Access without Appropriateness: Chicken Little in Charge?

Abstract

Health policy makers in Canada have swung between the twin poles of ensuring access and controlling costs. Recently, access has dominated. Reconciling these opposing ideals, rather than alternating between them, requires adding the concept of appropriateness, and recognizing that rapid access to unneeded care may do more harm than good. Several examples are given of resources wasted (and side effects endured) through inappropriate use, and a few modest suggestions for improvement are made.

How bad is Canada's healthcare system? If one were to judge from press coverage, it is in deep trouble – unsustainable, and forcing sick Canadians to wait for life-or-death care unless they are lucky enough to be able to travel south of the border. If one were to judge from comparative statistics, however, a different picture emerges: health outcomes are better than the average for OECD countries; costs are more or less under control, although with some worrying trends; and people receiving care seem relatively satisfied. F. Scott Fitzgerald wrote: “The test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function.” Health policy analysts stress three key goals – cost, access and quality – and suggest that trade-offs among them are inevitable. It goes without saying that there is always room for improvement, and indeed, several thorough reviews of the system have reached remarkably similar conclusions about the changes that need to be made: Expand the definition of insured services beyond the historically based emphasis on hospital and physician care to include, at minimum, targeted home care services, prescription drug coverage and better mental healthcare. Reform primary healthcare. Emphasize disease prevention and health promotion. Get better data. Reorganize care to improve efficiency and patient flow.

Ignoring, for a moment, the fact that there is no such thing as a Canadian healthcare system – it being neither national nor a system – health policy makers are confronted by Fitzgerald's intelligence test. We appear, indeed, to have responded by holding two opposed ideas, cost and access, but only one at a time, swinging alternately between them. For the past few years, one idea has dominated – access – which translates as, “There are not enough…” (doctors, nurses, fill in the blank). People have to wait too long for care. Expensive drugs are not provided at public expense. Chicken Little cries triumphantly throughout the land: The sky is falling! We need more! Governments responded in 2004 with wait time targets for five services: cancer care, cardiac care, hip/knee replacements, cataract surgery and CT/MRI scans.

Then, the second idea intrudes: Publicly funded healthcare is unaffordable. The system is not sustainable. We need cost control. A brave few point out that our current bottlenecks result from the success of earlier cost control efforts: per capita inflation-adjusted Canadian health expenditures actually dropped in the mid-1990s. Funders capped hospital budgets, providing, in turn, an incentive for cost shifting. Hospitals laid off nurses; the resulting exodus from the profession soon created a nursing shortage. Physicians had to battle for operating room time; those with less internal power (often the providers of elective surgery) found themselves on the losing side, leading to wait lists for their services. The backlash, in turn, led to the current focus on restoring resources. A cynical observer might suggest that there will soon be a similar backlash against the costs of the access agenda, particularly if the economy slides into recession.

Is our intelligence first-rate enough to reconcile these opposing ideas, rather than just alternate between them? A modest proposal suggests yes, if we allow in a third idea that too rarely enters the dialogue – appropriateness. It involves recognizing that more is not always better. Care can do harm as well as good. Rapid access to care that is not needed is not always wise.

Many years ago, Bob Evans (1984) noted that we do not want to buy healthcare, we want to buy health, but do not know what economists call the “production function” connecting them. Healthcare is not a normal consumer good. In general, markets do very well at distributing many kinds of goods, using price signals to balance supply and demand. If demand is high, price rises until enough consumers are priced out of the market to balance supply and the new, lower level of demand. If price is low, demand rises.

But healthcare adds in another wrinkle – need. And, taken seriously, this addition erodes the basic premises of markets. First, it adds another decision-maker – the expert. I decide what I want, but the healthcare professional decides what I need. Next, it destroys the mechanism of price signals. If I need care, should I get it, regardless of my ability to pay? If the answer is yes, then there is no limit to the cost that can be charged – “your money or your life” has a long history of successfully parting individuals from their cash. Conversely, if I do not need care, should I get it as long as I am willing to pay? Evans's insight is important – why on earth would I want to? Most healthcare is not pleasant to consume; there are risks and side effects. Our focus on balancing costs and benefits leads to an understandably negative reaction from people who disagree with the conclusion that they are not “worth” the cost of treating them. Don't we know that human life is priceless? A focus on balancing risks and benefits, in contrast, would make explicit what health professionals already know: Sometimes less is more. Treatment may do more harm than good. It is important to target interventions towards those for whom the benefits are likely to outweigh the harms.

