by Thomas H Lee
Attacking the measures can occur in several ways:
• Focusing the measures on processes rather than actual patient outcomes. The rationale for this approach is that many factors affect patients’ actual outcomes, including how sick they are and whether they take their medications and adopt healthy lifestyles. In any business, the ultimate focus should be on the customer’s outcome. But when the short-term focus is on passing judgment on clinicians, those clinicians want to be evaluated on the basis of processes within their control (e.g., prescribing a beta blocker for patients who have had a myocardial infarction). They also want those processes to be supported by evidence from rigorous research studies. The result is often a focus on process measures that do not differentiate among clinicians because everyone does them with a high rate of reliability.
• Shifting accountability from the group to individuals. Clinicians want to do a good job, but they see their jobs as individual efforts, not a team sport. They want feedback on issues for which they alone are responsible so that they can control their own fates. The problem with this perspective is that when many clinicians are involved in a patient’s care, no one accepts full responsibility for improvement.
• Identifying situations in which the measure could lead to unintended consequences. For example, critics of measurement often hypothesize about what would happen if 100 percent compliance with the behavior supported by the measure were achieved and recall individual patients who might have been injured if such perfection had been the result of the use of a measure. They articulate fears that physicians might avoid caring for the sickest patients or ignore patients who are not the focus of the measures. For example, they might contend that diabetes measures could cause physicians to ignore nondiabetics.
After poking holes in the measures, clinicians who are unhappy with performance measurement turn to the data used to bring measures to life. Again, the reality is that the data and the analyses are never perfect and cannot be used to prove most issues beyond a shadow of reasonable doubt. The data are rarely complete; for example, patient experience surveys tend to be completed by only about 20 percent of patients. Patients cannot be compelled to complete surveys, and there is always speculation about the differences between those who do and those who don’t.
Even if 100 percent of patients completed surveys, populations of patients vary in ways that affect the outcome. It is possible to adjust for some of the factors that influence patient experience and other outcomes, but it is impossible to know if every important difference in populations has been taken into account. Of course, collecting and managing the data needed for risk adjustment requires work and time. The bottom line: in my long management career, I have spent many hours and dollars on risk adjustment, but I don’t think I have made anyone happy—less angry, perhaps, but not happy.
A New Measurement Goal
But what if the goal of measurement is improvement in the pursuit of a noble shared purpose rather than punishment? In that case, many of the issues discussed above are viewed in a different light if not dismissed altogether. If the real goal is to be better next year than you are now, risk adjustment is not a stumbling block, assuming that your patient population is not changing in a major way. You are comparing yourself to yourself; no risk adjustment is needed.
The data sampling may be far from complete, but if there is no systematic difference in the way patients are identified and contacted, the populations are likely to be similar. Someone has to prove that they are different before anyone can blame changes (or lack of change) on differences in the population.
Measures may have inherent potential flaws, but if the goal is improvement, it is clear that not measuring is a terrible strategy. I like to remind physicians that we have to make decisions for patients with imperfect data all the time. There are some physicians who are paralyzed by uncertainty and cannot make a decision without ordering one more test, then another test after that, and so on. We know that uncertainty will always be with us and that the physicians who cannot act because of it are not the doctors we ourselves want to go to.
The same is true for improvement of healthcare delivery. The measures and data will never be perfect. We should make them as complete and as good as we can, but at a certain point, if the goal is improvement, we have to use the data that are available and try to get better.
When the goal is improvement rather than judgment, providers can use measures for which 100 percent is not the ideal target. In fact, measures—and the variation that they reveal—are most interesting and have the most impact when they capture the tension between two competing values. For example, when surgeons decide whether to operate on a very sick patient, they have to weigh short-term risks (the chances that the operation might cause the patient to die or have other immediate complications) against the potential long-term benefits (the chances that the patient will live a longer or healthier life because of the operation).
It is obvious that you don’t want to choose a cardiac surgeon with a mortality rate of 100 percent. It is less obvious that you probably don’t want a cardiac surgeon who has a mortality rate of zero—no patient ever dies—because that surgeon is probably emphasizing short-term safety over long-term outcome. That risk-averse surgeon may be declining to operate on patients who have a high risk of complications with surgery but who have an even greater risk of dying without the operation. The best surgical mortality rate might be somewhere close to the middle of the pack.
Here is an analogous example: Among patients with pain, it is obviously not a good idea to give narcotics to every patient, but it is also a bad idea to never give patients narcotics. If a clinician learns that his or her prescribing habits or his or her patients’ reports of pain control are vastly different from those of other clinicians, that clinician might have something to learn from colleagues.
