Physicians, Thou Shalt Ration: The Necessary Role of Bedside Rationing in Controlling Healthcare Costs

Peter A. Ubel MD
HealthcarePapers
Vol. 2 No. 2 2001 | Physician Rationing

Abstract:
Physicians are often asked to be "gatekeepers," determining their patients' access to medical therapies and technologies. At the same time, most physicians have been taught that they should act as patient advocates, pursuing patients' best interests regardless of cost. This paper reviews moral arguments ethicists have made for and against "bedside rationing." It argues that healthcare rationing is appropriate in order to help control healthcare costs, and that rationing decisions made at the bedside by physicians must be part of the rationing system. A system that attempts to control costs by mandating an elaborate set of rules would be burdensome, and many physicians would find ways around the rules anyway.

Physicians are deeply conflicted about their roles in cost-containment. Some of the conflict has to do with discomfort over the concept of "rationing," but they are also in conflict about much deeper issues. The author argues that patients can do with less than the "best" treatment and physicians must come to terms with this. Finally, healthcare systems need to signal physicians that it is acceptable for them to offer "less" to their patients in order to serve the greater good.

Ms. Johnson comes to her physician with symptoms of gastric reflux (GERD). Her doctor gives her a prescription for cimetidine, even though he knows omeprazole would be better at relieving her symptoms. He thinks the cost of this other medication is too high for it to be the initial treatment.

In an old Mel Brooks movie,Moses is seen coming down from the mountains with three stone tablets. He announces to his people in a stentorian thunder: "I come down with a copy of God's Fifteen . . ."- he fumbles one of the tablets, and it falls to the ground - ". . . er,Ten Commandments for how to live a good life."

This is pure speculation, but I would guess that somewhere on that broken tablet was a commandment that read: "Physician, thou shalt not ration!" I base my speculation on the tone many people take when debating the appropriateness of bedside rationing by physicians. Opponents of bedside rationing argue vehemently that physicians should never ration from their patients. For example, in a New England Journal of Medicine editorial, Howard Hiatt (1975) wrote: "A physician must do all that is permitted on behalf of his patient." In a similar vein, Dr. Norm Levinsky, chair of medicine at Boston University, has written that: "Physicians are required to do everything they believe may benefit each patient without regard to cost" (Levinsky 1984). Hiatt and Levinsky's statements are consistent with the traditional moral view that physicians should advocate for their patients without regard to costs. This view is treated almost as a theological truth in the United States. But I am a heretic. I think that it is sometimes appropriate for physicians to ration healthcare from their patients in order to help control healthcare costs. In this article, I briefly describe the "theology" of bedside rationing - why so many people think bedside rationing is immoral. I also discuss why I am a heretic. I think any method of controlling healthcare costs is doomed to fail, unless it is joined by some relaxation of physicians' advocacy duties. Finally, I discuss linguistic confusion about bedside rationing.Many people debate the appropriateness of bedside rationing without ever defining what they mean; this leads to disagreements about the morality of bedside rationing among people whose moral values are actually the same.

Theology 101: the Immorality of Bedside Rationing
Opponents of bedside rationing contend that it violates physicians' moral duties to advocate for patients' interests. In addition, they believe that bedside rationing would erode trust between patients and their physicians. Moreover, they hold that bedside rationing creates savings that do not necessarily go toward improving patient care for other patients. I elaborate on each of these objections below.

It has been long argued that physicians have a "fiduciary duty" to advocate for patients' interests (Morreim 1989). A fiduciary is someone who acts on behalf of those who otherwise may not be able to pursue their interests (Hall 1997). Physicians are seen as fiduciaries for patients because they have more knowledge than patients do. Patients may not know what is in their best interests or may not be able to pursue their best interests without a physician's help. Physicians also have fiduciary duties because patients are often dependent on them. Patients are emotionally dependent on physicians because their illnesses make them vulnerable; in addition, patients are legally dependent on physicians, who have been given powers to order medications and perform procedures that other people can not do. Finally, physicians' fiduciary duties arise because healthcare issues are often high stakes. These high stakes distinguish the doctor-patient relationship from otherwise parallel relationships, such as between an auto mechanic and a client. An auto mechanic has more knowledge about cars than a typical person and may have tools to repair a car that the average person could not afford. Although auto mechanics have moral duties, such as to be honest with their clients, their duties are not fiduciary, because the stakes are not high.

People often feel strongly about physicians' advocacy duties towards patients because they recognize that patients need to trust their physicians in order to receive good care (Goold 1998). Many aspects of healthcare depend on trusting interactions between patients and their providers. Patients who do not trust their healthcare provider may not tell the provider about the symptoms they are experiencing, or may not tell about high-risk behaviour they are engaging in. If worried that physicians were rationing healthcare from them, patients might lose trust in their physicians. Consequently, the quality of healthcare that physicians can provide to their patients would erode.

Some opposition to bedside rationing arises because of concern about who would really benefit if physicians rationed healthcare from their patients (Asch et al.; in submission). In the case presented in the introduction, a physician prescribed a less expensive medicine to a patient with reflux in order to save money. But whose money was being saved? If the patient would not have incurred most or all of the increase in costs of the more expensive medicine, then who benefits from this rationing decision? Many people think the money will simply go to a greedy insurance company or to the CEO of a managed-care company. These people contend that there is no moral justification for withholding the best care from patients, given that the money saved by bedside rationing will not necessarily benefit patients.

For these and other reasons, the traditional moral view is that physicians need to do what is in patients' best interests regardless of cost. But, as I stated above, I do not hold that view. Below I will describe why I oppose the traditional view of physicians' moral duties. First, however, I need to define what I mean by "bedside rationing."

Defining My Terms
Bedside rationing is a subset of healthcare rationing; in other words, it is one of many ways to ration healthcare. For the purposes of this article, I hold that healthcare rationing occurs whenever the healthcare system, or "society," allows patients to receive less than the most beneficial healthcare service. If a patient receives treatment A because of resource constraints, when a more expensive treatment, B, would have been better, then treatment B has been rationed from the patient. This rationing could have occurred because of bedside rationing - a clinician might have decided to prescribe A rather than B; or it could have occurred through market forces - the patient could have been asked to pay for either A or B and, thus, chose A; or it could have occurred through any number of other mechanisms.

