Priority Setting in Action

Purchasing Dilemmas

Frank Honigsbaum, John Richards, Tony Lockett

• Publisher: CRC Press
• Year: 2018
• ISBN:
  9781315345833
• Edition: 1
• Topic: Medicine
• Subtopic: Public Health, Administration & Care

The power of purchasers exposes the weaknesses of conventional thinking on the costs and benefits of priorities. Health policy analysts now have to develop rational criteria to support decisions in a process which may be inherently intuitive. This authoritative and practical text points the way towards clear choices in resource allocation and the implications of these choices on expenditure diverted among different health care programmes.

1 The international context

Priorities have long been set in health care, but seldom in an explicit or public manner. They were applied mainly in private by clinicians in the way they treated patients. Priorities assumed public view only in the form of waiting lists for treatment or restrictions on access to services such as renal dialysis.

In recent years, attempts have been made to conduct the process more openly. The State of Oregon in the USA led the way and stimulated developments elsewhere.

The United Kingdom followed as the result of pressures emanating from the introduction of an internal market in 1991. The act of purchasing health care was separated from its provision, and that led parties on both sides of the process to look more critically at the decisions taken. Purchasing authorities, in particular, began searching for ways to make the most of the funds at their disposal. They found it both necessary and desirable to make priority setting methods known.

This section describes what has been learned thus far. The way priorities are set in five leading countries is assessed: Oregon in the USA, The Netherlands, New Zealand, Sweden and the United Kingdom. This is based mainly on a report of a seminar held in Stockholm in 1993 and published under the title Priority Setting Processes for Healthcare¹. Additional information relating to Oregon has been obtained since.

Oregon

Oregon has set priorities across the whole spectrum of medical, dental and mental health care. An extensive list was prepared in the form of condition-treatment pairs covering all that health care could offer². The pairs were ranked in an order which, for the most part, put the most effective treatment at the top and the least effective at the bottom. The plan was devised to meet budget restrictions under America’s Medicaid provision for the poor, a system run by the states but subject to federal regulation. It was left to the state legislature to decide how far down the list Oregon could afford to go. However, because most of the money came from Washington, federal approval was needed before the plan could go into effect. On the list finally approved, 565 of 696 items were funded. This began operating on 1 February 1994, and covered only medical and dental care. Mental health care, along with conditions relating to the elderly and disabled, were not added until 1 January 1995. This integrated list contains 745 items, of which 606 are funded.

Attempt to use QALYs

Oregon tried three methods before its plan was finally approved. The first was based strictly on cost-utility, using the cost per quality adjusted life year (QALY) technique. If, for example, treatment for a particular condition costs £1000 and is expected to last ten years, the cost per year would be £100. But if only five of those years were of acceptable quality, then the cost would be £200.

The cost per QALY figure was used to rank the pairs in priority order, and it produced so many anomalies that the plan was nearly abandoned. Cost and outcome data were lacking for most of the exercise. Only well-defined procedures, such as cardiac artery bypass grafts, could be reliably assessed.

Ranking based on categories of care

The next method was based on broad groupings of care which were ranked in priority order. At the top of the 17 categories employed came life-saving treatments with a good quality of life, followed by maternity care and preventive services for children. At the bottom was a category containing treatments which would have little or no effect on quality of life.

Condition-treatment pairs were fitted into these categories using an algorithm based mainly on varying rates of mortality and quality of well-being. For example, within the first category came fatal acute conditions with treatment resulting in at least a 25% reduction in mortality during the five years following treatment and with at least 90% of surviving patients returning to a very high quality of life.

Ranking within categories was carried out by a mathematical procedure similar to the one used earlier but with less weight attached to cost and more to duration of benefit and quality of well-being. This still left some services in an awkward position, and a ‘reasonableness’ test was used to adjust the rankings. So many hand adjustments were made that a large element of guesswork and judgement went into the final order.

Quality of life ruled out

This method failed to win federal approval because it violated the rights of disabled persons under an Act passed in 1990. All consideration of quality of life had to be removed from the ranking process and that, combined with pressure from user groups, led to the upgrading of several controversial procedures.

The category method of classification also had to be abandoned because quality of life considerations had been used to set the order in which they were ranked. In its place, a new method was devised which is unique to Oregon and not conducive to replication.

In Oregon as elsewhere, very few outcome data were available, and this meant that a large element of subjective judgement went into the process.

Those who compiled evidence of effectiveness did not have the time or resources to make extensive literature reviews, and many findings were based on the experience of local specialists in the conditions under consideration. The list is being modified as additional data become available.

Nevertheless, however inadequate Oregon methods were, the list at least provides a template against which specific services can be assessed. Though cultural differences may prevail, it can offer guidance to priority setting everywhere.

The plan was finally approved by the incoming Clinton administration subject to several conditions, the main one being that no further treatments could be removed without federal approval.

The plan in operation

The plan has appeared to work tolerably well since it began operating in February 1994. This may be partly due to the coverage of controversial items previously excluded. The list reflects a cautious approach to rationing. Hard choices have been avoided, so much so that one wonders whether the plan will achieve the aim of its prime movers to shift resources from acute to chronic and preventive care.

