David R. Henderson  

Garett Jones on the Supply of Doctors

Tim Kane and Means-Testing... The GiveWell-Clemens Brainstor...

Which is the binding constraint?

My co-blogger Garett Jones posted this week on the supply of doctors, suggesting a new way of thinking about them. He used the idea of opportunity cost: if potential doctors face a high opportunity cost because of lucrative opportunities in the business world, then we won't get many new doctors.

Garett granted that doctors have succeeded in getting the government to restrict their supply but he seems to favor his new explanation.

Here's why I don't, and in laying this out, I'm making a point that his commenter, Mark V. Anderson, made.

There is a huge latent supply of people willing to be doctors. I know a number of them myself. One young man I know wants so badly to be a doctor that he took a year off to study for the MCATs. For every person admitted to medical school, there are many who are not. Presumably many of them could be good doctors. Some casual evidence: this young man was studying for his MCATs at our kitchen table and I started looking through the material he was studying: the physics of the spring, mechanical advantage, etc., things that I'm sure would come in very handy when he's trying to do a surgery. Other casual evidence: Navy officers I've taught at the Naval Postgraduate School who were enlistees in the Navy Medical Corps because they didn't cut grasp enough math or chemistry to get into medical school. One of them told me how he had once delivered a baby in Alaska with a flashlight in his mouth. [The enlistee's mouth, not the baby's.]

So why aren't these people, and many like them, doctors? Because of the aforementioned restriction of supply. Raise the opportunity cost of doctors and you won't get fewer doctors: you'll just get a shorter queue trying to get into med schools. Lower the opportunity cost of doctors and you won't get more doctors: you'll just get a longer queue trying to get into med schools.

Note: Commenter MingoV writes that the constraint is not due to the AMA restricting supply but is, instead, due to the limited federal funding of slots in medical schools. Even if that's true, that doesn't undercut my case that the binding constraint is slots in medical schools. Instead, MingoV is addressing why there's a binding constraint. But we don't need to take as given that the only way to become a doctor is to go to medical school. We're just used to that way because that's how it has been legislated for close to a century. One can certainly imagine motivated potential doctors learning to be doctors by apprenticing to existing doctors for a few years. That would expand supply without increased federal funding.

Coda: The binding constraint is the restriction on slots in medical schools.

Postscript: I'm not making fun of Garett's style by using his "Coda." Imitation, in this case, is flattery.

Comments and Sharing

COMMENTS (30 to date)
Justin Rietz writes:

Another anecdote: I have a friend who had a 3.5+ GPA at an Ivy league school who was incredibly driven to become a doctor. He applied to a large number of med schools and didn't get into a single one. He took a year off, volunteered at a medical research lab, studied to retake the MCATS, and worked on the side at a private, surgical medical practice.

He re-applied to medical school the next year, but this time to something like 50+ schools, if I recall correctly. He was rejected at all but one, at which he was wait-listed. Fortunately, they eventually let him in.

zdc writes:

The binding constraint isn't the number of slots in medical schools, it's the number of residency spots.

The number of medical school spots is not federally limited, and has increased considerably over the last several years. Allopathic (M.D.) enrollment for 1st years in 2012 was 19,517, compared to approximately 15,800 in 2001. Total number of 1st year slots at allopathic schools have increased almost 30% in the last 20 years...far greater than population growth. 15 new allopathic schools have opened since 2007.

The you've got the osteopathic schools (D.O.), which have increased enrollment at an even faster rate. D.O. schools had 5,800 1st years in 2012. Total osteopathic enrolment has increased from 7,822 in 1993-94 to 20,633 in 2011-12 -- on pace to triple over 20 years.

Then you've got the Caribbean schools which have another few thousand students, which have ridiculously large class sizes (they're for profit, Ross University had something like 600 students in it's 2010 entering class, double the size of the largest U.S. school class sizes).

All that to say, your coda, and MingoV's post, are incorrect. Not to mention other foreign medical graduates that come to train/work in the US.

And, it should be noted that as class sizes have increased, so have attrition rates. By expanding class sizes, you end up taking marginal students that wouldn't have previously been accepted. Not surprisingly, many of them can't cut it academically or clinically and wash out. Standards exist for a reason. Just because someone want to do profession A, doesn't mean they have the ability to successfully perform in such a role.

