^##	1.12 (0.27)	1.04 (0.25)	−0.07 (0.84)^***

Number of insured males/number of insured females; ^##Number of dependents/number of insured;^*** P<0.001 by paired t test (df's=1480); SD, standard deviation.

Table 2 shows the indices of health care costs for 2003 and 2007. The mean case rates for outpatient and dental services increased while the mean numbers of service days for inpatient, outpatient, and dental services decreased. The mean health costs for inpatient and outpatient services increased while those for dental services decreased.

Table 2.

Descriptive Statistics of Health Care Service Demand Indicators in 2003 and 2007

Variables	Minimum	Maximum	Median	Interquartile range	Mean	SD
2003 (N=1481)
Case rate for inpatient services	0.00	0.68	0.07	0.68	0.07	0.03
Case rate for outpatient services	2.23	29.22	4.74	26.99	4.77	0.96
Case rate for dental services	0.23	2.63	1.30	2.40	1.30	0.16
The number of service days for inpatient services	0.00	17.58	10.46	17.58	10.45	1.44
The number of service days for outpatient services	0.26	3.14	1.60	2.88	1.61	0.12
The number of service days for dental services	1.72	4.62	2.36	2.90	2.36	0.15
Health costs per person for inpatient services	0.00	825.20	256.33	825.20	264.48	81.11
Health costs per person for outpatient services	182.59	2,000.73	459.43	1,818.14	466.01	97.81
Health costs per person for dental services	27.46	318.48	166.33	291.01	165.56	21.81
2007 (N=1481)
Case rate for inpatient services	0.00	0.24	0.07	0.24	0.07	0.02
Case rate for outpatient services	2.20	11.73	5.02	9.53	5.05	0.74
Case rate for dental services	0.23	3.06	1.35	2.83	1.35	0.17
The number of service days for inpatient services	0.00	17.50	9.66	17.50	9.60	1.39
The number of service days for outpatient services	1.26	2.84	1.49	1.58	1.49	0.08
The number of service days for dental services	1.33	19.43	2.15	18.10	2.17	0.47
Health costs per person for inpatient services	0.00	1,031.59	276.59	1,031.59	287.12	97.79
Health costs per person for outpatient services	169.26	952.50	473.29	783.24	477.52	87.02
Health costs per person for dental services	22.23	287.50	154.71	265.28	154.38	19.55

SD, standard deviation

Table 3 shows the results of the multiple regression analyses conducted using case rates in 2003 and 2007 as dependent variables and the characteristics of health insurance societies in 2003 as explanatory variables. In both 2003 and 2007, the standardized partial regression coefficient was significantly high for average age and average monthly salary for inpatient, outpatient, and dental services. Expenditures for preventive health care services showed no significant correlation with case rates for outpatient and dental services in both 2003 and 2007. Expenditures for preventive health care services also showed no significant correlation with case rates for inpatient services in 2003; however, it showed a moderately significant negative correlation with case rates for inpatient services in 2007. The determination coefficients of the model were significantly high for inpatient, outpatient, and dental services.

Table 3.

Results of Multiple Regression Analyses on Case Rates in 2003 and 2007

2003 (N=1481)	Inpatient (β)	Outpatient (β)	Dental (β)
Expenditures for preventive health care services (log-transformed)	−0.053	−0.034	0.045
Average age	0.215^**	0.472^**	0.468^**
Number of insured (log-transformed)	0.001	−0.021	0.003
Gender ratio^# (log-transformed)	−0.032	−0.458^**	−0.220^**
Average monthly salary (log-transformed)	0.087^*	0.461^**	0.454^**
Dependent ratio^##	0.090^*	−0.009	−0.177^**
Adjusted R²	0.077^**	0.450^**	0.390^**
2007 (N=1481)
Expenditures for preventive health care services (log-transformed)	−0.073^*	0.001	0.018
Average age	0.317^**	0.433^**	0.415^**
Number of insured (log-transformed)	−0.005	−0.043^*	0.005
Gender ratio^# (log-transformed)	−0.053	−0.524^**	−0.267^**
Average monthly salary (log-transformed)	0.139^**	0.478^**	0.436^**
Dependent ratio^##	0.042	0.072^*	−0.064
Adjusted R²	0.135^**	0.488^**	0.349^**

P<0.05; ^** P<0.001; ^#Number of insured males/number of insured females; ^##Number of dependents/number of insured.

