Mental Health and Resources - Addendum

Since my last post, I've had a few opportunities to chat about the state of mental healthcare in the area, particularly what could be done to improve the situation. It's in the news a fair bit now, so it's come up in conversation a lot, particularly while on my Psych rotation. One idea that came up was increasing Psychiatrist work hours.

It's a valid thought. Psychiatrists do work fewer hours than most other physicians, especially hospital-based physicians. If any physician can be told that they need to work longer hours, it may be Psychiatrists. However, I'd like to push back against that notion for a few reasons.

First, while Psychiatrists work fewer hours than many other physicians, is that because Psychiatrists are working too few hours, or because other physicians are working too many? I'd argue it's the latter. Psychiatrist hours still average over 45 a week, not much below that of other specialties, particularly office-based practices. Given overall rates of physician burnout and unhappiness with work-life balance, asking Psychiatrists to work longer hours would likely help one problem by causing another.

Second, while no physician works optimally when tired, tired Psychiatrists can be particularly troublesome. Psychiatrists have to be careful not just with their clinical decisions, but their words as well. Emotional lability increases with exhaustion or chronic mental fatigue - a bad thing for all physicians, but particularly for one who has to stay mindful of their phrasing whenever they interact with a patient. With non-psychiatric patients, words matter of course, but the margin for error is much greater - a poorly phrased statement can typically be apologized for, clarified, or otherwise explained without enduring consequences. Not so with some psychiatric patients.

Lastly, there's only so much additional hours from Psychiatrists will do to help. It wouldn't help with inpatient services. It wouldn't help with community supports. It wouldn't help with long-term care availability. At best, it would help reduce outpatient wait times. That's still a desirable outcome, though I doubt that longer hours for Psychiatrists would be anywhere near enough. If wait times could be reduced from 1 year to, say, 9 months, that would be a positive change, but 9 months is still far too much time. 

It's a tough situation. As our society increasingly accepts mental health as an important priority, demand for mental health services will also increase. Realistically, all options should be on the table, including longer working hours for Psychiatrists. However, we need to avoid band-aid solutions and "work harder" is pretty close to my definition of a band-aid solution. Other approaches are necessary and, I believe, should be explored first before simply asking current workers to do more.

Work Hours and Handovers

A major argument against restriction of work hours for residents (or staff, though that gets less attention) is the notion of increased handovers. The basic story goes that when you have residents work fewer hours, or at least fewer hours in a row, they have to handover the care of their patients to other residents more often and since handovers are well-established as leading to error, more handovers mean more errors. It is thus hypothesized that any gains made on resident fatigue by restricting hours will be eliminated by increased problems associated with handovers.

When work hour restrictions were put in place in the US, studies on healthcare quality came back pretty equivocal, so the hypothesis appeared valid and has been seized upon by many with an interest in medical education, particularly those opposed to work hour restrictions. I accepted this theory as well, since it seemed to fit the evidence and was plausible in theory.

Now, acceptance of this theory didn't change my views on reducing work hours in medicine, since handovers are not an inherent problem, but rather one that could be addressed and improved. Moreover, since our current system necessarily involves a lot of handovers, fixing them should be a top priority anyway. As a result, any losses in care due to increased handovers associated with shorter working hours should, if anything, get better as time goes on, allowing any gains from those shorter hours to be realized.

However, the more I look at the problem, the less I'm convinced this theory holds much water in the first place, at least not in many cases.

For example, shorter shifts don't necessarily lead to more handovers for the majority of patients. Individuals on call for 24 hour shifts don't typically just take care of their own patients, they take care of patients on their "team", many of which are cared for by other team members. When I'm on call during my current rotation, I am responsible for up to 30 patients, yet most days I only care for 4-6 patients. Residents often take on a bit more work, but even then, they're unlikely to be familiar with more than 10 patients. They need to be updated - or receive handover - from the other members of the team. If we had two distinct shifts taking care of a set of patients, one day shift and one night shift, the total number of handovers would be virtually unchanged for the majority of patients.

Likewise, not all handovers are created equal - handing over a patient that both parties are reasonably familiar with is a much different situation than handing over a patient that the receiving party has never cared for before. This gets back to a more fluid concept of continuity of care which I've written about before, where continuity of care doesn't mean one physician caring for one patient, but rather one physician maintaining the same role in caring for that one patient.

So how do I explain the rather equivocal results from studies on resident work hour restrictions? My current belief is that work hour restrictions failed to achieve their process goal, which was improving resident fatigue. Through a combination of increased work intensity, scheduling loopholes, and well, simply ignoring the restrictions, residency programs simply found creative ways to continue to work their residents far, far too hard. We didn't see a change in patient outcomes, because we never really saw a change in resident working conditions, simply trading one inadequate resident work system for another.