Hazel Meade writes:

Why is federal funding needed for slots in medical school? Why can't students pay the tuition?

Next question: Why do we hand out students loans to everyone, regardless of what they plan to study, instead of focusing them on fields where there is a shortage of workers?

Restrict the supply of loans to study psychology (the therapy kind), and more people will choose to study medicine.

David R. Henderson writes:

The binding constraint isn't the number of slots in medical schools, it's the number of residency spots.
If you have evidence on that that you can link to, I would appreciate it.
Also, even if that's true, again, we shouldn't get stuck in an historical pattern chosen by government. There is no necessary reason, other than a legal one, for someone to have a residency spot before becoming a full-fledged doc.

Peter writes:

Incentives matter. The incentives to be a doctor, especially a general practitioner, are diminishing rapidly.

Les cargill writes:

A filter model, eh?

We presumably still have doctors who commit malpractice. Those then qualify as false positives of this filter

If this one-dimensional "gate" model holds, then it's still not restrictive enough, or there would be no ( or less ) of that.

Justin Rietz writes:


"And, it should be noted that as class sizes have increased, so have attrition rates. By expanding class sizes, you end up taking marginal students that wouldn't have previously been accepted. "

If class sizes are expanding faster than the population, we would expect a higher attrition rate. But I'm not sure why this is an issue. The important question is whether or not we are getting a larger number of qualified doctors, i.e. graduates, due to larger class sizes?

Eric Hosemann writes:

I think evidence for David's point can be found in the increasing popularity of physician assistant and nurse practitioner positions. These positions require a decent chunk of a doctor's diagnostic and "prescriptive" ability, so to speak, without the doctor-size schooling requirements. Pursuing them effectively bypasses David's binding constraint. I suppose it bypasses Garrett's to some degree as well. They are an attractive option for intelligent people not facing the opportunity cost of passing up an investment banker job.

zdc writes:

Here's background on federally funded GME spots.

"There is no necessary reason, other than a legal one, for someone to have a residency spot before becoming a full-fledged doc." -- Acutally, there is. Residency is where doctors learn the skills and gain the experience they'll need to take care of patients. Sure, I guess you could be a GP and treat coughs and colds without one, but residency is the 'apprenticeship' of medicine, where doctors (who earn their M.D. at the end of medical school, prior to residency) learn their craft. You don't finish up medical school 'knowing' how to be a neurosurgeon or an anesthesiologist. Residency is years of arduous hands on training, because that's how long it takes to learn the intricacies of modern medicine. Even after completing residency (for example, 5 years for an orthopedic surgeon), many physicians (percent depends on specialty) go on to pursue additional years of training in fellowships because there's yet more to learn.

While lumped together for the sake of discussion, it should be noted that 'doctors (physicians) are a heterogenous bunch in terms of scope of practice and skill set. While the libertarian approach to minimizing or doing away with licensing and credentialing requirements may make sense with regards to primary care access (what most people tend to think of when they're talking about physician shortages and such), it's not a reasonable approach to high-end, specialized care. While an NP or PA or even an educated patient can 'manage' straightforward day to day complaints like sprains, strains, coughs, and colds with results non-inferior to a physician, the same can't be said of more complex, critical medical care. Say you hit the ER with a terrible headache and it turns out it's due to a rupturing intracranial aneurysm. At this point, you don't have time to provider shop, and you're not going to get a second chance. You want someone that's dealt with similar scenarios hundred, if not thousands of times. Clipping or coiling an aneurysm is something that takes years of training to get comfortable doing, and is a complex procedure in which there is virtually no margin for error. Long, structured, and well-organized training programs serve to develop these hard-earned skill sets and experiences.

David R. Henderson writes:

I think you misunderstood my request. It wasn't for data on federally funded spots. It was for evidence that the binding constraint is the number of residency spots. As you might recall, that was your assertion.

Brent Buckner writes:

Lots of links out there indicate that the binding constraint is the number of residency slots
(e.g. http://www.healthleadersmedia.com/page-1/PHY-279699/Med-Schools-Boost-Enrollment-But-Residency-Slots-Threatened and
http://medmonthly.com/2012/01/we-need-more-residency-slots-but-who-will-pay/ )

Anecdotally, residency slots is the issue of concern that I have heard from Foreign Medical Graduates and those in that pipeline.