The definition I have proposed for healthcare rationing is consistent with how most health economists define the term. I have defended this definition elsewhere (Ubel and Goold 1998). Nevertheless, there is no single "best" way to define a complex term like healthcare rationing, nor is it crucial for me to convince you that the definition I propose is the best. Instead, I put forward this definition as a way to clarify the term as I discuss why I believe bedside rationing is morally acceptable.

Before discussing why I believe bedside rationing is an acceptable method of rationing healthcare, I need to note a fact that is perhaps obvious. If healthcare rationing is unacceptable, then bedside rationing is unacceptable. In other words, those who argue that healthcare should not be rationed not only disapprove of bedside rationing but also would disapprove of any other method of rationing healthcare. I will not argue against this view here. Others have argued convincingly, I believe, that there is a need to ration healthcare (Hall 1997; Callahan 1990; Eddy 1994). The proliferation of new technologies being offered to patients with a wide range of illnesses has made it impossible to offer every patient the best possible healthcare services in existence. Each day, new medications become available that, if they were free, would probably be offered to hundreds of thousands of patients. These medications are so expensive, however, we often hesitate to provide them to everyone that would benefit. Cholesterol medications, for example, are slowly diffusing towards a broader group of patients, but if they were as free as water, many people at relatively low risk of coronary heart disease might start taking them. To make this discussion manageable, then, I will ask readers to assume that some amount of healthcare rationing is necessary to help control healthcare costs. The question for this article then is whether any amount of this rationing ought to be done at the bedside by clinicians. In other words, is bedside rationing a legitimate form of healthcare rationing?

Not surprisingly, it is helpful to begin with a definition of bedside rationing. Susan Goold and I have argued that three conditions are necessary for a clinical action to qualify as bedside rationing: (1) the patient must be given less than the best available healthcare, (2) the best healthcare must be withheld in order to save societal resources, and (3) the physician (or clinician) must have control over the healthcare decision (Ubel and Goold 1998). For example, in the reflux case described in the introduction, the physician prescribed the less expensive reflux medicine in order to save society money. Hence, the case was an example of bedside rationing. A subtle change in the case, however, would change this classification. If the patient was responsible for the difference in cost of these two medicines, the decision would not necessarily qualify as bedside rationing, because the physician could potentially be ordering the less expensive medicine to save money for the patient. (In such a situation, the physician ought to talk with patients about how they want to spend their money.) Another change in the case would also eliminate it as an example of bedside rationing. If the patient's health plan required physicians to prescribe less expensive reflux medicines before prescribing expensive ones, the doctor's prescription would not be at her discretion and the health plan would be rationing the expensive medicine from the patient, not the physician.

To better understand what I mean by bedside rationing, it is helpful to think of alternative ways to ration healthcare. Healthcare can be rationed by ability or willingness to pay. Healthcare can also be rationed by formulary committees who decide that expensive reflux medicines are no longer available to all patients. A health plan may decide not to offer lung reduction surgery to its patients, and a government insurer may decide to limit PET scanners for its citizens. These are examples of administrative level healthcare rationing, but not examples of bedside rationing.

Can the Cost of Reflux Treatments Be Contained without Bedside Rationing?
Imagine a healthcare system that is trying to reduce the use of expensive proton pump inhibitors (PPIs), such as omeprazole, in patients with reflux disease. Imagine at the same time that all the physicians in this healthcare plan have vowed never to ration at the bedside. How would this healthcare system go about reducing the use of PPIs?

To begin with, if this healthcare plan did nothing to control PPI use and physicians were committed to providing the best possible care to their patients without regard to cost, then physicians would almost never prescribe the less expensive reflux medicines. PPIs are simply better reflux medicines than the less expensive H2 blockers. Some people might reject the idea that offering less expensive reflux medicines to patients is an example of rationing (Asch and Ubel 1997). They might argue that many patients do just as well with H2 blockers as with PPIs. For these patients, then, no benefit has been withheld if they receive H2 blockers first; if no benefit has been withheld, then no bedside rationing has occurred. This reasoning is faulty, however, because it utilizes an after-thefact evaluation to judge a prior-to-thefact decision. Prior to prescribing a reflux medicine, physicians do not know whether H2 blockers or PPIs will work better for a particular patient. However, patients' chances of successful reflux treatment will be significantly greater with PPIs. Indeed, if money were no object, there would be no reason (in most patients) to prescribe an H2 blocker instead of a PPI.

Because PPIs are superior to H2 blockers but more expensive, the healthcare plan has to find a way to keep physicians from prescribing them if it wants to save money on reflux medications. One way to do so would be to require that all patients undergo a trial of H2 blockers before receiving PPIs. How would such a requirement work? First, the health plan would need a system for documenting whether patients had already been on H2 blockers, so that they could receive PPIs after the H2 blockers failed. The system would also need to track whether patients had received H2 blockers from other healthcare plans prior to transferring to their new plan. In addition, it would need to develop a system whereby physicians could appeal and prescribe PPIs for patients who had taken H2 blockers on the outside or who had some "contraindication" (some medical reason they could not take H2 blockers). Preparing for such exceptions and appeals costs money. The health plan would need to spend money to develop information systems that could monitor the program. This in itself has resource implications and would have to be weighed against the amount of money that would be saved by reducing PPI prescriptions.

But the system would have to be even more complex than I have indicated, or it would create some clinical problems. For example, PPIs are important medicines for treating patients who have non-reflux related stomach problems caused by the bacterium H. pylori. Would physicians be able to prescribe PPIs for such patients? In addition, should the system be prepared to allow exceptions for patients who come in with "severe reflux symptoms"? If so, how should we define severe reflux disease? Once these exceptions are made, how would they be monitored?

