Comprehensiveness in Public Health Care

An Impediment To Effective Restructuring

Claude E. Forget October 24th, 2002

In recent years the public debate over the future of Canada’s health care system has intensified and given rise to growing frustration over governments’ apparent inability to bring forward tangible solutions. As the public awaits the much- anticipated Kirby Senate Committee and Romanow Commission reports, there is some doubt at to whether these latest initiatives will finally engender a true debate about the real issues. Because of the controversial nature of the topic, dis- cussion of health care reform and financing has for some time been mired in semantics and wishful thinking. Indeed, the underlying thesis of this essay by former minister of health for Quebec, Claude E. Forget, is that words can get in the way. This is particularly so when the words in question acquire the status of “principles.”

The focus of Forget’s analysis is “comprehensiveness,” one of the five prin- ciples underpinning the Canada Health Act. He explains that this “pillar” of our health care system was originally intended for very specific historical purposes that no longer exist, but yet the precedent holds. Thus we find ourselves in the rather surreal situation where government leaders pledge their unswerving com- mitment to maintaining a comprehensive public health care system and at the same time ponder the feasibility of adding long-term care and prescription drugs to so-called comprehensive, publicly insured services. Forget argues that the con- cept of comprehensiveness is misleading because it fails to recognize the funda- mental reality that the Canadian health care system is not, never has been, and should not attempt to be all-encompassing. In his view, to continue to uphold the principle of comprehensiveness sets an irresponsible agenda for the future and, given the ever-expanding and innovation-driven nature of modern health care, is tantamount to having “a system with a powerful engine and no brakes.”

Forget’s basic argument is that for the public health care system to endure it must be able to control its own evolution. Three things must be done in order for this to occur: (1) the principle of comprehensiveness must be renounced; (2) governments must establish the size of the financial envelope for publicly financed health care as a proportion of national wealth and not as a by-product of the system’s operations, as is currently the case; and (3) the financial envelope must then be used to satisfy a needs-based list of priority functions. This entails moving away from a supply-side approach to health care management (i.e., sup- ply of doctors, nurses, equipment, facilities, etc.) toward the establishment and implementation of strategic priorities.

The author identifies four distinct categories of health care functions to which he assigns an order of priority based on the need for a public sector role. Rather than focusing on primary care and emphasizing the “first line of services” as is customary in most health care reform discussions, Forget places this routine maintenance function at the bottom of his list of priorities as it is the only cate- gory that does not have the characteristics of a public good. He points out that primary and secondary services are by and large personal services, not servic- es to the community (contrary to the risk management function); they are insurable at an affordable cost for most people (contrary to severe long-term disability); and they entail none of the “public utility” features of ultra-special- ized R&D-intensive services. While Forget does see a continued role for gov- ernment in making health maintenance services accessible, his proposals do imply that what resources are available for routine maintenance would depend on what is left after other, higher priority functions have been funded.

The author recognizes that determining which health services should or should not be publicly insured will always be a delicate and difficult task, and many may not agree with his list of priorities. But regardless of how one sets these priorities, it is essential that they be defined on the basis of a clear ration- ale rather than historical precedent.

Introduction

Words, as they are used in the context of debates over the future of health care in Canada, no longer seem to have the meaning that was originally attached to them. This hinders truly effective involvement by the general public in govern- ment consultations, and it compounds the problems that arise in policy debates over what is already an extremely complex issue.

This dissonance between language and concrete reality is particularly man- ifest with regard to comprehensiveness, one of the principles to which ministers of health and even first ministers feel compelled to swear allegiance and unswerving commitment whenever they meet. This is all the more puzzling as in the same breath they discuss their desire to add such services as long-term care or prescription drugs to existing insured services. This seems to be contradicto- ry. How can one add to services that are “comprehensive”? Comprehensiveness is a binary concept: it either is or is not present. At most, it is present with very specific exceptions. When used as a vague attribute of something that could be “more or less” comprehensive, the notion has no value whatsoever.

A similar dissonance is revealed when analysts praise the health care sys- tem for its effectiveness because of its “single payer” characteristic that, in their view, makes possible large savings in administrative costs. Yet it is also known that the share of government resources in health care is about 70 percent and decreasing. Consequently, there must be payers other than government. Moreover, the controversies over the federal and provincial governments sharing the burden remind us that the “government” we refer to is an abstraction that is the result of treating as one what are in reality two payers.

My purpose here is not to entertain. The words used in discussions about the Canadian health care system do have a particular meaning that is the result of a long tradition, and they are incomprehensible outside that context. That context was shaped by the need to regulate large federal-provincial transfers and signal general political orientations. There was no effort at the time the concept was first put in cir- culation to spell out for Canadians what their entitlements were vis-à-vis the public system. Indeed, provinces were left wide discretion in spelling out these entitlements.

In policy discussions and public debates this dissonance is a source of con- fusion. When so-called basic principles are contradicted by day-to-day realities, the public understandably doubts the sincerity of those pronouncements and becomes disillusioned about the public health care system. Debates become blurred and confusion reigns when solutions are proposed. In this brief essay I will propose that the principle of comprehensiveness be abandoned.1 The concept of comprehensiveness emerged for historical reasons that are no longer relevant. It is inapplicable today and it would be irresponsible to propose that it is an attainable target.

