Showing posts with label health economics. Show all posts
Showing posts with label health economics. Show all posts

Tuesday, 13 June 2017

Ontario Binding Arbitration Framework

The OMA and the Ontario Government, against all odds, have come to an agreement. A preliminary one. On how to approach a real agreement. This isn't a physician services agreement, which is what would be needed to provide for a real update of physician funding schemes, but it's a start. Effectively, this sets the stage for negotiations and provides a framework to come to an agreement through a binding arbitration system.

In rejecting the previous tentative physician services agreement, I had two main concerns. First, that the presence of a defined, limited physician services budget with clawbacks on physicians as a whole put too much responsibility on individuals physicians for results beyond their power to affect. Worse, it set up an economic situation that would further encourage physicians to bill and practice in ways that advantaged them at the expense of patients, taxpayers, and their own colleagues. Second, it did not provide any protections for physicians against decision-making of the Ontario government, which is vitally important as physicians have minimal protections from current labour laws and no practical ability to strike in the event of a protracted labour dispute.

This agreement, the Binding Arbitration Framework, effectively has the Ontario government yielding the second point entirely. If this framework is agreed to, we have binding arbitration now and, by the looks of it, moving forward to future negotiations. It appears to be a reasonable solid arrangement, without much in the way of loopholes. The worst criticism I've read is that the government maintains policy-making rights, including which services should be funded which... of course they do. I doubt we could remove that in an agreement regardless of how it was written, thanks to the current Canadian constitution and the Canada Health Act. This was always going to be a loophole, but it'd be extremely inconvenient - politically and legally - for the government to take advantage of this under the proposed framework.

More interestingly, the framework touches on the first point of concern I had as well. It doesn't remove a set physician services budget, which I still find worrisome, but neuters its effectiveness considerably. First, it makes the budget subject to binding arbitration. Second, it explicitly dictates factors that an arbiter should consider, most importantly demographic changes. It does include consideration for the economic situation in Ontario, which is troublesome from an administration that has used any hint of economic weakness to justify public service cuts, but could end up being beneficial as well, as economic strength should work in our favour. Third, and most importantly, breaches in the cap on the physician services budget do not automatically result in punitive action against physicians. Rather, breaches will be subject to arbitration, where it can be determined to what extent, if any, physicians are responsible. Additionally, any punitive actions are subject to negotiation and arbitration, meaning more targeted cuts are possible, in contrast to the current approach of across-the-board cuts. I'd rather see the physician services budget cap removed entirely, but this is a reasonably well-tailored way to maintain its existence while blunting the more objectionable aspects to such a cap.

Some other objections to this agreement have been raised, however. First is that it generally precludes negotiation on the unilateral changes enacted since 2014. Yet, I haven't seen any indication this would be on the table - even the most favourable labour decisions wouldn't include significant back-payments and the OMA has little bargaining power for this time (due in no small part to its own actions). There's what's ideal and there's what's realistic; getting any meaningful compensation for the last 3 years during (failed) negotiations seems like a pipe dream more than anything. Second, the agreement limits ability to engage in strikes or job actions that directly affect patient care... which we couldn't (and shouldn't) do anyway. Job actions that negatively affect patient care in a meaningful way, including strikes, contravene recommended ethical guidelines, including those listed by the CPSO, our regulatory body. Furthermore, physician strikes don't work. The history of them in Canada is full of physicians losing badly. While recognizing a significant desire among Ontario physicians to engage in job actions, if this agreement prevents that, it'll only be saving us from ourselves.

I think I've painted a fairly clear picture of where I stand on this framework. It's not perfect, far from it. Yet, it gives Ontario physicians clear, long-term wins that looked unlikely even a few months ago. The compromises for those wins, while not insignificant, are tolerable. Critically, these benefits come before money is even on the table - I honestly figured that to gain some of these concessions moving forward, physicians would have to endure more financial pain in the short-term. To have them arranged in advance of the negotiations for the current physician services agreement would be a welcome surprise. I rejected the previous tPSA, loudly and proudly. I just voted in favour of this framework.

Wednesday, 18 January 2017

Obamacare and the Prospect of Repeal

As a Canadian medical student, it's worth keeping an eye on other healthcare systems. They can be a source of inspiration when advocating for changes to our own system, as well as a warning when other changes are being considered here that have not worked well elsewhere.

The US has a reasonably effective, but highly unequal and terribly inefficient healthcare system. This has been true for decades. It was true before Obamacare came in, it's been true since Obamacare was implemented, and it seems likely to be true in the near future as well. However, changes are on the horizon. US Republicans have made opposition to Obamacare to be a defining element of their party policy and they now have near-full authority to alter the American healthcare landscape as they collectively see fit.

In that light, I've become intrigued by this AskReddit thread I recently stumbled across, "US residents of Reddit: How will the repeal of the Affordable Care Act affect you...positively or negatively?". While hardly scientific, this kind of personal feedback offers a fairly unique glimpse into the effects of healthcare policy on individuals in real-time.

The answers seem to be falling into two categories, revealing both the strengths and weaknesses of Obamacare. On the beneficial side, there are numerous stories of individuals with significant, long-term medical conditions who would be medically or financially crippled without Obamacare. These individuals would be denied care or be forced to pay insurance costs effectively on par with their extensive medical costs if Obamacare didn't exist. In extreme cases, they would face bankruptcy or premature death due to these financial burdens.

On the other side is people, often younger individuals or families, facing insurance bills that eat up a substantial portion of their income and provide little benefit given their current health. The figures here are obscene in some cases, hundreds of dollars per month with high deductibles. Together, these conditions mean the insurance is useless for the day-to-day medical expenses of generally healthy individuals, functioning only as expensive disaster insurance in the case of an accident or unexpected illness.

These two outcomes are directly related. Obamacare's primary goal was to increase coverage, particularly for individuals with pre-existing conditions. These people have medical costs beyond what they can afford in insurance payments. To afford to cover them, premiums had to rise on other, healthier individuals. That's not very helpful for those healthier people, especially those without much money, as the cost of insurance goes up for them without any gain in benefits. So, some will choose to drop their health insurance, rightfully determining that the cost of maintaining their insurance is not worth their premiums. To prevent this logic from causing a spiral of higher costs and more people dropping out of the market, Obamacare mandated that everyone buy insurance. That means younger, healthier people buying health insurance more expensive than their needs to cover the costs of sicker, typically older individuals who pay less than their needs.

Virtually every country's healthcare system involves some form of these transfers, from the young and healthy to the older and sicker. It's effectively a requirement for a functioning healthcare system. The anecdotes from the AskReddit thread are therefore entirely unsurprising in general features - to move towards universal care, the proportion of medical costs covered by healthy individuals will almost certainly increase, in order to lower costs on sicker individuals.