I am trying to show, through examples, that rule-based rationing is problematic because the rules can very quickly become unmanageable. Perhaps just as important, rule-based rationing systems are susceptible to physician "gaming"- physicians interpret rules in ways that benefit their patients (Morreim 1991). A notable example of gaming is occurring in the state of Oregon, which, since the mid-1990s, has been trying to reduce its Medicaid expenditures through an explicit rationing plan. Medicaid is a U.S. healthcare program for poor people and is paid for by a combination of federal and state monies. Oregon was having a hard time keeping its Medicaid expenses in line while trying to maintain coverage for all the poor people who needed healthcare. One solution, as formulated by then legislator (and now governor) John Kitzhaber, was for Oregon to specify which healthcare services it would offer to Medicaid patients and which ones it would not. The state hoped to save money by withholding Medicaid services that were deemed less important than other services (Garland 1992). The savings garnered by not paying for such services could then be used to offer Medicaid to more patients.

Despite good intentions, the Oregon Medicaid rationing plan has not saved a dime, because physicians have found ways to get around the rules (Kilborn 1999). For example, if patients have multiple diagnoses below the funding line, physicians will get reimbursed for their treatment. Consequently, when patients come in with "below-the-line diagnoses," physicians almost always find several other "below-the-line" diagnoses in order to get reimbursement for their treatments. Through this and other loopholes, physicians have found ways to make sure patients get the treatments that are best for them. Indeed, physicians are notoriously good at gaming healthcare systems to get benefits for their patients. Thus, to the extent that healthcare plans try to tie physicians' hands to control healthcare costs, physicians wriggle free of the ropes.

What about a Simpler Type of Rule to Control Spending?
So far, I have discussed several rule-based methods of reducing PPI prescriptions. I have argued that these rules need to become more elaborate in order to control healthcare costs, but they can still often be overcome by physician gaming. How about a simpler rule: require gastroenterologist approval of every PPI prescription. Would this simpler rule control costs? Such a rule would have several advantages over more complex rules. It would allow for clinical judgments to be made about which patients really needed PPIs. No committee would need to create an official definition of "severe reflux disease." Instead, clinicians talking to each other on the phone could decide whether a patient really needed a PPI prescription. Such a rule allows for clinical judgments based on the specifics of individual patients. This contrasts with previous rules, which were meant to be applied to all patients.

Despite its advantages, this type of rule has pitfalls too. Most important, it could potentially overwhelm gastroenterologists with pages and phone calls about PPI prescriptions. At one institution where I worked, this plan was rapidly defeated when gastroenterologists told primary-care physicians to write down that they had GI approval any time they wanted to prescribe a PPI. The gastroenterologists were so fed up with receiving phone calls about PPI prescriptions that they found a way to defeat the system.

Let us shift our attention away from reflux disease for a bit and consider a common diagnostic test that has significant expense - CT scans.Who should be able to order a CT? Should all primary-care physicians be able to do so without prior approval? What about primary-care nurse practitioners? At one institution where I worked, nurse practitioners could order CT scans (and MRI scans, for that matter, a significantly more expensive test) without discussing this with a radiologist or primary-care physician. I saw patients who presented with new onset shoulder pain who were referred for MRIs by clinicians without anybody asking a radiologist if that test was indicated. In my clinical judgement, this is a wasteful practice. But how do we keep it from happening?

Healthcare systems could require that all CT scans and MRI scans be approved by radiologists. As with the gastroenterology example described above, however, such a policy would potentially overwhelm radiologists with such requests. Moreover, in many cases, primary-care practitioners have every reason to know that a scan is indicated.

Although it makes some sense to require clinicians to speak with radiologists before ordering extremely expensive radiology tests, I have concerns about a system that requires such conversations. First, such a system ignores many less expensive tests that, nevertheless, are ordered frequently enough that they cost healthcare systems a lot of money. Think of all the plain film x-rays that are ordered for low-back pain and for routine screening of lung fields that have almost no clinical value. Second, such a system imposes burdens on radiologists and other physicians who must now find time in their busy days to speak with each other, even in circumstances where the correct radiology test to order is obvious. This not only takes up these clinicians' time, but also ultimately costs healthcare systems money, because radiologists and other physicians are highly paid professionals. Third, the systems create a layer of bureaucracy in order to document that conversations have occurred between radiologists and other clinicians. Bureaucracies create hassles. Just as important bureaucracies cost money.

Instead of Bureaucratic Rules or Time-Consuming Approval Requirements, Why Not Give Physicians Feedback about Their Utilization?
There are alternatives to forcing clinicians to call each other on the phone to get approval for every lab test, radiology test, and expensive medication they want to order. One alternative is to give clinicians feedback, every few months or so, about how much they utilize expensive diagnostic tests in comparison to their peers. Research has shown that such feedback reduces physicians' utilization (Berwick and Coltin 1986; Schectman et al. 1991). Physicians hate feeling like they are outliers; they do not want to rely on expensive tests more than other physicians do.

Such feedback systems deserve a role in helping control healthcare costs. They do not require burdensome rules, but they do not avoid bedside rationing. If physicians are only interested in patients' best interests, they will completely ignore how their utilization patterns compare to other physicians. In fact, in a world without bedside rationing, they will continue to order what they think is best for their patients. Those who order fewer CTs and MRIs may have reason to wonder if they are ordering too few tests.

What about Avoiding Bureaucratic Rules by Implementing Capitation or Other Financial Incentives?
Asking physicians to bear some financial risk for the tests and medicine they order for their patients has also been shown to control healthcare costs. Under such "capitated" healthcare systems, physicians are given a certain amount of money to take care of their patients (Hillman 1990). Some percent of the money they spend caring for their patients is then taken from their salary. This encourages physicians to order fewer tests and referrals.

Many people have raised moral objections to capitation-reimbursed systems. I do not plan to discuss these arguments here. Instead, I want to make a simple point: capitation systems only control healthcare costs by encouraging bedside rationing. If patients' best interests were all that mattered, most clinical decisions would be unaffected by capitation: clinicians would still do what is best for their patients, regardless of costs.

What about Rationing with Practice Guidelines?
Some colleagues of mine in Michigan recently published an article on the cost effectiveness of performing routine retinal screening exams for diabetic patients (Vijan et al. 2000). The standard of care, up to now, has been to make sure that all diabetic patients have ophthalmology examinations each year to screen them for diabetic eye disease. My colleagues argued that annual screening is unaffordable for patients with mild diabetes; such a screening rarely prevents blindness, compared to screening every two to three years.

