Professor of Health Policy and Management; Director
Ye Zhu MD MPH
PhD candidate, University of North Carolina, Chapel Hill, USA
Oluseye Ogunmoroti MBBS MPH
Center for Global Health, University of Georgia, USA
Nancy Terwoord RN
Community Health Centers of Greater Dayton, USA
Sylvia Ellison MA
Center for Global Health Systems, Management, and Policy, Wright State University, USA
Akira Fujiyoshi MD MPH
Hirotsugu Ueshima MD PhD
Katsuyuki Muira MD PhD
Professor and Chair
Department of Health Science, Shiga University of Medical Science, Japan
Received date: 15 September 2012; Accepted date: 12 December 2012
BackgroundThere is a 42% lower cardiovascular disease (CVD) death rate in Japan compared with the USA. Do physicians report differences in practice management of CVD risk factors in the two countries that might contribute to this difference? AimsCVD risk factor management reported by Japanese versus US primary care physicians was studied. MethodsWe undertook a descriptive study. An internet-based survey was conducted with physicians from each country. A convenience sample from the Shiga Prefecture in Japan and the state of Ohio in the USA resulted in 48 Japanese and 53 US physicians completing the survey. ResultsThe survey group may not be representative of a larger sample. The survey demonstrated that 98% of responding Japanese physicians spend
cardiovascular risk factors, global health, guidelines, health systems, practice manage-ment
Physician management of cardiovascular disease (CVD) risk factors in the office or clinic setting is not well understood. Although various CVD risk factor guidelines have been promulgated (e.g. ATP III, JNC VII), their dissemination and implementation have not been especially successful or well-studied.
CVD death rates are dramatically lower in Japan than they are in the USA. The World Health Organ-ization (WHO) reports the age-standardised CVD mortality rate for men and women is 179 per 100 000 in the USA versus 103 in Japan. Therefore, Japan has a 42% lower CVD death rate than the USA. The non-age-standardised prevalence of hypertension in Japan is reported to be 47%, whereas in the USA it is 28.7%.[2,3] Hypercholesterolaemia has been reported as lower in Japan than the USA.[4–6] The prevalence of diabetes in Japan is 6.9%, whereas in the USA it is 10.7%.[7,8] Smoking in Japan has a prevalence of 51% in men and 10% in women compared with the USA where 23% of men and 20% of women smoke.[9,10] In 2000, the WHO released its most recent evaluation of health systems internationally. While the USA was ranked 37th among the health systems in this contro-versial report, Japan was ranked 10th.
The Japanese healthcare system is relatively inex-pensive in comparison with other high-income coun-tries. In 2003, healthcare spending in the USA was $5635 per capita, while in Japan it was $2139 per capita. The USA committed 15% of gross domestic product (GDP) to health care, while Japan spent 7.9%. Japan has universal health care, with an ‘employer mandate’ for those working and government subsidies for those not working. It is a high volume, high tech-nology, low-cost health system. Physicians are gener-ally in private practice, although some are hospital employees; they are highly respected members of the Japanese society. The system works through a series of private, not-for-profit insurance companies.
Between 2005 and 2006, the National Heart Lung and Blood Institute (NHLBI) created 12 enhanced dissemination utilisation centres, recognising the need to disseminate and implement research discover-ies by ‘taking the findings of clinical investigations and translating them into the practice of medicine at the community level’. The principal investigator of this study was one of the centre directors established by NHLBI. Members of this team have previously studied
how microsystem changes in physician offices im-prove the use of practice guidelines.[16,17] This group has compared management of CVD risk factors in US and French physicians in the microsystem of the office. We are now extending this study to compare US and Japanese physicians’ practice behaviours and attitudes in managing CVD risk factors.
There is little doubt that much of the difference in CVD death rates can be attributed to substantial differences between the two countries in the culture of eating, exercise and other social determinants of health. The two populations are genetically different as well as the health system, with one providing universal care to the population and the other not. Yet, the question arises if there are differences in how physicians man-age CVD risk factors. Even if differences are found, demonstrating a cause–effect relationship is challeng-ing and beyond the scope of this study.