For clinicians to perform well enough to meet some minimal standard, measures and data must sometimes be used for judgment. However, to drive improvement so that organizations can respond to competition, clinicians must have a different perspective. They have to believe that the goal is improvement. And the only way that they will give their all to the goal of improvement is if they share a noble purpose.
The Importance of Shared Purpose
Clinicians are willing to act on imperfect data and use measures for which 100 percent is not the ideal target when there is a compelling goal, such as the care of a patient. For healthcare organizations, there needs to be a similarly compelling shared purpose to put performance measures and data in perspective. If there is a goal for an organization that is as compelling as saving a patient’s life, personnel are unlikely to get stuck reviewing measures and data for reasons not to believe their implications. Instead, they will assess the best data available to them and try to get better.
That is why when one is considering the four models for social action from the work of Max Weber, the first that should be addressed is shared purpose. Shared purpose is different from the other models in that it is created through narrative, not data. It is developed through the stories that cause providers to feel pride or shame. These stories can be deconstructed, and the factors that provide intrinsic motivation to clinicians can be identified. Measures and data can then be used—despite their imperfections—via Weber’s other models for social action in pursuit of this shared purpose.
How does an organization develop a shared purpose? As Toby Cosgrove and I wrote in our 2014 Harvard Business Review article, the creation of shared purpose begins with the same steps used to build consensus in any organization: listening, demonstrating respect for different views, and creating processes through which everyone can be heard and contribute. However, healthcare has special challenges: the diversity of types of work to be done, the training of the personnel, and the organizational structures within which they work (e.g., employed versus affiliated physicians). Shared purpose has to transcend all the factors that divide provider personnel; after all, that’s what bein
g shared is about.
These discussions have to acknowledge that we have a natural tendency to dwell on our problems, such as rising costs, lack of access, and disappointing quality. We don’t give much time or energy to what we are trying to create or the possibility of a better future. But to help our colleagues move beyond grief and anger about what they might be losing as their organizations respond to the market’s imperatives for quality and efficiency, leaders have to shift the conversation to something different: something positive, something noble, something that transcends the concerns of any of us as individuals.
Only one goal meets those criteria: improving care for patients, which means meeting their needs and improving their outcomes. That focus has to be the core of any change agenda that demands sacrifice from clinicians.
At the same time, leaders must acknowledge the need for sacrifices. Individual clinicians will lose some personal autonomy as they work in effective teams, and that may or may not be offset by an increase in the group’s autonomy. There may well be losses of income. No leader should make promises that the status quo will be preserved and expect to have more than a few years before those words come back to haunt him or her. No leader should say that the organization will prioritize doctors’ interests over those of patients. Instead, leaders should emphasize that the journey will be difficult but the goal of high-value care is central to the professional pride of the organization as well as business viability.
Skeptical readers may question whether clinical personnel are too jaded by modern times to respond to lofty visions. I think that the opposite is true and that nurses and physicians in particular are driven by intrinsic motivations. The most vivid examples occur during catastrophes. During Hurricane Sandy and the Boston Marathon bombings, no doctors or nurses worried about compensation. In fact, the clinical personnel who were unable to get involved and directly contribute in some way were disappointed. They knew that they had missed out on something good, one of the rewards of being in medicine that transcend money.
What about when the catastrophe is longer term, playing out over years or more? Healthcare personnel can still rise to the occasion. One of my favorite examples is the campaign to save Grady Health System in Atlanta, which was losing millions weekly and was on the verge of bankruptcy for at least half of the decade 2000–2010. The question being asked about Grady, along with other safety-net hospitals around the country, was often not whether it would collapse but when and how and what would happen next.
Leaders from Grady and local government conceived a brilliant campaign with the tagline “Atlanta Can’t Live Without Grady.” One of its leaders, Matt Gove, told me how that campaign grew out of the need to communicate what was obvious to anyone who walked the halls of Grady or the streets around it: it would be a disaster for hundreds of thousands of people if Grady further deteriorated or closed.
The campaign started simply with photos of patients or prominent Atlantans with the Grady logo over their hearts under the “Atlanta Can’t Live Without Grady” tagline. Over time, the campaign evolved to include detailed and moving stories of people who were saved at Grady, each one with the quote “I Wouldn’t Be Here Without Grady.”
Can an organization be rescued by inspiration? A lot of other types of work were going on at the same time, of course, and help came from the outside, in part because of the campaign. Grady received more than $300 million in philanthropic commitments, and the state passed a hospital-provider fee that helped support institutions like Grady that serve large numbers of Medicaid patients. Clearly, foundations and politicians got the message that Atlanta couldn’t live without Grady.