If physicians only worried about patients' best interests, they would ignore my colleagues' work, because annual screening would still prevent more cases of blindness than less frequent screening. However, it is likely that in the near future, leading diabetes organizations will change their recommendations on how to screen patients with mild diabetes and recommend screening every other year in low-risk patients. These guidelines will probably have a significant influence on physicians' referral practices.

If physicians begin to follow diabetes society "guidelines" for how often to screen people for retinal disease, they will be engaging in bedside rationing. However, I expect that many physicians will not realize that they are rationing at the bedside when they follow these guidelines (Asch and Ubel 1997). I happen to think these guidelines are signals from society about how much money they want physicians spending to prevent rare illnesses. The decision is, however, still ultimately up to individual clinicians. Such guidelines will have no effect on clinical practice unless physicians are willing to ration at the bedside.

How Do I Justify My Heretical Support of Bedside Rationing?
As hinted above, I have a major concern with healthcare systems that ration without any reliance on bedside rationing: these rationing systems will be burdensome. I scratched the surface in discussing how a healthcare system might try to reduce the use of PPIs. I did not even begin discussing the similarly burdensome rationing mechanisms the system would need to control the use of expensive hypertension medicines, reduce subspecialty referrals, decrease the use of marginally beneficial lab tests, reduce the length of outpatient visits, or reduce the frequency with which physicians order follow-up appointments. A system that controls healthcare costs by creating elaborate rules around all these varied types of clinical decisions would be a bureaucratic nightmare and a clinical disaster.

Those who want to control healthcare costs must decide how they will trade off between blunt, obtrusive rules that completely delineate physicians' behaviours and some amount of bedside rationing that encourages physicians to reduce their use of marginally beneficial healthcare services (Welch 1991).

When I say I am in favour of bedside rationing, I mean the following: at times physicians need to relax their advocacy duties and give their patients less than the best possible healthcare services in order to save money for society. The entire rationing burden should not fall on physicians' hands. There is an appropriate role for administrative rationing mechanisms. In fact, many of the "burdensome rationing rules" I discuss above would be made much less burdensome if we could rely on physicians to occasionally ration at the bedside. For example, a healthcare system could ask physicians to prescribe H2 blockers whenever possible before prescribing PPIs. Such a guideline, handed down by a healthcare system or by a respected medical society, would help physicians remember that H2 blockers are still good medicines for many patients and that society needs to control healthcare costs by reducing the use of expensive PPIs. At the same time, this guideline would allow physicians to use their judgment about when to make exceptions to the guideline.

No system that relies completely on administrative rationing mechanisms will succeed in reducing healthcare costs. In addition, a system that relies heavily on willingness and ability to pay to ration healthcare is morally questionable (for reasons I will not go into here). Thus, the best way to ration healthcare is to have a mixture of administrative rationing mechanisms and clinicians engaging in bedside rationing, with a touch of willingness to pay on the side.

I recognize that there are moral problems with bedside rationing. The problems cannot be eliminated, but they can be reduced. For example, we need to make sure that physicians ration in ways that do not greatly reduce patient trust. I think this is achievable.We also need to do what we can to make sure that money saved by healthcare rationing (bedside or other) goes towards appropriate ends. The goal of healthcare should not be to maximize profits. We need to find ways to help physicians ration at the bedside so that they will not do it haphazardly or in a discriminatory manner. At the same time, we must judge bedside rationing the same way we judge democracy - by comparing it to the alternatives. In this case, the alternatives include burdensome rationing rules, many of which physicians would bend in their patients' favour, and increased use of out-of-pocket expenses to ration healthcare, which favours wealthy patients over others. Bedside rationing has weaknesses, but I think its weaknesses are worth accepting in order to avoid the weaknesses of alternative ways to ration.

What Do Physicians Think of Bedside Rationing?
I have all too briefly discussed the theology behind opposition to bedside rationing and the heresy of relying on bedside rationing to help us control healthcare costs.What do most physicians think about these issues? Do physicians hold to the good old religion and oppose bedside rationing or are they joining heretics like me in accepting its necessity?

Sorting out physicians' attitudes towards bedside rationing is tricky, in part because of linguistic confusion about what it means to ration at the bedside. When physicians disagree about bedside rationing, they could potentially be having one of two kinds of disagreement: (1) they could be disagreeing about the appropriateness of having physicians do less than the best for their patients, or (2) they could be disagreeing about the meaning of the word rationing. That is, physicians might agree with each other that it is appropriate to withhold PPIs from patients, but disagree about whether this is an example of bedside rationing. As an analogy, consider two people who are looking at an insect. In one case, they agree that the insect is a moth, but disagree about whether it is beautiful or ugly. This is similar to agreeing about what it means to ration at the bedside while disagreeing about whether such rationing is justifiable. In another case, imagine they agree that the insect is beautiful while disagreeing about whether it is a moth or butterfly. This is analogous to agreeing that bedside rationing is acceptable - that doing less than the best for patients is acceptable - while disagreeing about whether this is an example of bedside rationing.

To sort out these two ways of disagreeing about bedside rationing, David Asch and I presented 1,000 general internists in the United States with a vignette in which a hypothetical physician offers a less expensive and less effective colon cancer screening test to a patient in order to save money for society (Ubel 2000).We asked physicians whether the physician who ordered the less expensive colon cancer screening test was acting "appropriately" and whether that physician was performing "healthcare rationing."We found that physicians generally agreed that it was appropriate for the hypothetical physician to offer the less effective test to the patient. In fact, only 20% of physicians felt that ordering the less expensive colon cancer screening test was inappropriate. Physicians were completely divided about whether such a decision was an example of healthcare rationing. Forty percent thought it wasn't, 40% thought it was, and 20% had no idea. This suggests that physicians generally support the idea of making cost-quality trade-offs at the bedside. They are comfortable offering a decent screening test to someone, even though a slightly better and significantly more expensive test is available. They are not sure whether to call such a thing "bedside rationing."

This disagreement among physicians about what qualifies as bedside rationing should not surprise us. As I stated above, there is no single way of defining complex terms such as rationing or bedside rationing. And besides, when we conducted this survey, my elegant definition of bedside rationing had not yet been published! More important, rationing is a loaded term. People may support the concept of rationing without wanting to label it that way. I do not have a problem with that. I personally like to use the term rationing, because I think it forces people to consider the moral implications of their decisions. Other euphemisms might be easier to swallow, but they might make us less likely to notice when we are ingesting rotten food.