However, there is a problem in scale and precision unique to the US. American healthcare is terribly inefficient, costing far more than comparable countries. The premiums quoted for healthy individuals in the AskReddit thread are beyond what they could be or would be in other systems. Likewise, the burden of covering the costs is falling disproportionally on lower-income individuals relative to other countries. To put this in perspective, under the American system, I would have to buy the same insurance this year that I'm in school, with no income, as I would next year with a solidly middle-class salary, and then again when I'm a practicing physician in a few years with an upper class income. By contrast, in Canada, my healthcare costs would vary significantly with my income via taxes. Obamacare put some subsidies in to help balance things out, but these have proved to be inadequate in many cases.

Obamacare was deeply flawed and the AskReddit thread exemplifies who those flaws have affected unfairly. Republicans are eager to fix this problem of overly high costs for relatively healthy individuals and it seems clear that this can be accomplished. Whether they will pursue policies that satisfy that goal, without causing additional, more serious problems as a result, remains to be seen.

Friday, 9 September 2016

How Could We Reform Physician Compensation?

With the rejection of the tPSA by OMA members, we're back to square one in the negotiation with the Ontario government regarding physician compensation. It will likely be a long, ugly battle, just as it was before the tPSA came to light. Still, it's a good opportunity to delve into the discussion about what a reasonable adjustment to physician compensation could look like. After all, while I was strongly opposed to the tPSA, I'm not at all opposed to sensible efforts to save money on physician compensation, and I believe there are ways to do it which are in the best interests of patients and taxpayers which wouldn't unduly burden physicians themselves.

1) Implement Relativity

This came up frequently in the "Yes" side of the tPSA arguments and, despite being a "No" voter, I support this approach. Relativity in the context of physician payment negotiations in Ontario essentially means that physicians should be paid comparable amounts for the work that they do. More specifically, the argument is that if cuts are to occur, they should fall on those who are paid more than their counterparts with similar training and working conditions.

There are some specialties where physicians clearly earn more than is justified, in relative terms at least. I could name several specialties or subspecialties here, but I'll single out Ophthalmology, as it's one of the more commonly cited examples. Ophthalmologists have extremely high gross billings on average. Their high overhead reduces this number considerably, but their take-home income is still quite high. Their patients are generally low acuity with threats to morbidity rather than mortality, and their working hours are on the lower end, particularly for a surgical specialty. Importantly, there is no shortage of individuals wanting to work in Ophthalmology in Ontario - both job openings and residency positions are extremely competitive - so we could cut Ophthalmology compensation without significantly affecting our ability to fill available positions. Ophthalmology does involve complex, technical work with an important impact on quality of life for their patients and their pay should reflect that expertise. However, with outsized incomes, the overall Ophthalmology compensation could be cut without significant ill-effects on patient care, as could the income for a number of other specialties or subspecialties.

One point to mention here - in no way would I argue for equal pay for all practitioners, or even equal hourly pay. I'm going into Family Medicine (if I have any say) and expect to earn on the lower end of things for a physician. That's fair to me - I'm actively choosing to avoid long training times and long work hours with high acuity patients (outside of some ER shifts potentially). Some inequality in pay is justified.

2) Eliminate Cash Cows

Overpay between specialties is worth addressing, but so is overpay within specialties. Each specialty tends to have its own cash cows, procedures or practice arrangements that overpay relative to other activities. These cash cows skew practice patterns. The economics of healthcare do not work the same way they do for other products, but many of the same principles apply. In this case, if you pay doctors more for doing something, they'll do that something more often, even if that something isn't the best use of time or resources. It also produces some inequality within specialties, as practitioners who establish themselves with a certain high-pay procedure, or control the infrastructure necessary to perform it, end up with higher incomes than those without the privilege. Eliminating cash cows by pricing high-pay activities more modestly could provide some savings with a neutral or possibly even beneficial effect on patient health.

There are a few caveats here, however. In some cases, cash cows are actually a case of good incentives. Some activities are inherently undesirable for practitioners, yet necessary for patients. High compensation in these situations may be required to provide patient access. Another consideration is that cash cows can compensate for low-paying or even net negative activities. Our billing system is bad enough that some activities actually cost practitioners money, or pay so little that they aren't economically feasible on their own. Physicians often still provide these procedures or services because they're valuable for their patients, while making up for the lost time and money through higher-yield services. Taking away those overpaid activities means fewer practitioners can realistically afford to provide underpaid activities.

Bottom line is that adjustments to billing codes have a significant potential to save money on healthcare costs, but require careful consideration. There are a lot of secondary effects to changing billing codes that aren't immediately apparent to an outsider.

3) Strengthen Primary Care

Reducing physician costs in the provincial budget really can only be done one of two ways - reduce the cost per service provided by physicians, or reduce the number of services provided. The first two items discussed deal with the former. The rest of this piece will focus on the latter.

Primary care is the gateway to healthcare services, whether through Family Physicians or through Emergency Rooms. We rely on these practitioners to restrict access to specialists only to those who truly need them. The more that primary care providers can manage without calling on specialists, the less specialty services will be used. By strengthening the capabilities of Family Physicians, we might be able to save money on more expensive specialists, reducing how often patients require their care.

This is obviously easier said than done and in many cases, not a good idea. Primary care physicians don't see some conditions often enough to develop sufficient expertise and they shouldn't hesitate to get experts involved in these situations. Likewise, other expensive elements to healthcare, such as some diagnostics or interventions, are restricted to specialists to release, limiting their use and arguably saving the overall system a fair bit of money. Opening these resources up to Family Physicians would not be beneficial to patients or taxpayers.

Nevertheless, with reasonably wide variation in referral practices between Family Physicians, especially between locations, we do have room to cut back on referrals without sacrificing quality of care or adding costs to other parts of medicine. Emphasizing (and ideally compensating) continuing medical education, while strengthening the training for primary care providers (particularly in medical school!) would be my favoured initial methods. Incentive schemes to keep referrals to a minimum may be worth considering, if they can be carefully constructed not to punish necessary referrals.

4) Train Fewer Doctors

Ok, this is a can of worms, but let's open it just a little bit. Canada tried this approach and it largely didn't work. Back in the 90's, Canada's medical schools didn't keep up with population growth and in some cases, medical school enrollment was actively curtailed. Of course, physician services were still necessary, and ultimately, healthcare costs continued to rise unimpeded. Physicians, now scarce, saw their salaries rise, while non-physician providers like Nurse Practitioners and Physician Assistants slowly grew in number. Eventually, the deficit in physicians was too high and Canada reversed course, massively expanding medical school enrollment while opening up new residency opportunities for foreign-trained physicians. The idea that restricting physician numbers could save money was fairly clearly wrong.