This is usually easier said than done. Knowing whether benefits outweigh harms requires evidence. Who gathers it? One problem is that people get paid considerably more for providing a good or service than for advising that it is not necessary. Unsurprisingly, those with fiscal interests seem inclined to show that their services are beneficial (Lexchin et al. 2003). Gathering evidence takes time. Should sick people be expected to wait? If not, what should be done before the evidence is in? Evidence often shows that there is a sliding scale of benefit. How much benefit is worth buying? At what cost? People may disagree in their interpretation of the evidence. How much autonomy is appropriate, and who should decide?

In consequence, we have tended to ignore appropriateness altogether for some procedures. True, among the recommendations of the Romanow Commission (2002) was the obviously unlucky Recommendation 13: “The Health Council of Canada should take action to streamline technology assessment in Canada, increase the effectiveness, efficiency and scope of technology assessment, and enhance the use of this assessment in guiding decisions.” Clinicians are making a valiant effort to assess evidence, and have used it well in developing the wait time standards for cancer and cardiac treatment (Health Council of Canada 2007). (Not coincidentally, wait times for those conditions appear to be largely under control.) However, those for cataract surgery, hip and knee replacement, and diagnostic imaging (CT, MRI) have proven more difficult, and that appears to be where much of the Chicken Little dialogue is focused.

Why does the appropriateness dialogue have such little traction?

It is remarkably difficult to convince patients or physicians willingly to forgo therapy that they believe would be helpful “merely to save money.” In the current dialogue, efforts to target are equated with denial of needed care. Patient groups – either grassroots or provider-funded Astroturf groups, backed by the media – are quick to demand that third-party payers pay any price for the newest drug. These groups are found in most countries, making remarkably similar arguments, and few governments can resist such pressure. Decisions by Canada's Common Drug Review (Tierney and Manns 2008) that drugs do not offer sufficient benefit to warrant listing are quickly, and vociferously, denounced by both pharmaceutical companies and patient groups. Ontario's attempt to introduce PET scanning within the context of evaluating its effectiveness, rather than being hailed as a way of determining when this procedure is beneficial, has been denounced as limiting access.

Hang risks and benefits; the public assumption appears to be that no one should ever wait, and that more is always better. Consider the priority categories for joint replacement agreed to by the Wait Time Alliance (2005: 19), and its benchmarks:

• Priority 1: A situation that has the potential to deteriorate quickly and result in an emergency admission should be operated on within 30 days.

• Priority 2: A situation that involves some pain and disability but is unlikely to deteriorate quickly to the point of becoming an emergency admission should be operated on within 90 days.

• Priority 3: A situation that involves minimal pain, dysfunction or disability and is unlikely to deteriorate quickly to the point of requiring emergency admission should be operated on within 6 months.

Some might suggest that situations involving minimal pain, dysfunction or disability might not warrant surgery at all, let alone within six months, and one suspects that most surgeons would agree. That, however, was not the reaction of the focus groups: “Six months is too long to wait if you think about anyone supporting a family. How can they wait that long?” (Wait Time Alliance 2005: 20).

Shortening waiting periods for unneeded therapy that causes harm seems counter-productive. But the horror stories are almost exclusively of people denied expensive innovations that might benefit them. Somehow, the horror stories of people getting expensive innovations that proved inappropriate and even harmful do not carry the same weight. The lessons of Vioxx have not sunk in. One recent example is Bayer's Trasylol, a drug used to prevent blood loss during artery bypass graft surgery (Deber 2007). According to newspaper coverage, it cost roughly $1,300 per patient, compared with $11 and $44 for its alternatives (Pringle 2006). It accordingly had to be better, and an estimated 246,000 US patients received it in 2006 – most for off-label uses. What did this extra money buy? Two recent non-randomized studies told us: higher death rates, and higher risks of such serious side effects as kidney problems, heart attacks and strokes (Mangano 2007). The total cost for these worst outcomes was estimated at between $250 million and $1 billion. (One excellent source for keeping up with the published data is websites set up by lawyers sensing a new revenue opportunity).