Nevertheless, I am happy to define rationing in different ways, if that is what people want to do. Instead, what is more important to me is that physicians recognize that they do not currently pursue patients' best interests without regard to costs. I can find examples for almost any physician I know in which they are forgoing a marginally beneficial test or referral because of its expense. Different physicians have different thresholds, but I would guess even Norm Levinsky does not order a thyroid screening test on every patient he sees every few months. If money were irrelevant, even he would order more thyroid tests. Physicians have not done a good job of recognizing that they are making trade-offs between cost and quality. By failing to recognize these trade-offs, they are probably not doing a very good job of making them. If clinicians recognized the trade-offs they made every day, they could begin to look across their entire practices and see when they were trading off too much quality for not enough cost savings, and when they were not trading off enough.

Conclusion
Heated debates about the morality or immorality of bedside rationing have been missing the point.We are so worried about the loaded term "rationing" or about old-fashioned moral ideals that were developed in a time when healthcare costs were not nearly as high as they are that we are not facing up to the new reality.We need to control healthcare costs, and physicians must play a crucial role in helping society do so.

Society is still coming to grips with resource constraints in medical care, especially in the United States. Not surprisingly, many people are not sure who they think ought to be making rationing decisions. Clinicians are equally confused; they do not want to bear a disproportionate share of decision-making over rationing. On the other hand, most clinicians do not want to practise healthcare amid a sea of burdensome rules that limit their abilities to take care of patients.

Whether or not clinicians call it "rationing," they need to recognize that they have a crucial role in helping to control healthcare costs. The best way to control costs is for clinicians to relax their advocacy duties in conjunction with other rationing mechanisms. Clinicians need to recognize that patients can do with less than the best. Physicians need to come to grips, individually if not as a group, with what services they can withhold from patients. And finally, healthcare systems need to find ways to signal to physicians that it is okay to do less than the best for their patients in order to serve the greater good of the population.

About the Author
Peter A. Ubel, MD
Veterans Affairs Health Services Research and Development
Michigan Program for Improving Healthcare Decisions
Division of General Internal Medicine, University of Michigan

Dr. Ubel is a Robert Wood Johnson Foundation Generalist Physician Faculty Scholar, recipient of a career development award in health services research from the Department of Veterans Affairs, and recipient of a Presidential Early Career Award for Scientists and Engineers (PECASE).

READERS (with personal or institutional subscriptions) can access detailed commentary here.

Commentaries are by:

The Tragedy of the Medicare Commons?
Peter H. Barrett

Rationing Healthcare: The Appeal of Muddling Through Elegantly
David J. Hunter

Cutting Healthcare Costs without Rationing at the Bedside: Preserving the Doctor-Patient Fiduciary Relationship
Saul J.Weiner and Charles L. Rice

Bedside Rationing by Physicians: The Case Against
Arthur Schafer

The Need Is to Prioritize, Not Ration
Val Rachlis

Physicians Must Participate in Establishing Standards of Care
Gregory Powell

Advocacy and Rationing Are Compatible
Claude Gratton and Margaret Keatings

Final Response
The Author Responds: Putting Bedside Rationing Back into Perspective

Acknowledgements
The author acknowledges Julie L. Lucas for her assistance in manuscript preparation. References
Asch, D.A., et al. When Money Is Saved by Reducing Health Care Costs, Where do Physicians Think That Money Goes? (in submission).

Asch, D.A. and P.A. Ubel. 1997. "Rationing by Any Other Name." New England Journal of Medicine 336:1668-71.

Berwick, D.M. and K.L. Coltin. 1986. "Feedback Reduces Test Use in a Health Maintenance Organization." JAMA 255(1): 1450-54.

Callahan, D. 1990. What Kind of Life: The Limits of Medical Progress. New York: Simon and Schuster.

Eddy, D.M. 1994. "Health System Reform:Will Controlling Costs Require Rationing Services?" JAMA 272: 324-28.

Garland, M.J. 1992. "Rationing in Public: Oregon's Priority-Setting Methodology." In Rationing America's Medical Care: The Oregon Plan and Beyond, M.A. Strosberg, et al., eds.Washington DC: Brookings Institution.

Goold, S.D. 1998. "Money and Trust: Physician Incentives and the Doctor-Patient Relationship." Journal of Health, Politics, Policy, and Law 23(4): 687-95.

Hall ,M.A. 1997. Making Medical Spending Decisions: The Law, Ethics, and Economics Of Rationing Mechanisms. New York: Oxford University Press.

Hiatt, H.H. 1975. "Protecting the Medical Commons: Who Is Responsible?" New England Journal of Medicine 293: 235-41.

Hillman, A.L.. 1990. "Health Maintenance Organizations, Financial Incentives, and Physicians' Judgments." Annals of Internal Medicine 112(12): 891-93.

Kilborn, P.T. 1999. "Oregon Falters on a New Path to Health Care." New York Times. p. A1.

Levinsky, N.G. 1984. "The Doctor's Master." New England Journal of Medicine 311(24): 1573-75.

Morreim, E.H. 1989. "Fiscal Scarcity and the Inevitability of Bedside Budget Balancing." Archives of Internal Medicine 149: 1012-15.

Morreim, E.H. 1991. "Gaming the System: Dodging the Rules, Ruling the Dodgers." Archives of Internal Medicine 151: 443-47.

Schectman, J.M., E.G. Elinsky, and L.G. Pawlson. 1991. "Effect of Education and Feedback on Thyroid Function Testing Strategies of Primary Care Clinicians." Archives of Internal Medicine 151: 2163-66.

Ubel, P.A. 2000. Pricing Life: Why It's Time for Health Care Rationing. Cambridge,MA: MIT Press.

Ubel, P.A. and S.D. Goold. 1998. "'Rationing' Health Care: Not All Definitions Are Created Equal." Archives of Internal Medicine 158: 209-14.