Table 4 shows the results of the multiple regression analyses conducted using the number of service days in 2003 and 2007 as dependent variables and the characteristics of health insurance societies in 2003 as explanatory variables. In both 2003 and 2007, the standardized partial regression coefficient was significantly high for average age, average monthly salary, and gender ratio for inpatient, outpatient, and dental services. Expenditures for preventive health care services showed a significant negative correlation with the number of service days for inpatient and outpatient services in both 2003 and 2007; there was no significant correlation with the number of service days for dental services in 2003 and 2007. The determination coefficients of the model were significantly high for inpatient, outpatient, and dental services.

Table 4.

Results of Multiple Regression Analyses on the Number of Service Days per Case in 2003 and 2007

2003 (N=1481)	Inpatient (β)	Outpatient (β)	Dental (β)
Expenditures for preventive health care services (log-transformed)	−0.163^**	−0.103^**	−0.029
Average age	0.171^**	0.309^**	0.212^**
Number of insured (log-transformed)	0.166^**	−0.011	0.040
Gender ratio^# (log-transformed)	0.122^**	0.084^*	0.241^**
Average monthly salary (log-transformed)	−0.061	−0.119^**	−0.279^**
Dependent ratio^##	0.041	0.032	−0.029
Adjusted R²	0.108^**	0.138^**	0.134^**
2007 (N=1481)
Expenditures for preventive health care services (log-transformed)	−0.118^**	−0.097^**	−0.022
Average age	0.129^**	0.335^**	0.123^**
Number of insured (log-transformed)	0.182^**	0.022	0.018
Gender ratio^# (log-transformed)	0.146^**	0.107^**	0.157^**
Average monthly salary (log-transformed)	−0.057	−0.118^**	−0.257^**
Dependent ratio^##	0.058	−0.048	−0.038
Adjusted R²	0.093^**	0.144^**	0.077^**

P<0.05; ^** P<0.001; ^#Number of insured males/number of insured females; ^## Number of dependents/number of insured.

Table 5 shows the results of multiple regression analyses conducted using health care costs in 2003 and 2007 as dependent variables, and the characteristics of health insurance societies in 2003 as explanatory variables. In both 2003 and 2007, the standardized partial regression coefficient was significantly high for average age and average monthly salary for inpatient, outpatient, and dental services. Expenditures for preventive health care services showed a significant negative correlation with health care costs for inpatient and outpatient services in both 2003 and 2007; there was no significant correlation with health care costs for dental services in 2003 and 2007. The determination coefficients of the model were significantly high for inpatient, outpatient, and dental services.

Table 5.

Results of Multiple Regression Analyses on Health Costs per Person in 2003 and 2007

2003 (N=1481)	Inpatient (β)	Outpatient (β)	Dental (β)
Expenditures for preventive health care services (log-transformed)	−0.129^***	−0.112^***	0.023
Average age	0.358^***	0.501^***	0.513^***
Number of insured (log-transformed)	0.152^***	0.019	0.069^**
Gender ratio^# (log-transformed)	0.032	−0.281^***	0.010
Average monthly salary (log-transformed)	0.131^***	0.350^***	0.230^***
Dependent ratio^##	0.051	0.104^***	−0.203^***
Adjusted R²	0.208^***	0.400^***	0.289^***
2007 (N=1481)
Expenditures for preventive health care services (log-transformed)	−0.113^***	−0.075^**	−0.038
Average age	0.299^***	0.462^***	0.418^***
Number of insured (log-transformed)	0.126^***	0.032	0.081^***
Gender ratio^# (log-transformed)	0.051	−0.315^***	−0.018
Average monthly salary (log-transformed)	0.083^*	0.335^***	0.213^***
Dependent ratio^##	0.061	0.177^***	−0.105^**
Adjusted R²	0.153^***	0.391^***	0.205^***

P<0.05; ^** P<0.01; ^*** P<0.001; ^#Number of insured males/number of insured females; ^##Number of dependents/number of insured.

