Baker and Schwartz have provided an excellent overview of the theories and practices of wait time reduction, complemented by a summary of Ontario's plans to reduce excess waits for cancer care. In this commentary, I pursue a number of issues implicit in their analysis and prescriptions, and revisit the logic of the origins of and solutions to wait times.

People wait (involuntarily) for healthcare for two fundamental reasons: either there is not enough capacity in the system, or capacity is used inefficiently. While in Canada there is growing consensus that both causes are in play, public policy has until recently focused almost exclusively on dealing with perceived shortages. Ottawa has targeted billions of new dollars to address wait times (though how remains a mystery). Virtually every provincial government has periodically added new money - often in mid-year - to increase the number of diagnostic or surgical procedures in the hope of reducing both the number of patients waiting and the time that they wait. On the organizational front, the Saskatchewan Surgical Care Network is notable for its emphasis on inefficiency, fragmentation and a lack of transparency rather than resources alone (or even at all) as possible causes of unreasonably long wait times.

Baker and Schwartz have outlined proven strategies for using resources more efficiently, among which are system redesign, such as eliminating unnecessary steps; defragmenting the entry portals into the system; and more flexible access to resources available to larger groups of people in need. These measures are intuitively sensible and, one would think, relatively straightforward to implement. But reengineering redeploys resources, and redeployment often affects providers' (notably physicians') incomes. In many hospitals and health regions, physician access to OR time has largely depended on the accumulation of large numbers of people on their personal wait lists. Those with the largest lists often get the most OR time; there is thus an incentive to maintain a large list. If the median or maximum waits for a physician's patients are long, the proposed solution is invariably more OR time; if this is the case across many physicians and patient categories, the proposed solution is invariably more resources. While the story does not play out in this manner everywhere at all times, it remains more the norm than the exception. These incentives are exacerbated by the vagaries of feefor- service payment systems, which have always rewarded procedures over consultation.

It is important to note that in such circumstances, neither the lists, nor the physicians' judgment about who needs what procedure are subject to meaningful peer or other scrutiny. Despite claims that Medicare is highly regulated and severely managed, decisions about who needs and receives (particularly non-urgent) services are taken in a state of anarchy. This is more true of elective than urgent procedures, but there is a large body of Canadian research documenting huge variations in practice that persist even when known to administrators and practitioners.

The results are predictable: some patients will wait a very long time; some physicians will assemble large wait lists; there will be major variation in the indications for a procedure, and a general lowering of the threshold over time; and only energetic and savvy patients will be aware that wait times vary greatly by physician and institution. One would be hard pressed to design a system more likely to produce chaos, unfairness and constant pressure to expand.

There has been a more systematic approach where patients' lives are at stake. For heart surgery and cancer treatment, there is more monitoring of wait times and usually a reasonable attempt to serve people in order of need. Often there are standardized needs assessment protocols and target wait times (some evidence-based, others not). Even here, though, there are unarticulated and unresolved dilemmas turning on the issue of need.

It is simple logic that if the need for services and the supply of services match, over time and in aggregate, wait time issues should disappear. At times there will be unforeseeable clusters of need and personnel shortages or maldistributions that create temporary or location-specific problems, but the basic principle holds. If, in a given time period, the number of people newly presenting with a need equals the number of people the system is able to serve, wait times should be trivial (assuming backlogs have been dealt with). But this basic equation depends on a crucial development: consensus on what constitutes a legitimate need that the healthcare system can address at an acceptable cost with a reasonable prospect of a positive outcome.

These are, of course, highly contentious issues. New Zealand faced them bravely by instituting a point count system to measure need, and setting thresholds for entitlement to service in the public system. Those falling below the threshold score were ineligible for publicly financed care. When the threshold is set too high in the eyes of the public and/or providers, two responses are predictable. One is a demand for a private option, so that people can pay for and get service regardless of their point count. The second is a loss of confidence in the public system as seemingly unjust decisions and pitiably ineligible cases make their way into the media and shape public perceptions. New Zealand does have a private, parallel system, and the public has on occasion expressed dissatisfaction with the thresholds set. There have also been reports of "gaming" the system by subtle and not-so-subtle cues that encourage patients to report higher levels of pain or disability, thereby inflating the point count above the crucial threshold. Regardless of these imperfections, we owe a debt to New Zealand for attempting to bell an elusive and sharp-clawed cat, just as we learned a great deal from Oregon's heroic effort at rank-ordering services.

