Is the Canadian Health Care Model in Crisis? ACHIEVEMENTS AND CHALLENGES OF MEDICARE IN CANADA: ARE WE THERE YET? ARE WE ON COURSE?

Transcription

1 Is the Canadian Health Care Model in Crisis? ACHIEVEMENTS AND CHALLENGES OF MEDICARE IN CANADA: ARE WE THERE YET? ARE WE ON COURSE? Stephen Birch and Amiram Gafni Health care policy in Canada is based on providing public funding for medically necessary physician and hospital-based services free at the point of delivery ( first-dollar public funding ). Studies consistently show that the introduction of public funding to support the provision of health care services free at the point of delivery is associated with increases in the proportionate share of services used by the poor and in population distributions of services that are independent of income. Claims about the success of Canada s health care policy tend to be based on these findings, without reference to medical necessity. This article adopts a needs-based perspective to reviewing the distribution of health care services. Despite the removal of user prices, significant barriers remain to services being distributed in accordance with need the objective of needs-based access to services remains elusive. The increased fiscal pressures imposed on health care in the 1990s, together with the failure of health care policy to encompass the changing nature of health care delivery, seem to represent further departures from policy objectives. In addition, there is evidence of increasing public dissatisfaction with the performance of the system. A return to modest increases in public funding in the new millennium has not been sufficient to arrest these trends. Widespread support for first-dollar public funding needs to be accompanied by greater attention to the scope of the legislation and the adoption of a needs-based focus among health care policymakers. The primary objective of Canadian health care policy is to protect, promote and restore the physical and mental well-being of residents of Canada and to facilitate reasonable access to health services without financial or other barriers. Canada Health Act (1) International Journal of Health Services, Volume 35, Number 3, Pages , , Baywood Publishing Co., Inc. 443

2 444 / Birch and Gafni In the political climate of the past decade, considerable attention has been given to controlling or reducing public expenditures in Canada. In the case of health care, this has led to a steady stream of interest in reform and restructuring of provinces health care systems as means of accommodating these increased fiscal pressures (2). However, the objective of health care policy, and the principles of public funding for services delivered free at the point of delivery (in Canada referred to as first-dollar coverage for health care) through publicly administered insurance funded by general taxation that have emerged from this objective, have remained unchanged throughout this period (3). Support for the system is generally based on the findings of research that compares the achievements of Canada s provincial health care systems with the achievements of the much more expensive health care sector in the United States (4). Comparative analyses of Canada and the United States may provide important messages about the planning, management, delivery, and funding of health care services. However, the value of such comparative analyses as an indicator of the performance of the Canadian system depends on the two countries having similar health care goals. Marmor and colleagues (5) note that the features of many of the social programs, including health care, in the United States reflect the prevailing values of the U.S. population values that are not generally shared by other populations. If this is so, it is less useful to judge Canada s performance by reference to systems that are designed to pursue very different goals unless, of course, the U.S. system is performing as well as or better than Canada, even with respect to Canadian goals. More useful analyses could be based on comparisons with jurisdictions with similar goals to Canada s. But the results of such comparisons, although potentially more revealing than Canadian-U.S. comparisons, still focus attention on how we are doing compared with (some) others rather than how we are doing in relation to what we set out to do. In this article, attention is focused on assessing the performance of Canada s health care systems with respect to the stated objective of health care policy. We first describe briefly the development of health care policy in Canada, with its policy focused primarily on mechanisms that enable individuals to access care without payment. In the absence of price, however, the demand for care increases as individuals ability to pay no longer acts as a constraint on their demands for care. Overlooked in the development of this policy were the mechanisms required to ensure that, faced with increased demands for care, services would be distributed in accordance with the objectives of the policy the protection, promotion, and restoration of health of the population (or population needs). We then introduce a needs-based perspective on the performance of the system. As previous research indicates, systematic barriers to care in relation to need remain. Moreover, there is some evidence to suggest that the effects of these barriers are associated with differences in the planning, management, and delivery of services among provinces, and the barriers seem to have increased over time as fiscal pressures on the system have increased. Finally, in the case of