What does it mean to jettison comprehensiveness? It means, first, recog- nizing the gap between the totality of health services and the aggregate scope of publicly financed health programs. This gap is not trivial. It has always existed in the past, it exists now and it will continue to exist in the future.

Second, it means acknowledging that the carving-out of a large public role in health, whatever its original merit, is not optimal from the perspective of today’s reality. This is because this carving-out is based on a “silo approach” whereby health problems have been viewed from the supply side: first hospitals, then physicians and now drugs and home care, etc. All these, it should be acknowledged, are inputs into health maintenance and restoration programs that should be examined (and either covered or not) from the patient’s perspective. Moreover, the selection of insured services has been determined by the interplay between federal and provincial governments: while this dynamic is important, its results reflect past political configurations that may not be relevant today.

We need to move away from this heritage and, at least conceptually, rede- fine government’s role in health care. This, I will argue, must be done by refer- ring to the following concepts:

  1. user-centred models of health services
  2. needs-based prioritization of coverage under the public system

Comprehensiveness: Historically Narrow, Nonexistent Today and Foolish for Tomorrow

Comprehensiveness as a rule of the public health system goes back to the hospi- tal insurance system of the 1950s.2 In offering to share approximately 50 percent of the cost of a universal regime of hospital insurance, the federal government had a limited objective: some programs such as mental health and long-term care for TB patients were already fully under provincial fiscal responsibility. It did not make political sense, the thinking went, to share in the cost of already function- ing programs. Comprehensive coverage would therefore apply to all hospital serv- ices prescribed by physicians (i.e., services that were “medically required”) in this narrow domain. It naturally excluded physicians’ fees, even for services per- formed in hospitals (that coverage came later).

This rule of comprehensiveness had a particular use in settling what was and was not cost-shared between Ottawa and each of the provinces. Federal civil servants would check all hospital beds, on the spot, to ensure that those includ- ed for cost-sharing were neither psychiatric nor tuberculosis beds. Everything else that was prescribed by doctors, as long as it was performed within hospital walls, was shareable. Comprehensiveness was therefore a rule used by public ser- vants (whom this author supervised on the provincial side). For the public, at that early stage of the system’s existence, it meant only that they would no longer receive a bill from the acute care hospital into which they were admitted. At that time, the concepts of ambulatory care, home care and out-of-hospital drug cov- erage did not exist. Public health, psychiatric hospitals and sanatoria existed, but they were provided for publicly and separately, for reasons that had nothing to do with “comprehensiveness.”

In the latter part of the 1970s, the basis for cost-sharing between Ottawa and the provinces was modified. Ottawa calculated its contribution, using a block funding formula based on GNP growth, in a way that no longer required a precise accounting of shareable items. By then a historical basis for the required federal payments had been established, and the only important question to be resolved was how fast the federal payments would grow. Logically, the concept of comprehensiveness should have been discarded as irrelevant at that point: the provinces increasingly saw the federal payments as a contribution to their over- all health care expenses which, by then, were much greater than those Ottawa had hitherto shared.3 Those provincially provided services, however, were not any more comprehensive than the federally cost-shared component had been, even if their scope was wider. In fact, comprehensiveness was hardly more of a reality than it had ever been.

After the adoption in 1977 of the law that set up block funding for health services, the four principles of universality, comprehensiveness, portability and public administration, while formally still in force, were of questionable practi- cal and legal relevance in the view of the federal government itself. As the issues of over-billing and user charges emerged and the four hallowed principles were of no use to combat them, the federal minister of health had to introduce legis- lation, the Canada Health Act, to add a fifth principle, “accessibility,” and impose penalties on provinces that allowed these charges. The narrow focus of the Act was achieved at the cost of deliberate polarization of the issues.4 This was partic- ularly pernicious with regard to comprehensiveness. The Act reasserted a princi- ple that was manifestly honoured in the breach: the Canadian health care system was not comprehensive and never had been, at least in the ordinary sense of the word. No one had ever pretended it was and no one really cared. But the seeds of the confusion that persists to this day were planted. In addition, an unattain- able political agenda had been established. For some policy makers it became imperative that the scope of the Canadian system be extended to make reality conform to the principle. It was also imperative that this broadening of the sys- tem be carried out according to uniform Canada-wide rules, as had been the case at the time when uniformity (of a sort) was a requirement for cost-sharing, according to precise rules of eligibility.

This legislative initiative, by importing concepts that had made sense in a given but poorly understood historical context into a totally different one, took for granted decisions that had never been taken or openly debated. This has cre- ated the present, ludicrous situation whereby governments manifestly cannot cope with the increases in the costs of their present responsibilities and at the same time consider the feasibility of implementing new programs (such as long- term care and drugs).

Some in the political sphere (see, for instance, the Mazankowski Report) clearly perceive the reappraisal of the comprehensiveness principle to be inevitable. However, some aspects of this reappraisal, for example, referring the issue of the scope of covered services to experts, seem timid and insufficient. The belief that experts can bring a resolution may look attractive by analogy with the former rule for hospital insurance, according to which all “medically necessary” hospital-based services were covered. But this seems be a failure to appreciate that this rule, in its original context, was a rule for inclusion. At the time, federal decision makers had not delegated the responsibility for designating all that they did not want to insure, for example, tuberculosis, mental health, dental care, aesthetic surgery, psychology and chiropracty. Physicians, on the other hand, determine what treatment their patients will receive by means of a fundamental professional ethic: they must do everything that will not be harmful and could conceivably help. To determine the scope of social insurance, other considerations and other rules are required; experts can help, and indeed are essential, in their implementation, but they should not play a role in fundamental decisions that must be made publicly.