Well, sort of. The problem in the 90's was that Canada restricted physician numbers so much that baseline demand for services couldn't be met. That led to compensatory spending to make up that deficit. Now, however, physician numbers are much closer to the OCED average and still on the rise. We're starting to see some first signs of physician oversupply, with unemployment and underemployment on the rise, but this has been largely concentrated in resource intensive, hospital-based specialties where jobs can be restricted by constraining the availability of those supporting resources. This speaks to an inherent problem in funding physicians to meet patient demand - demand for healthcare services often expand with supply and physicians can create their own demand. Specialties that can finance their own infrastructure tend to have great job markets regardless of the local supply of physicians.

The example I always mention is Psychiatrists in Toronto. They are in far greater supply there than in smaller cities and rural areas, yet do not seem to have trouble getting work and make salaries comparable with national averages. They just take care of far fewer patients than their counterparts elsewhere, seeing them more often or for longer. This isn't necessarily a bad thing, but it clearly a more expensive arrangement for the province. This is hardly a unique situation. Most specialties, including Family Practitioners, could easily double (or more) the number of people in their clinics or procedure rooms if they had the time and resources to do so. As we start to see the number of physicians rise, and especially as we see more students going into fields like Family Medicine, Psychiatry, and Geriatrics, we could see physician billings substantially increase, as these new practitioners will not be nearly as constrained by public resources, and can cause - rather than simply respond to - increasing demand for healthcare. I do think we need to shift more physicians towards these specialties, which are likely still undersupplied relative to other specialties, but we are likely overshooting on overall physician supply. Cutting back on this oversupply would likely help the bottom line in the long run.

Wednesday, 24 August 2016

The Debate Under the Debate

One interesting aspect to come out of the Ontario tPSA debate is suggestions coming out from all corners about how to reform our healthcare system to address current funding shortfalls. Some ideas are new and innovative. Others have been around for some time and return to the surface when the debate about funding returns to prominence. This post is about the latter.

1) User fees paid to the government

One often-discussed option for easing the burden of healthcare on the public pocket is the idea of user fees. This represents money paid by patients for each interaction with the healthcare system, typically a small amount like $10. The idea is partially to help cover healthcare costs, but mostly to discourage patients from visiting physicians and thereby reducing demand for healthcare. Anyone truly sick will often happily fork over $10 for the service, but the hope is that unnecessary visits will be less frequent.

The problem here is that patients often don't know the difference between a necessary visit and an unnecessary one. We go through a lot of training to figure out that difference and it can still be a subjective call. Missed visits that are necessary can not only reduce health outcomes, but can increase costs in the long run as cheaper preventative or maintenance care is replaced with expensive interventions on illnesses that arise or progress that might not have otherwise. Reducing low-cost interactions but increasing high-cost interactions with the healthcare system are not a great way to save money.

User fees also represent a tax on the sick and disproportionally hurts the poor. So, for both moral and practical reasons, Canada has generally fought to avoid user fees paid to the government.

2) User fees paid to physicians

A bit of a twist on #1, with fees going to physicians themselves rather than the government. The notion here is that if physician incomes must fall, we should be able to supplement incomes with user fees. This idea has many of the same drawbacks as with user fees paid to the government, but with some caveats. Importantly, individual physicians presumably understand their patients' situations better than the government and can adjust the user fees accordingly. That might comprise forgiving fees for poorer or sicker patients, or perhaps increasing them for those clearly abusing the system. User fees might be thus better tailored to their original intent of reducing unnecessary visits.

The added downside to this set-up is that it creates a conflict of interest for physicians. Physicians would benefit financially regardless of whether the user fee they charge is helpful. User fees could be used as a means to punish patients who have real medical needs but are perceived as troublesome. The fees could even be used as a way to try to push patients out of a physician's practice by making it financially burdensome for the patient to see that physician.

3) Parallel private care

I could (and probably should) do an entire post about introducing privately-funded healthcare into Ontario. The argument for private care in the current setting is that it activates unemployed or underemployed physicians to ease the financial burden on public institutions while shortening wait times for important but non-emergent conditions such as hip replacements. I have no doubt that private healthcare would be good for physician employment prospects. By providing a mechanism for increased funding into healthcare, I also don't doubt that wait times would come down on average, though some people would benefit more than others.

However, I don't think there would be any relief on the public coffers. Fully-public healthcare is remarkably efficient, saving significantly relative to private care by minimizing administrative costs, advertising costs, and profits. Countries that allow significant private care (as opposed to private insurance that, with subsidies and regulations, functions very similarly to publicly-funded care) tend to have equally high public funding of healthcare. The US government spends about as much on healthcare as Canada's government. Private care tends to increase overall spending on heatlhcare without decreasing public spending or making significant gains in health outcome measures, albeit potentially better on some process measures. Notably, a lot of that extra money goes directly to physicians. Private clinics compete for the best physicians and are able to afford them at higher prices, which in turn puts pressure to increase physician salaries in the public sector. It's a win for us, but likely not a win for the economic or physical health of Ontario or Canada.

4) Salaried physicians

The government's objection to the current set-up is that they really don't have any final control on physician supply or utilization. They have some power, but ultimately any qualified physician can set up practice in Ontario and work as hard as they want, billing as much as they want, provided they can demonstrate their billings are legitimate. This leaves a lot of uncertainty for the government's finances with only some indirect methods - such as reduced residency positions - to influence physician billings. One solution to this dilemma is to simply replace the current model of physicians as independent consultants, to one where all physicians are employees, purposefully hired and placed on a salary. This eliminates any uncertainty in physician numbers or cost per physician, while providing guarantees for physicians themselves in terms of compensation.

I'm sympathetic to this approach, especially with the side-benefits that would undoubtedly be included for physicians. Working conditions would come under higher scrutiny, while pensions and benefits would be on the table, something not really possible in the current set-up.

There are some major downsides, however. Salaried physicians are not nearly as productive as non-salaried employees. There's far less of an incentive to push patients through when you get paid the same for seeing 15 patients a day as you do seeing 25. Making up that productivity gap would be extremely expensive, though some patients would undoubtedly benefit from the increased attention. Salaried physicians would also be difficult to administer. Physicians do a lot of ground-level management of their practices, whether it's paying for staff or arranging clinic space or ordering supplies. That's difficult to fit in a salaried model where physicians are more directly governmental employees and not independent consultants. The Ontario government could (and may want to) take over individual clinics, but that means a lot more administrative work to pay for, essentially taking on what physicians normally pay in overhead costs. This could be worked around depending on how a salary model is set up, but it wouldn't likely be a simple process.

