For several years, cost-conscious health policy makers in Europe and North America have proposed or instituted new and potentially powerful incentives for physicians to reduce inefficient drug prescribing. In Germany, such limits took the form of a yearly cap on drug expenditures in the national health insurance system (set at a previous year's level). German physicians had to compensate the government for the first DM 280 million ($A 233 million) they spent above the cap. In the United Kingdom, general practice fund holders were given a capitation prescribing budget. Any savings on prescribing could be reinvested in other patient services. In the U.S.A., some health maintenance organisations (HMOs) have considered or begun to provide both financial rewards and penalties in 'drug risk pools'. In these pools, groups of physicians receive, as extra income, a proportion of the savings from spending less than a prospective drug budget. As in Germany, the physician groups must also share in any over-expenditures.

The hope is that such economic incentives will induce physicians to reduce costs by choosing effective, low-cost therapies whenever possible, and by reducing use of unnecessary medications without adversely affecting the health of their patients. This hope rests on numerous heroic assumptions. First, the medical practice must be large enough to spread risk over many patients, so that individual prescribing decisions are not unduly distorted by the incentive structure. Second, prescribers must receive timely feedback on performance related to the budget, and specific advice on improving efficiency while maintaining quality. Third, physicians must know about drug substitution, not just in general, but for specific clinical situations where substitutions could save money or create unnecessary risks. Finally, physicians must know and be sensitive to differences in drug pricing, an unlikely proposition judging by research on this topic.

Are these assumptions valid? Could such policies sometimes encourage shifting of costs to other health care sectors or, even worse, threaten quality of care and ethical principles of doctor-patient relationships? Unfortunately, current information is scant. Nevertheless, we can comment on key areas that are potentially affected by these policies: costs, quality and ethics.

Do prescribing budgets reduce drug costs? It depends. Physicians are not immune to financial rewards and sanctions. After the German government threatened physicians with even modest financial penalties for exceeding the budget (an amount equivalent to a maximum of only 1% of income), estimated drug expenditures declined by about DM 2 billion (about 10%) during the first year of the budget cap.1,2,3 On the other hand, the incentives offered by the U.K. fund holding may have been no more effective than feedback regarding adherence to voluntary prescribing targets.4,5 It should be emphasised that drug costs represent a modest fraction of total health expenditures in many industrialised countries e.g. 8% in the U.S.A.6 and 10% in the U.K.7,8

Might reduced prescribing result in compensatory increases in other health expenditures? In the U.S.A., we call this endemic problem 'balloon squeezing' and 'cost-shifting'. For example, well-controlled studies of patient-level caps in Medicaid (e.g. a 3-prescription limit) showed that the modest government drug savings were dwarfed by increased costs of institutional services among chronically-ill elderly and schizophrenic patients, two high-risk groups whose experience with this policy was carefully studied.9,10 These effects are partly explained by physicians, family members and patients shifting care to where prescription coverage was maintained9 and probably, in some cases, by the adverse health effects of reduced access to medicines.10

The extent to which practice budgets produce unintended cost-shifting is still unclear because of the lack of data from well-controlled studies. In Germany, a hypothesis was that the drug cap increased referral of patients from primary care to specialists and hospitals which were exempt from the regulation. Preliminary estimates, based on unadjusted before and after comparisons, do indicate an increase in costs of specialists and hospital care that approximately equals the savings in drug expenditures.1,2,11 Similarly, case reports in the U.K. indicate that general practitioners sometimes ask patients recently discharged from hospitals with expensive medications to 'take the drug back to the hospital so that it can be used there'.12 While the extent of cost-shifting is unknown and definitely merits further study, economic theory and common sense suggest that compartmentalised drug budgets are likely to induce physicians to shift costs under certain circumstances.

A cost-containment measure is likely to have little impact on quality or outcomes of care if it can selectively reduce unneeded medications while preserving essential care. Unfortunately, the use of financial incentives and penalties without vigorous education is likely to be a blunt-edged sword that reduces necessary as well as unnecessary care. Previous controlled research has already shown that patient disincentives, such as severe cost-sharing and prescription limits, reduce use of essential agents, such as insulin, cardiovascular drugs and antipsychotic agents.9,10,13,14

The preliminary estimates from Germany show that much of the decreased drug expenditure associated with the budget cap represented increased generic prescribing and reduced use of ineffective preparations (e.g. peripheral/cerebral vasodilators). However, the use of essential medications such as insulin, drugs for Parkinson's disease and lipid-lowering drugs also declined by 11-33%.1 Such changes in medication use could clearly compromise the health of vulnerable patients, but the extent to which this occurs will require well-controlled research.

