Richard Smith: The ways in which hospitals can destroy health

Spending more and more on hospital care, means that you “crowd out” spending on other activities that do much for health, says Richard Smith.


I’m reluctant to watch medical programmes, but several people I respect urged me to watch the BBC’s Hospital, which tells stories from St Mary’s Hospital, London. I did watch the first episode, and it led to a spirited debate with my wife and reflections on the ways in which hospitals can destroy health (there are no doubt others).

The story that led to the debate was about a 60 year old man with the most enormous aortic aneurysm that on a chest X-ray seemed to fill half of his chest. He’d had several previous operations, and the tall, charming vascular surgeon who was going to operate on him said he’d been involved in half of them. The patient was calm, patient, and attractive.

We saw the patient first when he was admitted for what was going to be a huge operation, demanding a large team. It had taken a long time to find a date that worked for the whole team. The patient sat for all of one day hoping for a bed and did get one in the late afternoon. The next question was whether there would be a bed for him in the intensive care unit. It looked unlikely, but the next morning they took him to theatre and prepared him—only for the operation to be cancelled because there was no bed in intensive care. The patient was philosophical. Most people watching the programme no doubt thought: “How ridiculous. We clearly need to spend more on healthcare and have more beds.”

The man came back several weeks later, and this time he did have his operation, which seemed to go well. But at the end of the programme we were told that the patient had spent six weeks in intensive care and then died suddenly not long afterwards.

It was obvious to my wife that we needed more beds so that people didn’t have their operations cancelled. It wasn’t so obvious to me, and I tried to explain (to her and the majority of the population) why I thought what seems heartless.

I started by saying that when I was a medical student people with kidney failure did not get dialysed if they had diabetes or were over 40. I then suggested (although I don’t know) that it was probably only recently that the surgeons would have attempted such a heroic operation, which depended not only on the skill of the surgeons and anaesthetists but also on much else, including first class intensive care.

“What are you saying?” she said “That we should go backwards? You can’t do that. New things are developed, and they have to be offered to people. What would you have done with that patient? Told him to go home and wait for his aneurysm to blow, killing him instantly.”

I’d stated poorly; I tried to explain that I wasn’t in favour of going backwards but that we had to recognise that we can’t go on forever offering new treatments—because unfortunately the nature of innovation in healthcare is usually to increase rather than reduce costs. It’s the opposite of computers, phones, cars, and much technology where innovation reduces cost. If you insist on making all new treatments available to everybody, then you have to spend more and more on hospital care, meaning that you “crowd out” spending on education, housing, the environment, social care, public health, primary care, and community development—activities that do much for health even within the narrow definition of health as the absence of detectable disease.

She agreed that there had to be limits, and I explained about the National Institute of Health and Clinical Excellence (NICE) ruling that innovations had to be not just effective but cost effective, meaning that an innovation would costs less than about £30 000 for a Quality Adjusted Life Year. (I then needed to explain what a QALY is, and it’s something that may be familiar to health policy wonks but is a strange concept.)

“But what about this man? Shouldn’t he get his operation?”

I seemed to be avoiding the question. We discussed how much easier it is for NICE to rule on new drugs than on surgical operations and interventions that happen every day in the NHS, many of them cheap but also not very effective. The work up and the operation for the man might have cost about £100 000, and if he’d got by with, say, five days in intensive care that would have added another £10 000. The surgeon said before the operation that the patient might live many years, and if he’d lived four years feeling well that would have met NICE’s criterion. As it happened, he spent six weeks in intensive care at a cost of about £90 000 and died soon afterwards.

“Crowding out” is one way that hospital care destroys health. One other way is that it keeps people alive in a state considerably less than full health. As innovations continue we both consume more resources and keep still sicker people alive, and within the population there are more sick people, meaning that the health of the population is reduced.

Then hospitals are dangerous places. We have long known that something like 10% of people in hospital are damaged in some way by the experience—and about 1% are killed, making medical error one of the leading causes of death. Things are getting safer—but painfully slowly; and as the patients and their management become more complex the possibility of doing harm increases.

Finally, hospitals are important for breeding antimicrobial resistance to antibiotics, opening up the possibility that the heroic operations like the one on the man with the huge aortic aneurysm could no longer be contemplated—and that much of modern medicine would become impossible.

“But what would you have done about the man with the aneurysm? You haven’t answered my question.”

What would I have done? Once you are in the position of having the possibility of offering expensive or heroic treatment that might bring big benefits what can you do but offer it? The trick is to avoid getting into such a position.

My wife wasn’t convinced by my arguments.

Richard Smith was the editor of The BMJ until 2004.

Competing interests: None declared.