Conclusion

Funding physicians is hard. The economics of healthcare are complex and don't behave like other services or commodities. Any arrangement will come with some trade-offs. When faced with the dual priorities of optimal care for minimal dollar, balancing these trade-offs is no small feat and there are no simple answers. Unfortunately, simple answers always rise to the surface when the current system shows signs of weakness, as any healthcare funding system inevitably will. These approaches could have a place as one component to a broader plan, but likely aren't feasible on their own. A careful consideration of all policy options is necessary, understanding that no individual approach will likely be sufficient in isolation.

Sunday, 7 August 2016

Ontario tPSA - Reasons to Vote "Yes", Reasons to Vote "No".

I've been fairly upfront about my opposition to the tPSA as it currently stands. I believe it's a step in the wrong direction. That said, I've had a chance to read and listen to those who support the deal, and there are some very valid arguments to discuss from the "Yes" side. I thought I'd take a minute to lay out the points that got me closest to considering a "Yes" vote, as well as a bit more that I've posted previously about why I'm still not convinced.

The Best Argument I've Heard For The tPSA

There are good reasons to vote "Yes" to this agreement. The most convincing arguments I've heard have tended to follow this general train of thought:

1) The tPSA provides the most funding we can reasonably expect to get for physician services over the next 2 years and almost certainly provides the most funding overall for the next 4 years.
2) The tPSA provides us the best chance at binding arbitration for the next deal in 4 years' time, while providing meaningful third-party oversight for the remaining negotiations in the interim.
3) The tPSA provides incentives for the OMA (not individual physicians) to push for policies that hold down physician costs.
4) The tPSA provides some protections for physicians or physicians-in-training on what I would consider relevant but comparatively small side-issues.

I accept and agree with all these arguments. Combined, they make a reasonable case for voting "Yes" to the tPSA, particularly in the short-term. If I was nearing retirement, I'll be honest in saying I'd have a hard time voting this deal down. I think the OMA has made some major missteps in negotiating with the Ontario government and promoting its position to the public, but I'm doubtful that these would be ameliorated in any significant way by rejecting the tPSA or changing the OMA/negotiating team. For the next 4 years, as bad as this deal is, we're not going to get much better by rejecting it.

Why I Will Still Vote "No"

Put simply, I will vote to reject the tPSA because our healthcare financing problems - as well as my financial stake in the healthcare system - last longer than 4 years. And at the end of those 4 years, we could be in a very, very bad situation. We've never had a total cap on physician services before. Consenting to it now, even under some duress, makes it that much harder to eliminate later. Eventually it may be recognized as bad policy, but that could be a long way down the road, decades from now.

I've described before why a collective hard cap is such a bad idea, but let's put some numbers to the theory. To keep up with demand for physician services, we'd need to increase a collective cap on billings by 3-3.5% each year. To keep up with inflation as it currently stands, we'd need an additional 1.5-2% per year increase on the cap, plus an equal increase in fees for each billing code. In sum, we'd need a 5% increase in the cap, plus 1.5-2% increase in actual income just to break even in real terms.

It's incredibly easy for the Ontario government to spin this to the public as a 5% increase in salary, even if it's simply a cost-of-living adjustment plus funds for enough new physicians to keep up with population growth and aging. It's much harder for the OMA to explain why it's not a 5% raise. Binding arbitration would help, as any arbiter should understand these points well enough. However, the government's trump card - that they can't afford such a high increase for budgetary reasons - is hard to refute. There will always be a need to save money on healthcare, that's the nature of healthcare economics. Even a reasonable, informed arbiter may not believe a 5% increase in the cap is justified.

The problem is that we'd need a 5% raise in the cap each and every year, with each and every deal. If inflation or healthcare usage increases, that number has to go up. Anytime we fall below that, billing compensation drops and incomes drop even further as overhead costs are largely static and can't be easily reduced.

Rejecting the tPSA means short term losses for physicians, likely sizable ones. I'm comfortable with that. Ontario physicians on the whole make a bit more than we should when looking at national and international comparisons. The Ontario government does have a short-term budget problem that needs to be addressed and physician compensation is part of that. Ontario physicians can afford a hit to their incomes, even a reasonably substantial one, so long as we get back on a sustainable path. Accepting the tPSA means sparing ourselves short term pain but at the risk of larger long term losses, possibly permanently holding down physician salaries long past the point of reason or good economics.

Conclusion

When we talk about a sustainable healthcare system, we need incentives for all parties to push for sustainable funding mechanisms and to behave in taxpayers' and patients' best interests. The tPSA provides incentives for the Ontario government and individual physicians to do the opposite, giving only the OMA a loose incentive to push for reasonable policies. Bleeding physicians slowly is not sustainable, does not encourage physicians to behave responsibly with government funds, and in the long run will likely hamper access to important services for Ontario patients.

The cap will likely overwhelm future discussions between the OMA and Ontario government, distracting from more important matters. The "Yes" side has presented the tPSA as a pathway to more evidence-based policies in future negotiations. I don't believe that we're likely to get well-crafted policies by accepting such a poorly-constructed one as the backbone of this agreement.

One big flaw in my argument I must acknowledge - we could reject the tPSA and still end up with a global physician budget. If we don't win the right to binding arbitration and fail to rally the public to our cause, or if we win binding arbitration but fail to convince the arbiter, we could still be stuck with a global cap on physician billings. However, if we accept the tPSA, I believe we all but guarantee such a cap will exist for the foreseeable future. I cannot consent to that.

Wednesday, 20 July 2016

OMA Tentative Agreement

Quick(well, it was supposed to be quick) post for those in Ontario about the OMA's tentative agreement with the Ontario government on physician compensation. It came out of nowhere and is causing quite a bit of outrage from more than a few physicians, some directed at the government, some directed at the OMA itself.

The OMA will be holding a non-binding vote on the matter, which apparently I get to participate in, despite being a student. Based on the text of the agreement, my vote will be a definite "no" for two major reasons.

First, and you'll hear a lot about this from practicing physicians, is a lack of binding arbitration. Right now the government seems to be getting away with unilateral action. The OMA has sued to retain their right to binding arbitration, but lawsuits can be difficult to predict. Few labour laws apply to physicians and for a variety of reasons, we have no ability to undertake real labour actions. Some form of binding arbitration is about the only protection we can expect to protect our rights as workers.