Mm writes:

The rate limiting step is residency- which is difficult to increase rapidly in fields other than primary care. The notion that after medical school you are ready to practice comes from an alternate universe. Medical school prepares you for your training- it isn't your training. To train specialists you need not only the physical structures(hospital, clinics, expense equipment), but the teachers and a sufficient population of patients who require the care of specialists. Medical school provides the basic preparation to train & equips you with the knowledge base to handle future developments in your field. Medical school is basic training, residency is combat infantry school. You do not send people out as trigger pullers after basic, and you don't send MDs out to care for people after med school.

Roger Sweeny writes:

Both Mm and zdc seem to agree that after 4 years of college and 4 years of medical school, you are only ready to begin training to be a doctor.

["Medical school prepares you for your training- it isn't your training." Mm]

["Residency is where doctors learn the skills and gain the experience they'll need to take care of patients." zdc]

If that is true, there is something very wrong with pre-med and med school. I find it very, very, very, very hard to believe that 8 years of school is required just to get people to the point where they can begin to develop real doctor skills. If that is true of today's schools, they need major change.

zdc writes:


I think you misunderstand the process of medical education and the training of physicians. The subsequent posters hopefully provided some insight.

To address you original post directly:

"There is a huge latent supply of people willing to be doctors." That does not mean there is a huge latent supply of people who possess the personal attributes to become effective physicians. Even in the current system, a not insignificant percentage of people who make it to enrollment in a medical school (which you erroneously think is the binding constraint) end up never treating a patient. This can be attributed to everything from inability to perform the expected tasks to what I've seen most commonly (and what G. Jones correctly recognizes) -- high opportunity cost. Whether it's people opting out after earning an M.D. to go into consulting (a fair number of students from top tier medical schools end up earning an M.D. then heading off to greener pastures working for the likes of McKinsey and the like), taking an administrative post, working less than full time by choice (in 2011, 44% of female physicians worked less than full time), or quitting altogether because the work isn't work the pay. Just taking more students who think they want to be doctors(of marginal caliber who don't make the cut currently, and are therefor, on average, less qualified and more likely to washout than the current crop). There's a huge latent supply of people that'd love to play in the NBA, too...doesn't mean there are tons of people who could successfully compete at that level.

I'm glad you noted that your anecdotes were, at best, casual evidence to support your answer to the question of, "So why aren't these people, and many like them, doctors?" Just because some medic helped during a childbirth doesn't mean they have what it takes to become an OB/GYN any more than me making a free throw in a high-school gym means I've got what it takes to compete with Blake Griffin on a basketball court. Childbirth happens successfully without medical intervention more often than not. Just being present when things work like they're supposed to doesn't qualify you to be a physician.

And taking a year off to study for the MCAT may just as well make you less, not more qualified to be a physician. One test certainly does not a doctor make, but most students who matriculate successfully study for that exam without taking a year off, but rather, while living their life as usual, maintaining a full course load, engage in social and extra-curricular activities, and often working. I could take a year off and watch NBA game film and shoot around at a local gym all of my waking hours -- doesn't mean I'd have what it takes to compete at the NBA, or even college level.

The rate limiting step in the production of physicians is residency training, because without residency training, you're not able to acquire the skills to capably practice as a modern physician (with the provisio, as mentioned previously, that specialty medicine and primary medicine are different beasts altogether. Whether 4 years of medical school and 3+ years of residency to treat sniffles and aches and pains is another discussion...). Training residents is expensive, in both direct and indirect costs, the federal funding for residency spots varies by institution, but is on the order of $110k/trainee/yr. It would be prohibitively expensive for individuals to fund this training on their own, so without federal funding, spots would be few and far between. For example, a cardiothoracic surgeon spends anywhere from 7-10 years in residency in fellowship. For a student who already racked up the median debt for a medical school student graduating last year of $160k, good luck finding someone that would lend you the money to physicians to teach you and support yourself (and a family). Ergo, the binding constraint on the supply of physicians in the US is the number of residency training spots, which are practically limited by federal funding for said spots.

MingoV writes:
One can certainly imagine motivated potential doctors learning to be doctors by apprenticing to existing doctors for a few years.
Only people who think that medicine is a trade instead of a profession could believe such a statement. Could a high school student be apprenticed to a non-academic economist and reach PhD-level abilities in six years? I doubt it.

