Skip to main content
Management in Practice

Does our health system deliver value?

Competitive strategy expert Michael Porter, the Bishop William Lawrence University Professor at Harvard Business School, has shaken up the thinking around healthcare reform with Redefining Health Care, a book he coauthored with Elizabeth Teisberg, associate professor at the Darden School of Business. Porter talked with Q3 about his ideas on how to bring the right kind of competition to healthcare and developments since the 2006 release of his book.

  • Michael Porter
    Bishop William Lawrence University Professor, Harvard Business School

Q: Much of the national debate around healthcare revolves around issues of access. Your work focuses more on value in healthcare. What do you mean by value? Why is it the key issue?

Access is critical. We believe universal insurance is not just a matter of equity but also a matter of efficiency and effectiveness. In a system without universal coverage, a number of factors — cross subsidies, lack of preventative care, and patients presenting too late for care in expensive settings — lead to poor health outcomes and inefficient care.

However, the real challenge is not just access but dramatically improving the value of healthcare delivery. Value is defined as the patient health outcomes per dollar spent. Unless value substantially improves, the system will collapse under the weight of an aging population. Also, thinking about value is important because one of the most powerful ways to reduce cost is to drive up quality. In healthcare, better quality actually lowers cost, certainly in the long run and often even in the short run. Hence improving value, not cutting costs, must be the goal.

The essential question is how do we rethink the way we deliver, measure, and pay for care in order to align it better with value?

Q: Your book describes how healthcare delivery needs to be restructured. Why?

Today healthcare is organized and structured around interventions and specialties. Care is a series of interventions in which the patient goes from department to department. But value is not created by a single intervention or single specialty but is the collective outcome of the entire process of care for a medical condition. A medical condition is an interrelated set of patient circumstances best addressed in an integrated way. A medical condition includes common co-occurrences. Diabetes, for example, often gives rise to eye issues, vascular issues, and kidney issues.

Care for a medical condition should be organized into integrated practice units (IPUs), which include all the necessary skills and specialties. The IPU model involves integration not only across specialties, but also across the cycle of care. For any medical condition such as breast cancer, there is a cycle of care that involves a long series of activities stretching across months and years. In the breast cancer example, surgery may be one part of the cycle, but it is a small part in spite of the attention that it receives in the current system. Under an IPU model, doctors will shift from practicing a specialty to organizing around medical conditions, which in turn will move affiliations away from traditional departments toward the network of health practitioners who are jointly responsible for care cycles.

True structural change is necessary in healthcare delivery, not incremental improvements to the current system. A number of organizations are applying lean thinking or Six Sigma approaches to healthcare [See an article in Q3 about one such effort]. However, if you don’t change the organization away from departments and specialties, process improvement cannot make much of a difference.

Q: You argue that mandatory measurement and reporting of results is perhaps the single most important step. Could you explain that?

Healthcare is suffering from a catastrophic absence of results measurement, which is what we should really care about. Today, we add up the total cost of healthcare but can’t relate these costs to the outcomes achieved. If value is the goal, we need to measure how much value we are actually delivering.

There is increasing attention to measuring “quality” but quality is often defined incorrectly. The only true quality is outcomes: How well does the patient do? But most quality efforts in healthcare today are focused on the process of care instead of outcomes. Does the caregiver follow evidence-based guidelines? How pleasant is the service experience?

Process measurement can be beneficial, especially within provider organizations, but it is a dangerous way to control the system. Processes of care are complex and need to be tailored to patients’ circumstances. There is constant innovation. So a focus on process compliance creates serious risks of producing sub-optimal care and retarding innovation. Rather than try to improve healthcare by controlling the way doctors practice medicine, we need to measure how well patients do. This is the only way to truly equip doctors with the knowledge and motivation they need to improve.

Outcome measurement needs to relate to the unit of value. Today, outcome measurement tends to focus on discrete interventions, not on the whole care cycle for the patient’s medical condition.

Universal outcome measurement already occurs in limited areas. For example, in organ transplantation and in vitro fertilization, there are universal outcome measurement systems. In both cases, these grew out of federal legislation designed to address scandals or abuses. These cases have proven that outcomes measurement is feasible, and that it is a powerful force in driving value improvement.