The goal of this study was to seek to identify differences in physician outpatient management of patients with CVD risk factors. Might some of those differences highlight some of the differences between the US and Japanese healthcare systems? Could those systems differences explain part of the differences in CVD death rates? Finally, does the availability of universal access to care in Japan, with less adminis-trative complexity than the USA, allow physicians and patients to adopt professional and culturally pertinent approaches to the management of cardiovascular risk factors that could possibly improve outcomes of care. We conducted an internet-based survey of physicians in both countries. Our expectation was not to answer these complex questions, but rather to raise them, identify some possible differences and promote a more definitive future study to attempt to seek more complete information.
Preliminary information was collected by local inter-views and an on-site study of physician practices. Then an internet-based survey was conducted. Nearly all US physicians were primary care physicians (pri-marily family physicians), while the Japanese physicians were in a variety of specialties; the addition of other specialties reduces the ability of the authors to reach accurate conclusions.
Japanese physicians were surveyed through the Shiga Prefecture Medical Society, the equivalent to a county medical society in the USA. Shiga Prefecture in south central Japan is fairly representative of the country, with many demographic characteristics being in the middle third of all Japanese Prefectures (in-cluding population density and prevalence of hyper-tension, diabetes and smoking in men).
Most US physicians surveyed were members of the Ohio Academy of Family Practice, although primary care physicians in other US states were also included. Ohio is in the eastern-central part of the USA. It is generally considered representative of the USA; its overall health ranking is 36 of 50, placing it in the middle third of US states.
This study used a small convenience sample of participants and cannot be claimed to be a represen-tative sample of all physicians either in the USA or Japan. A brief multiple choice survey (44 questions) was made available through the internet. The data was collected on Survey Monkey. The survey was written in English and translated into Japanese. Physicians were not paid any fees for completing the survey.
Forty-one questions were multiple choice, three were free-text. The physicians were contacted by mail and no follow-up communication occurred. There were four sections to the survey, emphasising the following themes: physician demographics, practice characteristics, management of CVD risk factors and health policy.
The survey asked each physician about their prac-tice behaviours, for example, how often they typically saw a patient with hypertension and hyperlipidaemia for on-going management (e.g. every two weeks, every six months) and how much time they typically spent with a patient with one of these problems (e.g. 2, 10, 15 minutes).
Physicians were asked if they used electronic medi-cal records and what percentage of their patients were successfully treated to national or regional guidelines for hypertension and hyperlipidaemia. The survey asked what factors influenced their selection of medication in treating patients with risk factors for CVD and what their primary focus was while caring for these patients (lifestyle modification, medication or both), if they had a reminder system for on-going care, if they routinely provided patient education (and what per-centage of visit time they estimate they spent on patient education), if they provided patient education ma-terials, if they had easily accessible information on current guidelines, if they routinely followed guide-lines, and if they had Peer Review of their outcomes, either internally from a colleague or externally from the government or insurance company. The survey asked if race, cultural issues or lack of health insurance affected the care of their patients, and if they were influenced by the government or insurance companies in making medical decisions. They were asked if their personal income or the revenue brought to their practice group was influenced by their clinical out-comes in CVD risk management and if so how much it affected their income (e.g. 20, 40, 60 and 80%). This study was approved by the Institutional Review Board of both Wright State University and the University of Georgia. Participants provided informed consent by participating in the study.
Data were collated and responses of the Japanese physicians were compared to the US physicians (system to system) in parallel to CVD death rates using SAS version 9.1 for analysis. Nominal data were analysed using the chi-squared or Fisher’s exact test, as appro-priate. Wilcoxon’s test was used to test the differences of physician visit time between these two countries. Ordinal data were analysed using the Cochran– Armitage test for trend to seek differences in physician practices between the nations. A P-value of < 0.05 was considered statistically significant for all statistical tests.
Forty-eight Japanese physicians and 53 US physicians completed the survey. This was a convenience sample. The responses came from a small and non-represen-tative sample of physicians in each country. However, the number responding in each group was similar and statistical significance was sought between the two groups of physicians. Japanese physicians were older and many were not primary care physicians. In the USA, primary care physicians were more fully repre-sented (Table 1).
