Screening for Breast Cancer: Deja Vu All Over Again

Here is an excerpt from an interesting NY Times article. Read it all, it has a great punchline!

Dr. Leon Gordis, the chairman of the expert panel that advised the National Institutes of Health on mammograms last week, is a veteran of controversy.


He was a member of a panel that examined medical complaints by veterans of the Persian Gulf War, another that looked into allegations that food additives make children hyperactive and one that evaluated the safety of the nation's blood supply in the early days of the AIDS epidemic.


But Gordis, an epidemiologist at the Johns Hopkins University School of Medicine, said nothing had prepared him for the venomous reaction his panel got when it said in a report that it had no reason to recommend routine mammograms for women under 50. The reaction, he and others said, says more about the politics and psychology of breast cancer than it does about the science behind the committee's decision.


The panel was asked whether routine mammograms could prolong the lives of women in their 40s. There is abundant evidence that when women who are 50 or older have mammograms every one to two years, they reduce their chances of dying from breast cancer by about 30 percent. But whether women under 50 would benefit from similar screening has been uncertain.


After spending six weeks reading more than 100 scientific reports and then hearing 32 presentations in a two-day meeting, the group decided that there was not enough evidence that women in their 40s would benefit to advise them to have the X-ray test as part of routine health screening. The panel said women should weigh the risks and benefits of the test and decide for themselves whether they want it.


Barely had the words come out of Gordis' mouth Thursday morning when the audience began muttering and people began rushing to the microphones to rebuke the group, whose members sat looking stricken under the barrage. Prominent radiologists castigated the committee, with some accusing it of bias and others say the panel was condemning American women to death. One of the radiologists, Dr. Daniel B. Kopans of Harvard Medical School, said the committee's report was ``fraudulent'' and should not be released to the public until it was ``corrected.''


Dr. Richard D. Klausner, who, as director of the National Cancer Institute, had asked that the panel be convened, rushed to the hallway to use a public telephone after Gordis read the statement. In an interview there, he said he was ``shocked'' by the conclusions, adding that he disliked their negative tone. He said an advisory board to the cancer institute would review the decision this month.


Some breast cancer patients who are convinced that their lives were saved by mammograms said they felt betrayed by a report that questions the usefulness of these X-rays of the breast in younger women.


People who were not at the meeting also have chimed in. Dr. Bernadine Healy, dean of the College of Medicine at Ohio State University and a former director of the National Institutes of Health, said that although she had not read the report, she was shocked by the panel's conclusion. ``I am very disturbed that a group of so-called experts challenged the notion of early detection,'' she declared. ``What they are saying is that ignorance is bliss.''


Before the week was over, Gordis said, he had been summoned by Sen. Arlen Specter
of Pennsylvania to testify before Congress on the panel's report.

The punchline? This article is from 1997.

Hospital Operating System: Part 1

Hospital Operating Systems can transform hospital operations... So, what's a Hospital Operating System?

I have argued that we ought to transform hospital operations because doing so will have dramatic financial and quality impact and may reduce the number of avoidable deaths due to medical and pharmaceutical errors.

I have also suggested that this transformation starts with installing a Hospital Operating System not with process improvement.

So, what exactly is a Hospital Operating System?
(Disclosure: my company develops and sells a Hospital Operating System solution)

A Hospital Operating System would be comprised of people, processes and software that work together to achieve the goal of optimal patient-flow: moving all patients from arrival through discharge with quality and efficiency. More specifically, an effective Hospital Operating System will

• be a human-machine system composed of people, IT systems, standard operating procedures, and executive mind-set
• interconnect all processes relevant to patient-flow
• be cybernetic

I will talk about the first two characteristics in future posts, but with respect to that last one - the really geeky sounding one - please consider the thermostat.