I have tried to show that in the past, the concept of comprehensiveness had a precise sense which, in today’s environment, has been lost. Now the con- cept is misleading, because it is neither a good approximation of reality nor a fea- sible short-term target. For the more distant future, comprehensiveness is an irre- sponsible agenda. Modern health care is driven by research and development: a commitment to comprehensive coverage would legitimate any expansion driven by either inherent incentives or managerial weaknesses in an ever-expanding, innovation-rich field. In other words, it would be a system with a powerful engine and no brakes. Today, what is effectively covered by the public system no longer corresponds with what should have priority but simply with what is pos- sible and can be done inside established structures.

Almost all the participants in the current discussion around health care are at pains to proclaim their attachment to our universal, public health insurance system and their hope to see it maintained. I personally have no other agenda. However, for a system to endure, it must be able to control its own evolution. To make this possible, the following elements seem essential:

  1. Acknowledge that the publicly financed system cannot now or in the future cover the totality of the field, and renounce the principle of compre- hensiveness.
  2. Determine the financial envelope to devote to publicly financed health services that Canadians would consider desirable and feasible. This enve- lope, which should be proportional to national wealth and should there- fore grow with it, could also evolve over time in response to other clear and transparent rules. I believe it is of fundamental importance that this envelope be determined outside the health care system as opposed to being a by-product of its operations, which seems to be the case at pres- ent. For the public system (which, it should be remembered, is only part of the overall health sector), revenue must determine expenditures, not the other way around.
  3. Use the financial envelope to satisfy a needs-based list of priority functions. To proceed in this way would, I suggest, result in very substantial changes to the range of services covered involving subtractions and additions. It would therefore require a significant transition period. The present system is an inheritance that reflects a mixture of popular preferences, techno- logical assumptions and economic and political happenstances, and it needs to be revisited.

A User-Oriented Examination of the Main Functions of Health Services

Having accepted that public resources are and will remain only one part of the total resources dedicated to health prevention, maintenance and restoration services, it is easy to say that public resources should be allocated to the most important needs. The difficulty arises when one attempts to explain what this might mean.5

As a first step, it seems a good idea to get as far away as possible from the traditional supply-side debates about health care services: the supply of physi- cians, nurses, equipment, facilities, information systems and drugs, etc., is a dif- ficult and important issue. All of these inputs serve a large variety of purposes or needs, and it is the assessment of these needs, from the perspective of the “pub- lic” (or users or consumers), that raises the most interesting questions and the most important choices for coverage. While a needs assessment seems both desirable and perhaps obvious, this has not been the basis upon which our pub- lic system has evolved. Coverage has evolved by including one class of inputs after another, more or less irrespective of the health problems that might be addressed! This is still the approach today: one only has to think about the sug- gestions being made for the coverage of drugs or long-term care! It does not real- ly make sense to encourage a belief that covering all types of (prescription) drugs, for all types of problems and beneficiaries and in any situation is a rational way of framing a policy issue in a context where public resources are limited, there is uncertainty as to drug-usage monitoring and there is a culture of drug usage and marketing driven in part by lifestyle choices.

Health care can be analyzed in terms of four distinct functions, each addressing a discrete sub-universe of needs and each different from the other three in terms of their public or private nature (in the sense of the economic the- ory of public goods), the structure of incentives present (would consumers be tempted to over-utilize a particular service given public free-of-charge provi- sion?), their insurability (by private sector insurers), the nature of their cost func- tions (constant or declining costs) and their relevance to population health or distributive goals. The four functions are the following:

  1. Managing health risks
  2. Coping with severe, irreversible disabilities
  3. Solving acute health problems
  4. Providing routine maintenance

These functions are listed in order of their priority as I see it.6 This, as I will elaborate further, has distinct implications when seen in conjunction with the principle that revenue must determine expenditure.

The First Function: Managing Health Risks

There are many determinants of morbidity and mortality that can be con- trolled. Typically, this is not done by interfacing with individual patients (although as we will see, there are exceptions), but by acting on the general physical and social environment.

Recently, most developed countries, Canada being no exception, have experienced difficulties, some would say crises, with regard to drinkable water supplies (for instance, the deaths that resulted from contaminated water in Walkerton, Ontario), atmospheric pollution or contaminated blood.7 Infectious diseases, far from being a thing of the past, have reappeared as a concern. Examples are HIV-AIDS and Creuzfeld-Jacob disease, both of which seem to imply a possible leap from animal species to humans. Globalization, with the concomitant massive population exchanges, is making tropical diseases a grow- ing phenomenon in temperate countries. Increasing population densities and concentrations of infected people in health facilities increase risk by speeding up bacteriological mutations and sometimes increasing virulence.8 Sanitary control over the food chain is no longer a simple procedure of standard-setting and inspecting local processing facilities, but also – in the context of international trade in foodstuffs – raises the issue of traceability of supplies across borders and whether it is possible to control the production process. Finally, there is also the threat of bio-terrorism.