I can't dismiss handovers as a factor entirely, since I've seen first-hand how poor handovers can cause substantial problems for patients. Yet, when it comes to work hour restrictions, I'm coming around to the conclusion that they're used a bit as a red herring, part of a rhetoric thrown out to justify maintaining the status quo on work hours, without much conclusive evidence to demonstrate that the status quo should be maintained.

High Unemployment, Long Hours

The current job market for physicians is a bit weird right now. There are several specialties with a combination of high unemployment/underemployment (by physician standards), but long work hours for those employed and generally high pay. Many surgical specialties fall into this category. I've expressed that this is a rather absurd situation, as physicians who likely need shorter hours and have the per-hour income to afford an associated decrease in take-home pay from working less have plenty of people to pick up additional work yet who aren't being utilized.

I haven't had a good reason for why this isn't happening, so I've assumed that physician stubbornness (or perhaps greed) in these specialties is to blame.

I stumbled across a random comment that may provide a better explanation, essentially boiling the situation down to bargaining power. Hospitals and universities in Canada would generally prefer to hire the minimum number of specialists necessary to get the work they have done. In specialties with more physicians than positions, the hospital (or physician group) has additional leverage to insist on longer hours - after all, if one physician doesn't want to handle a standard load, there are plenty of other physicians who would gladly do the work.

Similarly, in specialties where the number of positions vastly exceeds the number of available physicians, it's the doctors who have the bargaining power. The is a major need for more psychiatrists, yet they tend to work relatively short hours. Same for Emergency Medicine, which has reformed itself to have comparatively light schedules, albeit with shifts at all times of the day, and has a very favourable job market.

Granted this doesn't explain everything - Radiation Oncology has low average work hours, decent compensation and a terrible job market. Patient population, culture within the specialty, and funding models obviously have an impact on the work hour and employment situations. Nevertheless, it's an interesting theory that I thought I'd share, one I intend to keep in mind as the job market in certain medical specialties evolves.

Public Opinion on Resident Work Hours

Stumbled across this article while researching a completely unrelated topic. Basic idea is that public opinion on resident work hours is rather strongly in favour of suggested restrictions and even is in favour of far more restrictive work hour limits. American-focused, but still relevant here in Canada.

I'm not a huge fan of making policy decisions - in medicine or otherwise - based primarily on public opinion. After all, people can make poor decisions and large groups of people are not guard against poor decision-making. More specifically when it comes to resident work hours, the public probably lacks a degree of perspective necessary to formulate a well-informed viewpoint.

However, the disparity is striking, and demonstrates how out-of-touch the practice of medicine is with the rest of society. While physicians debate the merits of an 80 hour average week, with more than a few physicians coming out strongly against such restrictions, the majority of non-physicians would be in favour of a 60 (or even 50!) hour average week with 80 hours maximum in a week. These opinions were fairly broadly-held, not dependent on demographics or ideology.

What's currently being discussed in medicine in terms of work hours borders on what "normal" people would consider insane. Perhaps these long work hours are necessary due to other factors. Still, even if it is a necessity, it's increasingly hard to maintain the pretense that these hours should be considered manageable by any common standard. Patients don't think their physicians should be working such long hours - if we continue with the current work weeks, we'd better have a very good justification as to why it's necessary they do.

Work Hour Restrictions

Continuing on with my recent theme of super-long hours in the hospital, partially for myself, but mostly for my residents, it's worth mentioning the big elephant in the room for Canadian residency programs: resident work hour restrictions.

Long work hours have been identified as a potential cause of a lot of problems in medicine, both for providers and receivers of that care. The obvious solution then is to have providers - in particular the overworked residents - spend fewer hours on the job. While most countries with similar residency systems have implemented some form of work hour restrictions, Canada has not outside of Quebec. As such, it remains a topic of rather passionate debate, with fairly strong, opposing viewpoints from those at all levels of training.

Canadian practitioners are most familiar with how work hour restrictions came to be in the US, which were first implemented nationwide in 2003. The initial impetus for this change came from the death of Libby Zion, who, under the care of two overworked junior residents, was given a medication that when combined with her normal daily medications, resulted in a life-threatening condition that was not caught and eventually resulted in her death. The argument for work hour restrictions was fairly straight-forward: tired residents cannot deliver optimal care, so capping their hours (along with other restrictions on the distribution of those hours) will lead to more rested, more capable residents who would minimize mistakes for patients. There is more than enough evidence to support the notion that overworked or exhausted individuals are not the greatest workers and are prone to mistakes

The other justification, not always mentioned prominently in the debate, was to maintain the health and well-being of those residents. After all, residents are people too, who shouldn't be worked to the point of exhaustion. There is also quite a bit of evidence that long work hours contribute to medical or personal problems.