Working a thousand miles to the north in Boston, I heard about the Grady campaign because I learned that it changed the culture of the hospital, that morale improved and personnel turnover declined. Again, many factors surely played a role besides the articulation of a shared vision of what Grady meant to Atlanta. But the fact is that Grady is still there, and I am sure the campaign had something to do with it. The relevant question at many safety-net institutions is not whether an organization can be rescued by inspiration; it’s whether an organization can be rescued without it.
Adding “Shared” to the Purpose
One basic problem that healthcare leaders often describe in creating a sense of shared purpose is the “shared” part. Physicians in particular, they say, don’t want to work in teams. They don’t see themselves as part of an organization even if they are employees of a hospital or healthcare system or their own physician group. There is a fantasy among some that employed doctors follow orders. Anyone who has worked in an organization with employed doctors knows that they are doctors first and employees second—a distant second.
As doctors, they view their duty to patients as preempting other obligations. This next statement is a broad generalization to which some exceptions can probably be readily identified, but here goes: my strong belief is that the vast majority of physicians won’t do things that they know are counter to their patients’ interests no matter how great the financial incentives are. I am not naive; if financial incentives nudge them in that direction, physicians may talk themselves into believing that some tests or treatments might be beneficial for patients even if the evidence is thin, even lame. But their ego identity as advocates for patients is a more powerful force than financial incentives in many settings. Major financial incentives to do things that have no connection to meeting patients’ needs tend to have mixed effectiveness at best, but it takes only modest incentives to nudge clinicians to do something that they know to be the right thing for patients. Often it takes no financial incentive at all.
The implication of this dynamic is that doctors’ and nurses’ intrinsic motivations provide a path to meaningful change. The leaders of healthcare organizations can engage their personnel by putting the focus on patients and their suffering, which every clinician agrees trumps all other concerns on a short-term basis. The goal is to make that prioritization endure, for every patient, on the weekends as well as the weekdays, in the middle of the night, throughout the year.
The creation of a shared purpose begins with stories, not data. These are the stories that crystallize how providers want to see themselves. They are the cases that bring providers pride or bring them shame. Accordingly, discussions with clinicians about reorganizing care should never begin with talk of contracts and compensation or numerical trends in issues that seem related to excellent patient care only indirectly. For example, waiting time for patients is an important business issue for every healthcare organization and a major source of annoyance and anxiety to patients. But clinicians tend to think of that issue as something that is not as important as doing as good a job with the patient right in front of them and do not readily see how that complex issue is under their control.
Sometimes a single case is sufficient to galvanize clinicians. In 2008, a patient called Cleveland Clinic’s urology department and asked for an appointment. His problem was difficulty urinating, and he was given the next available appointment, which was two weeks away. That evening, he was in the emergency department because he could not urinate at all. The patient was not permanently harmed, but he certainly experienced several hours of exquisite discomfort.
Cleveland Clinic’s leaders discussed the case, and one of them asked, “Do we want to be the type of organization that doesn’t even try to figure out if patients should be seen right away?” With that perspective and with the embarrassment resulting from one case, they decided that their current appointment system was intolerable and they had to try to do better. Cleveland Clinic ultimately introduced a same-day appointment policy in which every patient who calls for an appointment is asked if he or she would like to be seen that day. If patients do, they may have to drive an hour to an available clinician, but they are almost always seen. Cleveland Clinic now handles about 1 million of its 5.5 million annual patient visits on the same day, and this policy has helped the clinic expand its market share. I
t was not a marketing executive’s brainchild; it came from clinicians’ resolve to do better over a single case.
To engage clinicians and other personnel, statements of shared purpose must be simple and direct. An example that I admire is Cleveland Clinic’s simple statement “Patients First.” Distributing 40,000 lapel buttons with those two words was one of the first acts by Toby Cosgrove when he became CEO in 2004. Many other organizations had considered that kind of statement before and since and struggled with the implications (“Doctors Second”). But addressing rather than ignoring the implications is the reason such a shared purpose is so critical to business success.
After all, one of the key lessons about strategy described in Chapter 2 is that organizations have to have clarity on their definition of success. They have to know the goal. A mission statement that reflects multiple purposes, all of which are noble (research, teaching, and patient care is the classic three-part mission for academic medical centers), sounds good but does not provide a framework for guiding tough decisions. Organizations that want to survive and thrive in a healthcare market driven by competition on value have to make decisions.
As forward-looking organizations recognize the need to make choices and articulate a clear top priority, the needs of patients always come first. It is a dynamic that makes me proud to be in healthcare, in which we have lots of good people who want to do good things for patients. That dynamic also happens to be good business thinking: meeting the needs of patients as well and as efficiently as possible (i.e., value) is the one goal that every stakeholder in healthcare supports.