If you want the practice of medicine to become the application of medical technologies, then the apprenticeship route will work. If you want clinicians who fully understand how the body works, how its metabolism functions, the signs and symptoms of dysfunctions, the exams and tests needed to make diagnoses, and the understanding of why treatments work (or fail) and what side effects may occur, then you need a training model that includes lots of basic science and medicine courses and lots of time with supervised care of patients. The latter is our residency and fellowship system. Its advantage is that each resident gets overseen by many physicians instead of being apprenticed to just one.

Steve Sailer writes:

Milton Friedman wrote about the artificial squeeze on the number of doctors in "Capitalism and Freedom:"


David R. Henderson writes:

Interesting and often cogent reasoning. I'm wondering, though, whether you're willing to answer the question I asked you. If you recall, it was whether you can give a cite that establishes that residency slots, and not medical school slots, are the binding constraint.

MichaelM writes:


I don't think anyone wants anybody but an extremely highly trained neurosurgeon performing brain surgery on them...but how much of the work that a neurosurgeon does is actually this extremely skills intensive stuff?

The same questions applies to all medical fields, and the healthcare delivery industry in general. What you get when you restrict entry isn't necessarily fewer highly trained specialists, but a little of everything else. Perhaps each specialist should have a small army of slightly less trained specialists following him around -- similar to residents I imagine -- who should each have a small army of highly experienced tradesmen following him around. People who are there essentially to make sure that the extremely valuable time of each layer of people in the hierarchy is spent on the most valuable tasks. We see this already to an extent with the system of doctors, residents, nurses, and other healthcare personnel, but perhaps the industry is being held back by the contours of medical licensing.

I couldn't really tell you what a market for healthcare without mandatory licensing systems would like, but you seem to be running with a gimped version of the fundamental Smithian insight into the operation of markets by supply meeting demand: productivity advances are made by the extension of the market on the demand AND supply side. By restricting entry on the supply side, you retard the market process as much as if you restricted the number of patients doctors could serve.

You can make the argument that we just need to create new licenses as new possible positions show up, but then we get into the different, but related, argument of burying the entrepreneurial process under a mountain of red tape. Do you think regulatory bureaucracies can keep up with the rapidly changing technological workplace of the modern healthcare facility as quickly as individual doctors can?

Think of the libertarian proposal as less turning the doctor's profession into an unregulated free-for-all and more turning doctors into entrepreneurs. You can still have standards -- the entire global manufacturing industry operates within a set of mostly voluntary standards -- you just need to consider leaving behind restrictive barriers that turn the whole business into nothing so much as a guild system.

zdc writes:


Not sure what sort of citation you're expecting.

If you triple the number of medical school slots next year, 4 years hence you'll have done nothing to change the number of independently practicing physicians in the United States. You could increase the number of medical students in the US 40-fold to 1 million per year, and still, you would not end up with more capable, practicing physicians 4 years later. You'd have a bunch of people who knew some of the basics and could do a physical exam, take a history, maybe perform some very basic procedures. But they wouldn't be able to provide the vast majority of medical services that occur in hospitals across the country safely, effectively, or efficiently. They need additional years of focused training to develop the highly specialized skill sets requisite to successfully treat patients. That training is organized into a system called residency. So, the rate limiting step in the production of physicians fully trained for independent practice is residency spots.

Like pretty much everything else, residency spots are a limited resource. You need to have an adequate number of mentors/teachers (attending physicians) to instruct the residents, you need an adequate number of patients with varied disease such that they can learn the profession (each patient only has one gallbladder, or appendix, etc, so it can only be taken out once, which means you need a critical mass of patients to provide trainees enough opportunities to learn), you need money to pay all the people involved in the process, etc. Even if you had infinite financial resources, you couldn't double the number of residency spots tomorrow, because you need enough patients to learn on and enough teachers to mentor and instruct the residents.

You could put all 300-whatever million residents of the US through medical school over the next 4 years. At the end of that time, you'd have no larger a number of capable neurosurgeons, orthopedic surgeons, etc than you do today. Therefore, the number of medical school spots is not a binding constraint upon the supply of fully trained, functional physicians able to practice independently.