Prescribing budgets and medication risk pools could pose worrisome conflicts between the interests of clinicians, patients and health organisations. The smaller the practice size, the more likely that physicians will face the economic consequences of individual prescribing decisions. Ideally, drug prescriptions represent the culmination of a doctor-patient interaction marked by trust and confidence. Recasting prescribing in a way that offers personal incentives to physicians not to prescribe may have important symbolic and ethical consequences. Consider the case of a recently discharged elderly patient with acute myocardial infarction and left ventricular dysfunction. Prescribing relatively expensive ACE inhibitors and cholesterol-lowering drugs carries a strong statistical probability of reducing future hospitalisation for cardiovascular disease. At the margin, will some physicians avoid prescribing such drugs because they are at risk for the added cost of medications, but gain no economic benefit from an avoided hospitalisation? Since the payer wishes to decrease both prescribing and preventable hospitalisations, do physicians and administrators have conflicting interests? Will the addition of personal economic incentives undermine patients' confidence and trust in their clinicians' decisions?

Such direct conflicts of interest could have enormous negative effects on public opinion. Consider a front-page headline on the German drug budget: 'Doctor refuses Berlin woman cancer medicine'.15 Might stronger financial rewards result in even more stark headlines such as 'Physicians profit from denying life-saving heart drugs to elderly woman'? Clearly, most physicians would work tenaciously to avoid such situations; however, even the appearance of such conflicts would be damning.

Recommendations for policy
Previous research on pharmaceutical policy and quality of care supports a cautionary attitude towards incentives related to prescribing budgets. The limited evidence, while documenting drug savings in some cases, suggests that there may be offsetting increases in expensive substitutes, including hospitalisation and referrals to specialists. Another possible effect includes reductions in health status among vulnerable populations whose independent functioning depends on access to effective medications.

Government and private health plans should not implement prescribing budgets widely until their impacts are better understood. Well-controlled research should carefully measure both intended and unintended effects, and determine whether (and under what circumstances) budgets should exclude specific high-risk populations such as elderly patients with chronic illnesses that benefit from expensive, but cost-effective, medications. In addition, we need to know how to educate physicians and provide the necessary feedback about prescribing. Incentives need to be structured to prevent possible perverse effects like withholding cost-effective drugs that prevent hospitalisation. In the absence of such data, adoption of simplistic policies that are not sensitive to the needs of vulnerable populations represents a poor gamble. The outcomes may be unacceptable to patients, clinicians and health care policy makers alike.



  1. Vorderwulbecke U. The pharmaceutical industry and the drug sector in Germany. Statement, October 11, 1993, Tubingen, Germany, at the workshop on `The health care sector in Germany'.
  2. Schoffski O. Consequences of implementing a drug budget for office-based physicians in Germany. Pharmaco Economics 1996;10(2 Suppl):S37-S47.
  3. Henke KD, Murray MA, Ade C. Global budgeting in Germany: lessons for the United States. Health Aff Millwood 1994;13(4):7-21.
  4. Stewart-Brown S, Surender R, Bradlow J, Coulter A, Doll H. The effects of fund holding in general practice on prescribing habits three years after introduction of the scheme. Br Med J 1995;311:1543-7.
  5. Wilson R, Walley T. Prescribing costs in general practice fund holding [letter]. Lancet 1995;346:1710-1.
  6. Levit KR, Lazenby HC, Sivarajan L, Stewart MW, Braden BR, Cowan CA, et al. National health expenditures, 1994. Table 10. Health Care Financing Review 1996;17(3):230.\r\n
  7. Department of Health. Health and personal social services statistics for England. London: Government Statistical Service, 1995.
  8. Office of Health Economics. OHE compendium of health statistics. 9th ed. London: Office of Health Economics, 1995.
  9. Soumerai SB, Ross-Degnan D, Avorn J, McLaughlin TJ, Choodnovskiy I. Effects of Medicaid drug-payment limits on admission to hospitals and nursing homes. N Engl J Med 1991;325:1072-7.\r\n
  10. Soumerai SB, McLaughlin TJ, Ross-Degnan D, Casteris CS, Bollini P. Effects of a limit on Medicaid drug-imbursement benefits on the use of psychotropic agents and acute mental health services by patients with schizophrenia. N Engl J Med 1994;331:650-5.
  11. Munnich FE, Sullivan K. The impact of recent legislative change in Germany. Pharmaco Economics 1994;6(1 Suppl):S22-S27.
  12. Crump BJ, Panton R, Drummond MF, Marchment M, Hawkes RA. Transferring the costs of expensive treatments from secondary to primary care. Br Med J 1995;310:509-12.
  13. Soumerai SB, Avorn J, Ross-Degnan D, Gortmaker S. Payment restrictions for prescription drugs under Medicaid. Effects on therapy, cost, and equity. N Engl J Med 1987;317:550-6.
  14. Lohr KN, Brook RH, Kamberg CJ, Goldberg GA, Leibowitz A, Keesey J, et al. Use of medical care in the Rand Health Insurance Experiment. Diagnosis- and service-specific analyses in a randomized controlled trial. Med Care 1986;24(9 Suppl):S1-S87.\r\n
  15. Arzt verweigert Berlinerin Krebs-Mittel. Berliner Zeitung 1993 Feb 4.

Stephen B. Soumerai

Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care, Boston, U.S.A.

Dennis Ross-Degnan

Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care, Boston, U.S.A.