Practice management behaviours
There were significant differences between the Japanese and US physicians who participated in the survey in practice management of CVD risk factors (Table 2). Patient visits are different in terms of both patient volume and visit time. Two thirds (66%) of the Japanese physicians in the study saw more than 125 patients per week. This is many more than the US physicians, 74% of whom said they saw between 25 and 125 patients per week (P < 0.0001). When asked how long the typical office visit lasts, 98% of Japanese physicians claimed <10 minutes, while 76% the US physicians claimed to spend 10–20 minutes per visit. Only 23% of the US doctors reported spending <10 minutes with a patient, and only 2.1% of Japanese physicians claimed to spend >10 minutes (P < 0.0001) (Figure 1).
The differences in follow-up time for appointments between these two countries were significant. The vast majority of Japanese physicians (87%) preferred to follow-up after <3 months, while only 32% of the US physicians did so. By contrast, 68% of the US phys-icians chose to follow-up after >3 months, whereas only 13% Japanese physicians did (P < 0.0001). Most Japanese physicians (61%) reported spending <30% of the time educating patients, while the majority of US physicians (60%) contended that they spent >30% of the visit on patient education (P = 0.043). Differ-ences in prescription renewal frequency between the two countries were significant. Most of the Japanese physicians (83%) prescribed medicine for <1 month, whereas 75% of the US physicians prescribed for 1–6 months (P < 0.0001).
The attitude toward CVD guidelines usage was different between the two groups. The US physicians were more likely to use guidelines (P = 0.0413). Fifty percent of them reported using guidelines frequently (>80% of their patients) with CVD risk factors, whereas only 20% of Japanese physicians reported use of guidelines in >80% of their patients. Although both groups used guidelines, their attitudes toward general and strict usage varied (P = 0.009). Most (69%) US physicians said they followed the guidelines carefully or strictly, while only 43% of the Japanese physicians said the same. Most Japanese physicians (57%) said they used the guidelines more generally or not at all, compared with 31% of the US physicians. When selecting medications, Japanese physicians thought guidelines (38%), personal experience (26%) and journals/textbooks (21%) were the most important factors, while US physicians tended to think price (45%), guidelines (39%) and personal experience (13%) were the most important factors (P < 0.0001).
National social factors impact questions
Financial rewards for performance at any level were reported by 36% of US physicians and only 15% of Japanese physicians (P = 0.0004) (Table 3). Almost all of the Japanese physicians (96%) reported that > 80% of their patients were covered by private or public insurance, while only 23% of the US physicians made the same assessment (P < 0.0001).
Over half (53%) of the US physicians thought social factors such as racial, economic and cultural issues affected patient care, whereas 21% of Japanese phys-icians thought so. Sixty percent of Japanese physicians thought that health care was generally the same to all patients in their own practice, whereas only 17% of US physicians reached the same conclusions (P < 0.0001). Generally, physicians’ opinions about whether they could practise freely were quite different in these two groups (P < 0.0001). Just under half of the Japanese physicians (43%) thought they could practise freely, whereas only 9.6% of the US physicians felt that they could practise without interference from the govern-ment or insurance companies. As many as 81% of US physicians thought they could not make decisions freely, whereas 34% of Japanese physicians similarly felt they were not free to make clinical decisions.
Nearly half of Japanese physicians (47%) versus 9.6% of US physicians reported that >40% of their patients had blood pressure cuffs at home and took blood pressure recordings frequently themselves (P = 0.0001) (Figure 2). More Japanese patients reportedly knew their blood pressure and cholesterol levels. Most (70%) Japanese physicians reported that >40% of their patients knew their levels, while only 34% of US physicians reported the same (P = 0.0002) (Figure 3). As many as 75% of the US physicians believed their patients did not follow good lifestyle practices, while 36% of Japanese physicians felt the same (P < 0.0001) (Figure 4).
The limited sample size and recognition of this being a convenience sample means conclusions are tentative. More definitive data are needed by either using a much larger sample or by direct observation of physicians in the management of CVD risk factors.