Discussion

In both 2003 and 2007, average age had relatively strong correlations with case rates, number of service days, and health care costs for inpatient, outpatient, and dental services. In both 2003 and 2007, average monthly salary was significantly correlated with case rates, number of service days, and health care costs for inpatient, outpatient, and dental services. This result occurred because, as pointed out in previous studies, both average age and average monthly salary show strong correlations with case rates, number of service days, and health care costs for inpatient, outpatient, and dental services. Thus, the older the age and the higher the monthly salary, the higher the frequency of utilization of health care facilities.^22,23

Expenditures for preventive health care services showed significant negative correlations with both the number of service days and health care costs for inpatient and outpatient services in 2003 and 2007. This result can be explained because the number of days of providing outpatient services is considered to be related to the severity of illness.¹⁸ Health insurance societies have spent a significant amount of money on preventive health care services, and as a result have curbed health expenditures for inpatient and outpatient services. Chronic diseases require longer treatment periods and have greater health care costs than mild diseases.⁶

In both 2003 and 2007, there were no significant correlations between expenditures for preventive health care services and case rates, number of service days, and health care costs for dental services. Presumably this is because expenditures for preventive health care services were limited for dental preventive care as health care expenditures in dentistry are relatively low.⁶

Preventive health care services in Japan have contributed to preventing lifestyle-related diseases, which presumably reduces health care expenditures for outpatient and inpatient services. The correlation coefficient between expenditures for insurance services in 2003 and 2007 was as high as 0.867, which suggests that it was consistent for the period. The fact that health insurance societies provide preventive health care services in accordance with well-established policies is considered to influence the indices of health care costs. Preventive interventions can be undertaken by individuals, the health care system, or community programs.⁴ However, physical activity as a personal intervention induces net savings by improving one's health,²⁴ while counseling by clinicians to persuade people to exercise is of uncertain effectiveness and cost-effectiveness.²⁵ Preventive health care services provided by health insurance societies are targeted to motivate individuals to change their lifestyles and health counseling is usually provided by nurses, registered dietitians, and health fitness professionals.

This study has a number of limitations. First, the data used in the study do not provide specific information regarding disease type or severity. Because each data point represents a mean value per insured person covered by an insurer, an ecological fallacy may have occurred. Therefore, the study could not control confounding factors such as trends in economic conditions, lifestyle, workplace environment, and education level, which may be related to the incidence of diseases because of the study design.²⁶ Second, annual workplace health checkups may be covered by employers under the Industrial Safety and Health Act. However, health insurance societies provide health guidance and preventive health care services based on the results of those health checkups. In addition, societies could provide more comprehensive health examinations covered by expenditures earmarked for preventive health care services. Finally, only those health insurance societies that were members of the National Federation of Health Insurance Societies and whose complete data were available in annual reports for 2003 and 2007 were included. However, this study's sample included 1481 of the 1640 societies operating in 2003. In addition, the demographic changes in the societies reflected those occurring in the entire Japanese population.

In 2008, the Japanese government mandated that insurers provide certain health examination and guidance services for metabolic syndrome. Therefore, the preventive health care services provided by health insurance societies before 2008 were different from those provided after 2008. This study evaluated the impact of expenditures for preventive health care services before 2008. Thus, the impact of expenditures for preventive health care services after 2008 should be evaluated in future research. It is also necessary to conduct long-term randomized controlled trials to determine any causal relationships.

The results of the present study, using cross-sectional and longitudinal analyses, suggest significant negative correlations between expenditures for preventive health care services and health care costs. These results are considered to support the effects of preventive health activities promoted by health insurance societies to reduce health care costs.

Footnotes

Author Disclosure Statement

Drs. Babazono, Maeda, and Lkhagva, and Ms. Gao and Mr. Nishi declare no conflicts of interest with respect to the research, authorship, and/or publication of this article. The authors receive no financial support for the research, authorship, and/or publication of this article.

References

World Health Organization. Ageing. Available at: http://www.who.int/topics/ageing/en/index.html. Accessed June 17, 2012 .

Morgan

, Cunningham

. Population aging and the determinants of healthcare expenditures: The case of hospital, medical and pharmaceutical care in British Columbia, 1996 to 2006. Health Policy, 2011; 7:68–79.

Evans

, McGrail

, Morgan

, Barer

, Hertzman

. Apocalypse no: Population aging and the future of health care systems. Can J Aging, 2001; 20:160–191.

Woolf

. A closer look at the economic argument for disease prevention. JAMA, 2009; 301:536–538.

The White House. Executive Order 13544: Establishing the National Prevention, Health Promotion, and Public Health Council. Available at: http://www.whitehouse.gov/the-press-office/executive-order-establishing-national-prevention-health-promotion-and-public-health. Accessed June 17, 2012 .