Vijan, S., T.P. Hofer, R.A. Hayward. 2000. "Cost- Utility Analysis of Screening Intervals for Diabetic Retinopathy in Patients with Type 2 Diabetes Mellitus." JAMA 283(7): 889-96.

Welch, H.G. 1991. "Should the Health Care Forest Be Selectively Thinned by Physicians or Clear Cut by Payers?" Annals of Internal Medicine 115(3): 223-26.

How Well Protected Are Canadian Research Participants: Who Knows?

From Dorothy Pringle, the Editor-in-Chief of the Canadian Journal of Nursing Leadership:

The human research enterprise in Canada is large and growing. It spans a wide range of fields that includes political science, sociology, anthropology, education, social work, nursing, epidemiology and medicine among others. Funding comes from many different sources, including the charitable research foundations such as the Heart & Stroke Foundation and the Canadian Cancer Society, hospital foundations, the three national funding councils - the Canadian Institutes of Health Research (CIHR), the Social Sciences & Humanities Research Council (SSHRC) and the Natural Sciences and Engineering Research Council (NSERC) - and the major provincial funding organizations in Nova Scotia, Quebec, Alberta and British Columbia among many others. Not all research is funded. This is particularly true of the humanities and student research across all disciplines.

In Canada, all research on human subjects must be approved by a research ethics board (REB). The major funding bodies will not release funds until an REB has approved the study. In Canada, the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (TCPS) (Medical Research Council of Canada, 1998) dictates research ethics. It was developed over the 1990s by the Tri-Council Working Group under the auspices of the three national funding councils (the Medical Research Council preceded CIHR as the health research council signatory) and was finally adopted in 1998. By December of that year, all institutions that conducted research were expected to be in compliance with its tenets, or be in the process of becoming so. Since that time, many organizations have invested significant dollars in staff and in information technology to enhance their capacity to undertake credible reviews. Research ethics offices in universities and hospitals, in particular, have grown in size and expertise to respond to the increase in studies that have to be reviewed and to demonstrate the increased expectations of accountability. So, is there a problem, and if there is, what is it? The answer is, we don't know, and the existing system does not make it possible for us to know. The current state of research participant protection in Canada has evolved over time and in response to the numerous policies and expectations set by Canadian and international research funding bodies and other authorities. These policies are not necessarily congruent, and it is virtually impossible to be in compliance with all of them. In fact, the TCPS pertains only to research conducted through funding from the three national research councils, although most organizations stretch that to apply to most of the research coming to their REBs. There are no standards for education in research ethics for investigators, graduate students undertaking research, REB members and chairs, or research participants; for example, some REBs have no members with expertise in law, privacy legislation or qualitative research yet are reviewing proposals that require some or all of these competencies. There are no standards for the operation of research ethics offices in terms of resources, compensation for or acknowledgement of REB members' work, and length of terms for REB members. Proportionate review is still a thorny issue, particularly for researchers in the social sciences and humanities, who believe, with some justification, that the TCPS is biased toward medical or health sciences research. They contend that the level of risk to participants in most of their research studies is much lower than that of medical research in which participants' health and well-being may be jeopardized, and their studies do not require the same scrutiny accorded medical research. The TCPS addresses this and calls for risk to be judged appropriately to the circumstances of the research, but REBs are left to develop their own standards on this issue. Researchers proposing studies that involve collecting data in several (and sometimes dozens) of sites face a requirement for their research proposals to be approved by an REB in each site; this can take months and even years. On the other hand, while there are gaps in policies, concerns about under- and overzealous application of policies that do exist, and angst about multiple REB approvals, Canada has not seen the major breaches in research ethics that have arisen and cost lives in other jurisdictions. In the United States, some universities have had their entire medical research program shut down and all federal funding withdrawn because of grievous breaches to research ethics. However, there is the shared view that "the governance of research in Canada is fragmented and uneven - many players overseeing many other players through the use of many instruments" (Experts Committee 2008: 23) and, at the same time, no way of knowing how well research participants in Canada are protected. Various groups have been and currently are trying to address these issues. The Interagency Advisory Panel on Research Ethics (PRE) was created in 2001 by the three federal funding councils to continue to elaborate the TCPS policies. The PRE has been working for several years on revising the TCPS, including the section on proportionate review, and is supposed to release its revisions for consultation before the end of 2008. Unfortunately, work from the PRE has been slow in developing, and the panel must live with the fact that its sponsors are in conflict of interest when it comes to developing policies to cover research that they fund. This requires an arms-length relationship. NCEHR, the National Council on Ethics in Human Research, is a voluntary organization of individuals with an interest in promoting research ethics and protecting research participants. It is funded by CIHR, Health Canada, the PRE and The Royal College of Physicians and Surgeons of Canada (RCPSC). NCEHR, which has been on the scene since 1989, has developed well-regarded educational programs and a site-visit program for organizations to assist in improving their participant protection programs. Additionally, in 2003 NCEHR proposed an accreditation program that it would operate. However, NCEHR has not been able to secure funding for the proposed program, is frequently limited in its reach by lack of funding, and faces the same problem as the PRE in that its funds come in part from the body that funds research, reducing its arms-length status. SSHRC proposed a type of oversight program, called a public assurance system, as a way of dealing with its constituency's problems with the TCPS. This program was never embraced because it was seen as having "no teeth." In frustration, after a meeting called by NCEHR in June 2005 to review the penultimate draft of its report describing its proposed accreditation program, The Royal College of Physicians and Surgeons convened a meeting to try to deal with the lack of progress in developing an oversight program for research ethics. It invited Health Canada, the three funding councils, the Association of Universities and Colleges of Canada (AUCC) and subsequently eight other organizations including the Association of Faculties of Medicine of Canada to form a "coalition" of sorts, called the Sponsors' Table. The main raison d'être was to establish an expert committee to look into a range of governance models for the oversight of ethics in human research and to explore issues including implementation and funding" (as cited in Moving Ahead, 2008, p. 15) The Experts Committee (of which I was a member) had among its members scientists from the humanities, social and health sciences, including those who investigated research ethics as their area of scholarship, research ethics program administrators and legal experts, and was chaired by the late Arthur Kroeger, a highly respected former federal deputy minister (Experts Committee 2008). The committee's mandate was to provide advice on developing a system for human subject research participant protection in Canada that would address the issues of concern. After nine months of meeting and consulting, the committee prepared a draft report, circulated it for consultation, revised it and submitted the final report, Moving Ahead, to the Sponsors' Table at the end of March. A sad note is that Arthur Kroeger died shortly after the report was completed. The Expert Committee acknowledged both the strengths and weaknesses in the Canadian system but reached the conclusion that a new, independent organization was required that would take responsibility for the oversight of research ethics programs, including policy development, the establishment of educational standards, and the development and operation of an accreditation program for participant protection programs in Canada. These three functions were seen as interdependent, with one influencing the other. It was proposed that the Canadian Council for the Protection of Human Research Participants be established under the Canadian Corporations Act so that it would be at arm's length from all funding bodies. Because of fiscal realities, it was further proposed that the implementation of the Council be staged, with accreditation coming first, then policy, then education. The Sponsors' Table has acted on some of these recommendations; it has established working groups on policy, education and accreditation and has secured funding to support them. It has not embraced the recommendation of the independent Council at this time, but neither has it rejected it, noting that a number of operational issues must be resolved first. It is not clear under whose auspices an accreditation program would operate or how the most concerning aspects of conflict of interest related to policy and accreditation would be resolved. In the view of the Experts' Committee, it is appropriate for the Sponsors' Table to further the development of an oversight system but it is inappropriate for it to operate such a system. It does not have broad representation of organizations and it perpetuates the problem of conflict of interest because the funding councils are among its members. Nursing has not been a major player in this national drama. Nursing organizations are not represented at the Sponsors' Table, whereas organizations representing physicians are, and The Royal College of Physicians and Surgeons has played a leadership role. This is not acceptable. Nurses are now major participants in the research enterprise in this country. Nurse executives are responsible for knowing that the REBs under their auspices are staffed appropriately, have the requisite expertise among the members and are conducting appropriate proportional reviews. Deans and directors of graduate programs need to ensure that their students acquire the knowledge of research ethics that will allow them to conduct ethical studies as students, and as investigators following graduation. Nurse researchers use and are dependent on having robust and highly ethical participant protection programs available to them to review and approve their research proposals. All of us - researchers, nurse clinicians, nurse administrators and nurse citizens - need to have confidence in these same research participant protection programs. We cannot at this time. It is essential that nursing get involved, first by securing representation on the Sponsors' Table. At least one (but it could be more) of the national nursing organizations needs to take the lead on this, with a representative that is knowledgeable about issues related to research participant protection programs. Let's step up to the plate on behalf of all those individuals who contribute to nursing and all other types of research. About the Author Dorothy Pringle, OC, PhD, Editor-in-Chief References The Experts Committee for Human Participant Protection in Canada. 2008. Moving Ahead: Final Report, Ottawa. Medical Research Council of Canada. 1998. Tri-Council Policy Statement. Ethical Conduct for Research Involving Humans. Ottawa.