Without that assurance, even if this deal proves to be acceptable in the short-term, it leaves open the door for significant long-term reductions in fees over time. Death by 1000 cuts, as it were. If that were to occur, the only recourse would be for physicians to leave the province or profession, neither of which is good for physicians, patients, or the government. I worry about physicians having too much clout over their own income and we certainly shouldn't be given free reign, but some defense against competing interests is necessary. Absent any others, binding arbitration is a deal-breaker.

Second, the Physician Services Budget (PSB) has to be eliminated. Despite all my protestations above, if I had to choose between adding binding arbitration and getting rid of the PSB, I'd choose removing the PSB. The idea behind the PSB is logical at first glance - physician billings make up nearly 10% of the provincial government and are growing at a rate faster than inflation or tax receipts. If they could be effectively capped, the government would save a significant amount of money and put itself in a much better financial situation moving forward.

The problem with the PSB is that while it provides a cap, it does not provide any mechanism for it to be achieved effectively. Expenses for physician billings go down if physicians bill for less procedures or get paid less for each billing. The PSB ostensibly gives physicians an incentive to collectively bill less, in order to avoid any clawbacks. Unfortunately, because the cap applies to physicians collectively, and each individual physician cannot change the actions of physicians as a whole, individual physicians retain the same incentives as before, namely to bill as much as possible. The PSB does nothing to reduce quantity of billing.

So, the PSB instead becomes about reducing cost per billing by using the clawback mechanism. An effort to reduce average cost per billing on its own isn't too objectionable, but the approach taken with clawbacks is where trouble occurs. The clawback hits all services equally, from services that are paid more than they should be, to ones that are undercompensated. Theoretically the new agreement includes a mechanism for the OMA to tailor any clawbacks to economic circumstances, but no details or assurances have been provided on that front. Additionally, because the clawback could be variable, it makes financial planning for physicians very difficult. It's like buying groceries, leaving the grocery store, then finding out that the price you were given in-store changed and you suddenly owe more for some or all of your items. Makes it hard to plan a food budget!

Lastly, the PSB and clawback punishes physicians for factors out of their control. The number of patients go up? Well, then physicians get paid less. Patients require more services, either because they're getting sicker or even getting the extra help they need (think mental health services especially here)? Physicians get paid less. If physicians need an income reduction because we get paid too much, or certain services are too richly compensated, I'm fine with that. It's when it happens indiscriminately and due to factors unrelated to physicians' performance or ability that I object.

Money is not a major factor for me in my career. I'm choosing a relatively low-earning career path and am quite happy with that. I wouldn't object to a reasonable reduction in my future income either, given national and international comparisons. However, the process matters and the impacts of those cuts on patients, and the financial well-being of the system, do matter.

If I do get a say on this, I will be voting "no".

Friday, 22 April 2016

Investment vs Consumption in Healthcare

It's no secret that healthcare budgets these days are under stress. As provincial governments look with concern to current healthcare costs, as well as to potential future healthcare costs with our graying demographics, there's plenty of incentive to minimize spending within the medical community.

While I disagree with some of the approaches being taken, it's a worthy goal - not everything in healthcare is cost-effective, and there's good reason to think that freeing up money for other social expenses could be more be more beneficial than many direct medical expenditures. It's also an opportunity to dig into the role of medicine in our society and what we actually expect for the rather large sums of money we throw at our healthcare system.

Healthcare comprises both elements of investment (expenses which are expected to yield an economic return) and consumption (expenses which are not expected to yield an economic return but which hopefully confer benefits to longevity or quality of life). That's a bit too sharp of a split - in reality, all healthcare should be improving the quantity and quality of life of patients and most expenses tend to have some sort of an economic return, even if that return falls below the medical expense and wouldn't be worthwhile as an investment alone.

However, the degree to which we priority each element - investment vs consumption - tells us a fair bit about how we view healthcare's role. These priorities in turn can dictate how much we spend on healthcare overall.

Right now, in Canada and much of the rest of the western world, the focus is on consumption. When people rally for better access to medical services, it's usually for the immediate impact on patients' quality of life. Likewise, when analyzing an intervention's merit, we tend to focus on cost-effectiveness, using metrics like cost per Quality-Adjusted Life Year (QALY). These capture quality of life gains, and analyze all medical interventions as though they were consumptive in nature, but generally ignore economic returns on expenditures.

It's difficult to view and analyze healthcare from an investment perspective, however. For one, it's really hard to accurately measure return on investment when it comes to medical interventions - QALYs or similar metrics are much easier to determine. Second, doing so requires making judgments on the economic worth of individual people (or groups). Most of us are pretty uncomfortable thinking of human lives in economic terms. Physicians especially try to avoid these judgments, as they can devalue certain groups, especially the vulnerable and those in need of the most help.

Yet, there are some broader programs that make no direct judgment on individual worth, yet would have clear beneficial investment components in addition to their meaningful improvements on quantity and quality of life. Mental health is a major one. It hits adolescents and younger adults frequently and is a significant impairment when it comes to productivity at work or school. Individuals who are not currently working are also at high risk for mental health problems, particularly the elderly with dementia and/or depression, of course, but compared to many other medical services we provide, the benefits from investments in mental health more frequently affect individuals who could be more economically productive if optimally treated. When we start making the hard choices about which programs and services are worth funding, it may be worth putting a higher emphasis on economic returns - it might make our tax dollars go a little bit further and lessen the burden of our healthcare expenditures.

I'm rambling a bit, so I'll wrap up by saying this: we want to be healthy and our medical system is a major part of that. However, healthcare does more than just improve health and other social services play a big role in improving health overall. It's worth taking a step back every once in a while to reevaluate our priorities in medicine, to be critical about what how best we can serve the societies we live in, and how some parts of our healthcare system - even the helpful parts - might not be worth the cost when compared to alternatives.

Friday, 11 March 2016

Mental Health and Resources

I'm on my Psych block now. I'll give my final thoughts on the rotation when I'm further into it, but for now, I'd like to take a minute to express my deep frustration with the availability of medical services in this country and my area in particular. Long story short, we need more beds, more services, and more resources in general.

I have plenty of frustrations about the medical system in general. That's kind of the point of this blog, to get those frustrations out constructively. A lot of my concerns relate to the actions of physicians as a group, as well as physician-dominated organizations including medical schools. I think we, as a profession, hold a lot more blame for the problems in medicine than we're generally willing to admit, and that we have a lot more power to push forth positive change than we currently exercise. I try to advocate for internal changes within the profession of medicine ahead of or in conjunction with changes outside of the medical profession. I especially try to avoid lobbying for more money to be thrown at any particular problem, because resources are finite and while money can paper over a lot of structural problems, they don't solve them either. Those structural issues need to be tackled regardless and doing so may remove the need for more funding.

