Accountable care organisations in the United States have piloted a physician led model of care that improves quality, costs and outcomes, writes Alison Moore
The idea of one organisation holding a budget for the whole of a defined group of patients’ care and working with other providers to drive cost and quality improvement is not new. But over the last few years there has been increasing interest in both the UK and the US about how this model can work in practice.
‘While the initial hospitalisation may have been surgical, what drove the readmission was medical and really was better managed by the primary care physician’
In the US, accountable care organisations have been reporting benefits from models that focus on promoting healthy lifestyles and prevention, early intervention if chronic diseases develop, and case management of the most complex and costly patients.
They are usually provider led and in the US have strong physician leadership - the person at the top has responsibility for both quality and financial outcomes across a patient population.
- Neil McKay on effective leasership for change
- Chris Ham on ‘bilingual’ clinical leaders
- Read more in this series about delivering lasting change within the NHS
Benefits can include reduced spending on secondary care, decreased unjustified variation in patient care and improved outcomes for patients.
One of the leaders in this has been Geisinger - a not-for-profit organisation that offers health insurance, as well as running primary care facilities and hospitals in Pennsylvania. It was one of the organisations accepted to run a pilot ACO under the Affordable Care Act.
Geisinger’s approach to providing and funding healthcare had for some years focused on integration and reducing “downstream” costs.
Geisinger Health Systems was founded by Abigail Geisinger. The daughter of a wagon maker, she married an iron magnate, who funded a hospital in Danville, Pennsylvania. This opened in 1915 with a young doctor, Harold Foss, as its first surgeon. He remained with the company until 1958, by which point the hospital had 300 beds, 500 employees and admitted more than 11,000 patients a year.
Since then Geisinger has expanded enormously. It now provides medical care to a population of 2.6 million and has 448,000 members of its healthcare plan. It has an integrated approach, and runs facilities from primary care clinics through to a teaching hospital. Through xG Health Solutions, it helps other healthcare providers adopt some of its innovative approaches. It is nationally and internationally recognised for its integrated approach, quality and service.
A key component was the development of the Proven Health Navigator - a model of enhanced primary care that communicates and shares decision making with patients and those around them.
Its other four elements are: population management; a “medical neighbourhood” which offers care in the most appropriate setting; measuring quality and cost outcomes; and a reimbursement system that is values based.
Moving to this model of care has been challenging, and required significant changes in working practice and payment systems. Geisinger has achieved this shift by concentrating on clinicians’ values.
“We have committed to focusing on quality - physicians and nurses get behind that,” says Dr Thomas Graf, Geisinger’s chief medical officer for population health and longitudinal care service lines.
“The first thing is to tease out what is a real quality concern [for healthcare professionals] and what is another concern disguised as a quality concern.
“In our scheme the approach has not been that we are going to reduce costs or to focus on them, but we are going to improve quality - 90 per cent of the time that improves costs as well.
“At some point we will run out of quality improvements, which will also reduce costs but I think that is a long way off.”
Culture is crucial, adds Janet Tomcavage, who leads value based care strategic initiatives for Geisinger.
“The culture of the organisation that we are working with needs to be a culture that is willing to change and is willing to think differently about how they manage people, patients and the population, and the accountability they have.”
An example of this is when a patient with comorbidities leaves hospital, she suggests. Does primary care re-engage with them as soon as possible? Geisinger patients are offered an appointment in the first few days with the aim of avoiding readmissions.
“The idea that primary care is responsible for care post-surgical admittance was a new one. It had physicians saying: ‘Why do I have to see them post-surgery?’” says Ms Tomcavage. “If you have a hip replacement, but also [chronic obstructive pulmonary disease] or heart failure, the stress of that can impact significantly on the co-morbid conditions. The primary care provider is best suited to evaluate and manage this.
“Primary care providers soon realised that, while the initial hospitalisation may have been surgical, what drove the readmission was medical and really was better managed by the primary care physician rather than the surgeon.”
A second point is physician leadership: doctors are more likely to be found in overall leadership roles than in the UK. Chief executive of Geisinger Glenn Steele is a doctor, for example. But Ms Tomcavage - whose background is nursing - also stresses the importance of teams at the operational level as the system moves from a physician delivered model to a physician led but team delivered approach. Physicians have two roles, she suggests, “complex medical decision making and patient relationships, but it takes a team to manage a population”.
One of the initial investments Geisinger made was in case management: nurses who would work with the most complex and challenging patients identified as at greatest risk of unplanned admissions. It provides one such nurse for every 1,000 Medicare (older people) or 3,500 Medicaid patients (Medicaid is the government funded healthcare programme for people on low incomes). Patients are given a number to call if they start to deteriorate and the case manager will work with them in whatever setting they are in.
Geisinger has also pioneered the use of data to drive improvements and changes in medical practice. But Dr Graf points out what is needed is data that can then be used by doctors and nurses to drive meaningful change. He sees data as an accelerator, not an end in itself.
