Can user charges make health care more efficient?BMJ 2010; 341 doi: https://doi.org/10.1136/bmj.c3759 (Published 18 August 2010) Cite this as: BMJ 2010;341:c3759
- Sarah Thomson, deputy director12,
- Thomas Foubister, research officer in health policy1,
- Elias Mossialos, director12
- Correspondence to: S Thomson
Sarah Thomson, Thomas Foubister, and Elias Mossialos explain why charging patients for health services we want them to use makes little economic sense
The United Kingdom’s new government has promised not to cut public spending on health. It has also said that the NHS in England will have to make more than £20bn (€24bn; $30bn) in efficiency savings over the next few years if it is to meet patient demand.1 This may lead to calls for an expanded role for user charges—both to bring more money into the health system and to make health care more efficient.2
Most healthcare systems require patients to pay something at the point of use (table⇓). Some countries apply user charges to all health services; others, like England, use them more sparingly (box 1). In this article we explain why user charges may undermine efficiency, and show how a few countries are developing a more sophisticated approach to user charges to ensure this does not happen. We also ask whether the English NHS should follow their example.
Box 1: User charges in the English NHS
User charges in the NHS generate around £1bn a year, equivalent to about 1% of the NHS budget in 2009-10.4
Patients pay £7.20 for each prescription (73% of the average prescription cost in 2008)5 or £104 for a 12 month prepayment certificate or £28.25 for 3 months
Exemptions for children, older people, people on low incomes, and those with certain medical conditions mean that only 11.4% of prescriptions are charged for5
The maximum copayment for a complex course of treatment at an NHS dental practice is £198; most courses of treatment require copayments of £16.50 or £45.60
Children, people on low incomes, and women who are pregnant or have had a baby within a year are exempt
Despite these exemptions, NHS charges create financial barriers to access, particularly among poorer people and people with chronic conditions. Half of the Citizens’ Advice Bureau clients surveyed in England and Wales in 2000 reported difficulties in affording NHS prescription charges, and over a quarter failed to get all or part of a prescription dispensed during the previous year because of the cost.6 In 2006 the House of Commons Health Select Committee described the system of user charges as “a mess”4 and criticised the policy for its piecemeal development and lack of comprehensive underlying principles
User charges and efficiency
Two arguments for expanding the role of user charges in the NHS might be made, both relating to the potential to get better value from the resources available. The first argument is that user charges can help make up for shortfalls in public funding. This is an efficiency argument because the implicit claim is that in the current economic climate, social welfare will be enhanced by substituting patient payments for tax financing. There is no theory of fairness in tax policy that permits targeting the ill in this way, and we will not consider this argument further.
The second argument is that user charges make people more discerning in their healthcare choices.7 When health care is free at the point of use, patients seek care for as long as there is some benefit to be had—however small, and irrespective of its cost. Imposing a charge will encourage people to avoid care that is low value or not cost effective (costs more to provide than the benefit produced is worth), freeing up resources to provide more high value care. As a result, the amount of health the healthcare system produces will grow relative to the resources available.
For this to be the case, however, several assumptions must hold. People must possess the information needed to make the right choices; they must be able to understand the information they have; and they must be able to make rational choices based on that information. These assumptions do not hold in health care.8
Research shows that although people do reduce their use of health care when faced with a charge, they are unable to distinguish low value from high value care. The strongest evidence for this comes from the respected RAND Health Insurance Experiment, a large randomised controlled trial carried out in the United States in the 1970s and early 1980s.9 The RAND study found that people across income groups who faced a user charge reduced the use of effective care almost to the same degree as they reduced the use of ineffective care.10 This important finding has been reinforced by subsequent studies based on natural experiments.3
Advocates of user charges point out that in spite of reduced use of effective care, the RAND study showed no adverse health effects for people who are not poor.2 This is hardly surprising, given the study’s design and timing. The trial was not long enough to capture the longer term effects of user charges. It also excluded people older than 62 and those too disabled to work. In addition, as the architects of the RAND study and other researchers have noted, many conditions that were untreatable or acute in the 1970s are now considered manageable, so the range of treatment potentially subject to charges is far greater today than at the time of the study.11 Their message is that the low level of adverse health effects found in the RAND study should not be used to justify user charges.