3 Canada s Health Care Policy / 445 some preventive services, the inequities in the distribution of services seem to be as great under Canada s policy of care for all, free at the point of delivery, as under the patchwork of programs and mixture of funding provisions for the same services in the United States. We present some possible explanations of the apparent limitations and deterioration of the performance of Canada s health care policy. In particular, the achievements of the program of first-dollar public funding seem to have been compromised by the focus on specific health care programs, the declining role of the federal government in financing health care services, and the absence of considerations of need in issues of resource allocation. We conclude that the most recent developments offer little reason for optimism not only are we not there yet, but there is reason to believe we might no longer be going in the intended direction. HISTORICAL DEVELOPMENT OF THE MEDICARE PROGRAM IN CANADA Under the 1867 British North America Act, health care in Canada was one of several low-cost items deemed to be a responsibility of the provinces. As the cost of health care rose over time, provincial governments turned increasingly to the federal government for financial support. Following World War II, concerns with unequal access to health care led several provinces to introduce public hospital insurance. This led to increasing pressure on the federal government to get involved in health care, but the articles of confederation restricted what role it could play (6). Federal action concentrated on setting explicit conditions to be met by provincial governments to qualify for federal cost sharing of provincial health care programs. In 1957, the Hospital and Diagnostic Services Act laid down conditions for federal funding of hospital insurance programs. This was extended to physician services under the 1965 Medical Care Act. In 1984 these acts were consolidated under the Canada Health Act. The Canada Health Act (1) set out explicit objectives for health care policy. Under the Act, policy levers remain confined to the conditions that provinces are obliged to meet to qualify for federal financial support. The five conditions require that the entire population of a province be covered by the provincial plan; that the plans cover a comprehensive range of hospital and physician services; that coverage be portable between provinces; that the plans of each province be administered by a public authority; and that the terms of access to services be the same for all in the province and based on the notion of medical necessity. The level and form of cost sharing is not specified in the Act. At the time of the legislation, provinces received approximately 50 percent of the cost of their programs from the federal government once the qualifying conditions were satisfied (7). This incentive was sufficient for each province to introduce its own legislation to ensure that the conditions were met. For example, physicians must opt out of billing a province s plan entirely in order to charge patients for

4 446 / Birch and Gafni services covered by the Act. In this way, physicians cannot extra-bill patients for insured services. This Medicare program arose from concern over the prohibitive cost of health care and the importance of insurance against the costs of care being made available to all members of the population (6, 7). Poorer groups were observed to make less use of health care services than better-off groups, even though they suffered more than their fair share of sickness. Inequalities in ability to pay for health care were seen as a cause of inequalities in use. Poorer groups were unable to demand the same level of care as better-off groups because they did not have the same capacity to pay for care. In this way the problem was viewed as a demand-side issue. The price barrier to medical care was removed by creating provincial monopolies for buying hospital and physician services (what are referred to under the Act as the insured services). Federal cost sharing was the lever to get each province to outlaw all aspects of private payment for these insured services. Little attention was paid to the role of supply factors in determining the distribution of services. The levels and distribution of supply, as well as the mechanisms for the planning, management, and delivery of services, were inherited from the pre-medicare era. Public payment of private providers was based on physician fee schedules set by provincial ministries in negotiation with medical associations and cost reimbursement of hospitals determined by each hospital s activity. Ability to pay was proscribed as the mechanism by which services were shared among competing demands. Nothing was put in its place to serve this important economic role. It was implicitly assumed that, in the absence of prices, the increased demands for care would be served in ways that did most to protect, promote, and restore the health of the population. In other words, available health care resources would in some way follow needs for care. Those with poor health or at greater risk of health problems would receive priority in the allocation of health care resources. Conventional wisdom indicated that the Medicare program was achieving the objectives set out in the Canada Health Act. For example, the former federal minister for health, Monique Begin, claimed that under the medicare program in Canada everyone has equal access to quality health care (8), while Evans argued that the Canadian approach has been very successful in equalizing access to health care services (9). Bob Rae, then premier of the country s largest province, Ontario, went further, claiming that under his government s stewardship people will continue to have access to the best medical care system in the world (quoted in 10). But what is the evidence on the performance of the Canadian system with respect to the accessibility principle of allocating resources in line with medical necessity? Research on the effect of user charges on the population distribution of care indicates that the Medicare program has led to a more equal distribution of care among income groups. The role of patient payment in service utilization was analyzed by Beck and Horne (11, 12). They studied the effect of