David R. Henderson writes:

Not sure what sort of citation you're expecting.
A citation that backs up your claim, which you've essentially repeated in your latest comment, that the binding constraint is on residency slots. I'm not saying you're wrong and you clearly are sure of yourself. What I'm looking for is some evidence for your claim. If you don't want to offer any, that's fine. But simply repeating your claim doesn't advance the discussion.

zdc writes:


You initial sentence, "...but how much of the work that a neurosurgeon does is actually this extremely skills intensive stuff?" gives away your abject lack of any practical knowledge regarding the delivery of healthcare in the present day United States -- yet somehow that doesn't stop you from opining on high as to how it could be improved upon.

The multi-layered system you espouse already exists. There's primary care docs, then specialists, then sub-specialists, then you have tertiary referral centers (places like Hopkins and Mayo) where you can find sub-sub-specialists who may be experts in treating a single disease.

To talk about Smithian free markets in the setting of healthcare is folly at best. A truly free market in healthcare would let many patients die at the doors of the ER (take a typical urban ER where half or more of the patients have no means by which to pay for their next meal, let alone resources to afford the liver transplant they drank themselves to, or for the complex surgical care, ICU time, and rehab and physical therapy they'll need after getting shot during their drug deal gone bad). May as well clean out all the NICU's, too...not many families can afford million dollar preemies out of pocket.

Medical outsiders think delivering medical care is like making widgets. It's not.

Milton Recht writes:

The current structure of schooling, training and laws creates significant barriers to entry to those who would like to be doctors, limits the availability of medical services, increases costs and unnecessarily forces individuals to use doctors.

Look at actual usage of medical services, and see where doctor involvement can be eliminated or significantly lessened. For example, eliminate the need for prescriptions for commonly used non-addictive medicines. Why does a woman need a prescription for birth control? Or a man for ED medicine? Why can't someone with a bad sore throat get an antibiotic for strep without seeing a doctor? or antibiotic drops for an eye infection such as pink eyes?

Expand the role of nurses, nurse practitioners, surgical assistants, and pharmacists and without the need that they be under doctor supervision or part of a physician practice.

Reduce immigration restrictions on foreign trained doctors and reduce requirements that they do US residencies. Instead require posting of medical schooling and residency and let the public choose.

Just like we have different levels of college education, Associate, BA, Masters, PhD, let's do the same for medicine. Can also do the same for medical training as we currently have, e.g., Board Certification. For example, EMTs are not doctors but are adequately medically trained to handle emergencies.

The US can easily eliminate unneeded demand for doctors and increase the availability of competent medical providers through legal and licensing changes. It would allow those that want to medically help people do so at many levels of training and it would eliminate unnecessary demand for expensive overly trained medical personnel.

These and other changes would go a long way to eliminate barriers to entry and the supply demand imbalance between medical services providers and medical services users. The changes would also reduce costs and eliminate the guild effect of the medical profession.

zdc writes:

Below you'll find links to several in-depth, heavily foot-noted reports. I don't know if you'll find the exact phrase 'residency spots are the binding constraint with regards to physician supply' because such is readily understood as a given, fundamental fact by anyone remotely familiar with the topic of the physician workforce.

Maybe this will suffice (from the opening line of the executive summary of the first link below -- State and Managed Care Support for Graduate Medical Education: Innovations and Implications for Federal Policy July 2004 (PDF - 38 pages). Council on Graduate Medical Education, US Dept of Health and Human Services):
"Significant changes in the financial climate of State governments may adversely influence their continuing support of graduate medical education (GME - i.e., residency training) and the resultant physician supply.





You may find some of the other material in the hrsa reports quite interesting, as well.

Ghost of Christmas Past writes:

@DRH: Here are the links that zdc didn't have handy:



Neither of them empasizes one additional subtlety (I'll try to find a link for this next and get back to you): some healthcare policy wonks want to reduce, not increase, the supply of physicians, because they do not believe that increased supply will meet demand at a lower price point.

The first link above does mention this concern:

Concerns also have been raised that the greater the number of physicians, the more health care is supplied to people, including much care that is unnecessary or even harmful.

(Harmful to the budget as opposed to the patient is left unstated.)

Some wonks believe that increasing the number of doctors will simply increase the number of doctors on the government's payroll-- that is, the wonks assume that the price of doctors can only rise and that demand is infinite, so from a budgetary standpoint the wonks think the only way to "contain costs" is to restrict the supply of doctors.