Ministry of Health, Welfare and Labour, Japan. Kokumin Eisei no Doko (Health State of the Nation, Annual Report). Tokyo: Health and Welfare Statistics Association; 2012:84–99.

National Federation of Health Insurance Societies. Health Insurance, Long-term Care Insurance and Health Insurance Societies. Tokyo: Kenporen; 2006:46–49.

Pignone

, Saha

, Hoerger

, Mandelblatt

. Cost-effectiveness analyses of colorectal cancer screening: A systematic review for the U.S. Preventive Services Task Force. Ann Intern Med, 2002; 137:96–104.

Maciosek

, Coffield

, Edwards

, Flottemesch

, Goodman

. Solberg LI. Priorities among effective clinical preventive services: Results of a systematic review and analysis. Am J Prev Med, 2006; 31:52–61.

10.

Solberg

, Maciosek

, Edwards

, Khanchandani

, Goodman

. Repeated tobacco use screening and intervention in clinical practice: Health impact and cost effectiveness. Am J Prev Med, 2006; 31:62–71.

11.

Kahn

, Robertson

, Smith

, Eddy

. The impact of prevention on reducing the burden of cardiovascular disease. Diabetes Care, 2008; 31:1686–1696.

12.

Cohen

, Neumann

, Weinstein

. Does preventive care save money? Health economics and the presidential candidates. N Engl J Med, 2008; 358:661–663.

13.

Russell

. Preventing chronic disease: An important investment, but don't count on cost savings. Health Aff (Millwood), 2009; 28:42–45.

14.

Bemelmans

, van Baal

, Wendel-Vos

, et al. The costs, effects and cost-effectiveness of counteracting overweight on a population level. A scientific base for policy targets for the Dutch national plan for action. Prev Med, 2008; 46:127–135.

15.

Galani

, Schneider

, Rutten

. Modeling the lifetime costs and health effects of lifestyle intervention in the prevention and treatment of obesity in Switzerland. Int J Public Health, 2007; 52:372–382.

16.

Jacobs-van der Bruggen

, Bos

, Bemelmans

, Hoogenveen

, Vijgen

, Baan

. Lifestyle interventions are cost-effective in people with different levels of diabetes risk: Results from a modeling study. Diabetes Care, 2007; 30:128–134.

17.

Ramachandran

, Snehalatha

, Yamuna

, Mary

, Ping

. Cost-effectiveness of the interventions in the primary prevention of diabetes among Asian Indians: Within-trial results of the Indian Diabetes Prevention Programme (IDPP). Diabetes Care, 2007; 10:2548–2552.

18.

Rasmussen

, Thomsen

, Kilsmark

, et al. Preventive health screenings and health consultations in primary care increase life expectancy without increasing costs. Scand J Public Health, 2007; 35:365–372.

19.

Richter

, Kraft

, Kleinwechter

, et al. Effects of a telephone intervention in patients with type 2 diabetes. Dtsch Med Wochenshr, 2008; 133:2203–2208.

20.

Roux

, Pratt

, Tengs

, et al. Cost effectiveness of community-based physical activity interventions. Am J Prev Med, 2008; 35:578–588.

21.

Babazono

, Kuwabara

, Hagiihara

, Nagano

, Ishihara

. Do interventions to prevent lifestyle-related diseases reduce healthcare expenditures? A randomized controlled clinical trial. J Epidemiol, 2011; 21:75–80.

22.

Babazono

, Kuwabara

, Hagihara

, Yamamoto

, Hillman

. Does income influence demand for health services despite Japan's “Health Care for All” policy?. Int J Technol Assess Health Care., 2008; 24:125–130.

23.

Lkhagva

, Gao

, Babazono

. Does co-payment rate influence the relationship between monthly salary and health care service demand among the insured of health insurance societies in Japan?. Popul Health Manag, 2013; 16:58–63.

24.

Hatziandreu

, Koplan

, Weinstein

, Caspersen

, Warner

. A cost-effectiveness analysis of exercise as a health promotion activity. Am J Public Health, 1988; 78:1417–1421.

25.

Eden

, Orleans

, Mulrow

, Pender

, Teutsch

. Does counseling by clinicians improve physical activity? Summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med, 2002; 137:208–215.

26.

Thayer

, Fabius

, Frazee

. Research and development in population health. In: Nash

, Reifsnyder

, Fabius

, Pracillo

, eds. Population Health: Creating a Culture of Wellness. Sudbury, MA: Jones & Bartlett Learning; 2011:257–270.