Not so much when it comes to mental health. The state of mental healthcare in my region is receiving a lot of scrutiny, justifiably, due to a number of frankly horrible situations that have occurred recently or not-so-recently. A lot of these criticisms have been directed towards the government, particularly the provincial government, and I largely agree with those criticisms. Some of the criticism has been directed towards the hospital I'm training at, the physicians and other healthcare workers involved in care for those with mental health, as well as the school I attend. For once, I can't agree with these criticisms. The hospital has clearly tried to prioritize mental health. Employees and physicians have largely just tried to make the best of a bad situation, accepting what is objectively a poor situation because no better alternatives exist. Even the school has gone out of their way to increase the number of Psych residents to help with the workload and encourage more Psychiatrists to stay in the area.

When valuable 1-on-1 therapy is virtually impossible to achieve without insurance or significant independent funding, when long-term care facilities are routinely unavailable, when the mental health ward is operating above 100% capacity (even after accounting for those waiting for long-term care), when wait lists to see an outpatient psychiatrist are on the order of months or years... even monumental changes in the organization of services aren't going to cut it.

Psychiatry does have its share of problems that can be laid at the feet of physicians, of that I have no doubt. However, it's pretty clear that more resources are needed. I don't see mental health being substantially improved without more inpatient beds as well as more supports both in and out of the hospital.

Sunday, 27 September 2015

Confusing Situations in Medical Specialty Job Markets

I've had a busy week, hence the low posting (more on that later, hopefully). However, I wanted to take a minute to discuss a phenomenon in the modern medicine job market that is endlessly baffling me: specialties with physicians who are simultaneously overworked and underemployed.

Underemployment in quite a few specialties has become an increasing concern. Resource-intensive specialties, especially surgical ones, are having the worst time. Neurosurgery, for example, has a notoriously terrible job market - across the country, positions are rare and demand ludicrously high qualifications. Yet in many of these specialties, practitioners are working insanely long hours, often have high stress levels, and in many cases, are rushing through patients at a rate that limits good patient care.

I can't help but see an obvious solution to this set of problems - let the overworked staff take on a little less responsibility, hire the desperate-for-work newly graduated physicians to take on some of the slack. Lower levels of stress for everyone involved, right?!

Yet, there are a lot of reasons this doesn't happen. Money is a big factor. Increasing the number of physicians working fully means higher health care costs. The physician's compensation is one part, but the support for these physicians is even more expensive, especially when it comes to surgeons. One figure I was quoted was that a new Cardiothoracic surgeon costs the health care system $1.5 million a year, but only a third of that or so goes to the surgeon. The rest goes to the administrative staff, the nursing staff, the research staff, and additional services needed to support that surgeon's work.

Now, if these new physicians were simply taking over work formerly done by other physicians, a good portion of the cost increase for additional physicians would be defrayed by lower costs of those other doctors. However, physicians generally don't want to work shorter hours - after all, many have the option, but choose not to. Each physician has their own reasons for the long hours, but here are the ones I've seen more-or-less directly.

1) Habit. Residents work insane hours. Consultant physicians are therefor used to these insane hours. They may be working 65 hours a week, but they're used to 80, so they think they're cutting back, even if they're still working rather long hours.

2) Career advancement. Physicians are, almost by definition, ambitious people. Being a fully trained doctor with employment is only the start in many fields. To move up means making an impact, which means more responsibilities, both in clinical and non-clinical settings. Those all take time. Even when the extra work is non-clinical, that crowds out the clinical time, leading to some jam-packed clinics.

3) Money. It matters for governments, but also for physicians. For everyone really, but there are some important financial factors specifically for physicians. Our careers start later with a substantial amount of debt. Most physicians are not salaried - they have to cover their own pensions, benefits, insurance, etc. Many also cover their own supplies, infrastructure, or support staff, meaning it costs a fair bit of money just to have the practice open.

Physicians are also notoriously poor money-managers. A good number of physicians don't save enough for their retirement and of those who do save enough, many invest poorly. High incomes but low wealth characterize many physicians' financial situations - they make a lot, but habitually spend most of it.

There's undoubtedly some greed in there as well. That's not meant to be a judgement, we all are greedy in our own ways and physicians are no different. Conspicuous consumption - buying expensive things to show you can buy expensive things - also seems to be a hallmark of a high-earning physicians.

Whatever the motivation, working less for physicians means earning less, so there's a strong incentive to keep hours up.

4) Preference. One thing I've noticed in my first bit of clerkship is that while long hours can be incredibly draining, I feel that drain mostly at home, not at the hospital.

At the hospital, while there's a lot of work, much of it quite stressful, there's also a lot of gratifying aspects. The work is intellectually stimulating, socially engaging, and can have some clear positive effects on others. There's a fair bit of direct, positive feedback - rarely from other physicians, mind you (unless you're a student!) - but certainly from patients. I said before that physicians are, almost to a person, ambitious. Most are also egotists. Getting thanked, being told that you're smart, or that you're a wonderful person simply for doing your job is like a drug. It's a drug that's very hard to give up.

At home, with rarely enough time to pull your share around the house, to be social with friends or family, to spend time with your partner, there's almost a recipe for failure. I'm incredibly lucky to have a supportive SO - which seems to be the major factor in keeping life together as a physician with long hours - but if there's a place I feel like I'm not quite living up to expectations, it's at home. This doesn't seem to be an uncommon sentiment among physicians and it creates a destructive cycle prone to workaholism. Physicians feel good at work, bad at home, so they spend more time at work. That makes work feel better and home feel worse, which in turn pushes these physicians to spend more time at work and so on.

The sum of all these effects is older physicians working longer than they need to, longer sometimes than they want to, all while younger physicians struggle to find enough work. While efforts are certainly underway to change this dynamic - financial counselling, more flexible work schedules, encouraging physicians to spend more time with their family - all of this does help to get physicians to take more time for themselves and in specialties with job shortages, frees up some work for new graduates. However, where these initiatives seem to be having the greatest effect are on those younger physicians looking for work, as well as on physicians in specialties that don't have shortages.

Point of this ramble (and wow, much longer post than I was expecting) is that while it's laudable to go the extra mile for your patients, long hours don't always equate to doing the best for your patients - or yourself.