5 the imposition of a user fee on physician services in Saskatchewan before the Canada Health Act. The fee was associated with a reduction in use of these services. The authors noted that this reduction was considerably greater among poorer families. Moreover, the reductions were not temporary delays in utilization; the income-based differences in use persisted over time. With total public spending on health care remaining relatively stable, this represented a redistribution of resources from the poor and less healthy groups in the population to the wealthier and healthier groups. Enterline and colleagues (13) compared physician use in Montreal before and after the introduction of Medicare. Before the abolition of private payment, use was positively correlated with household income. This relationship was reversed following removal of user payments. Over the longer period, utilization was found to be independent of income. Similarly, Barer and colleagues (14) found that, in Ontario, in the absence of direct charges to patients under Medicare, the incidence of hospital and physician services was independent of income. Moreover, the volume of services received was greater among poorer groups. Badgely, in a broad review of 30 studies, concluded that the introduction of Medicare had been associated with the progressive redistribution by class in the use of health care (15, p. 662). The claims about the success and, in some cases, superiority of the Canadian approach of nationalized payments and prohibition of private payment are based on the findings of these and other studies showing that Medicare was associated with a redistribution of services by income group. It is ironic to note that the objectives of the policy, as stated in the Canada Health Act, were not concerned with the redistribution of services by income but with the deployment of health care resources in accordance with need, irrespective of ability to pay (16). Insofar as factors other than ability to pay might be expected to affect utilization, such factors warrant inclusion in any consideration of the performance of the Canadian health care system. A NEEDS-BASED PERSPECTIVE ON MEDICARE IN CANADA Canada s Health Care Policy / 447 We noted above that the Medicare program was based on the notion that payment for services at the point of delivery is a barrier to access to medically necessary care. By abolishing user prices, services would be made accessible on the basis of medical necessity irrespective of the particular approaches used by provinces to the organization of health care delivery. Although medical necessity was not defined in the legislation, the conception of need has been interpreted consistently by governments at both the federal and provincial levels in terms of the prevailing level of health or level of risk to health of an individual or group (16, 17). Those with poorer health, or facing greater risks to their health, have greater needs for services that might be expected to improve their health or reduce

6 448 / Birch and Gafni their risks. Although there is no single objective measure of health status, alternative approaches to measuring the differences in health among subgroups of the population show a high level of correlation (18). These range from the more objective but narrowly defined measures such as life expectancy and mortality rates to the more subjective but broader measures such as self-assessed health status. Here we draw on the findings of research performed over the past decade to shed light on these issues. What role has been played by need in explaining the distribution of services in the population? Has the role played by need been associated with interprovincial differences in the organization of the delivery of services? Has the role of need changed as the fiscal pressures on provincial health care programs have increased? Finally, is the degree to which need is met, or resources are allocated in accordance with relative need, greater in Canada, where services are free at the point of delivery, than in systems where users are responsible for the cost of services? Many studies have shown that, in terms of service utilization, the poor are not disadvantaged under Medicare, leading Badgely to conclude that the program has been associated with a progressive redistribution by class in the use of health care (15, p. 662). However, little if any consideration had been given to differences in need for care among income groups. Manga (19) noted the potential paradox of a distribution of services that is independent of income also being inequitable in terms of access to care in relation to medical necessity. Comparisons of income-based gradients in the incidence of utilization with corresponding gradients in the incidence of activity-limiting health problems help to illustrate the issue (20). The data in Table 1 are taken from the General Health Questionnaire, a survey of a random sample of the non-institutionalized population in 1985 (21), soon after reinforcement of the principles of Medicare under the Canada Health Act. Dental care is not an insured service under the Act and so is largely funded privately. Individuals in the poorest households were half as likely to have visited a dentist as those in the wealthiest households. Moreover, there is a clear gradient between the richest and poorest groups. The pattern for family physician services is different. The probability of having visited a family physician was the same for all income groups confirming the results from other research studies. Turning attention to need, we see that individuals in the poorest group were 2.5 times as likely to have had an activity-limiting health problem compared with individuals in the richest group. So, although the poor were more likely to report a health problem than the rich, they were not more likely to report having used health care. As Manga (19) noted, providing access that is independent of income may not be sufficient to provide access that is in line with need. The data in Table 1 represent simple cross-tabulations of income with use and need that fail to control for other factors that might be associated with utilization (16). Birch and colleagues (20) therefore analyzed the relationship between utilization and income using a range of other sociodemographic variables in order to identify factors that might impede or facilitate access to care within groups with