Thermostats are interesting. Actually I should say, the thermostat in your house is part of a system that controls the temperature in a very interesting way. It

• allows a goal temperature to be set 
• senses the temperature of the environment 
• utilizes heaters and air conditioners to change the temperature 
• contains simple rules describing what signals should be sent to the heaters and air conditioners when the temperature deviates from the goals 

The thermostat participates in a nice, self-correcting system which tweaks the environment through its effectors - heaters and coolers - based on comparing the goal temperature with the information it gets from its sensor (thermometer). It does its best to maintain the temperature despite fluctuations in the number of people in your house, the outside temperature, etc.

By way of contrast, consider the simplestat. The simplestat attempts to keep your house comfortable by turning the heat on from 5:00 a.m. to 9:00 a.m. and again from 7:00 p.m. to 11:00 p.m. While this device might work perfectly well in Atlanta in March, it will be a disaster in August. And, it will be a disaster precisely because it is not influenced by the environment - it completely ignores the ambient temperature, for example.

The simplestat is an example of a closed system and, as such, it is doomed to perish. The thermostat, on the other hand, is an example of a cybernetic system meaning it interacts continuously with its environment and is capable of self-regulation. And, this is really important, no system can persist if not cybernetic.

Hospitals often do the hard work of process improvement and craft excellent processes with respect to a moment in time. However, these optimized processes are typically embedded in closed systems that live in an inherently dynamic environment, so most often the efficacy of the processes degrades over time. The closed systems are brittle cracking under any type of environmental change and making any subsequent improvement costly.

Software and IT systems do not necessarily help. Significant attention has been paid to the idea that process improvements are not sustainable if they depend on manual procedures, so technology has been used to memorialize, support, facilitate and enforce processes. Ironically, the technology used is frequently itself a closed system. Perhaps because of its closed nature, the cost of changing an IT system's behavior is typically very high so responding to environmental change becomes prohibitively expensive. Thus, the IT system acts as a sort of digital cement unable to bend with change resulting in the degradation of process effectiveness.

Ideally one should use technology, but that technology should be cybernetic and be part of a larger hospital-wide cybernetic system that contains, orchestrates and provides visibility to the goal of moving all patients from admission through discharge. Instead of digital cement, the software component of a Hospital Operating System would be more like digital rubber allowing a wide-range of flex to accommodate a steadily changing environment

Such a system would regulate itself with respect to preserving its operational goals. Stated more formally (and in all of its geeky glory), a successful Hospital Operating System will be cybernetic. It will

1. allow operationally relevant goals to be set 
2. contain sensing mechanisms whereby information relevant to the operational goals is registered.
3. contain effector mechanisms whereby the system acts on its environment.
4. contain transformational rules or procedures whereby information received from the sensors is compared with information about the goals and error-correcting signals produced to modify the behavior or the effectors
   

Hospitals are too complex and dynamic not to have an operating system, and a Hospital Operating System needs to be cybernetic, adapting to the dynamic hospital environment and supporting future process improvement efforts.

Is Joe Flower wrong?

The November 3rd The Health Care Blog features an op-ed piece by Joe Flower entitled Why "free market competition" fails in health care.

Mr. Flower is a speaker, author, and healthcare futurist studying change and the future in healthcare for nearly 30 years. So, given that expertise, it is with much trepidation that I respectfully disagree with him and question whether he has argued his case.

Mr. Flower says, "In trying to think about the future of health care, thoughtful, intelligent people often ask, ‘Why can’t we just let the free market operate in health care? That would drive down costs and drive up quality.’ They point to the successes of competition in other industries. But their faith is misplaced, for economic reasons that are peculiar to health care.”

The structure of his argument seems to be the following:

Free market competition fails in healthcare for economic reasons that are peculiar to healthcare

1. True medical demand is wildly variable, random and absolute. Some medical needs are mandatory for life; and, life-saving and life-giving medical need has no correlation with ability to pay
2. Medicine is so complex and difficult we depend on the seller (the doctor) to tell us what we need. Whether we buy or not usually depends almost solely on whether we trust the doctor and believe what the doctor says.
3. The benefit of medical capacity / infrastructure accrues even to those who do not use it.