It is not all bad news. Increasingly effective tests are making screening for some illnesses that can benefit from early diagnosis possible and cost effective. Examples are breast cancer in women over 50, prostate cancer in men and learn- ing disabilities in children. Public health is acquiring a cognitive dimension. Information about health-promoting diets, exercise, the potential harm of sub- stances such as alcohol, tobacco and other drugs, the efficacy of treatment and the side effects of medication are not only contributing to healthier lifestyles but are becoming significant accessories to the effectiveness of more personalized health services. Information on the safety of immunization programs can be cru- cial, and information on the performance of health care providers will progres- sively become a requirement. Proactive programs to enhance safety in the work- place, on the roads and in sports arenas have demonstrated their usefulness.

These activities clearly belong to the category of services that economists label as “public goods.” They are health services without patients and for that rea- son, if the state did not provide them, it is highly unlikely that they would exist at all. Possibly because of this characteristic, they are deprived of media visibili- ty and a strong political constituency except in times of crisis, and even then the situation changes only for a short while.9 Controversies that arise in the wake of spectacular breakdowns suggest that this function is probably badly neglected and underfunded. Yet there is arguably no other functional group of health serv- ices that bears as close a relationship to overall morbidity and mortality,10 and it should therefore be the first priority.

While these services are of primary importance, the rigour with which they are deployed and managed should not be taken for granted. It is valid to question the quality of their performance and the effectiveness of their outcomes. To say that risk management is the first priority in health care does not necessarily mean that a lot more public resources should go into it, regardless of the answers to those questions. Nor does it mean that increasing the public resources that are put into risk management should be ruled out. Governments that provide all the financial support for public health care are also responsible for the services, and they are therefore the judges of their own performance. This is not an entirely sat- isfactory situation: a genuine distance has to be created between those two roles.

The Second Function: Coping with Severe, Irreversible Disabilities

People affected with mental, sensory and physical handicaps and some severe illnesses for which no effective treatment is known experience devastating impacts on their quality of life, their earning capacity and their very ability to live autonomously. These severe and irreversible disabilities can also impose a dis- tressing burden on the affected individuals’ families, which could eventually become the source of yet other problems. Very old age is not among the disabil- ities listed here, but it is often concomitant with them.

The purpose of health services in such situations is to provide support that will help patients cope with their disabilities in a condition that approximates, as far as possible, a normal lifestyle. By definition no healing is possible, so all the rel- evant services can be described as “palliative services.” It is often possible to facili- tate mobility and autonomy, or to substitute for it, and to provide ways of main- taining links with the community in the natural environment or in a high-quality institutional or semi-institutional environment. This involves controlling pain, making prosthetic appliances accessible and hopefully, through it all, minimizing the feeling of isolation, powerlessness and anxiety. This is a long and expensive list.

Over a lifetime, however long or short it may be, the cost of these personal services is crushing for all but the wealthiest. This is very much what economists would label a “private good,” but one that is unaffordable and uninsurable outside a compulsory and practically universal insurance program.11 The large individual cost is compensated for by the low incidence in the population at large. On grounds of efficiency, the case for a public, government-run social insurance model only applies to the financial system, not to service delivery.12 Historical recognition of the priority status of this category of needs is an observable fact: governments early and widely assumed responsibility for mental institutions and the long-term care of tuberculosis, before the advent of national, publicly funded health care sys- tems. Disabilities resulting from wars also gave rise to publicly funded programs in the countries affected. The nature of senile dementia, together with the increasing demographic importance of the aged who are confronted with it, is leading to polit- ical pressure for a more generalized approach to long-term palliative services.13

Although there is an efficiency rationale for public insurance for long-term disabilities, there are also subsidiary distributive considerations. Failing public provision of palliative services for all instances of permanent disability, the extreme inequality of conditions created by them would militate in favour of at least some public support.

Is there a deficiency today in the provision of publicly funded, long-term palliative services? The answer is obvious. Modern societies have frequently turned their backs on institutional services as an answer to those problems, with- out dealing with the vacuum thereby created. Our cities, with their homeless and beggars, although there are no doubt other factors, testify to this deficiency.

As societies become wealthier, they can afford to buy more of this kind of protection against dreaded risks, but this requires a public debate such as that likely to be generated by the Romanow and Kirby recommendations. However, the public debate about this function of the health care system tends to be muted because the most direct potential beneficiaries do not have a very strong voice, while the general public faces a very small or very remote risk of being affected by such diseases. Public decision-making over this issue is therefore somewhat like decision-making related to public goods: it does not happen of its own accord but requires an initiative and a campaign of persuasion.

The Third Function: Solving Acute Health Problems

Since the beginning of the 20th century medicine has embarked on prob- lem-solving, borrowing from science its rigour and methodology. Severe illness- es are targeted as so many problems that can be resolved: medical successes, at first very modest, are gradually increasing. In this way the field has given rise to vast expectations.

These expectations do have a basis in reality, but that reality is far less than the totality of medical and hospital services. An illness can almost always be named and its symptoms often alleviated; in a smaller number of cases it can also be cured, or at least it is possible to sustain or restore the impaired vital function as well as a reasonable quality of life. When I refer here to activities targeted at solving severe health problems, I am referring essentially to this last group of activities.