Is the PHR Just a Tool for the Wealthy and Healthy?

Neil Seeman

Is the PHR like the "executive physical"? Some have suggested that personal health records (PHRs) - a system whereby patients enjoy custodial rights to their secure health data anytime, anywhere - is a privilege that appeals just to the wealthy, healthy or "worried well."

Where's the Evidence? From what I can tell, there is very little substance to this argument in the current peerreviewed literature. Yet I have heard it from those unimpressed by travelling PHR corporate merchants. One often-cited study, by Weingart and colleagues (2006), found that "younger, healthier patients are most likely to make use of this technology." But that study looked at one portal, PatientSite, used at a Boston teaching hospital (Beth Israel Deaconess) and affiliated community practices. It had a small sample size, and the authors acknowledged that "results may not be generalizable to other portals or health care systems, practices, or patients." Most important: patients were eligible to enrol in PatientSite only if one of their physicians had enrolled, and recruitment efforts had focused on primary care doctors. The Weingart study, important as it was, looked at data from 2000-2004; consumer behaviour, especially among those suffering from chronic illness, has changed since then. And the new model of the PHR is available to anyone, anywhere, whether or not his or her physician is enrolled. A Tool for the Rich? The notion that only the wealthy will benefit from PHRs can be easily dismissed; the same argument was made against the telephone, the television and the toaster, yet all these innovations are now taken for granted by the middle class and all but the very poorest among us. PHRs through Google or Microsoft HealthVault are free to anyone, and at least 75% of Canadians have Web access from home. I predict (although it is hazardous to make predictions in healthcare) that in five years, the percentage of Canadian homes without regular PHR engagement will be lower than 2% (about the same proportion of homes without a television). Just as Wikipedia re-invented the encyclopaedia and the Web decimated newspaper ad revenues, the ubiquitous PHR will re-invent our relationship with the health system. Keep in mind that Facebook, now the third-most popular website in Canada, was in its infancy five years ago; change occurs rapidly in the world of Web-based consumer engagement. In a recent paper in Healthcare Quarterly, Kevin Leonard and colleagues (2008) make the analogy to automated teller machines (ATMs). Most of us can remember a time when there were no ATMs, when all transactions required long teller queues. And now, as former US House Speaker Newt Gingrich once pointed out at an OHA keynote address, who among us - be honest - has not fidgeted while waiting in a two-minute ATM line? Banking expectations have radically changed, for the rich and poor alike. Just for the Healthy? Let us deal more seriously with another argument: the idea that the PHR - as proselytized by Google Health, Microsoft HealthVault, Revolution Health and more than 200 smaller competitors - is a gimmick so that the already healthy can maintain their health. After all, PHRs do not keep people away from emergency rooms or from acute care when they're really sick. We heard this argument against Health Savings Accounts (HSAs). The two arguments, however (and the supporting data), are different. HSAs, according to their most partisan defenders, reconnect the patient and the provider with the true cost of care. And patients thereby start to make wiser decisions - or so the argument goes. We gain insights like: "Whoa, we have increased our use of generic medications, we should shop around for the best deals." Or, "Maybe an MRI costs a little bit more at this hospital versus that hospital." And so, according to this theory, we start to use the health system more strategically, perhaps through a high-deductible, catastrophic plan, the deductible being a cash account that accrues interest-free. Yet the HSA is hardly a panacea. If crafted unwisely, it potentially creates more burden for providers and payers, and can lead to increased spending on wealthy consumers to the detriment of the poor. Although this is a very different issue from that of PHRs, it is instructive. In the case of the PHR, there is every reason to assume that the highestintensity users of PHRs will be those suffering from chronic disease - a swath of society that transcends socio-economic class. Research has shown that chronic disease communities thrive exceptionally well on social networking sites. Remarkably, these communities, on sites such as Facebook and MySpace, are more active than celebrity or sports communities (Seeman 2008a). Bottom line: the healthy (and their providers) go to Google for generic health information; the poor, the rich, the middle class - those with chronic disease - seek out, on a daily basis, more niche-oriented Web communities, like diabetesmine.com. The same phenomenon will occur with the PHR. My own more recent research has shown that the "stickiest" health-related blogs cater to niche, chronic illness populations (Seeman 2008b). The users, chronic illness sufferers, will serve as free ambassadors and champions for tools like Microsoft HealthVault and other PHR platforms, building their own specialized add-on applications, which will then position them to blog about their own PHR experience. A Whatcom County, Washington, consortium that developed a PHR as part of a Robert Wood Johnson Foundation project implemented