7 Canada s Health Care Policy / 449 Table 1 Relative risk of health care use and need by household income, Canada, 1985 Relative risk, by family income, $ thousands In previous year: < >50 Visited dentist Visited family physician Had activity-limiting health problem Source: Birch and Abelson (16). the same levels of need. After controlling for need and other sociodemographic variables, they found that variations in income were not significant in explaining variations in use of family physician services. In a corresponding analysis of hospital service use, Newbold and co-workers (22) found that, after controlling for need and other sociodemographic variables, admission to hospital was greater among low-income households, but once admitted, length of stay was independent of income. Further analysis of family physician utilization found that variations in use among individuals in poor health were not explained by any of a considerable number of variables. Birch and colleagues (20) found that, for those in other levels of health, higher education, social support, and female gender were significantly associated with a greater probability of using physician services. In other words, no barriers to equal use were found for those individuals in greatest need. However, for those in less need, which is not the same as no need, these factors seem to play an important role in facilitating use within groups of individuals with similar levels of need. The abolition of private payment was not sufficient on its own to ensure that access to services was needs-based. The same research group (23, 24) explored the extent to which the achievements of Medicare were robust to differences in the organization of health care under the same public funding arrangements. A comparative analysis of the use-need relationship in Ontario and Quebec, provinces with substantial differences in the planning, management, and delivery of health care, found that household income was important in explaining the relationship between use and need in Quebec but not in Ontario (23). In other words, source of funding alone was insufficient to liberate utilization from income considerations. Similarly, organizational elements change in response to research, innovation, and changes in the fiscal climate over time, even though the source of funding remains the same. By the late 1980s, concerns with the escalating costs of the Medicare program led to greater attention being paid to cost-containment strategies in terms of controls on the quantity of supply (e.g., hospital beds and

8 450 / Birch and Gafni physician numbers) and on total expenditures (e.g., hospital global budgets, physician billing thresholds and clawbacks) (25). Eyles and colleagues (24) compared the patterns of use and need in 1991 with those in 1985 as a way of examining whether the achievements of single-source public funding are resistant to these organizational aspects of the system. They found that income-based patterns for the use of services (Table 2) remain similar to those observed for Dentist utilization shows a strong income gradient, but physician services are independent of income. The income gradient for need increased, however, with all income groups reporting a greater incidence of activity-limiting health problems in relation to the top 20 percent of the income distribution than in In other words, increases in the income-based differences in need over time were not reflected in income-based differences in utilization. These authors went on to show that, after controlling for need, individuals in employment and older age groups were more likely to use services than the unemployed and younger age groups. This suggests that some groups were less affected than others by the general belt tightening in provincial health care systems and its manifestations need was not the invisible hand policymakers had assumed it would be. Finally, Katz and Hofer (26) estimated income-based differences in the use of screening for cervical and breast cancer in Ontario, where services are funded entirely from the public purse. These were compared with corresponding incomebased differences in the use of the same services in the United States, with its patchwork of funding sources and predominance of private payment. Population average rates of screening were similar in both countries except for mammography, which was more likely to have been performed in the United States than in Canada. In both Ontario and the United States, however, highincome women were approximately twice as likely to have had a Pap smear or a clinical breast exam as low-income women, after controlling for age, education, and marital status. In the case of mammograms, the income-based differences Table 2 Relative risk of health care use and need by household income, Canada, 1991 Relative risk, by family income, $ thousands In previous year: < >60 Visited dentist Visited family physician Had activity-limiting health problem Source: Eyles et al. (24).

9 Canada s Health Care Policy / 451 were about the same as for Pap smears and clinical breast exams in Ontario, but were greater in the United States. However, the differences between Ontario and the United States were not statistically significant. First-dollar public funding of health care, this suggests, is not on its own sufficient to ensure that medical necessity drives a health care system. Based on comparisons of screening services among women in Canada and the United States, first-dollar public funding may not even be necessary to achieve the distribution of services achieved in Canada. Not having to pay directly for a service seems to have little influence on its own on whether women in different social groups use these services. TOWARD AN EXPLANATION OF THE PERFORMANCE OF MEDICARE We argue here that changes in the health care environment over the past decade have been associated with a reduction in the effectiveness of federal cost sharing as a means to promote the objectives of the Canada Health Act. In this way, the current provisions of the Act are unlikely to be sufficient to address the deterioration in performance of the Canadian model of health care funding. Under the Canada Health Act, the federal government s role and influence is restricted to its contribution to the cost of provincial program. Over time, the level and share of the federal government contribution has diminished. For example, in Ontario, expenditure on insured programs, physicians and hospitals, fell by 2.5 percent between and Over the same period, federal transfers to the province for the support of all social programs (including health) fell by 32 percent (27). Table 3 presents data from the Canadian Institute for Health Information s report on national health expenditures in Canada (28), showing trends in the level and share of federal government contributions to insured services. Although the downward trend in the level of federal transfers was reversed at the turn of the century, the proportionate contribution to Table 3 Federal transfers and public health care expenditure on hospitals and physician services, Canada, to a Hospital and physician services expenditures, $ millions Federal transfers, $ millions Contributing federal transfers, % 33,812 8, ,029 6, ,246 7, ,752 8, Source: Canadian Institute for Health Information (28). a Projected expenditures.