An important characteristic of these activities is that they are constantly evolving. This evolution, well reflected in the media, is what inflates expecta- tions. It also inflates costs and makes their progress unpredictable and difficult to control. The process that drives this evolution cannot be separated from the services themselves: the problem-solving function of the health care system is a “work-in-progress,” and this has significant consequences for the inclusion of these activities in the public system.

The ways in which we currently account for the costs of this problem-solv- ing capability obscure rather than illuminate the true nature of the cost function for ultra-specialized services. For instance, although the capital intensity of these services has increased enormously, we do not typically include the cost of capi- tal used in their production, nor is this element incorporated into the financial system underpinning health care institutions. Moreover, we do not fully account for the training costs, which are only partly reflected in personnel costs. We keep a separate account of research costs, but this ignores the fact that even if research costs can be (imperfectly) distinguished from clinical costs, the research activity cannot really be separated from the problem-solving activity that defines those services. Finally, the structure for control and evaluation, also an indispensable ingredient, carries a cost. If all the costs were aggregated, it would emerge that the problem-solving function in health care has the characteristics of a public utility, a factor that has many important implications.

Public utility pricing and regulation is an issue that economists have debated for over a century and a half. This is not the place to go into this theory. Suffice it to mention here that a public utility is a firm whose marginal costs are smaller than its average costs and for which, therefore, competitive pricing (at marginal cost) would inevitably lead to a “natural monopoly.” Taking all the costs of problem-solving ter- tiary services into consideration, it is clear that the (marginal) cost of treating one more patient, i.e., the cost of narrowly defined “clinical activities,” is far less than the total costs (including infrastructure, training and process control activities) divided by the total number of patients. In other words, there are, in creating and sustain- ing the ability to solve difficult health problems and doing it dynamically in an inno- vative environment, very large fixed costs that are independent of the number of patients effectively treated: this is what directly leads to the public utility paradigm.

The first implication of this is that the financing mechanisms used to pay for such activities cannot rely on “competitive” pricing. The occasional proposals for hospital financing to be changed from global lump-sum budgets to a system of payments by results, such as number of episodes of treatment, would be dif- ficult to implement because of the extent of the non-clinical costs.14 When there are fixed costs, their allocation among concurrent activities is necessarily arbi- trary. Indeed, the use of the DRG (Diagnostic Related Groups) system to pay for hospital services in the United States and several other countries has created a potential problem of severe underfunding for teaching hospitals.15 The source of the problem goes beyond the presence of fixed costs and also results from the fact that marginal clinical costs keep changing because clinical activities themselves are part of the research and development activities of such institutions.

The second implication is even more important. We know that a public utility need not be government owned or publicly funded. Regulations are anoth- er appropriate tool used to make sure that a natural monopoly does not abuse its power over consumers. Regulations normally deal mostly with pricing. In a uni- versal, free-to-the-user system such as public health care, there is no price to reg- ulate. What needs to be regulated is the R&D process itself. The complexity of regulating that process is such that public responsibility may very well be, on efficiency grounds, the only feasible approach.

The regulatory framework that government should contemplate with regard to the problem-solving function of health care services would not be restricted in scope to tertiary care institutions. Having funded the highly spe- cialized core infrastructure needed to support that function, government has an interest in seeing to it that this huge cost is spread over as large a clinical base as possible, in order to take full advantage of declining average costs. Government therefore would be expected to try to contain “mission inflation” by institutions that stand outside that core infrastructure. In other words, this public utility notion of the role of tertiary institutions leads, for reasons of both efficiency and effectiveness,16 to a high degree of concentration of tertiary pro- cedures in a very small number of extremely well-endowed centres of excel- lence. This is in turn leads, more or less consciously and effectively, to the con- cept of a health care system whose components each have a clearly delineated role: relatively numerous community clinics and hospitals delivering estab- lished procedures according to well-defined guidelines to patients without sig- nificant complications, and a very limited number of core institutions that develop innovative approaches and evaluate complicated cases while still offer- ing a very wide spectrum of services.

The need to regulate the R&D process is linked to the often expressed con- cern that health care services should be science-based and every procedure should be evaluated giving due recognition to outcomes. I would qualify these concerns by adding that rigorous rules, which are essential for evaluating a procedure, cannot necessarily be used to determine whether to include that procedure in insurance coverage. We must recognize that a procedure has to be included in coverage if it is to be evaluated in the first place, and it has to retain that status at least as long as there is an ongoing process of improvement and development. Now this is not a minor qualification, since the process of developing solutions to serious health prob- lems is constantly evolving over a wide spectrum. This is why good controls over the innovation process are more important than any attempt to control public health care costs by exclusionary approaches (i.e., by de-insuring services).

There must, therefore, be very demanding rules requiring this R&D process to take place in the appropriate environment. Public coverage must be co-extensive with the presence and effective implementation of these rules and strong internal control mechanisms.

