Abstract
What is an integrated health care system?
My answer is based on long and careful observation of Kaiser Permanente, and some observation of other integrated delivery systems (IDS). 1 First, integration is a culture based on shared values and goals, a culture of teamwork and mutual support, and a culture in which the Medical Group accepts a shared responsibility for the care of patients. The shared values include clinical excellence, continuous improvement, customer service and recognition that, one way or another, the costs of the Program are paid by its members so affordability is important. The goals are to be a clinically and financially successful program that also contributes to the health of the communities it serves. Teamwork means that all the doctors, generalists and specialists, as well as the other health professionals, are partners in the same enterprise, sharing the same premises, with specialists there to support the generalists with readily available advice and help.
Second, in an integrated system, the incentives of the providers and the needs and wants of the patients for high quality, affordable care, are aligned. What is good for the members is good for the Medical Group and vice versa. An important vehicle of this alignment is that all components of the program, particularly hospitals and medical groups are paid out of a common revenue stream, so the doctors have an interest in using hospitals wisely, including minimizing members’ need for hospital by strong outpatient programs.
The third is patient-centered care; that is, multidisciplinary teams focus on the best interests of each patient, and not on specialty or departmental interests. I think the fact that the program's revenue comes from satisfied patients who have a choice helps to keep people focused on patients.
Fourth, physician leadership is very important. Physicians are the natural leaders and physicians will not follow if their leaders are not effective and respected. Kaiser Permanente puts a great deal of resources into leadership development, including for physicians.
Fifth, integrated systems coordinate care across settings – home, doctors’ offices, outpatient, inpatient, skilled nursing facilities and rehabilitation centres – with a flow of all necessary information and careful attention to handovers, so that care can be delivered in the least costly appropriate setting.
Sixth, integrated systems practice ‘population medicine'; that is, they study their whole enrolled populations and prioritize preventive interventions where they will do the most good in terms of patients’ health outcomes. They integrate personal and public health, emphasizing health promotion and disease prevention. They also try to integrate somatic and mental health, though their mental health responsibilities generally do not extend to long-term institutional care.
Integrated systems have always had shared, longitudinal, comprehensive, medical records and in recent years, they have led the way in deploying electronic records which are turning out to offer a huge advantage in terms of cost, accuracy and timeliness of information transfer. Integration means one organization delivering all or most of a patient's medical care. Usually, there are exceptions for referrals outside the system for specialized care of kinds that the system is not staffed to provide, such as organ transplants. But in those cases, they work with the outside providers to organize smooth transfers to and from the referral center and transfers of information. Integrated systems usually have shared practice guidelines to keep physicians up to date and to reduce unwanted practice variations. Quality and effectiveness are likely to be higher when all members of the team are working to implement the same guidelines.
Integration is a matter of degree. Even the most extensively integrated systems, like Kaiser Permanente in California, refer outside and pay for services such as organ transplants. Some systems have integrated care, but not finances. The Geisinger Health System in Pennsylvania, for example, is generally paid fees for services, though their doctors are paid salaries.
System integration confers many benefits. First, IDS can be and are publicly accountable for cost and quality, an attribute very much needed by purchasers of health services. Generally, integrated systems lead in public quality reporting, though it is almost impossible to measure the quality of individual doctors in most cases. Sample sizes are too small. (It is possible for a few fairly discrete procedures like CABG.) Second, integrated systems use fewer resources than non-integrated models. The RAND health insurance experiment found that Group Health Cooperative of Puget Sound produced high quality care using 28% fewer resources than the surrounding uncoordinated, fee-for-service sector, and they did it in the absence of competition and cost-conscious customers. The actual cost differences will vary with market conditions.
Why have integrated delivery systems not spread?
If integrated delivery systems are such a good idea, why have they not spread to all or most of America? The history goes back to the late 19th century and much of the 20th century when the organized medical profession insisted on, and fought for, the traditional system of free choice of doctor, insurance plans that do not discriminate among doctors and uncoordinated, unmanaged fee-for-service solo or small practice. The American Medical Association and the County Medical Societies fought organized medicine using political action, ostracism, boycotts and other coercive tactics.2,3 Their preferred model is economically advantageous for doctors because it obviates economic responsibility for costs and prevents economic competition among doctors. 4 We still see traces of this in our present culture – the idea that ‘if it costs less, it must be inferior'. It took very courageous and determined doctors to buck this force, but a few did.
This political resistance was reflected in the design of the Medicare program, the Federal Government's health insurance program for the aged and disabled. It was based on fee-for-service, the dominant paradigm of the 1960s. If President Johnson had tried to include alternative models of payment, he doubtless would have intensified the opposition of the medical profession. The experience of the past 40 years has shown that Medicare is extremely difficult to change. Too many doctors and others have too much financial interest in the present inflationary system. There have been proposals to reform Medicare in ways that would permit IDS to compete, including the Bipartisan Commission on Medicare in the late 1990s, led by Senator John Breaux and Congressman Bill Thomas. They proposed to replace fee-for-service Medicare with what they called ‘premium support payments’ in a cost-conscious consumer choice model that would permit real competition and value-for-money. This good idea was not accepted at the time, I think, for short-term, political reasons. But something like it is almost inevitable if the Federal Government is to avoid unprecedented tax levels or soaring fiscal deficits.
