Myths of Health Reform

With some health reform still a possibility, Paul Krugman writes what seems to be an extended plug in favor of the current plan under consideration. He cites three "myths" about health reform. Interesting read.

Article here.

Multitasking and Nurses

A recent article from The Chronicle of Higher Education online, cites studies that have repeatedly shown that we are not built to multitask.

Contrast this with a recent article (sorry, don't have an ungated version) from a journal published by Joint Commission on Accreditation of Healthcare Organizations (JCAHO) on interruptions and multitasking among nurses. (Hat tip to Sal Bognanni of the Iowa Health System for sending me this paper.) The most surprising finding in the paper was that none of the many errors the researchers found were due to interruptions or multitasking. The authors point out that the many interruptions and frequent multitasking of nurses hold a potential for increased errors, but their data don't support for this claim at all.

Are nurses just superhuman? Did the study miss something?

Health Insurance Options

Options in the trading sense: the right to buy or sell something in the future. A cute idea in this article from Managed Care. The authors suggest creating an options market for individual services in health care. So, since I have kids and a trampoline, I should buy an option on pediatric orthopedics. But since there's no history of diabetes in my family, I might roll the dice and not buy an option on diabetic care.

I say this is a "cute" idea because it's clever, but totally ignores how people actually behave. Forget the complexity of options trading, how many of us are involved enough in managing our investments that we have even thought about buying/selling options? Remember, people have long thought that putting money into their 401k was important, but taking the time to figure out how to configure their investments proved prohibitive.

Who knows, maybe I'm ready for it. We're done having kids, who wants to buy our OB option?

Does Health Insurance Improve our Health?

I'm sure I'm not the first on the scene, but I was reading The Atlantic this morning--I'm still stuck in the hard-copy world--and found an interesting article by Megan McArdle.

McArdle questions an assumption that, so far as I know, no one else does: is health insurance good for us? McArdle's article is polemical, to be sure. But she points out something very important: the data on the benefits of health insurance are, well, shaky.

And there's even some surprising evidence against the efficacy of one of the most powerful insurer in the US--Medicare. McArdle quotes researchers at University of Michigan.

Medicare increases consumption of medical care and may modestly improve self-reported health but has no effect on mortality, at least in the short run.

To be honest, I'm not willing to throw health insurance and health care under the bus, even given problematic evidence for/ against it. But I am glad that someone is questioning mostly unexamined assumptions about health care.

Standards for EHR, Part 2

In a previous post I noted a problem that arises from our not having an agreed-upon forum for setting standards for health information. To recap, computers have settled standard interfaces like USB ports, whereas health information has competing standards. In this post, I'd like to mention a couple of other issues.

The first issue is economic. Keyboards, for instance, would be a lot more expensive if there were no USB standard. Each keyboard and its computer would have to be custom made to fit each other. Among other things, this would mean that there would be no competition among keyboard makers, which would drive prices up. Keyboards aren't terribly expensive, so this might not seem like a big deal. But the EHRs that hospitals often buy cost millions--even tens of millions--of dollars. Both initial costs--in the broadest sense--and switching costs are low for a USB keyboard (or mouse or hard drive). But for EHRs, no standard means no competition. And no competition means astronomically high initial costs and generally prohibitive switching costs. Put another way, my doctor may not be able to afford to have his system talk to the hospital's.

The last issue I'll mention is that without an information standard we have a severely limited ability to compare the effectiveness of various treatments. For instance, it has taken more than 30 years to get a modicum of clarity about the relative efficacy of angioplasty vs. bypass surgery. Lacking apples-to-apples comparisons (or the willingness to do randomized controlled trials, which would mean some people would get no intervention after a heart attack) we’ll continue to have a devil of a time figuring out what works.

So, I'm hoping for an IEEE for health care information. But I'm not holding my breath--the IEEE took nearly 80 years to emerge as the standard for standards.

Standards for EHR, Part 1

We take for granted that any device with USB connection will work with our computers. We owe this clarity to the IEEE (Institute for Electric and Electronics Engineers). This professional organization has been setting standards for lots of things since 1963, including the USB ports on our computers.