in 2002 found that chronically ill patients became a kind of PHR sales force, enlisting friends and family to join (Sprague, Lisa. 2006). There is no reason (other than anti-corporate animus) to suggest that poor people have a lesser understanding of their chronic health conditions than do the wealthy or middle class, or that they have a harder time communicating their issues to a PHR. (PHRs can offer multi-lingual, real-time translation, though admittedly, the technology needs to improve.) Conversely, the people most likely to flock to PHRs are historically underserved and stigmatized populations who now make greater use of online communities than do others. This includes the mentally ill (including young men who are unaccustomed to talking about their mental illness) and populations who are physically dislocated, such as Aboriginal peoples. In a survey by the Markle Foundation (2005), people who perceived themselves as healthy were less likely than all other groups to use a PHR. As one PHR entrepreneur advised me, "The problem with the PHR is that the healthy ignore them, so that many marketing schemes look to make them appealing to disease-specific clusters of patients." The Real Role of PHRs Keep in mind that the PHR is not intended to solve poverty, a multi-factorial problem, or other more systemic drivers of chronic illness. It is meant to improve consumers' understanding of health issues; to give them more access to and autonomy over their health; to support more timely, appropriate preventive services; to support home monitoring for chronic diseases; to support the continuum of care; to avoid duplicate testing; to increase access to providers via e-visits; and to reduce adverse drug interactions and allergic reactions. John Snow and the PHR As many have noted, engagement requires more than usability. The platform applications, whether via Google or Microsoft or other tools, must provide sustained value. But my instinct is that the PHR is akin to John Snow and the water pump. During the cholera outbreak in 1854 in London, Dr. Snow analyzed the geography of water supply and mortality patterns in Soho and found a disproportionate number of cholera cases within a few blocks of a single water pump on Broad Street. At his behest, the pump was removed. This action has been hailed as one of the first public health interventions of the modern era. Although Snow's microscopic examination of a sample of the Broad Street pump water was not conclusive, his studies of the pattern of disease were convincing enough to persuade the local council to disable the pump. The same can be said about PHRs. The evidence is hardly conclusive, but we do know that consumers are keen to engage with the Web and share secure personal health information that they own. The burden of proof against the PHR should lie with those who challenge consumers' good judgment. A related side note: In a future column, I look forward to tackling another myth about the PHR: that it threatens privacy rights. Lawyers, not consumers, are saying this. Consider this statement by Nicolas Terry (2008): "personal health records are dangerously flawed adjuncts to or substitutes for provider-centric records, and while lacking many of the touted quality or cost-reduction benefits of oft-criticized electronic health records they pose substantially higher levels of risk regarding security, privacy, and confidentiality." Sure, identity theft and privacy risks are always legitimate concerns, and there is a role for government to play in ensuring the security of electronic personal health information. But law should be in the service of innovation that matters to chronic disease sufferers throughout the world. (To be continued…) About the Author Neil Seeman is Senior Resident, Health Strategy Innovation Cell Massey College, University of Toronto Adjunct Professor of Health Services Management Ryerson University Toronto, ON References Leonard, K., D. Wiljer and S. Urowitz. 2008. "Yes, Virginia, There Are System Benefits to Be Gained from Providing Patients Access to Their Own Health Information." Healthcare Quarterly 11(4): 64-68. Retrieved January 4, 2009. < http://www.longwoods.com/product.php?productid =20094&cat=563&page=1 >. Markle Foundation. 2005 (October). "Attitudes of Americans Regarding Personal Health Records and Nationwide Electronic Health Information Exchange." Retrieved January 4, 2009. < http://www.markle.org/downloadable_assets/ 101105_survey_summary.pdf > Seeman, N. 2008a. "Web 2.0 and Chronic Illness: New Horizons, New Opportunities." ElectronicHealthcare 6(3): 104-10. Retrieved January 4, 2009. < http://www.longwoods.com/product.php?productid =19506&cat=524 >. Seeman, N. 2008b. "Inside the Health Blogosphere: Quality, Governance and the New Innovation Leaders." ElectronicHealthcare 7(3): 101-8. Retrieved January 4, 2009. < http://www.longwoods.com/product.php?productid =20298 >. Terry, N.P. 2008 (August 22). "Personal Health Records: Directing More Costs and Risks to Consumers?" Social Science Research Network. Retrieved January 4, 2009. < http://ssrn.com/abstract=1248768 >. Weingart, S., D. Rind, Z. Tofias and D.Z. Sands. 2006. "Who Uses the Patient Internet Portal? The PatientSite Experience." Journal of the American Medical Informatics Association 13: 91-95. Retrieved January 4, 2009. < http://www.jamia.org/cgi/content/abstract/13/1/91 >. Sprague, Lisa. 2006. "Personal Health Records: The People's Choice?" Lisa, Nation Health Policy Forum Issue Brief - No. 820. Retrieved January 4, 2009. < http://www.nhpf.org/pdfs_ib/ IB820_PHRs_11-30-06.pdf >