A more serious barrier today is employment-based health insurance (EBI) which covers most Americans. American employers have Balkanized the market for health care and have proved themselves completely incapable of creating a competitive market. Let me repeat for emphasis: in America we do not have a competitive market in the usual sense of informed, cost-conscious, consumer choice. The key problem with EBI is that the great majority of employers do not offer their employees cost-conscious choices of delivery system. If they do offer choices, as some do, they offer few choices, creating a huge entry barrier for new systems that would like to form and compete. Most employers consider themselves too small to offer choices, so they lock their employees into fee-for-service. Insurance companies resist competing within employment groups and they offer employers lower premiums if they can remain the sole source of insurance for the employer. Some larger employers do offer some choices, but they pay some 80–100% of the premium of the plan of the employee's choice, thus blocking the ability of the IDS to market its superior value-for-money. Ideally, we ought to have a much more open market in which any delivery system meeting basic standards of quality and coverage could enter the market and compete, so that existing systems would face competition from innovators.
Some very large employers, such as the States and the Federal Government, and some universities that have a large work force in one location, do offer employees a reasonably wide choice of health insurance plans, including some that are parts of integrated delivery systems, and the employers contribute a fixed dollar amount toward purchase of the plan of the employee's choice, so that the employee choosing a less costly plan gets to keep all of the savings. Integrated delivery systems do very well in these models. For example, 81% of the employees of the University of California and of Stanford University have chosen HMOs based largely on multispecialty group practices; 94% of the employees of the State of Wisconsin have done the same. It is reasonable to suppose that if everyone had such choices, even higher percentages would choose integrated delivery systems because they would be in more locations convenient to where more people live. In any case, it is not necessary for 100% to choose IDS. On the contrary, responsible consumer preferences should be respected.
Could the NHS develop integrated health care delivery?
How could the NHS develop more integrated health care delivery, meaning mainly the close collaboration of primary care doctors and specialists? In my 1998 visit to the NHS, I visited some total purchasing pilots (TPPs) in which fund-holding GPs had pooled their resources in order to create capabilities that could supply more appropriate care for their patients at less cost. One thing that was particularly good about this was that it was a voluntary association in which like-minded GPs could get together to do the right thing. They did not have to wait for the approval of GPs who did not share their vision. And the management and organization of their models grew organically. I recall thinking that these TPPs could become the foundations for integrated delivery systems if they progressively included more and more secondary care specialists who wanted to work with them.
I was very disappointed to see fund holding and TPPs abolished for partisan political reasons in 1997 and then the creation of required Primary Care Trusts for all GPs in England. This move was counter-productive because whole new management systems had to be invented, and because it was not organic and voluntary.
I strongly agree with several insights offered by Chris Ham in relation to England's NHS (Scotland, Wales and Northern Ireland have gone a different route): ‘Policy makers need to resist the temptation to prescribe a single approach and to focus instead on encouraging the development of integrated care using the means that appear most appropriate in different contexts'. 5 ‘Primarily [integration] is about relationships between people. These relationships are not informal friendships. They have to be worked on and built professionally if clinical integration is to be meaningful and sustained through good and bad times.' 6 Like Professor Ham, I would ‘encourage emergence of integrated systems, based on networks of like-minded GPs working together to provide and commission care for the populations they serve. This would be like the total purchasing pilots…but could go further…' 7
The question of the day in England is whether or not there must be competition among IDS. I think competition is important. Kaiser Permanente's leaders and staff know that their revenues come from satisfied members who have a choice. 8 Their many service improvements were driven by competition. Much of this would not have happened, or not have happened so fast and so well, without it. Among the improvements of the past decade are: ‘open access’ or same-day appointments; convenient, urgent care for after-hours problems; secure e-mail with a personal physician and usually prompt replies; ability to make appointments by e-mail; laboratory tests sent the next day by secure e-mail; electronic prescribing, including pharmacies on premises (the doctors are able to tell their patients, ‘I am now printing the label, pick up the prescription at the pharmacy on your way out of the medical centre’); and electronic health records available on a secure web site. The physicians have always been salaried, plus a small bonus related to their ability to manage hospital costs. Some years ago, they modified the formula to include measured patient satisfaction, measured by a follow-up questionnaire soon after the patient visits a doctor. They also measure teamwork and reward it.
I appreciate there are thinly populated parts of England where competing PCTs would not be possible. But there are many large, urbanized areas where competing PCTs would be possible. The performance and innovations of the PCTs in the competitive areas could be used as benchmarks for the non-competitive areas. It does not have to be either/or for the whole country.
Conclusion
Many patients, including the most costly ones, present with multiple complaints and symptoms resulting from a multiplicity of chronic (long-term) conditions. This makes it all the more important that the knowledge and perspectives of the different specialties be brought to bear on the patient's problems in a seamless and economical way, and a treatment plan be designed and executed with these multispecialty perspectives integrated into a coherent and practical whole. Clinical integration is an important advance over the uncoordinated care typical of traditional models.