Therefore healthcare is not responsive to classic economic supply-and-demand, and a free market model is a bad model.

If you read my blog, you know that it is my belief that a free market healthcare system is necessary, and its absence is a root cause of what ails us. So, the fact that I disagree with his thesis is, perhaps, not surprising. With respect to his op-ed piece, however, I struggle to even follow his line of reasoning.

First point: It is true that some people get cancer and others don’t. It is also true that one's need for life-saving and life-giving treatment is not related to ones' ability to pay. However, I see no evidence that these facts disqualify a free market healthcare system from being the best idea. To my eyes, they simply underline the importance of high-quality yet low-cost services. And, I can think of no better way to fulfill this need than by utilizing free market forces.

The best way to minimize the number of people who cannot afford the healthcare they need is by having consumers participate in a system that drives down price and increases quality through consumer dollars rewarding skill, innovation and outcomes.

Furthermore, there is every reason to believe that a free market approach would provide business models for true health insurance which – like fire, flood or disability insurance – would offer reasonably priced calamity coverage by amortizing risk over large populations of people who find value in having insurance.

Finally, having a free market healthcare system does not forbid or even discourage public- and private-sector safety nets. The citizenry has generally not minded the use of their tax (and charitable) dollars to take care of the truly needy – America gives more than twice as much as the next most charitable country by dollar and percent of GDP. When capitalism has driven down costs and increased quality, we best utilize those tax and charitable dollars.

Mr. Flower’s argument, to my mind, is an example of arguing with extreme examples as a way of refuting a philosophy. In this case, the philosophy in question is that a free market healthcare system would be good; and the extreme example is “a free market healthcare system, divorced from any other complementary government or private endeavor, in a world where no one placed any value in safety net, charity, intervention or incentive.” In a robust debate regarding the usefulness of a free market healthcare system, one ought to examine whether the free market philosophy offers a better outcome than its alternatives, not discredit it with an unrealistic, extreme and soulless example of a free-market world.

Second point: Mr. Flower says, “Medicine is so complex and difficult we depend on the seller (the doctor) to tell us what we need. Whether we buy or not usually depends almost solely on whether we trust the doctor and believe what the doctor says.” While this might be an accurate observation of the world we live in (we do not have a free market healthcare system), it says nothing about whether healthcare must be this way, nor is this statement relevant to whether a free market healthcare system could work. He is confusing a result of our current system with an intrinsic quality of healthcare which makes it incompatible with a free market system.

Mr. Flower’s observation is likely to be the result of cultural attitudes in combination with not having a free market healthcare system. The original Oath of Hippocrates, circa 400 B.C, includes this very telling line, “I will impart a knowledge of the Art to my own sons, and those of my teachers, and to disciples bound by a stipulation and oath according to the law of medicine, but to none others. (emphasis mine).”

Culturally, we have tended to treat medicine as an arcane art available to only the few who complete the arduous initiations and trying apprenticeships. Add to this the fact that the consumer has only limited power (because the forces of free market capitalism have been disabled), to vote (with their dollars) for better communication and education, it is not surprising that medicine tends to feel unknowable by the lay person.

However, even despite these handicaps, there is good evidence that educating patients can produce informed consumers of healthcare (one of many examples: Computer-based patient education has been shown to be very effective in improving knowledge and clinical outcomes - Lewis, D. Computers in Patient Education, Computers , Informatics, Nursing 21(2):88-96, 2003).

WebMD has made complicated treatment and outcome information very understandable and, if our system allowed consumers to vote with their dollars, clear communication would be valued and improved. Imagine if care delivery organizations and providers competed for consumers along this dimension. Indeed Emmi Solutions (disclosure: I am the former CTO of Emmi Solutions), Krames, Med-IQ, WebMd, Relay Health and countless payor-initiated efforts have already made great strides in disputing the “medicine is too complicated to produce educated consumers” claim – all while seeking to make a profit in the context of the free market paradigm.