10 452 / Birch and Gafni expenditures on these services continued to fall. Moreover, a recent federal government commission on the future of health care in Canada noted that this declining share of federal government funds is expected to continue in the future. Thus, with federal cutbacks in transfers to provinces, an increasing responsibility for the cost of the insured programs is being shouldered by the provinces. The pressures to reduce budget deficits that lay behind the federal cutbacks are not confined to federal government, however. Provincial government resources are squeezed by their own deficit-reduction plans and by the increasing cost to provinces of the insured program. The double-edged sword has caused provinces to look to reduce the demand on insured programs as well as to find other funding sources for health care programs not covered by the federal legislation. The diminishing contribution of federal government to the cost of insured programs also reduces the opportunity cost to provinces of not meeting the conditions of the Canada Health Act. The outcome of these various influences can be seen in Table 4. Between 1992 and 1996, real health care spending in Canada rose by 3 percent. However, the contribution of public funds fell from 74 cents in every dollar to 71 cents per dollar. Both population size and gross domestic product (GDP) increased during the same period. As a result, public expenditure on health care per capita fell by 6.5 percent, notwithstanding the aging of the population, and the proportion of the country s production allocated by governments to health care fell by 12 percent over the same period. In other words, the federal government s commitment to the principle of first-dollar public funding for insured services held firm, but its willingness to support this commitment with dollars was ebbing away. The improved fiscal situation at the turn of the century saw both provincial and federal governments loosen the purse strings somewhat. Total health care spending increased substantially between 1996 and 2001, but the share of GDP spent on health care increased only marginally, remaining well below the 1992 level, and the contribution of public spending on health to total health care expenditures continued to decline. Table 4 Health care expenditure, Canada, Total, 1997, $ millions % Public % Private Per capita, public % GDP, public 74, , , , , , Source: Canadian Institute for Health Information (28).

11 Canada s Health Care Policy / 453 Despite the prohibition of private payment for insured programs, private payment has made up an increasing share of total spending on health care. As public spending on health care per capita diminished in the 1990s, private spending continued to increase at a rate of 3 percent per annum. Because of the provisions of the Canada Health Act, this shift toward private spending is not spread equally across services, as shown in Table 5. Private funding of hospital and physician services is proportionally very small, given the provisions of the Canada Health Act, representing services that are deemed not medically necessary and hence are outside public plans or services provided by physicians who have opted out of Medicare. In the case of hospital services, the private contribution fell steadily between 1992 and However, the private contribution to physician services has increased steadily over the same period. Other services have seen an increase of around six percentage points in the contribution of private funds during this period. The implications of this trend are more profound when considered alongside trends in the nature and scope of health care services (Table 6). For example, in 1992, 53 cents in every public health care dollar was spent on physicians and hospitals. By 2001 this had fallen to only 44 cents per dollar. During this period, the proportion of public health care expenditure absorbed by hospital care fell by almost 20 percent but the proportion spent on prescribed drugs increased by 43 percent. Similarly, Naylor (29) reported a 50 percent increase in the proportion of health care public spending on home care between 1987 and In other words, an increasing proportion of publicly funded health care services delivered to Canadians falls outside the categories covered by the Canada Health Act and an increasing proportion of expenditures on these services is being contributed by private payment. Provinces have the discretion to apply the Canada Health Act principles to these community-based services but not the federal resources to help support them. Provincial capacity to take a greater responsibility for the cost of community-based (and hence non- insured ) services has Table 5 Private share of health expenditures by selected category, Canada, Category Hospitals Physicians Other professionals Prescribed drugs Other institutions All categories Source: Canadian Institute for Health Information (28).

12 454 / Birch and Gafni Table 6 Distribution of public health care spending by selected category, Canada, Category % Increase (decrease) Hospitals Physicians Other professionals Prescribed drugs Other institutions (19.9) (12.0) (3.1) Source: Canadian Institute for Health Information (28). been eroded by provinces increasing role in funding the insured programs and the increasing utilization of community-based programs. At the same time, de-hospitalizing service delivery provides a way for hospital managers to off-load demands from their tight budgets. Once discharged, patients must pay for many aspects of care that the hospital budget would otherwise have been responsible for. In some cases, relocation of care in the community leads to competition for care from other public agencies. For example, physiotherapy services provided to hospital inpatients are included as insured services under the Canada Health Act. Increased pressure on hospital budgets and reductions in inpatient lengths of stay have led to a shift in the balance of physiotherapy services to outpatient or community settings. Provision of physiotherapy services in these settings is not an insured service under the Act. However, individual provinces can incorporate these services under their provincial health care plans. Because the Act does not apply to these services, however, there is no prohibition on providers (i.e., physiotherapists) providing the same services under other funding arrangements. For example, in Ontario, the Workers Compensation Board can pay rates for physiotherapy services that are higher than those provided by the provincial insurance plan in order to reduce waiting times for their clients and to accelerate return to work. The effect is the same as if the competitor was a private individual paying out of pocket, or a for-profit insurance company. Scarce health care resources are allocated on the basis of the source and level of payment, as opposed to medical necessity or ability to benefit the patient. Patients who rely on the provincial plan for their physiotherapy services receive less priority than patients whose care is funded under other arrangements and hence is not limited to the rates provided by the provincial plan, irrespective of medical necessity (30). Because this happens outside hospitals, the Canada Health Act is unable to prevent this distortion in the allocation of scarce health care resources. The Ontario