Friday, 26 June 2015

Concierge Services

There's a growing movement, more in the US than in Canada, but it's popping up here too, where physicians - typically family physicians - charge individual patients directly for their primary care needs. In exchange for a yearly or monthly charge, patients get a litany of services, often including basic lab work, imaging, and physiotherapy, in addition to their routine doctor visits. Insurance is essentially taken out of the equation.

In Canada, this is technically against the rules in many respects - physicians are forbidden from charging patients for their care while also charging the government for that care. It happens, but either the providers just hope the regulatory colleges don't get wind of their actions, or they add on extra services and claim the charge is for those extras (it's legal to charge patients for services not covered by government insurance).

In the US, however, it's perfectly legal and becoming increasingly popular. Family doctors love the set-up, as they get to spend more time with each patient. In addition, they have no real incentive to order more services for a patient (actually a bit of a disincentive to do so) and overall, their incomes seem to go up. Many patients like these set-ups as well, as they get more dedicated time with their physicians.

The downside is cost. Seeing fewer patients for more time each while earning more money means that without significant increases in efficiency, concierge medicine is more expensive than traditional fee-for-service or insurance-based care. In the US, there is one major source of increased efficiency - lower administrative costs. Most estimates I've seen put direct administrative costs in the US at about 15%, well above that of other countries. The mishmash of various insurers and their interest in reducing payouts accounts for a good portion of that. By eliminating all or most interactions with insurers, US physicians can save themselves a lot of time and money, which can go towards more time with patients, lower costs for patients, or higher incomes for themselves.

These cost savings appear to be significant. While initial concierge services were incredibly expensive, available only for the super-rich, emerging clinics seem to be charging in the range of $1500 a year per patient. This is still beyond the ability of low-income individuals to pay, but after regular insurance payments (now mandatory due to Obamacare statutes), middle-class patients may be able to manage these costs. If insurance companies discount their prices to account for concierge services, as some are, they may become more affordable. At the moment, however, the majority of patients in concierge practices are quite wealthy.

$1500 is also a substantial sum for primary care alone. Per person healthcare costs in the US and Canada are well above that, but are concentrated on a set of high-intensity patients who receive a substantial amount of non-primary care. If the US were to spend $1500 on every person's primary care - or even only $1000 - their overall healthcare costs would skyrocket above their already immense values.

In Canada, the situation is more clear-cut. With universal health insurance, concierge services are by definition an extra, with extra costs associated. There's virtually no room for increases in administrative efficiency, as one major benefit to universal health care is low administrative costs. That may seem surprising, given the stilted bureaucracy of Canada's health care system, but most of the money ultimately does go to providing care.

It will be interesting to see how the situation evolves as time goes on, especially in Canada. Currently colleges are cracking down on known instances of concierge medicine, but sporadically and often only in response to complaints. More importantly, while our public universal healthcare is well ingrained in the culture of the public, many physicians would prefer being able to work outside the system, or believe that private care has a role alongside public care. There may be pressure to allow for parallel private medicine in Canada, including concierge services.

Sunday, 15 March 2015

Paying For Healthcare (Part II)

A while back, I talked about the importance of trying to keep healthcare costs as low as reasonably achievable. Today, I'll try to delve into the much more difficult discussion of how that might be accomplished.

Reducing healthcare costs, or at least slowing the rise in those costs, is a tricky proposition. These costs have grown faster than inflation for a reason - controlling healthcare expenditures is hard! Certainly, if I had an amazing insight into how to accomplish the goal of reduced healthcare costs, I wouldn't be posting it on a blog - I'd be selling my consulting services to the government or insurance companies or hospitals.

However, when you break it down, there's really only two ways to keep costs down. Healthcare can be broken down as a set of discrete services, so we either need to reduce the total quantity of services provided, or we need to reduce the cost per service provided.

Reducing Quantity of Services Provided

The principle here is fairly simple. If you want to spend less money on say, antibiotics, simply prescribe fewer antibiotics! Or better yet, prescribe no antibiotics, ever. Of course, antibiotics have value - many infections are untreatable without them. Not using them at all would be a very bad idea.

Yet, there are some situations where antibiotics are unnecessary or even harmful. Simple, acute upper respiratory infections, for example, are often viral, for which antibiotics do nothing. Prescribing antibiotics for such an infection is very likely to be useless (there's also some evidence that even with bacterial infections, they're not very useful either, though this is a somewhat more complex area of discussion). Eliminating prescriptions of these antibiotics would save money without significantly impacting health. The Choosing Wisely Campaign attempts to address this sort of waste. Both American and Canadian campaigns have sprung up to encourage physicians - and to a lesser extent, patients - to be more discriminating about what medical services they actually necessary and when to employ them.

There are also some more limited instances where a service in healthcare is entirely unnecessary. Apparently we used to intentionally put people into comas by overdosing them with insulin. That wasn't a great idea... Eliminating useless or harmful services also saves money. While it's much harder to tell that a particular diagnostic test or therapy is completely useless for all people in all circumstances, there are almost certainly some of these services being employed in mainstream medicine today. Identifying these practices is worthwhile from both a medical and economic perspective.

Reducing Cost of Services Provided

While lowering the quantity of services is relatively straight-forward as a concept, reducing costs of each service is not.

What goes into the cost of a healthcare test or treatment is complex, but again, we can break it down into components - personnel, supplies, and durable equipment (including facilities).

1) Personnel

Healthcare takes people to work. Lots of people! People-heavy fields are expensive though and healthcare is no exception. That almost everyone in medicine is generally well-educated adds to this cost. There are a wide variety of professions within the healthcare system, but I'll focus on physicians. Physicians are moderately numerous and get paid a lot. Even small percentage reductions in numbers or salaries of physicians can meaningfully reduce healthcare costs.

Having fewer physicians is a non-starter. Each physician and there are already widespread concerns about overworking these providers. Physicians will increase significantly in number as medical schools have increased their class sizes to keep up with demand. That leaves salaries. Again, physicians receive good compensation. After overhead but before tax, most physicians make $200,000 a year. A few make a bit less, many make significantly more.

Reducing physician salaries, while hardly a sustainable strategy for minimizing healthcare costs, can be part of the solution. That's what the Ontario government did.

2) Supplies

There are numerous supplies that get used once in medicine and then tossed or consumed. There are often good reasons for this - we're not going to reuse needles, for example - but nevertheless, there are certainly instances where we overuse supplies or use supplies that are more expensive than necessary. For example, I did an observership in a clinical teaching unit (CTU) which naturally had many learners. We saw nearly a dozen patients, all of whom had some sort of communicable disease, but most without anything really all that serious or likely to spread. Yet, all 5 or so of us doing rounds put on full protective equipment, even though half of us just stood in the back, well away from the patients. Meanwhile family members pop in and out of the room with no equipment whatsoever. Me wearing that equipment is hospital policy and that policy exists for a good reason - but in this case I probably cost the hospital at least $20 without much in the way of appreciable benefit to them, the patients, or myself.