Practice management behaviours
Japanese physicians appeared to have higher work volumes. Those responding to the survey saw more patients in a week and spent less time on each visit. During the limited patient visit time, the Japanese physicians spent less time educating patients but they tended to follow-up in a shorter period and prescribe more medication to patients. Virtually all medical care is paid the same nationally to all providers as a uniform fee schedule on a fee-for-service basis in Japan. As physicians make relatively less revenue per visit than US physicians, there may be economic forces promoting short and frequent visits. Japanese patients see their physician more frequently than do US patients. For CVD risk factor management, more intensive care leads to better outcomes. Do more frequent, short visits lead to better outcomes in CVD risk factor management? A more comprehensive study is needed to address that question. The Japanese physicians surveyed report less interest in guidelines relying on their own experiences in patient care.
National social factors impact questions
Japanese physicians in this survey believed they prac-tised more freely than US physicians. They did not need to worry about insurance for most of their patients, and were less influenced by social factors like racial, economic and cultural issues. In Japan, the universal health insurance system was established in 1961. Japanese citizens are covered comprehensively and exclusively by national medical insurance and social insurance. Patients cannot pay privately for the care they receive. Patients have guaranteed access to any health providers without additional fees. Finally, the providers are reimbursed by the government though a fee-for-service process.
The physicians responding to the survey in both countries agreed that their patients were not too involved in managing their own risk factors, but Japanese patients appeared to be more involved in their care management as perceived by their phys-icians: more Japanese patients reportedly have blood pressure cuffs at home; more Japanese patients were reported to know about their levels of blood pressure and cholesterol; more Japanese patients were reported to follow ‘good’ lifestyles.
In Japan, health care is not only an individual issue but also a matter for the whole family. Japanese patients rely more on family and physician authority and place less emphasis on patient individualities. Therefore, patients might be more compliant to phys-ician directives. In Japan, nearly 62% of elders live with their children and it is culturally believed that children should provide care for their elders at home. Therefore, when a patient is found to have hyperten-sion, either the patient or their family may purchase a blood pressure cuff and take the blood pressure regularly.
Any conclusions reached by this study must be validated either by a larger study, a more representa-tive sample, or direct observation of physician prac-tices. The authors recognise that this study, using a convenience sample, provides observational evidence only. The nature of the survey cannot result in com-prehensive conclusions demonstrating cause-effect relationships. There were only 47 Japanese physicians and 53 US physicians in our study and it is possible that these small numbers may not be representative in either country. In addition, the groups of physicians in the two countries were of different ages and specialties. Recall versus direct measurement of practices always has limitations in accuracy. Results are based on a physicians’ survey not observation of their practices. Finally, cultural differences could influence survey responses.
Physician practice behaviours and the factors affecting practisingin Japan and the USA were compared focusing on CVD risk factor management, and attempting to seek evidence from the health system perspective for why CVD death rates are so much lower in Japan than the USA. The primary reasons are most likely unrelated to the medical care system and instead are likely to be a function of diet, public health, culture and the social determinants of health. Yet, there were clear differ-ences in the physicians’ responses to the survey ques-tions in how they reported managing CVD risk factors. The Japanese physicians reported seeing patients more frequently but for shorter visits. This could be a better process for chronic disease management. Japanese physicians reported practising more freely, with fewer factors affecting practice including social, systemic and financial issues. With a ‘single-payer’ system of care, Japanese physicians perceived less interference by insurance companies or the government than did US physicians. The freedom to practice by these Japanese physicians may contribute to less use of practice guidelines. Japanese patients may be more involved more in care management and more actively follow physicians’ advice.
The authors would like to thank Natalie Arford and Colleen O’Brien for their support in the manuscript preparation. Dr Claude Lenfant, has played a critical role in focusing the hypothesis and establishing the bi-national contacts. The Shiga Prefecture Medical So-ciety and Dr Sadahiro Asano, the former president, provided the contacts with local physicians in Japan.
This project was approved by the Institutional Review Board of both Wright State University and the Uni-versity of Georgia.
Not commissioned; externally Peer Reviewed.
Dr Schuster has received Funding from Pfizer for other project efforts. There was no outside financial support for conducting this study. This paper has not been submitted elsewhere for publication; duplicate publi-cation has been avoided.