I tell physicians that failing to measure outcomes is the greatest self-inflicted wound of the medical profession. How can you get better if you don’t measure how well you do? If you don’t measure outcomes, how can you resist second-guessing and intrusive oversight? You are simply inviting administrative micromanagement by health plans and outside parties. Hopefully more and more physicians will start to understand that outcome measurement is something that they need to lead, rather than resist.

Q: What is the role of competition in healthcare?

There is a lot of competition in healthcare, especially in the U.S. Unfortunately, we have the wrong kind of competition on the wrong things. The current system is characterized by what Elizabeth Teisberg and I call zero-sum competition. That is, competition to divide value rather than a competition to increase value. For example, when a patient is required to buy a consumer-driven health plan for which they have to pay a substantially higher percentage of the cost, that actually creates little value but shifts cost from the health plan and the employer to the consumer. That is zero-sum competition, not positive-sum or value-creating competition. When a health plan cuts a doctor’s payment for performing a particular procedure by 10%, that just re-divides value between the provider and the health plan. Competition in healthcare has a bad name because it is not about value for the patient. Worse yet, cost-shifting, restricting services, and using bargaining power actually destroy value, driving up costs and worsening outcomes.

Competition is a very powerful tool for improving value, particularly in a complex service like healthcare. But we have to structure the competition to compete on value. Pay for performance, the latest effort to utilize incentives to improve healthcare, is designed to reward physicians who hit certain benchmarks with a higher price for their services. But those benchmarks are the same old process benchmarks. “Did you prescribe aspirin at the right time?” rather than “How did the patient do?” That is not the way to introduce competition.

The best way to introduce competition is competition for patients based on results. Providers that can demonstrate excellent results in particular medical conditions should be rewarded with more patients. Those organizations will be able to grow, gain market share in the medical condition in their region, and open multiple locations across regions, as is now beginning for some truly excellent providers. This will substantially increase value in the system.

Today we have hyper-fragmentation in healthcare delivery. There are thousands of hospitals that all provide virtually the same services. They do so without adequate scale, without dedicated teams, and without adequate facilities, and many don’t produce very good results.

Value-based competition would almost inevitably lead to a reduction in the number of providers offering care for each medical condition. One of the outcomes of value-based competition would be greater volume and scale per provider in each medical condition, allowing greater resources and expertise in diabetes care, or migraine care, or breast cancer care, or whatever the medical condition.

Q: When many people hear competition, they immediately think of a consumer-driven model: Give the information to the consumer and they effortlessly migrate to the best option.


I wish it were that easy. If it was, we would have fixed the healthcare system a long time ago. I have serious concerns about the consumer shopping metaphor. Without the right information and the right organization of care-delivery, it is unrealistic to believe that consumers could or should direct their own care.

While the patient has to be an active participant in their care, the way care is delivered and paid for must fundamentally change. Consumer-driven healthcare also has a dark side, which is cover for cost-shifting. Many consumer-driven plans are the same old health plan except that now the consumer has to pay higher deductibles or higher co-pays.

Q: What would the mechanisms be for helping consumers choose the providers offering the best value?

We think that the patient’s doctor, whether this is a primary-care physician or specialist, is much better equipped to play the role of coordinating or directing care than the consumer. But the patient’s doctor lacks the information to play that role well today.

Outcome measurement will help physicians practice medicine better, but also help the referring physicians refer better. Only slowly will patients actually use the data.

The patients need to be as informed as possible and involved in their care. They need to adhere to their medication, do their exercises, and prepare for their surgery. In healthcare, the product is co-produced by the doctor and the patient. However, the system is not currently structured to actually engage patients. A focus on adherence to care and measuring adherence is all but missing in the discussion of the consumer’s role in healthcare, which I find unfortunate.

Q: Are there forces moving us toward a higher-value system?

Today things are getting sufficiently bad that most agree that change is needed. Even physicians are finding the current system so uncomfortable and unsatisfying that they are open and receptive. But we are still having a big debate over what the right changes are.

The employer community is riveted on change. They are now beginning to understand that the cost of poor health is much greater than the cost of health benefits. So, rather than minimizing the cost of health benefits, they ought to think about maximizing the health of their workforce.

Many providers are starting to change the way they deliver care. At Harvard Business School we have been developing a body of case studies on organizations like the Cleveland Clinic, M.D. Anderson Cancer Center, and the Joslin Diabetes Center that are embracing fundamentally value-driven models.