Brent Buckner writes:


One pager:

[Hat Tip: Commenter Cynthia at http://marginalrevolution.com/marginalrevolution/2013/03/facts-about-doctors.html ]

Steve J writes:


Wow some very interesting comments here.

@Brent Buckner

What the hell we limited the number of residency slots in 1997! There are times it seems like requiring a basic understanding of mathematics/economics/etc should be a requirement for politicians.

Floccina writes:

@zdc when economists and medical statisticians look at the performance of NP's, PA's and midwives they see that they get the same results as doctors. Therefore I see no reason to not make it much easier and quicker to become an MD.

It looks to me like the current system is so excessive that it is even bad for those that make it to MD and for the patients.

Devon Herrick writes:

Physicians are prone to make the claim that if it were easier to become a doctor, we'd all run the risk of being treated by a mediocre doctor. It is hard not to sympathize with this argument; we all would prefer to be treated by a talented doctor rather than someone with mediocre skills.

There is a distinct skill set involved in becoming a doctor that doesn’t necessarily perfectly correlate with being a good doctor. By that I mean that there are probably people who would be good doctors, but who lack the skill set (or resources) required to endure the process of becoming a physician. Granted, the current regimen of undergraduate pre-med, then selective medical school, finalized by lengthy residency requirements does ensure only highly-intelligent, highly-motivated people become doctors. But that's not to say there aren't more efficient ways to ensure quality.

Another thought comes to mind: our system of medical education is so costly that we error by making a physician’s time too valuable to spend with patients. As a result, physicians cannot afford to spend very much time during a patient encounter analyzing patient problems (third-party payment exacerbates this problem). A less-talented provider, who is not so time-constrained, could in some instances possibly provide better care by spending more time with the patient. There is some evidence of this: nurse practitioners (who are intelligent and highly trained, but have not gone through as extensive a training process as physicians) have higher customer satisfaction ratings than physicians because nurse practitioners spend more time with patients and listen to what the patient has to say. The current system doesn't work very well for conditions that require patient education to ensure patient compliance.

zdc writes:


"when economists and medical statisticians look at..." And from the definitive coda issued by an economist at the start of this thread, we can surmise that just because an economist offers up an opinion of medical care the delivery thereof, even if supported by statistics, doesn't make it so.

Sure, mid-levels can provide quality care to patients. When they're treating a limited selection of patients in controlled settings. But there's a huge selection bias -- the tough cases either intentionally excluded by pre-set criteria if the study 'proving' their efficacy occurs in the setting of a trial, or they're excluded de facto if the numbers are drawn from real-world retrospective data, because complex cases seek a higher level of care. The fact that a midwife can have similar results to a physician associated with managing the pregnancy and delivery of a healthy 20-something who carries to term isn't surprising. Billions of babies have been delivered successfully throughout history without interaction with ANY medical provider. But if it's a high-risk pregnancy, or any complicating factors arise, suddenly everyone is looking for an MD to help. You'd be shocked to know how many deliveries supervised by midwives don't go as planned and show up at the local ER leaving the 'overtrained' (as you suggest) MD to pick up the pieces.

"Therefore, I see no reason to make it easier and much quicker to become and MD." Aside from the fact that an MD does not a capable provider make (see above posts. You could have everyone in the US spend the next 4 years in medical school, and you still wouldn't have a single new capable neurosurgeon, orthopedist, etc. at the end of that time). So, you think you can design a better system. How about something concrete from you. I'd love to see how you think (expert that you are, you're seen studies that say certain subsets of patients have good outcomes no matter who sees them) someone, let's say a neurosurgeon, should be trained. How would you identify quality candidates for training, select a place for them to train, pay for their training, ensure adequate supervision during training, and ensure that they had achieved such skill as to be released into the wild to operate on someone such as yourself.

Can't wait to see your solutions!

zdc writes:

@Devon Herrick

You're an astute observer. You concluding sentence highlights an overwhelmingly truth of great importance.

One issue I would bring to your attention, you mistakenly equate quality of care with patient satisfaction. (provide better care by spending more time...nure practitioners...have higher customer satisfaction ratings). While the metrics obsessed loved patient satisfaction scoring, such scores have little, if any relationship to the quality of care received.

A NYTimes op-ed (by a nurse, no less) highlights the fallacy of such in a brilliant piece entitled 'Hospital Aren't Hotels'. It's an insightful read.


Comments for this entry have been closed
Return to top