“I would emphasise the change in people systems as being more important,” he says. For example, Geisinger found that fewer than one in 10 patients were getting all the preventative interventions available. Clinicians were shocked by this and got involved in redesigning the system to increase that number to more than 80 per cent. This has been done in a way that minimises the impact on frontline staff - for example, a centralised office arranges routine screening tests.
Many patients for whom Geisinger is responsible - either through an ACO or through its health insurance - will be treated at some point in a hospital or clinic that is not directly run by the company.
Getting the financial model right for payment for all parties involved is crucial. A pure fee for service model would not promote the type of improvements in quality the company wanted at primary or secondary level. And payments have been increasingly linked with quality measures - hospitals that reduce complications get more money, for example.
In the ACO structure, all participating hospitals and physician groups share in the improvement in quality and savings created by more “upstream” care delivery, that reduces the need for downstream high cost “rescue” treatment.
Ms Tomcavage says physicians are often suspicious at the start but become enthusiastic.
“In our model it’s an upside risk - there is no downside. It has been a good approach for us because we have seen improvements in quality and we have seen reductions in total cost of care. But it is hard work. When stressed, providers often go back to what they know.
“At the end of the day you are going to get what you pay for. If you pay for work units, you are going to get more work units. The payment needs to be aligned to where you want your outcomes to go.”
In the long term, as prevention and early intervention bear fruit, there could be fewer referrals to hospital and some hospitals will lose out.
Dr Graf points out that “babyboomers” are just entering the Medicare system, and are of an age when they may require more secondary care. This will prevent a dramatic fall in hospital use in the next few years but he foresees problems for smaller, more rural hospitals that are dependent on medical admissions in particular.
“That means the hospital has to transform. One solution is to look at providing a more community based service and services in the home. They may need half the beds they used to. Can they downsize and find a sustainable future?”
Geisinger runs some hospitals itself but also commissions care from other hospital providers.
“The first thing is helping them culturally understand - helping them see what are the things they need to do more of in the future. Show them a way to understand their role,” says Dr Graf.
He cites the example of a small rural hospital Geisinger acquired that is 10 miles from its coronary care centre.
“While high risk patients need to be seen in the coronary centre, low risk patients ought to be seen at the local hospital - yet many are still sent to the main centre. The company is trying to change referral patterns to encourage better use of both of the facilities.
“It is tuning the system to the patient - and something that we have yet to master.” It is about having all the best information for the clinical staff about the status of the patient, and matching it to the resource need - and then pushing that resource to the patient or getting them to the best site of care within the system, he adds.
“Information, organisation, discussion and systematic flawless execution - not incentives - is what will drive performance. If patients feel you understand them, understand their needs, and believe and see you actively taking care of them, the process will succeed.”
Reducing variation has been an important part of the drive for quality. When a patient has surgery, for example, a large number of standardised procedures must be followed - such as administering antibiotics at defined times before surgery.
Geisinger measures how often all of this is done - and there can be 40 measures for heart surgery. Doctors can choose not to follow these but are expected to justify when they are deviating from what is recognised as best practice.
The Institute for Healthcare Improvement’s “triple aim” of addressing cost, quality and patient experience have been taken on board by many US healthcare organisations. But Dr Graf adds another: the experience of the healthcare professionals involved.
‘If patients feel you understand their needs, and believe and see you actively taking care of them, the process will succeed’
“We need to think about that element of the professional experience - the professional experience improvements we have seen are fantastic. That has a lot of other effects. They become more efficient.”
Part of this has been ensuring that, as far as possible, work is done by the most appropriate person. In diabetes care, for example, Geisinger has identified nine elements of good routine care. Increasingly doctors are concentrating on just the three for which their skills are most important, and other staff are doing the remainder.
Engaging patients is one area where ACOs have been criticised for slow progress - although most health systems are still struggling with the most effective way to do this. In the US patients are more likely to have access to their own records and test results than in the UK. Geisinger has been using text and phone reminders for routine screening, and also sends birthday cards to patients - congratulating them on steps they have taken to improve their health and suggesting more they can do. About 10 per cent respond and make appointments. Follow-up phone calls are made to those who have not responded.
But Dr Graf ponders whether patients may need more incentives to become engaged in improving their own health. One option in an insurance model is guiding patient choices through different levels of co-payment.
“That steers them away from the less efficient high cost providers,” he says. But giving patients more information in an understandable format can also push them in certain directions. Geisinger is carrying out research into patient motivations around engagement and what would work best.
It is difficult to be definitive about the positive outcomes of the ACO model - nationally, many of the programmes in which there were initial savings were operating in high cost hospital areas. However, Geisinger supported another ACO in Maine, which adopted many of its approaches and achieved a 5 per cent fall in costs against an average of 1 per cent across the 32 pioneer ACOs.
In West Virginia, where it has also worked, admissions were cut by 19 per cent, readmissions by 36 per cent and emergency room attendances by 13 per cent.
Some of the individual elements of the Geisinger approach have also led to improved clinical outcomes - the focus on diabetes care is associated with a reduced risk of myocardial infarction, stroke and retinopathy, for example. Overall, Proven Health Navigator has led to cost reductions of 4.3-7.1 per cent. And there is growing evidence of improved outcomes for patients.