To illustrate what such an approach could entail, here are two elements of a possible process-control framework:

  • Before commercialization is allowed, drugs must undergo a series of suc- cessive evaluations, namely clinical tests defined as phases 1 to 4. Medical and hospital procedures could be subjected to a similar requirement. This would, of course, be done for the system as a whole and not for each insti- tution separately, and it would relate to well-defined procedures. It must also be borne in mind that the process of innovation in question refers to a new procedure from its emergence as a concept to the point where it becomes stabilized as an accepted routine.
  • Procedures being developed and evaluated would only be covered by the regime if, and only if, they were performed in an appropriate insti- tutional environment. In phase 2 (phase 1, drug evaluation with healthy volunteers, being largely inapplicable here), the only acceptable envi- ronment would be the originating tertiary institution. The procedure would be labelled “experimental” and disclosed as such to patients. A rigorous evaluation protocol would be followed. That protocol would include the normal methodological safeguards (control group, random- ization, etc.). In phase 3, the procedure, now labelled “innovative,” could be performed in other tertiary environments and with the appro- priate protocol involving a much greater number of patients and a longer evaluation period. If conclusive, the procedure would then become a “recognized procedure” (phase 4) with specific guidelines for its appropriate application.

In all cases and for all phases, such problem-solving activities would be covered by the public health care system only if they were carried out in an insti- tution that had a structured program for quality maintenance and improvement that included, for each program or department, well-defined quality objectives, multidisciplinary teams to lead the process, performance indicators related to the objectives and a performance above a given threshold.

Constant innovation is the best characteristic of the health services in their “problem-solving” mode, but it is also the factor that makes the growth in their costs least comprehensible and hardest to control. In the euphoria that it gener- ates, anything seems possible, no matter how it is done. Public sector involve- ment must support and make accessible to all, without reservation, those proce- dures that are carried out seriously, rigorously and that ultimately work. The con- sequences of this approach are clear: de-insure whatever fails that rigorous test! It is impossible to say how much would be de-insured according to that pre- scription: one’s intuition is that it could be substantial and that significant economies could result. However, public and clear endorsement of high stan- dards could rebound to increase costs.

There is no question, in this scenario, of prohibiting professionals or insti- tutions that believe such requirements to be superfluous or needlessly severe from offering their services on a private basis. Their patients would clearly be put on notice that it is their own (more or less blind) trust that justifies this course. I do not believe that society should feel queasy were this to happen.

The Fourth Function: Providing Routine Maintenance

Routine maintenance, as a function of the health care system, is performed in a variety of settings and a variety of ways. It is carried out largely in primary and secondary care facilities (e.g., public and private clinics, health profession- als’ offices and community hospitals). However, it also reaches into so-called ter- tiary care facilities where a very large number of maintenance services, in addi- tion to the problem-solving activities, are carried out. This function encompass- es the vast majority of encounters between the health care system and patients, but constitutes a much smaller proportion of the total value of health services.

The maintenance function is not limited to patients’ visits to physicians and other health professionals for treatment of minor ailments or diagnostic ref- erences to more specialized resources. It also includes in-patient services such as low-risk deliveries or appendectomies. Also included in this category is long- term monitoring and treatment of medical conditions such as moderately high cholesterol, hypertension and diabetes for low-risk patients when no com- pounding factor is present. The unit cost of these treatments is relatively modest because simple, affordable technology is used outside the innovation-rich prob- lem-solving context previously described. However, this is not a static universe of services: it evolves over time as the results of the R&D activity stabilize and become more widely diffused. It can also evolve as any other economic activity in terms of organization and style of delivery.

It has been customary in the health care debate to start off by discussing pri- mary care or “the first line” of health services. Naturally, they are the ones we tend to be most familiar with because they are the most frequently used, and they can be considered as the “entry point” into the system. Many studies of the health care system deal at length with that segment of the system. It is my long held view that government involvement in primary care should not deprive individual patients of their influence over the way in which such services are made available.17 It is encouraging to see that this view seems to have gained some acceptability. The cur- rent distinction between the financing of primary care and service delivery and the possibility that the government might allow public funds to follow the patients by supporting them in their use of the services rather than by directly supporting the service providers justifies some hope that patients will one day truly control their own health care instead of being, as now, its passive and powerless subjects.

That being said, it is telling that the routine maintenance function (which largely overlaps with primary and secondary care) is at the end of this list of pri- orities for public funding. There is a rationale for this ranking: this is its position in reality, in a sense, because this category includes the bulk of health services that governments have chosen not to cover and that are presently privately financed. Public health insurance coverage excludes the services of pharmacists, psychologists, chiropractors, acupuncturists, dentists (to a large extent) and also out-of-hospital prescription drugs (even in Saskatchewan and Quebec, private financing accounts for about 1/3 of costs).18 It is interesting that we do not see any of the controversies and problems of availability that sometimes rage around insured primary care services arising with respect to excluded services.

There is no question that government should remain involved in making health maintenance services accessible by contributing substantially to their funding. However, if we agree that governments must also determine the total public resources that they are able and willing to commit to health, it follows that what will be available for routine maintenance will be what is left after other, higher priority functions have been funded. One can naturally take issue with the order of priorities set out in this paper: remember that the other func- tions are essentially “public goods” in one way or another, whereas the fourth and last category is not. By and large, primary and secondary services are per- sonal services, not services to the community (contrary to the risk management function); they are insurable at an affordable cost for most people (contrary to severe long-term disability); and they entail none of the “public utility” features of ultra-specialized R&D-intensive services. This is the only segment of the health care system where government’s financial involvement is exclusively underpinned by an income distribution justification which could be addressed through other means.