The argument against work hour restrictions was equally straight-forward: experience is the only way to make a competent physician, so work hour restrictions mean missed learning opportunities and less-competent physicians. There is also a large body of evidence for this argument, particularly in surgical specialties. You can't be competent with a surgical procedure until you've done it over and over again, so missed opportunities to scrub in add up.

It's now been 12 years since the first round of American work hour restrictions went into place and they've received a fair bit of scrutiny. The result? On all three points of consideration - patient outcomes, resident fatigue, and resident competency - the answer seems to be... mmmaaaaayyyybe...? In all respects, there seem to be studies which say work hour restrictions hurt, some which say they help, some which say they didn't make much different. On the whole, the answer seems to be that residents provide the same level of care, are just as tired, and just as competent as they were before the work hour restrictions.

Many of these studies come with fairly significant problems. With strong opinions on the subject being rather prevalent, bias in study design or result interpretation is a major concern that is difficult to address. Another large limitation is that actual hours worked before and after is rarely investigated. Some studies simply looked at results before and after the work hour restriction policy was put in place, without considering what effect that policy actually had on hours worked. Others used self-reported hours from residents which, as later investigations demonstrated, were often falsified to make it look like programs were in compliance with the regulations when they were routinely going over limits. Hard to see a change in outcome without a change in inputs.

A further criticism was that the work hour restrictions were still fairly permissive - 80 hours per week averaged over 4 weeks was the major cutoff. That's an average of about 11 and a half hours each and every day of the week, including weekends - hardly a light schedule. There were some other restrictions, such as 10 hours off between shifts, shifts that couldn't be longer than 24 hours (with an additional 6 for handover/education), one day off every week (on average) that were meant to make things a bit more palatable, but we're still not talking anything close to a 9-5 or even 7-5 job with weekends off.

There are also a lot of confounding factors when it comes to studying these interventions, which haven't been well addressed in previous literature. First is the challenge of handing over patients. Medicine sucks at transferring care between practitioners - it's a major opportunity for errors. When residents are prevented from working long shifts, it leads to more transfers of care, and more opportunities for mistakes to happen. The obvious response here is "fix the transfers", but that's easier said than done, so it remains a valid confounder. Another wrinkle in the intended effect of work hour restrictions is that residents don't always sleep when given more time. With 80+ hour work weeks, residents weren't just lacking sleep, they were lacking time to exercise, socialize, eat well, do domestic duties, spend time with their partner and/or children, whatever was important to them outside of medicine. When given the chance to spend less time at the hospital, sleep wasn't always the priority.

Unofficial duties also pop up outside the official work week. Every resident studies on their own time. Many participate in educational projects or research. "Homework" of a sort is pretty common. From my own experience, part of the challenge of dealing with 11 hour workdays is that I do 1-3 hours of studying, schoolwork, or EC tasks when I get home in the evening. In an environment where career opportunities are much less available than in the past, the impetus to go above and beyond, including using spare time, may eat up some extra time off.

When I started looking into the issue of resident work hours, my incoming opinion was that work hour restrictions were a painfully obvious solution to a painfully obvious problem. Given the available evidence, flawed as it may be, it's hard to maintain that view - work hour restrictions are not the disaster the detractors predicted, but neither is it a magic bullet. There are too many other factors to resident fatigue that need to be addressed in conjunction with work hours. We should be trying alternative approaches to reduce resident fatigue and evaluating them as intensely as we have work hour restrictions. In conjunction with or perhaps even in place of work hour restrictions, alternative approaches may provide the fatigue reduction many residents clearly need. It's hard to fathom that someone can work over 80 hours a week and be well-rested - there's simply no substitute for a good night's sleep - but I'm open to the idea if I can see some convincing evidence.

However, what are the alternatives? Safe transportation has come up, to prevent tired residents from driving home, but this is expensive and for many residents, costly in terms of time - a 15 minute drive home means more sleep than a 45 minute bus ride, for example. Another option, though not one I'd consider ethical, is medication. Some stimulants have been shown to improve wakefulness for tired workers, but sets a very dangerous precedent if it were to become compulsory. Past that, most discussions on this subject don't often provide realistic alternatives to improving resident fatigue. We should be open to any reasonable solutions to improving resident fatigue, but without workable alternatives to consider, effective implementation of resident work hour restrictions may be the only option for the moment, even if as a first step. Work hour restrictions may not be sufficient, and may not be optimally implemented in some jurisdictions, but they just might be necessary. Canada should consider an evidence-based approach to reducing resident working hours.