Published: 22/04/2002, Volume II4, No. 5902 Page 41 42

Payment by results transformed the landscape of US healthcare. Here Stephen P Dunn offers some lessons the UK would be wise to heed

In the 1970s and early 1980s, the US health system was breaking down; costs were spiralling out of control. US policy-makers responded by introducing two competing payment systems to control the costs of its seniors' health insurance programme Medicare.

In 1983, the original paymentby-results system - a fee-forservice prospective payment system (PPS) - was integrated into Medicare. Around the same time, US legislators also allowed managed-care organisations to enrol Medicare patients for an annual fee in a 'primary care trusts compete for patients' type reform.

The healthcare system was transformed from a pay-as-yougo cottage industry into a fiscally responsive and dynamic corporate machine with all the good and bad this entails. The private sector quickly copied.

Insurance companies transformed their own payment policies and they promoted the managed-care philosophy.

People were paid by results - by activity, not quality. Managed care began to take off.

The lessons of this transformation will need to be learnt if the UK is to avoid similar outcomes. Before the prospective payment system was introduced, US hospital margins were low.

Hospitals were reimbursed for a fair share of their Medicare costs, so if Medicare patients were responsible for 40 per cent of a hospital's activity, the care programme contributed around 40 per cent of (reasonable) costs.

This cost-based reimbursement was complex, resulted in unpredictable payments and spending for providers, and provided poor incentives for efficiency. From the government's perspective, there were no incentives to manage costs. From the providers' perspective, there was little ability to make surpluses.

Under PPS, instead of paying hospitals for the costs of an admission under a diagnosticrelated group (categories of patients based on diagnosed illnesses), standard payments are made for each type of admission.

Each hospital gets a nationally fixed and representative price for its admissions. If they can save costs, they keep the difference. If it costs more, they 'eat' the difference.

Numerous adjustments have been made to this national tariff.

Payments were introduced to help offset the costs of the most expensive cases and allowances were made for teaching and rural facilities; exactly the type of payment policies and market forces adjustments envisaged under payment by results.

The theory was that these adjustments should make it difficult to claim the price is not fair.

These incentives resulted in the health sector becoming more corporate and dynamic, the entrepreneurial response that the UK government seeks to foster.

There were waves of hospital reorganisations. Hospitals responded quickly to the incentives and shortened length of stay. They recognised the importance of discharge planning and diversified into ancillary services such as home care and durable medical equipment.

Coding also became imperative and clinical involvement in the financial stewardship and decision making of hospitals became important. DRGs will further underpin US initiatives to pay for quality and performance.

But it has not all been plain sailing. PPS has been blamed for early and inappropriate - quicker and sicker - discharges, that it has closed hospitals and killed patients. This may be extreme exaggeration but even so, the lesson is clear. Hospitals will respond to incentives and push any system to its limits.

More a pricing than a payment system, DRG is complicated and vulnerable to legislative tinkering.

Having a fixed rate of payment helps financial planning and the certainty that brings, and many hospitals in the US like this.

However, PPS updates and adjustments are the way the US government seeks to control its budget. As a result, its manipulation of the payment rate undermines its predictability.

Some hospitals argue that this undermines long-term planning.

It is claimed that DRGs do not properly accommodate new clinical technologies, which can skew the medical and surgical management of conditions. And sometimes the price does not fall as length of stays fall. In response, Medicare sometimes creates new categories of payments and allows add-on payments or passthrough costs when new technology is used.

Here, the UK can learn lessons.

It is not that PPS is inherently flawed. Messing around with the tariff and associated incomes flows on a yearly basis should be avoided. From a planning perspective, the tariff and its associated adjustments should be re-based or re-weighted every three to five years. In the intervening years, the annual increase should ideally be formula-driven and announced in advance. And at the very least it should include a pre-specified inflation measure, plus a technology uplift minus an efficiency gain. Such transparency will facilitate planning.

The UK may want to consider separate tariffs that distinguish between whether technology is used or not. Unless the UK devises a sensible system for dealing with new technology, the payment system may hinder the drive to implement and universalise the best. This has to be a major concern.