Conclusions

The importance of renouncing “comprehensiveness” as a principle derives from the realization that to do otherwise is to deny a fundamental reality: the Canadian health care system is not now, never has been and should not even attempt to be all-encom- passing. To retain the pretense of “comprehensiveness” is to irresponsibly refuse to adjust citizens’ expectations to the reality of government commitments.

Health services, which are such a dynamically growing sector of activity, cannot be allowed to determine the overall level of taxation. Canadians must, through their political institutions, establish a balance between health services and other societal pursuits they care about. If this balance were to progressively shift in favour of health services so that government expenditures for health were to grow from approximately 6 percent to 7 percent of GDP, so be it! This would be an important decision that would require much political courage. Whatever the balance that is struck, the public decisions necessary require more intellec- tual honesty and rigour than have been displayed so far. They also require a bind- ing commitment from government in order to provide predictability and stabili- ty to the health care system.

Finally, priorities need to be defined. This paper presents a way to do this that is, I believe, plausible. Interestingly, the fourth function, providing routine maintenance, is also the area in which the issue of user incentives for over-uti- lization arises. Drawing the line between insured and uninsured services will always be a delicate and difficult task, especially when one attempts to produce reasons instead of just invoking history. That nettle must be grasped.

  1. Recent comments on “comprehensiveness” reflect dissatisfaction with the concept without going so far as rejecting it out- right. For instance Monique Bégin (2002, p. 7) suggests defining comprehensiveness so that it “extends to much more than hospitals and doctors” (which it already does, as far as provincially defined health programs are concerned, to patients the only meaningful reference point!) while suggesting that it should also be restricted to evidence-based practices. The Mazankowski report (Premier’s Advisory Council on Health, 2001), the Clair report (Commission d’étude sur les services de santé et les services sociaux, 2000) as well as the Senate enquiry chaired by Senator Kirby all support the notion of expert panels “to review and make decisions on services that should be publicly insured” (Standing Senate Committee on Social Affairs, Science and Technology, 2002b, vol. 5, p. 33) and, by implication, on what should not be. None of these authors explain why “comprehensiveness” should remain a principle when they advocate moving away from it.
  2. The Hospital Insurance and Diagnostic Services Act, 1957, c. 28, s. 1 (sections 5 and 6) referred to all “the in-patient serv- ices and out-patient services to which resi- dents of a province are entitled to under provincial law … ” However, the second reading debate held on April 4, 1957 (Government of Canada, 1957) made it abundantly clear that the only exclusions envisaged at the time were mental health and tuberculosis services, and they were the main focus of opposition comments. Even nursing homes were deemed share- able by the federal health minister at the time, Paul Martin, Sr. provided they were licensed as hospitals by a province (1957, p. 3123).
  3. In the second reading debate preceding the adoption of the Medical Care Act, 1966-67, c.64, s. 1 (section 14 and 15), the then minister of finance, Mitchell Sharp is quot- ed as saying to the Tax Structure Committee in Ottawa on September 14, “It is our view that in respect of the three large and continuing programs which come under provincial jurisdiction, federal conditions should be withdrawn…This would be accomplished by escalating the adjustment payments after a given date, on an objective basis unrelated to program costs…Once programs like these have been established, no government would discontinue them …” (Government of Canada, October 13, 1966, p. 8640).
  4. The story of the political struggle to have the CHA adopted has been told by the main actor, the health minister of the time, Monique Bégin (1987) and a more detached evaluation of the CHA can be found in her recent paper (2002, p. 7).
  5. The need to optimize resource allocation in health services as a prelude or concomitant to increased funding is discussed in a recent paper by Donaldson, Milton and Currie (2002). The only criticism of this approach is that, like most health policy discussions, it tends to assume that the health sector is one more or less homogeneous bundle of services, whereas the present paper makes the point that the health industry is a cluster of differentiated sectors, with different “pro- duction functions” and sets of inputs, out- puts and beneficiaries that largely do not overlap. This raises allocation issues among those sectors which marginal cost-benefit analysis cannot easily overcome.
  6. I make no apology for describing these four “functions” and ranking them. Naturally, there is much room for debate here about the number of functions, their description and their ranking. The point is that there is an alternative to unattainable comprehensiveness that can be based on setting strategic priorities rather than through the micromanagement of a “bas- ket” of procedures or individuals services.
  7. See the Report of the Commission of Inquiry on the Blood System in Canada, 1997.
  8. In his brilliant book, Ewald (1994) uses the tools of evolutionary biology to account for the higher virulence of some bacteria within health facilities compared with the lower virulence of related strains in the general environment.
  9. Most discussions of health policy under- score the importance of prevention and public health. A recent example can be found in the report of the Standing Senate Committee on Social Affairs, Science and Technology (2002a, vol. 2, ch. 4).
  10. Life expectancy statistics, possibly because there are few, if any, other global measure- ment of health services outcomes, have been abused as indicators of the “effective- ness” of health services. While clearly rele- vant to preventive and public health, albeit with a very long time lag, they are proba- bly not relevant to an assessment of the effectiveness of curative services, hence the attempt to circumvent this objection using quality-of-life adjusted statistics.
  11. Some types of disability are indeed private- ly insurable and are effectively insured for the benefit of employees of large firms with well-developed social benefits packages. However, universal coverage through pri- vate insurance is inconceivable: only peo- ple who are in the active labour force at the time will be allowed into a private insurance program; coverage may be limit- ed in duration; at-risk individuals will be screened out in the underwriting process, an eventuality that is likely to augment with the improvement of screening for genetic predisposition. See Cowley (1992).
  12. There used to be discussions about the need for a universal disability insurance program in Canadian policy circles. These discussions faded over the 1990s, along with most left- of-center concerns, even though the Canadian Labour Congress (CLC) kept bear- ing the torch into the late 1990s. Federal- provincial discussions have also kept the issue ticking over. See Federal/Provincial/Territorial Ministers Responsible for Social Services (1998). However, the talk no longer seems to be about a universal insurance pro- gram but is more narrowly focused on barri- ers to mobility and child protection.
  13. In a special report, the Quebec Auditor General (2002) observes an alarming gap between desired and actual service levels in nursing homes for the aged. The report notes especially their high contingent of senile and confused patients.
  14. Proposals have been made from time to time to finance hospital services according to performance as measured by the number of standardized treatment episodes. These proposals have not been implemented in Canada; this is in contrast with a number of other countries that have implemented them, starting with the United States decades ago. The most recent official refer- ence to this financing approach can be found in the Quebec government’s 2002 Bédard Report and the supporting studies, including one by CIHI. Estimating the determinants of average cost for each of hundreds of standardized diagnostic cate- gories is fraught with numerous technical difficulties of a practical and conceptual nature. In the context of the present paper the main conceptual flaw of the model envisaged for Quebec consists in the assumption that all types of acute care hos- pitals share a common cost function.
  15. This is discussed in a report from the Prospective Payment Assessment Commission (1997) and Reuter (1997).
  16. It has long been known that effectiveness of many medical procedures is volume- dependent: better outcomes tend to be asso- ciated with the total number of analogue procedures performed in a given institution.
  17. The author of this essay co-authored a book on this subject See Forget and Jérôme-Forget (1998).
  18. Standing Senate Committee on Social Affairs, Science and Technology (2002a, vol. 2, table 2.5).