Other research shows how user charges can contribute to avoidable increases in
healthcare costs over time. Introducing user charges in one area of care—for example, outpatient prescription drugs—can have a squeezed balloon effect, initially lowering expenditure on drugs but increasing the use of other services such as half day or full day admissions to community mental health centres,12 nursing home admissions,13 and emergency care.12 14 This is because people may forgo necessary treatment, fail to adhere to treatment, or opt for free (but expensive to provide) care to avoid paying user charges. Total spending can increase if patients present in high cost settings with conditions that could and should have been treated (more cheaply) much earlier on.
The research evidence shows that the efficiency case for user charges outlined above does not hold. But it might still be possible to design a user charges policy that enhances efficiency. This could be done by building on the primary incentive effect of user charges—to forgo care—in such a way that patients only forgo low value care. The obvious course of action is to remove responsibility for decisions about value from users of health care and apply charges to low value care alone. A few countries, including the United States, France, and Germany, are developing such a value based approach (box 2). A review of their experience shows the rationale for and effects of value based user charges.
Box 2: Value based user charges
Value based user charges aim to steer patients away from low value health care (care which costs more to provide than it is worth in terms of clinical benefit) and towards high value health care. Cost effectiveness might be considered for the following categories, singly or in combination:
Specific products such as angiotensin converting enzyme inhibitors, β blockers, and statins
Care relating to specific conditions such as diabetes, asthma, and hypertension
Specific services such as preventive screening or immunisation programmes
Specific patterns of use such as obtaining a referral to specialist care or adhering to a care protocol
In countries where user charges are widespread a value based approach would involve abolishing or lowering charges for high or moderate value care. In countries with no user charges, it might involve charging for health care that is indisputably of low value.
Value based user charges in the US have emerged in the past 10 years against a backdrop of widely applied and rapidly rising user charges that have done little to stem the growth of healthcare spending or to enhance efficiency in the use of health services.15 Several insurers now offer enrolees reduced user charges for drugs prescribed for specific conditions (such as diabetes, asthma, and hypertension) or for specific groups of drugs (such as angiotensin converting enzyme inhibitors, β blockers, and statins).16 17 Others are taking a more focused approach, offering diabetic patients reduced user charges for selected services that prevent or lower long term complications associated with diabetes.17
In the US, value based user charges are an improvement on steering patients by cost alone. Almost all insurers offer tiered drug formularies with reduced charges for cheaper (often generic) products. Although these may lower costs, they do not necessarily enhance efficiency because patients continue to face a financial barrier to cost effective drugs (whether generic or branded). Applying user charges across the board also risks undermining efforts to improve quality, such as chronic disease management and pay for performance. It makes little sense for patients to face barriers to the care these programmes are established to encourage. Early evaluation of value based schemes suggests promising results in terms of health gain and cost savings.17
Researchers have recently estimated that abolishing user charges for high value health care, keeping them unchanged for health care of moderate or unknown value, and increasing them for low value health care, would increase the health benefit produced by the US health system without increasing costs or overall out of pocket payments.18 They also found that using cost savings from value based user charges to subsidise insurance coverage would result in a 31% increase in benefit (measured in life years gained) among those without health insurance.