13 Canada s Health Care Policy / 455 Ministry of Health recently announced its intention to delist physiotherapy services from the list of services covered bv the provincial health insurance plan, which, if implemented, would increase the barriers to accessing services for those patients not covered by the Workers Compensation Board. Where the relocation of care is not protected from alternative sources of payment by the federal legislation, ability to pay, closely correlated with incomes, becomes the method by which services are allocated. Income-based inequalities in access to these services are perpetuated by unequal access to private insurance coverage, much of which is employer-based. For example, Naylor (29) reports that the proportion of individuals with private coverage for prescribed drugs among the highest-income group was more than ten times that among the lowestincome group (Table 7). NEED AND THE HEALTH CARE SYSTEM OF THE NEW MILLENNIUM So what are the prospects for change? In 1994 Prime Minister Chrétien appointed a National Forum on Health to advise the federal government on innovative ways to improve the health system and the health of Canada s people (3). Its recommendations generally parallelled the overall thrust of the population health program of the Canadian Institute for Advanced Research (31), with the main emphasis on issues beyond the health care system. When attention did turn to health care, its recommendations were largely conservative and in many cases unclear. It recommended continuation of first-dollar public funding for medically necessary services but did not address the difficult issue of what constitutes medical necessity (17). It suggested that current levels of funding were sufficient and should be the target for the future, and that maintaining health care expenditures at the current observed level of around 10 percent of GDP is sufficient to support access to needed services. However, setting expenditure in line with Table 7 Private insurance coverage for drugs by income group, Canada, 1995 Income, $ thousands > <20 % Private insurance Source: Naylor (29).

14 456 / Birch and Gafni GDP essentially means affordability of, or willingness to fund, health care is independent of the determinants of the need for care. For example, if population growth outstrips GDP growth, under this approach health care expenditure per capita would fall. But even if GDP, populations, and needs per capita grow at similar rates, a policy of maintaining the rate of health care expenditure at current levels does not address the issue of the public-private mix in this rate of expenditure. Private health care expenditure is largely outside the control of governments. Hence, a target for health care expenditure as a proportion of GDP involves limiting or expanding public expenditure, which in principle can be used to achieve Canada Health Act objectives, in accordance with changes in private expenditure which is more difficult to manage in accordance with the Act s objectives. For services such as prescribed drugs and home care, this would be achieved by the National Forum on Health s proposal to extend the scope of the Canada Health Act to these services, hence outlawing private payment. However, dental and optical care, for which expenditures rise at a greater rate than publicfunded services, would remain privately funded and hence demand determined. Finally, the forum proposed a significant role for evidence-based decision-making and a shift toward client-based funding for primary care in order to increase the productivity of health care spending in a budget-conscious environment. This portfolio of recommendations did not address two important points. First, the impact of cost containment is not spread equally among the community. Instead, some groups, most noticeably those with higher incomes and better education, seem to be able to survive cost containment better than their less prosperous peers, even within the context of a publicly funded system. It may therefore be difficult to sustain sole-source public funding together with equitable access to care if better-off groups are unwilling to share the consequences of cost containment with their less prosperous and less powerful peers. Second, the notion of medical necessity as a rationing device is at odds with systems of physician and hospital remuneration that are largely throughput-based (i.e., payments relate to the number of services, not the level of need of individuals receiving those services) and with evidence -generating research that is service focused (it asks which services work on average in a population? not what is the most appropriate way for dealing with an individual patient given that individual s particular circumstances? ) (17, 32, 33). Medical necessity is an individual-focused construct in which need for care depends on the cultural, social, environmental, economic, and physiological circumstances of the individual. An evidence-based health care system takes a homogeneous approach to addressing what are essentially heterogeneous problems, which risks producing increased inequalities in health through the tendency to find solutions to the health problems of the better-off groups (33). More recently, two reports on the future of health care in Canada have called for increases in public expenditure on health care, with an increased share of that expenditure being contributed by the federal government (34, 35). Much of this