Third point: Mr. Flower’s third point seems to be that because people who don’t pay for healthcare get some benefit, a free market healthcare system will fail. He says, “This is the infrastructure argument. Every part of health care, from ambulances and emergency room capacity to public health education to mass vaccinations to cutting-edge medical research, benefits the society as a whole, even those who do not use that particular piece." In this one sentence, Mr. Flower references a number of  free market industries that are part of the healthcare system (commodity, transportation, manufacturing, etc.) as a way of saying that free market forces will fail healthcare. Admittedly, I just don’t understand this part of his argument, but I would point out that 

1. Having a for-profit news industry is good for the entire community. I get many benefits of living in a society which includes this industry even if I don’t buy or consume news. And, even though I get this benefit without paying for it, it survives in a free market system.
2. Having free market transportation industries (airlines, trucking, shipping, etc.) is good for the community because it subsidizes the cost of important infrastructure for which we all benefit and provides for lower-cost goods. Even if I never travel and walk to shop for everything I purchase, I get massive benefits from those industries without sabotaging their ability to thrive in our mostly free market system.

I do not see in Mr. Flower’s argument the identification of anything that is intrinsic to healthcare that argues against a successful free market approach. Could someone please point out my error?

Stop Improving Processes if You Want to Transform Hospital Operations

Imagine air-traffic control using nothing but telephones, pagers and white-boards. Seems like a bad idea, doesn't it?

It might work for a few dozen simultaneous flights but how about a few hundred or the 4,710 commercial flights that are in the air as I write this? In this Flintstones' world, I imagine, flying is unpredictable, expensive and, relative to our world, dangerous.

Stepping into Bedrock, would you suggest that they fix their problems through process improvement initiatives focused on improving taxiing, refueling and flight-planning? Or, would you recommend the use of modern technology and install a state of the art air-traffic control system that ensures all flights fly at their best possible rates with regard to safety and efficiency? I'm betting on the later.

Well, with respect to flowing all patients simultaneously from arrival through discharge we are basically in Bedrock. So, instead of improving sub-processes of patient flow why not tackle the root of the problem by installing an enterprise-wide hospital operating system?

While it’s true that a hospital may have some technology assist – hotel-like software helping manage the "booking" and cleaning of rooms and facilitating transports – most have nothing that approaches a system whose aim is ensuring all patients flow simultaneously at their best possible rates with respect to length of stay, service times, quality, safety and resource consumption.

Just as bad air-traffic control has dire consequences, so does not having this hospital operational control. I believe that there is evidence that its absence leads to tens of billions of waste annually and probably contributes to the 44,000 avoidable deaths due to medical and medication errors.

The truth is that today’s modern hospital is too complex, with too many simultaneous transactions not to have a system that is responsible for patient-flow, yet most often they have none. As a result hospitals tend to operate as a collection of independent departments that compete for limited resources: clinicians, patient beds, wheelchairs, medications, IV pumps and other essential diagnostic and treatment resources. Most attempt to treat this systemic problem by repairing its parts, but the treatments tend to be outside of any global system-aim or "organizational consistency of purpose", as W. Edwards Deming might say. Thus, state of the art process improvement methodologies are applied to one department at a time, but process improvement which rests on the shoulders of siloed operations and technology won't be transformative and is difficult to sustain.

Without an enterprise control-system, process improvement initiatives are unlikely to provide meaningful, sustainable, enterprise impact. If you believe that it is vital to transform hospital operations, start by installing a hospital operating system. And then start improving processes.

In future posts, I will describe the characteristics of a successful hospital operating system, and I will offer a few case studies of their success.

There are these two young fish swimming along and they happen to meet an older fish swimming the other way, who nods at them and says "Morning, boys. How's the water?" And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes "What the hell is water?"

David Foster Wallace