The major consumable supply in healthcare, however, is pharmaceuticals. We spend a lot on drugs. And a fair bit of that expense is likely unnecessary. Too many name-brand drugs ordered when generics would be just as effective. Physicians may prescribe too many antidepressants, too many anti-psychotics, too many antibiotics, too many ADHD medications, and too many statins, just to name a few. Fewer drugs prescribed means lower healthcare costs.

If you can still treat people effectively without using as many medications or other supplies or other supplies, costs are likely to fall. This is sometimes a bit hard to untangle from the above discussion of simply reducing services and there is a bit of crossover here, but conceptually, the goal is to provide the same treatment (i.e. addressing the patient's medical concerns), only with cheaper, fewer, or no drugs.

3) Durable Equipment

Last, but not least, we have the stuff that sticks around for a while. This can range from something as simple as my stethoscope, to the giant hospitals where I will do most of my training. There's not a lot of low-hanging fruit here to pick on, because generally durable equipment are fairly big, one-time expenditures that last for at least a decade, so a fair bit of thought gets put into them. I spent a ton of time researching what stethoscope was right for my needs. Likewise, no one builds a hospital on a whim.

If there are costs to be saved here, it's likely in the utilization of these durable goods and finding ways to encourage practices that draw on lower-cost durable goods, which often goes hand-in-hand with reductions in personnel or supplies costs. For example, outpatient clinics generally require less infrastructure than inpatient wards. Doing more outpatient work saves on infrastructure costs (though it saves a lot more in day-to-day costs). An MRI machine costs well over a million dollars, while ultrasound machines cost less than $100k - if we can get away with doing more ultrasounds and fewer MRIs, then we probably save on equipment costs by buying fewer MRI machines.

Closing Thoughts

I've given a fairly simplistic rundown of some of the major ways in which healthcare spending could be reduced. While I hope I've given a good sense of what could be done, I haven't really touched on the how these approaches could be accomplished. That requires a much more complex discussion of government policy, economics, legal considerations, and human psychology.

I also haven't discussed the drawbacks to slashing spending - and there certainly are some. Whenever you try to cut away the bad expenditures, you risk cutting away some good ones as well. It's undeniably true that antibiotics are over-prescribed in some cases, but people often need antibiotics! There's always a chance that in stopping needless use of antibiotics, some people will not get them when it is appropriate.

Similarly, every action has unintended, sometimes undesired secondary effects. Cutting physician salaries saves money, but if it reduces access to the healthcare system, it may result in sicker people who require more expensive services down the road. There are certainly methods to anticipate and minimize unwanted secondary effects, but no change is made in a vacuum and it can be hard to predict what the full effect of a policy change will be in 10+ years.

It's a complicated discussion and while I've tried to simply the problem, there are no simple answers. Yet, as a physician-in-training, I believe it's an important one - for the sake of my profession, my future patients, and for my community as a whole.

Wednesday, 4 March 2015

Paying For Healthcare

Recently, a number of events have happened - both in the wider medical community and in my little bubble - that involve how we pay for healthcare. It's a difficult subject, one that elicits a wide variety of opinions and positions.

The Fundamental Dilemma

Much of the controversy on healthcare funding comes from a relatively simple conflict, derived from two widely-held positions.

1) We value health and life more than money when it comes to us or our loved ones.

2) We value money (in the form of lower taxes and insurance premiums) more than health and life when it comes to the vast majority of people we haven't met.

Basically we care about individual health more than collective health, but since the collective is made up of individuals, we get stuck between a rock and a hard place - constant pressure to reduce total costs along with constant pressure to increase individual costs. Since many decisions on spending get made on the patient- or individual-level where the pressure to increase spending is greater, it tends to win. Hence, ever-increasing healthcare costs.

Affordability

The catch comes that many individuals can afford to pay more for healthcare. When they have the option to do so for themselves (but not others) they often do so. In the US, you see this in luxury care for those able to afford it, and medical bankruptcy for those unable to afford it. In Canada, you see this in the push for a dual-tier system, where the wealthy would be able to pay extra money for more care than provided for by the public system. For those concerned with more equal access for individuals of all socioeconomic backgrounds, the push is more for increased taxes to pay for higher-quality heathcare for all.

Mo' Money, Not Mo' Health?

I work in healthcare. I would stand to benefit significantly if we suddenly spent more on healthcare, whether through a private, dual-tier system or through increased government expenditures. Yet, more money devoted to healthcare doesn't necessarily mean better health.

The Canada-US disparity in funding expresses this fairly easily. The US pays far more for their healthcare than Canada does, yet has overall health outcomes which are, at best, equivalent. Part of that is having great care for those who can afford it, yet very low-quality care for a group who cannot afford reasonable care. Fixing that disparity could push the US ahead of Canada in standard health metrics, but even so, not by all that much (and it might not even do that).

The thing is, that much money could have a real, meaningful impact on people's health - but not if it's spent in the healthcare system. The further into this field I get, the more I realize how poor many of our diagnostic tests and therapeutic interventions are. Some are incredibly effective, but many of a middling efficacy, while a few other may not work at all. There are some interventions that we know work really well though - these are improved diet, increased exercise, quitting smoking, reducing stress, longer/improved sleep, and moderating alcohol consumption. The healthcare system is slowly beginning to focus on these highly important lifestyle factors, but because they're part of a person's lifestyle, they're affected by much more than just healthcare services.

The difference in healthcare spending between Canada and the US is about 6% of GDP. If Canada decided it could afford to spend an extra 6% of its GDP on improving health - roughly $100 billion each year - I'd much rather see it go to efforts to improve those metrics that we know work. Getting people to eat better or exercise more has always been a difficult challenge, but with $100 billion each year, I bet we could come up with a few things that'd have an impact.

This brings me to my main point - there's a reason to control costs in healthcare, because healthcare isn't the same as health and it's certainly not the same as life. We want good healthcare and I want to be part of providing it. But I want good healthcare because I want people (including myself!) to be happy and prosperous. If the costs of healthcare hinder those efforts by minimizing spending on other worthy pursuits more than they help directly by improving health, it's not really beneficial. There are diminishing returns when pumping more money into healthcare, especially when that additional money doesn't have a clearly established, useful purpose.

Bottom line - controlling health care costs, whether public costs paid through taxes or private costs paid through insurance premiums and out of pocket, is an important goal, even for those who want the healthiest people possible.