Electronic health records (EHRs, which includes "EMRs," where M is for "Medical") don't have similar clarity. There are standards, but they're currently competing for supremacy. (See this paper for a summary of the standards.) That is, there are standards, but no standard for standards like the IEEE.

Consequently, the system that my primary care doctor uses may not (in fact, probably doesn't) "talk to" the system at a hospital or even to the one used by the orthopedist my doc referred me to for ankle pain.

To be clear, I don't need my doctor and the orthopedist to use the same EHR, I just need their systems to talk to each other. The USB analogy is useful to illustrate what I mean. I can buy any brand USB keyboard--Logitech, Microsoft, Targus, Apple, etc.--and it will work with any computer that has a USB port. EHRs, however, have no such standard.

For this post, I just want to mention one disadvantage of competing standards: the orthopedist has to cover lots of things my primary care doc already covered. The redundancy is not just a hassle, it's costly and dangerous. The orthopedist and I both waste time and brain power that could otherwise spent on a more accurate diagnosis of my problem or on treating another patient. What's worse, if I'm unlucky enough to wind up in an emergency room, I may get treatment that fails to take into account important facts about my health.

More to come.

Studer and "Accountability"

Paul Levy, CEO of Beth Israel Deaconess Medical Center in Boston, has a post today on the problems of accountability. Levy cites a book by Quint Studer and argues that "it is not only impossible to hold people accountable in an organization, but trying to do so is a misallocation of managerial attention."

Readers of this blog may remember that I wrote two posts (Post #1 and Post #2) disputing the usefulness of another of Quint Studer's central ideas: hire (and keep) the right people (while getting rid of the wrong people, of course).

To be fair, Quint Studer neither owns this idea, nor is he an outlier in endorsing it. In fact, when I was still doing research at HBS, we found that hiring the right people was topmost in the minds of nearly every manager.

Likewise, holding people accountable is another cherished piece of conventional wisdom about leadership. Levy's post not only does a good job of laying it to rest, but provides four good questions that can substitute for all the hand-wringing managers do over accountability.

Confusing Mammogram Recommendations

This article from The Atlantic is the best explanation for what - to me, the casual news consumer - seemed like contradictory statements about revised mammogram guidelines.

I remember feeling like one day I read that the guidelines for routine mammogram screening had changed and then, the next day, I read that although the guidelines had changed, they hadn't. (It turns out, as you'll see below, my memory reversed the order of these events.)

Here is the link to the US Preventative Task Force's Recommendation (USPSTF), which revises the recommended age for routine mammograms from age 40 to age 50, published in the November 17, 2009 Annals of Internal Medicine. The day before, however, The American Cancer Society's Chief Medical Officer released a preemptive statement.

The American Cancer Society continues to recommend annual screening using mammography and clinical breast examination for all women beginning at age 40. Our experts make this recommendation having reviewed virtually all the same data reviewed by the USPSTF, but also additional data that the USPSTF did not consider.

So which is it? What are these "additional data"? And what's the harm of being extra cautious? Questions like these are what I think Crewdon does such a good job of clarifying.

Not Quite the Four Noble Truths

Buddhism has its four noble truths.

US Special Operations has its "four enduring truths," which I thought apply (mutatis mutandis) to leaders/managers.*

1. Humans are more important than hardware.
2. Quality is more important than quantity.
3. Special operations forces (SOF) cannot be mass produced.
4. Competent SOF cannot be created after emergencies occur.

Just a few notes. First, I'll be stealing this and replacing SOF with leaders. Second, notice that quantity matters, just not as much as quality. Finally, real leaders cannot be mass produced--no matter what HBS or Stanford or Wharton say. Creating a leader is a batch-of-one process.

*(With the wars that we're in, I'm reading Robert Kaplan's Imperial Grunts: On the Ground with the American Military, from Mongolia to the Philippines to Iraq and Beyond (Kindle Edition), where I found these stated.)

Computers Haven't (and Won't) Save Health Care

A study at Harvard Medical School concludes that--surprise!--computers don't save money. And they don't improve the quality of care.

The study's lead author, Dr. David Himmelstein, says that "computer systems are built for the accountants and managers and not built to help doctors, nurses and patients." This won't shock clinicians or anyone else who has seen these systems deployed. (See my earlier post on a related topic.) But it does run counter to conventional wisdom and the claims of some existing research on clinical information systems.