The "What is a real need?" debates have generally featured so-called elective surgery, where quality of life, rather than life and death, hangs in the balance. But they apply equally to lifeand- death situations as well, with a different twist. In such situations - cancer being a prime illustration - it is not the need that is in question, but the prospects for addressing it successfully. Science and medicine have not defeated most cancers: for many, interventions do little to alter outcomes, and in some instances the treatment diminishes quality of life without prolonging it appreciably. On a strictly utilitarian basis, one could convincingly argue that the costs of many healthcare interventions greatly exceed their benefits. Yet we are not strict utilitarians, and prosperous societies may thoughtfully decide to provide some costly and intensive treatments even when the odds against success are depressingly long. Rather than taking a New Zealand-type approach and draw a firm line that separates the eligible from the ineligible, in Canada we fudge the question in two ways. Either we make people wait (push them down the priority list), or we perpetually expand the eligibility pool by adding resources, in effect buying our way out of uncomfortable choices.

The "Canadian way" is not entirely an indefensible form of muddling through. Often it may be reasonable to lower thresholds for intervening - for example, if it proves safe to perform heart surgery on people over 80, adding years or quality of life, one would be hard pressed to argue against it. But some thresholds, particularly on the diagnostic side, have lowered dramatically without any evidence of improved management or outcome. Some of this utilization growth involves relatively inexpensive technologies, such as ultrasound, but much has occurred at the high end, particularly CT and MRI scanning. The threshold is not lowered as a result of transparent deliberation based on cost-benefit analysis and other measures of justice, effectiveness and efficiency. It is lowered by the collective but uncoordinated decisions of providers who find it almost irresistible to expand the use of technologies that pose no risk to patients and entail the marvels of human ingenuity.

Here is where flow optimization and system redesign successes meet their match, and in a sense sow the seeds of their own unravelling. Take the example of MRI. When practitioners knew that machines were few and capacity limited, they reserved referrals for cases where there was genuine diagnostic uncertainty and real urgency. Over time, the technology embedded itself into common practice and expectations, the referral criteria loosened and waits lengthened. Governments installed more machines. But adding capacity permanently solves the wait time problem only if indications for use remain roughly the same. They don't. If the new capacity is intended only to clear the backlog, the number served must exceed the number of newly presenting cases. This indeed clears the backlog, but also creates excess capacity, which in Canada is viewed as an embarrassment, if not a scandal. It is quickly absorbed by using the technology in new categories of cases, new line ups appear and the cycle repeats. So it is, mutatis mutandis, with cataract surgery, joint replacements, and so on.

I am not suggesting that there is no unmet legitimate need or that all thresholds are now "unreasonably" low. The point is that to solve wait times systemically and durably, there must be a serious discussion of need and its sister concept, appropriateness, in all their dimensions. This can be treacherous terrain, but there is no turning away from it. Otherwise, as the system successfully reduces wait times, "need" will be redefined downward - sometimes wisely, sometimes not - and avoidance of the central issue will be no more successful than Neville Chamberlain's policy of appeasement prior to World War II. Paradoxically, if we are not prepared to discuss needs and establish thresholds (on a principled and compassionate basis, with nuance and flexibility), leaving the wait time problem intact may be the best option. Vexing though they may be, long waits do discipline choices and behaviours, and the perception or reality of an overburdened system creates de facto thresholds - no doubt variable, unstudied and unfair.

In a just and well-ordered system, all waits should be insignificantly long, and medicare should serve all those with legitimate and addressable needs. The easy work is defining "insignificantly long"; the hard but important work is defining "legitimate and addressable need." We've circled the dilemma for too long; the prospects for a real solution to wait times depend on our confronting it.

About the Author

Steven Lewis is a health policy consultant and adjunct professor at the Centre for Health and Policy Studies at the University of Calgary.