Auditor General. Report to the National Assembly for 2001-2002: Highlights, vol. 1. Quebec City: Auditor General, 2002.

Bégin, Monique. “Revisiting the Canada Health Act (1984): What Are the Impediments to Change?” IRPP Thirtieth Anniversary Lectures Series, presented in Ottawa on February 20, 2002.

Bégin, Monique. L’assurance santé: plaidoyer pour le modèle canadien. Montreal: Boréal, 1987.

Comité sur la réévaluation du mode de budgétisation des centres hospitaliers de soins généraux et spécialisés (Bédard Committee). La bugétisation et la perform- ance financière des centres hospitaliers. Quebec City: Ministry of State for Health and Social Services, 2002.

Commission d’étude sur les services de santé et les services sociaux (Clair Commission). Emerging Solutions: Report and Recommendations. Quebec City: Ministry of State for Health and Social Services, 2000.

Commission of Inquiry on the Blood System in Canada (Krever Commission). Final Report of the Commission of Inquiry on the Blood System in Canada. Ottawa: Canadian Government Publishing, 1997.

Cowley, Alex. Long-term Care Insurance – International Perspective and Actuarial Considerations. GeneralCologne Re Publication, no. 22 (November 1992).

Donaldson, Cam, Craig Milton and Gillian Currie. “Managing Medicare : The Prerequisite to Spending or Reform.” Commentary, no. 157. Toronto: C.D. Howe Institute, 2002.

Ewald, Paul W. Evolution of Infectious Diseases. New York: Oxford University Press, 1994.

Federal/Provincial/Territorial Ministers Responsible for Social Services. In Unison: A Canadian Approach to Disability Issues. Ottawa: Human Resources Development Canada, 1998.

Forget, Claude E. and Monique Jérôme- Forget. Who is the Master: A Blueprint for Canadian Health Care Reform? Montreal: Institute for Research on Public Policy, 1998.

Government of Canada. House of Commons Debates, Vol. 8. Ottawa: Queen’s Printer, 1966.

Government of Canada. House of Commons Debates, Vol. 3. Ottawa: Queen’s Printer, 1957.

Premier’s Advisory Council on Health for Alberta. A Framework for Reform: Report of the Premier’s Advisory Council on Health (Mazankowski Report). Edmonton: Government of Alberta, 2001.

Prospective Payment Assessment Commission. Report and Recommendations to the Congress. Washington, DC: PPAC, 1997.

Reuter, James A. The Balanced Budget Act of 1997: Implications for Graduate Medical Education. New York: Commonwealth Task Force on Academic Health Centers, 1997.

Standing Senate Committee on Social Affairs, Science and Technology. The Health of Canadians – The Federal Role: Interim Report, vol. 2, Current Trends and Future Challenges (Kirby Report). Ottawa: Senate of Canada, 2002a.

Standing Senate Committee on Social Affairs, Science and Technology. The Health of Canadians – The Federal Role. Interim Report, vol. 5, Principles and Recommendations for Reform (Kirby Report), Part 1. Ottawa: Senate of Canada, 2002b.