To evaluate the claims as well as previous research, Himmelstein and his co-authors looked at the data used to make the positive claims, as well as other, larger datasets. What they found is that there is no correlation between clinical information systems and providing high-quality or low-cost care.

Himmelstein again.

For 45 years or so, people have been claiming computers are going to save vast amounts of money and that the payoff was just around the corner.... So the first thing we need to do is stop claiming things there's no evidence for. It's based on vaporware and [hasn't been] shown to exist or shown to be true.

Does Apple Really Think they can "Disrupt" Health Care IT?

This article--their own press about the topic--makes me wonder if Apple isn't looking to get into radiology? Given that designers, filmmakers and others that need high quality images have been faithful to Apple for so long makes me wonder if radiology isn't the perfect inroad for them.

Strategy for an Uncertain Future

A colleague of mine was asked a really interesting question recently about strategic planning. I thought that his question and her response merited a blog entry on the topic.

Strategy and planning (hereafter, planning) ought to serve one of two ends, execution or learning, but not both. Choosing the right mode will save us a lot of pain. (So-called "strategic planning" is often a mixture--we figure out where we want to go and chart a course to get there--which, as experienced people can tell you, often yields mixed results.)

First, I think it is important to note that planning-to-execute is well suited to "linear" problems--relatively static current state, known input-output relationships, predictable future influences, etc. Planning-to-execute is also the default mode of planning for most organizations. (I think the reason for this is that as a species we are prone to underestimating uncertainty. But that's another subject for another entry.) In many organizations planning-to-execute, it is the only mode of planning.

My experience, though, is that most of the challenges organizations face are non-linear--sensitive to initial conditions, messy or unknown input-output relationships, uncertain future influences, etc. Thus, organizations would be better served if planning-to-learn was their default mode.

Boeing has been suffering over just this issue in recent years. After Airbus outdid Boeing in 2005, Boeing decided it would transform itself to become the first producer of "all-composite" commercial aircraft. ("All composite" is in quotes because in reality they merely inverted the ratio of aluminum and composites from roughly 50% aluminum/15% composites to 15% aluminum/50% composites.)

To plan for this transformation, Boeing used their familiar and very effective planning processes. The unfortunate result has been lots of trouble. Rather than drag you through all the details, I thought I'd cite one fact alone: the 787 was scheduled to enter service in May of LAST year, and currently looks as though it might enter service in Q4 of NEXT year.

I'm not trying to pick on Boeing here. They are, in fact, better at planning than most any organization around. In this case, though, they used the wrong process for their situation. Given the discontinuity between sourcing, designing, and actually making an "all-composite" airplane, they needed a different process for planning--one designed to discover what they didn't know, to clarify emerging problems, and to develop new plans based on recent learning.

The lesson I draw from their experience--and that of many others--is that we ought to choose the right tool for the right job. And because I believe that most of the challenges we face are less certain than we think, I also believe that our default mode ought to be planning-to-learn. After all, given our current proficiencies, when the need arises we can easily pull out our planning-to-execute toolbox.

Checking the Right Boxes but Failing the Patient

I enjoyed this piece from a doc at UC San Diego. Here's an excerpt.

None of these interventions [electronic records/prescriptions, pay-for-performance, etc.], however well meant, address a fundamental problem that is emerging in modern medicine: a change in focus from treating the patient toward satisfying the system. The effects of focusing physicians’ attention on benchmarks and check boxes are not, I think, to the patient’s advantage.

I have heard nearly those same words from many, many physicians. The essence is that they know things are changing, but they feel like they are losing the critical human-to-human interface. Most efforts to improve the quality and safety of health care have created an increasingly worrisome gap between doctor and patient.

No physician I know claims to be smarter than the volumes and volumes of clinical evidence available. Most feel like the well-meant (and often successful) interventions to improve patient care don't serve doctors' ability to help their patients, but require that doctors serve them. And in a world where doctor attention is finite, this matters.

The question, then, I see them raising is: couldn't these interventions have been (and can't future ones be) designed in such a way that they both improve the quality and safety of care, and also increase the amount of time and attention docs can give their patients, rather than putting the two imperatives at odds?