Unfortunately, we still are not seeing enough progress on outcome measurement. There are too many bottom-up efforts that are creating confusion and duplication. We need some national leadership which is not yet forthcoming.

There are some promising signs from the federal government. One of the principles in value-driven competition is to align reimbursement with value. You don’t pay for interventions; you pay for care cycles. You don’t pay the surgeon separately from the anesthesiologist; you pay for the bundle of services that addresses the patient’s medical condition. There have been increasing discussions about experiments to better align reimbursement with value and experiment with bundled reimbursement models.

The health plan community is changing, but much too slowly. There are not enough health plans being proactive in new reimbursement models, new measurement approaches, or new contracting relationships.

My greatest optimism comes from the sheer number of experiments and initiatives that are starting to address value, not just the same old quick fixes and mindless cost-cutting.

There were initially some strong critics of the value-based approach who argued for a consumer-driven or pay-for-performance model. The value-based perspective is increasingly accepted by the leading institutions. It is really a matter of accelerating actual change in the organizations.

Q: Has your perspective changed on any issues since the book has been out?

Our conviction has only grown, but we have placed more and more focus on implementation.

One of the most common criticisms of our work is that patients don’t have just one medical problem, but often multiple problems. The integrated practice unit model of care for diabetes, or breast cancer, or any other medical condition, as we described in detail in our book, includes the common co-occurrences for that condition. A diabetes IPU has not just endocrinologists but nephrologists, cardiologists, vascular surgeons, ophthalmologists, and others in a single care organization because many diabetes patients have these issues, which are directly connected to their diabetes.

What if this diabetic patient gets cancer? Well, this patient will also need to be cared for by a cancer IPU for whatever cancer he or she has contracted. There will need to be coordination between the diabetes unit and the cancer unit, but that coordination is vastly easier in the integrated practice structure than it is today where the patient has 20 different doctors in 20 different departments performing hundreds of separate interventions in a sequential process. If the patient was cared for by an integrated diabetes unit, then the team captain for the patient’s diabetes care would be aware that the patient has cancer and would be able to efficiently coordinate with the team for the patient’s cancer care. Today, coordination of care is poor. The IPU model will help, not hurt.

I don’t want to sound like we know everything, because we are learning every day. We have sharpened our thinking, but are only more convinced today that a value-based approach to healthcare delivery is actionable, and achievable, and that excellent results will follow.

Q: It seems like one challenge is coordinating the bottom-up micro changes in delivery that require some top-down macro-level changes. How are you working on that?

I believe that we need to start with the practitioners, the people on the ground who are delivering care. How could the hospital be organized better? How should we measure outcomes in that condition? Elizabeth Teisberg and I each devote the majority of our time to practitioners, whether they be health plans or providers.

Some policy changes could help to speed the rate of change, such as in outcome measurement, Stark laws, corporate practice of medicine laws, IT standards, reimbursement rules, and Medicare reform. We are stepping up our involvement at the policy level. So far in this election cycle, the discussion is still primarily about insurance. We believe that the fundamental problem is delivery. There are signs that the debate could be changing, so we will be ready.

Q: Does your work translate to other countries?

In advanced economies, a number of countries have universal coverage and good access to primary care. But surprisingly, the delivery issues are pretty much the same everywhere. We have done extensive work on Sweden, Finland, Germany, Holland, Japan, and Taiwan, among others. Organization around specialties and interventions, lack of care-cycle thinking, lack of really integrated care, failure to measure outcomes, and misdesigned reimbursement for providers are typical everywhere.

In developing countries, I have done a substantial amount of work with Jim Yong Kim and Paul Farmer here at Harvard Medical School to examine healthcare delivery in resource-poor settings, which involves additional issues and challenges.

It is striking that what little work there is on comparative healthcare systems tends to be at a very high level of abstraction, using very broad metrics. There have been few studies that look at the guts of care delivery.

By studying other national systems, we expose areas where a country has advanced. For example, in Germany, a change in the law has given rise to a number of interesting integrated practice-unit-type structures. We are in the process of pulling together a collection of papers about the non-U.S. systems that synthesize what we have learned.

For more on this topic, read an essay on comparative international healthcare from Jennifer Baron '05, a senior researcher at the Harvard Business School's Institute for Strategy and Competitiveness.

Interview conducted and edited by Ted O’Callahan