Things have gotten sufficiently out of hand that in 2008 the New York Times started running a series entitled “The Evidence Gap: High-Priced Promise,” with the tag-line, “Articles in this series will explore medical treatments used despite scant proof they work and will consider steps toward medicine based on evidence.” The first example the newspaper selected fits within one of Canada's five wait time targets – specifically, CT scanners to produce detailed images of the heart (Berenson and Abelson 2008). The article described a US clinic asking to buy a machine that would cost them $1 million. But the doctors can pay for the equipment by doing enough CT angiograms (at $500 to $1,500 per test), and over 150,000 such scans were done in the United States in a single year, for a cost of more than $100 million. Such tests are not without risk; the dose of radiation is large enough to increase the lifetime risk of cancer. And the evidence that they benefit most patients is not there. But lobbying from patient and physician groups has ensured that US Medicare continues to pay for them.

In Ontario, nearly $100 million has been spent to increase the supply of CTs and MRIs. More machines have been bought, and more scans have been paid for. The government pulled in providers to determine how best to proceed (Trypuc et al.2006a,b,c). In response to greater capacity, utilization has soared. What is worrying is that the clinical benefit is often problematic. For example, a recent ICES analysis by You et al. (2008) reviewed hospital charts; the authors found that the most frequent reason for ordering a CT scan of the brain was headache, and less than 2% of those scans showed an abnormality. These tests also led to other tests in 25% of cases. How low a yield is appropriate? Scans involve radiation – regardless of the questions of wise use of resources, at what point does the clinical harm (e.g., increased cancer risk) outweigh the potential benefit?

The second story in the New York Times series looked at the cancer drug Avastin (Kolata and Pollack 2008). This drug has long been a poster child for the horror stories of people denied care. It costs up to $100,000 per year per patient; its sales are over $3.5 billion per year ($2.3 billion in the United States). And studies show that it prolongs life by only a few months, if that. Is it worth it?

Another Canadian example, following enormous pressure, newspaper stories about dying patients, lawsuits and involvement by the Ontario Ombudsman, the Ontario Ministry of Health and Long-Term Care agreed to send patients with colorectal cancer to the United States for another expensive drug, Erbitux. According to a story in the Globe and Mail, Ontario paid $32 million for 418 patients to have the infusions, alone or in combination with another drug, over a three-year period starting in 2005 (Priest 2008). Further research, reported at a 2008 meeting, suggested that it was useless for 40% of patients. The Globe story, however, conveyed this as yet another example of nasty government refusing to pay for something – in this case, not immediately making available (and paying for) a newly developed $500 test to see whether the drug would work. Little mention was made of the money wasted, and side effects endured, by premature adoption of a therapy with insufficient evaluation.

So, what might be done? One might hope that Chicken Little will gain some perspective and celebrate what works well. Other suggestions also come to mind. The first is to focus on appropriateness. This does not mean that the call for evidence should be used to block innovation; we should innovate, but also evaluate. Ideally, new therapies would be applied within the context of trials and registries, so that evidence could be collected and subsequent use targeted towards those for whom such treatments would do more good than harm. Another suggestion is to negotiate the price to be paid for these therapies, particularly for drugs, rather than agree to pay whatever companies wish to charge. Another is to shift the dialogue from cost–benefit to risk–benefit. Finally, it might be nice to focus on outcomes, with the recognition that we may end up paying less attention to diagnostic imaging and elective hip/knee surgery and more to prevention, home care and the other reforms that keep being suggested, but have so far not been implemented.

With luck, the title of this column might become yesterday's news, with Chicken Little settled happily in her barn. It is high time that health policy stopped being guided by fictional poultry.