15 Canada s Health Care Policy / 457 increase would be used to support an extension of national-level programs beyond the current insured services. These would include publicly funded national-level programs for home care services and prescribed drugs, supported in part by earmarked federal government dollars. However, neither report proposes that this be achieved by incorporating these programs as insured services under the existing Canada Health Act and hence subject to the five principles of the Act. Instead the proposals would simply add to the patchwork of programs that currently defines health care services in Canada. The federal government has yet to adopt these proposals, preferring to give priority to programs aimed at reducing waiting times for hospital-based services. Moreover, the premier of Ontario has already indicated that any federal government transfers earmarked for specific services such as prescribed drugs would be used, at least in part, to free up existing provincial government expenditures as opposed to supplementing those expenditures. The Canada Health Act set out to ensure that access to health care would be based on medical necessity in relation to the protection, promotion, and restoration of health of Canadians. Despite changes in the fiscal climate, population expectations, and the demographic structure of the population, commitment to this objective has remained firm. Successive reports on health care in Canada have strongly supported the objective and, in the case of the National Forum report, have called for an expansion of the scope of the Act to incorporate other services (3). So, are we there yet? Or are we any nearer to getting there? We have argued that we still have some way to go. Moreover, recent trends, together with policy responses to those trends, may have blown us off course. Over time, differences in need, as proxied by differences in health status, have played a diminishing role in explaining the distribution of services in the population. The development of provincial approaches to cost containment have paid little attention to the implications for the distribution of services in the population but have been associated with an erosion of satisfaction with the health care system among both providers and consumers. In early 1999, physicians in British Columbia found that their levels of activity had exceeded the caps imposed on billings by the provincial government. Faced with providing additional services for no additional reward, physicians closed their offices for a week. Tight controls on nursing pay levels together with the contracting out of nursing services to agencies led to nurses strikes in four provinces in the first half of In Ontario, concerns about physician service expenditures led to the appointment of a panel to recommend which medical services should be delisted from the provincial plan, using the medical necessity criterion (36). The panel had the task of finding $20 million of savings in physician expenditures by identifying services that should not be covered under the provincial plan. Once services are delisted, physicians are free to provide them privately to paying customers. Items delisted ranged from tattoo removal to in vitro fertilization, while services considered but not delisted included general anesthesia for noninsured dental

16 458 / Birch and Gafni procedures. The exercise focused on what a physician could not bill the provincial plan and paid no attention to what the physician would do instead. However, the trend in physician billing remained steady and the $20 million savings did not materialize, suggesting that the physician time freed up by not providing publicly funded tattoo removals and IVF was used to provide other services that remained publicly funded. In order to achieve the desired savings, physicians would have had to spend less time providing services under the public plan and hence bill the public plan less. But they seem to have spent the same amount of time providing publicly funded services, simply billing the public plan differently. This led to further delisting exercises in 1998 and In the latest delisting, attention was focused on other professional groups, possibly in recognition of the need to ensure that more listed services are not simply substituted for the now delisted services. In this case, physiotherapy, chiropractic services, and eye examinations by optometrists were delisted. The increasingly restricted content and service-based focus of the system risks undermining support for the system as the average Canadian discovers that the public purse has been used to pay for sex changes, deemed medically necessary but confined to a tiny proportion of the population, but not for child dental examinations or treatment. It is difficult to see how these arrangements can be justified on the basis either that those receiving sex changes are in poorer health than children with dental problems or that the sex-change procedure represents a greater probability of improvement in health than the dental treatment. Instead, these policies seem more consistent with resource considerations the resources associated with the expected demand for sex changes being much less than the resources associated with the expected demand for child dental examinations and treatment. To those members of the public without access to employer-based schemes for coverage of noninsured services, alternative options such as medical savings plans risk becoming an attractive alternative. Under these plans, individuals have more flexibility in deciding which services are to be funded. But this leads further down the road to a demand-driven system, without consideration of the relative merits in terms of expected improvements in the health status of the population of alternative uses of health care. In Ontario, a Hospital Restructuring Commission was charged with the task of rationalizing hospital facilities in the province. The commission recommended large reductions in the number of hospital beds through the closure of many hospitals and merging of others. At the same time, waiting times for hospital services have increased substantially. In some cases provinces are busing residents to the United States for urgent care. For example, the Ontario Ministry of Health has adopted the Canadian Association of Radiation Oncologists recommendation of a maximum wait for radiation treatment of 4 weeks. But the wait for radiation services in Ontario in early 1999 was between 7 and 14