The plan may prove more controversial if further cuts are forthcoming. The State of Oregon faces severe budgetary problems, and more items may have to be removed from funding. Whether this will be accepted by the federal government is uncertain.

Financial pressure may also come from the addition of mental health services in 1995. All but eight of the 49 items added are funded, and the high cost of some could force further restrictions in list coverage.

Another problem may arise if the list is not effectively monitored. Services are being provided mainly by managed care organizations, which receive capitation payments based on the estimated cost of included treatments. But unless these providers supply detailed data on each item, it will not be possible to tell whether the treatments paid for are actually supplied.

Problems in methodology

The removal of quality of life considerations has far-reaching implications and was resented by those who prepared the Oregon plan. They would have preferred to operate the list which failed to win federal approval.

However, they may be ready to accept a modification in the way quality of life is gauged. In Oregon, disability assessments were not made by those who were disabled. In future, disability may be based on an average of the scores of those who are disabled and those who are not.

In place of QALYs, disability life years (DALYs) may be used. With QALYs, no value is given for the extra years of life at a lower quality; DALYs take account of full life expectancy or length of treatment effectiveness regardless of quality.

Clinicians may still withhold covered services if they judge outcome to be poor, and that assessment could include quality of life considerations. More explicit restrictions could be developed from the clinical guidelines being slowly formulated (a maximum of eight every two years) under legislative direction.

A major weakness of Oregon methodology was the failure to take account of needs assessment. The state, like America in general, has a serious obesity problem, yet the two services relating to obesity were not funded. Decisions were made largely on the basis of effectiveness, and trials relating to obesity procedures have not produced the necessary evidence. Yet the need for some form of treatment is so pressing that funding should have been forthcoming.

Public values were meant to inform decision making, and an impressive list was prepared, taken from community meetings. But the meetings were not representative of the public and no method was devised to relate values to ranking. Subjective judgements were used instead.

Equity was not given a high value ranking despite the fact that the main aim of the plan was to extend Medicaid coverage to all below the poverty level. This value, as well as community compassion, does not appear to be as strongly held in Oregon (or America generally) as elsewhere.

The Netherlands, New Zealand and Sweden

The Oregon plan has inspired priority setting efforts elsewhere, but development has been slower. The comprehensive list compiled in Oregon is the only one that covers all of health care.

Oregon may have been able to move more rapidly because its plan applies only to the poor. Elsewhere, whole populations are involved and public feeling for equity may be stronger than in America. In this section, we shall examine methods in The Netherlands, New Zealand and Sweden.

The Netherlands

Methods proposed here come closest to the Oregon model, but they have yet to be implemented. In 1992, an official committee called for the creation of a priority list on Oregon lines using four criteria to produce a basic package of care. Treatments would have to pass through the following sieves before they would be included:

• Is the care necessary from the community point of view?
• If so, has it been demonstrated to be effective?
• Is it also efficient, using such methods as QALYs?
• Can it still be left to individual responsibility?

Individual patients and doctors might take a different view of necessary care, but the committee called for the community approach in the first sieve. It defined health as the possibility for every member of society to function normally. This enabled it to stress services for the chronic sick and vulnerable groups such as the mentally ill. As in Oregon, the committee felt the need to curb high-technology treatment in the acute sector.

The second sieve could eliminate many treatments if applied strictly. Rigid tests of effectiveness apply to only a small proportion of health care. This could frustrate protection for the chronic sick and vulnerable groups since many of the services involved have yet to prove their worth.

The third sieve deals with cost-utility and will again be difficult to apply because of the limited data available.

The fourth sieve is a novel one and leaves some treatment to be financed by patients themselves.

The final package is not likely to be as restricted as these sieves suggest since it will be determined by political considerations. Compulsory insurance now covers only 60% of the population, but another government committee recommended that it be extended to everyone. Such a system would have to offer ample provision. The committee called for 85% of treatments to be included in the basic package, but even this was not enough for the government, and it raised the proportion to 95%.

Some action has been taken; but a basic care package still has to be defined. Public opinion has tended to resist service restriction, and a three-year programme has been proposed, aimed at influential consumer groups, to show the need for hard choices. This is one of the strongest features of the Dutch model and it has been implemented.

New Zealand

Here, methods deviated most sharply from Oregon. An official committee was asked to define an essential core of services but, in view of the country’s long history of universal provision, it approached the subject cautiously. Existing services were assumed to be suitable but the committee thought changes might be made in the way resources were allocated. Before priority setting began, it sugges...

Table of contents

• Cover
• Title Page
• Copyright Page
• Contents
• List of contributors
• Participants
• Foreword
• Preface
• 1 The international context
• 2 The Southampton experience
• 3 Lessons learned
• 4 Emergent themes
• 5 Outstanding issues
• Appendix A Local health problems
• Appendix B Options for change and indicative costs
• Appendix C Marginal value
• Appendix D Health benefits
• Appendix E Other benefits
• Appendix F Stakeholder consultation results
• Appendix G Expert panel: personal profiles
• Index