France has recently applied the value based approach to visits to general practitioners and specialists. Because it considers direct access to specialists to be a low value use of resources, adults who obtain specialist consultations without a referral from their regular doctor (usually a general practitioner) now have to pay higher charges for each consultation (50% of the cost rather than 30%).19 The higher charges do not apply to patients with chronic or serious conditions, who have traditionally had free access to treatment relating to their particular condition(s).20 However, since 2006, these patients have free access to doctors only if they adhere to a coordinated care pathway protocol devised by their doctor and the health insurance fund.19 21 In 2008, around 80% of diabetic patients were covered by protocols.22
User charges have a smaller role in Germany than in the US and France. In 2004 the German government introduced a €10 charge for the first visit to a doctor in every quarter and for subsequent visits without referral.23 Initial evaluation suggested that the charge led some patients to avoid or delay seeing the doctor.24 Since 2007 these charges have therefore been waived for patients who register with a family doctor and obtain referrals to specialist care (as opposed to accessing specialists directly), on the grounds that registration and referral will improve coordination, reduce duplication, and lead to fewer hospital admissions.25 By the end of 2007 over 5.3 million patients had enrolled, particularly older people and those with chronic conditions.25 Some health insurance funds also waive user charges for patients who enrol in disease management programmes.26
User charges and the NHS
The three countries discussed above are trying to implement a user charges policy that steers patients away from low value care and towards high value care, rather than putting patients in a position where they must gamble on their own health. Should the NHS follow suit and adjust its user charges policy accordingly? There are two factors to consider: problems with the value based approach and the presence of alternatives.
A value based system of user charges can be expensive to design and administer. It can also be risky because of uncertainty about the clinical effectiveness—let alone the cost effectiveness—of many healthcare interventions. Given that individual patient characteristics can affect the clinical or cost effectiveness of an intervention, to get things entirely right may be impossible. So although a value based policy is a clear improvement on traditional user charges, it can only ever be partial in scope if major cost and serious risk are to be avoided.
In the US, France, and Germany, the ultimate aim of adapting policy on user charges has been to achieve outcomes the NHS already achieves not through user charges but through supply side controls. The use of financial incentives (free care) to register with a regular doctor, enrol in disease management programmes, encourage adherence to medication, and obtain referral to specialist care is intended to tackle the problems associated with highly fragmented delivery systems and fee for service provider payment—for example, lack of continuity, poor coordination, duplication, and inappropriate prescribing and use of specialists. These are problems the NHS has largely managed to avoid or redress through its strong primary care focus, general practitioner gatekeeping, policy on generic prescribing and, more recently, through the Quality and Outcomes Framework, which includes incentives for disease management.27 28 Thus value based cost sharing operates in contexts where alternative ways of achieving particular ends are absent—perhaps because they are less feasible.
Because there are costs and risks associated with value based user charges, and because the NHS has other tools for achieving the same ends, it is hard to see how expanding the role of user charges would enhance efficiency. A step in the value based direction would be positive —for example, it does not make sense to spend a great deal on financial incentives for general practitioners to improve outcomes for people with asthma, heart disease, and mental illness (through the Quality and Outcomes Framework), yet erect barriers to treatment by making these patients pay for prescription drugs. A charge could remain for low value care, but if care is indisputably of low value, why provide it at all?
The current economic climate requires tough decisions about NHS funding and provision. Because we value fairness alongside health gain, revenue for health care should be raised equitably. And because we want to get the best possible return on the money we put into health care, user charges should not be allowed to undermine efficiency by impeding access to high value care. The NHS is fortunate in having recourse to a larger set of policy instruments than many other countries, which makes user charges hard to justify. If the UK government’s concern is to enhance efficiency in health care, the best starting point would be to dispense with user charges altogether.
This paper is part of an occasional series prepared in conjunction with the European Observatory on Health Systems and Policies (www.healthobservatory.eu)
Cite this as: BMJ 2010;341:c3759
Contributors and sources: ST, TF, and EM have recently written an overview of healthcare financing in the European Union. ST and EM have extensive experience of European health systems through their involvement with the European Observatory on Health Systems and Policies. ST, TF, and EM devised the article. ST wrote the first draft. TF and EM contributed to subsequent drafts. All have read and agreed the final version. ST is the guarantor.
Competing interests: All authors have completed the unified competing interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare no support from any organisation for the submitted work; no financial relationships with any organisation that might have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Commissioned; externally peer reviewed.