Fixed payments have also created strong incentives to swindle the system. Upcoding patients' conditions meant that hospitals could be paid for more expensive services than they actually provided. This was done by shifting a patient's DRG to one that yields greater reimbursement, such as 'upcharging' a two-minute visit to diagnose a cold by intentionally using codes for a more serious ailment such as bronchitis.

The Center for Medicare and Medicaid Services has recently clamped down on fraud and excessive outlier payments; around 300 hospitals chronically gamed the system for $2bn a year for four years in a row.

Whole service industries have mushroomed to deal with compliance and fraud - particularly covering the financial, legal and IT areas. Clinical coding is audited to identify systematic gaming. There are also significant rewards for whistleblowers who report fraud - up to around a third of the amount recovered.

The NHS will need to monitor compliance and coding quality, possibly as part of Healthcare Commission visits. But the UK will also need to devote more resources to this than the US currently does if it is to prevent fraud. At the moment, only 30,000 of 11 million inpatient Medicare discharges are investigated.

The complexity of the system, combined with the low probability of being audited, means the likelihood of fraud being detected is low. The US system is ripe for abuse. The UK should attempt to avoid learning this the hard way.

The US experience with managed care also offers some lessons. In the 1980s, Medicare began enrolling people in managed-care plans as a way of controlling costs while improving quality.

Managed-care organisations would receive an annual (capitated) premium for defined population that was (actuarially) equivalent to the amount of services they would provide to the same defined population through traditional Medicare.

The theory was that health maintenance organisations would be able to take a wholesystems view and ensure that individuals received the most appropriate care in the most appropriate settings.

This should all be cheaper than the fragmented fee-for-service Medicare. The government recognised this and set the capitated payment at 95 per cent of what Medicare would normally pay, saving the government money and creating contestability with the more traditional fee-for-service Medicare.

Studies have found that Medicare managed-care organisations were even cheaper than the traditional programme; costing up to 85 per cent of what Medicare would normally pay, even though they received 95 per cent. In 2000, the government found that payments to Medicare managed-care organisations were 13.2 per cent more than they would have incurred in the traditional programme.

It was also found that patients enrolling in managed-care organisations tended to have fewer chronic conditions than those remaining in traditional Medicare, and that patients with chronic conditions had a higher rate of un-enrolling from Medicare managed-care organisations than patients without a chronic condition.

Some managed-care organisations were creamskimming and making profits by avoiding those who needed care.

Although there were rules designed to limit cherry-picking, they devised slick techniques to encourage and enrol healthy, low-cost people and discourage the expensive sick and disabled from joining.

Nevertheless, better managedcare organisations have also been able to take more of a wholesystems view. They tend to be better at chronic disease management and exhibit better outcomes than the more fragmented fee-for-service programme. The UK needs to recognise such nuances in implementing payments by results. Managed-care plans can succeed at 'care management' but making them compete for patients can lead to perverse results.

Letting PCTs compete for patients would be a recipe for disaster. PCTs should continue to be responsible for their defined populations. This aids efficiency by minimising risk selection, gaming and cost-shifting on the part of PCTs - the downside of managed care - and allows them to take a long-term view and form rational judgement about the payback from prevention - the upside.

Overall, the US system has worked. Hospitals progressed, margins improved and Medicare kept a handle on costs relative to the private sector. The system injected a much-needed dose of cost-containing incentives into healthcare and demanded much more from its administrators and practitioners.

But no-one says the system is ideal. So perhaps no-one should expect payments by results to be, either. Rather, we should aim to ensure that it enables the NHS to do the right things. We need to appreciate one of the big lessons from the US; that once the system is up and running it is difficult to change.

Attention should also be paid to the costs of administering such a system properly. Of course the Medicare programme is famously cheap by US standards, but that doesn't capture the provider costs in making claims and other compliance costs.

The fact these costs are capped by central government ties administrators' hands and creates opportunities for fraud and gaming. If the government doesn't address these issues, the health select committee will.

Stephen P Dunn is a research fellow at the Institute of Health Policy Studies, California University, San Francisco, and a senior adviser at the Department of Health.