17 Canada s Health Care Policy / 459 weeks. In response, Ontario residents are being sent to Buffalo, New York, at a cost of $16,000 per treatment to the Ontario taxpayer. The same treatment in Ontario currently costs the taxpayer $3,000 (37). Elsewhere, private free-standing clinics offer a privately funded alternative to waiting for particular treatments. For example, the Consumers Association of Canada reported on the distribution of provision of lens replacement surgery for cataracts in Alberta province (38). In Calgary, cataract treatments have been moved out of hospitals to clinic settings. Physicians use a state-of-the-art foldable lens implant that requires smaller incisions and hence entails less risk to the patient and a shorter recovery time. However, the local health board pays only for the standard lens. For about $250 per eye, cataract patients can jump the queue and get better-quality care. A few hundred miles away in Lethbridge, Alberta, all lens replacements are performed in hospitals using the state-of-the art lenses. Because the procedures are hospital based, the provisions of the Canada Health Act apply and the cost of the lenses cannot be passed on to the patient. Similar compromises exist in hip replacements; a 2.5 year wait can be avoided by asking for the titanium hip, the extra cost being met by the patient (39). Whether these factors affect the protection, promotion, and restoration of health of Canadians is unclear. Interprovincially there seems to be reasonable access with respect to health, at least as measured by life expectancy at birth. The Report on the Health of Canadians noted that provincial differences in life expectancy have diminished over the past 40 years to the extent that those born in British Columbia, with the highest life expectancy, have less than three years advantage over those born in Newfoundland, with the lowest life expectancy (40). Compare this with corresponding data for residential areas in the city of Winnipeg, Manitoba. Children born in areas with the highest incomes can expect to live 11.3 years longer than those born in areas with the lowest incomes (41). Although these differences in life expectancy may not be caused by the health care system, it seems that current health care policy does little if anything to ameliorate them. Moreover, the lack of attention to relative levels of need, as measured by relative levels of health, in the management, planning, and delivery of services could potentially increase social inequalities in health. For example, Alter and colleagues (42) analyzed the use of cardiac procedures and survival outcomes by area of residence for all patients with acute myocardial infarction admitted to hospital in Ontario between 1994 and They found that waiting times for procedures were 45 percent lower and use of procedures 23 percent higher for patients from the highest-income neighborhoods than for patients from the lowest-income neighborhoods, after controlling for patient age, sex, severity of disease, and a range of factors relating to the attending physician and institutional setting. Similarly, after controlling for these other factors, outcomes favored patients from the higher-income neighborhoods. A betweenneighborhood difference of $10,000 in median income was associated with a

18 460 / Birch and Gafni 10 percent difference in one-year mortality in favor of higher-income neighborhoods. This survival advantage was consistent across age groups and for shorter survival periods. In addition to the possible effects of health care policy on population health, consideration should also be given to the impact on the public s acceptance of and confidence in the health care system. A recent survey of satisfaction with health care in five countries found a remarkable loss of public confidence among Canadians over the past decade, both in absolute terms and relative to the changing level of confidence in other countries (43). Problems with access to specialist services were the largest source of concern. Although concerns about the system were common to all groups, analyses by income group indicated that the better-off groups had been less affected than poorer groups in terms of accessibility becoming more difficult (44). This supports the findings of the utilization research reported above that increased fiscal pressures seem to have a greater impact on poorer groups. Similarly, in a three-country study of individuals waiting for cataract extraction, more than 38 percent of respondents in Canada indicated they were willing to pay to have the surgery performed within one month, compared with 28 percent in Spain and 17 percent in Denmark (45). The corresponding proportions willing to pay higher taxes to eliminate waiting times were 15 percent, 12 percent, and 24 percent, respectively. In the 1999 federal budget, the government allocated $11.5 billion to increases in federal government support of provincial health care programs, spread over the next five years. This represented an increase in real expenditure of 1.5 percent per annum, approximately the level of increase required to bring expenditures back to early 1990s levels. Similarly, provincial governments announced increases in provincial expenditures on health care. Expenditure per capita in Ontario in rose to $1,725 per capita following two years of reductions in real per capita expenditures (27). But this restoration of spending was not accompanied by action on the structural features of the system and, in particular, on aligning the mechanisms for the allocation of health care resources with the needs-based objectives of the Canada Health Act. The recent increases in spending might make things look a little rosier on the surface, but this may simply cover up the more fundamental problems that place the long-term stability and success of the Canadian system at risk, as individuals increasingly are left to go outside the Medicare program and hence make private or third party arrangements for funding health care services. No, we are not there yet. We might not even be on the right course for getting there. Acknowledgments We appreciate helpful discussions with John Lavis (McMaster University) and Alina Gildiner (University of Toronto) on the relocation of physiotherapy services to the community, and the comments of an anonymous referee.