Is it acceptable for people to be paid to adhere to medication? NoBMJ 2007; 335 doi: http://dx.doi.org/10.1136/bmj.39286.422639.BE (Published 02 August 2007) Cite this as: BMJ 2007;335:233
At first glance the idea of paying patients to take therapeutic drugs seems absurd. After reflecting on the high cost of non-adherence to individuals and society, however, it may not seem such a ridiculous idea after all. But don't be fooled—your first thought was right.
Given the known costs of non-adherence, paying certain people to take their drugs may look like sound economics. If health professionals are willing to take on a coercive role, and society is prepared to pay a hefty price, which includes considerable loss of personal dignity and privacy, the practical problems can be overcome. But payment for adherence is never the answer, because it creates perverse incentives and undermines the therapeutic alliance between patients and doctors that is needed for successful long term health care.
For anyone concerned about the consequences of non-adherence, the possibility of paying certain people to take their drugs looks worthy of consideration. The cost of non-adherence to individuals and society has been documented extensively. It includes the value of the wasted drugs; the high cost to the National Health Service of avoidable events such as heart attacks and strokes; and preventable complications of long term conditions such as diabetes. With infectious diseases like tuberculosis, the public health consequences can be grave indeed. At the end of the 1980s, Rovelli et al showed that 18% of post-transplant patients did not adhere to anti-rejection drugs.1 In the non-adherent group, 91% of patients lost the transplanted organ or died, compared with 18% in the adherent group, showing how tragic and costly non-adherence can be.
A large proportion of non-adherence is intentional—people make a conscious (even if, from a medical perspective, illogical and ill informed) choice not to take the prescribed treatment. These are the patients for whom payment seems like a good option. There is no sense in paying people who want to take their drugs but cannot because they don't understand the instructions, can't get the tablets out of the packaging, or can't remember when their dose is due. For intentional non-adherers in particular high cost groups, the economic case for paying for adherence looks attractive. But unless we can identify the people who would not otherwise adhere, we will end up paying everyone.
Paying for adherence, whether in the form of cash or non-financial benefits, creates all the wrong incentives because we cannot screen out people who would adhere anyway. By introducing payment, voluntary adherence will disappear. After all, why should patients agree to take medicines for nothing if they can be paid? Many doctors already complain that patients demand too many drugs. Prescriptions that come with payment attached can only make this worse.
So paying everyone to take their drugs is clearly a non-starter. But there may be particular groups of patients, whose non-adherence is so costly to themselves or society (or both), that payment makes clear economic sense, even if we have to pay them all. Patients with infectious tuberculosis may be an example. However, even here the disadvantages of financial incentives outweigh the benefits.
Medicine taking behaviour is driven by the necessity-concerns equation.2 I may choose not to take drugs for treating my tuberculosis because they have unpleasant side effects3 4 5; because they affect how I feel about myself; or because I don't trust doctors or drug companies and I'm worried about immunity,6 dependency, or the possibility of long term harm. Simply offering me money is unlikely to change my views. On the contrary, it may convince me that my suspicions were justified (why would you pay me to do something that is in my interests?) and make me want to get the cash without taking the pills.
As soon as money is introduced into the equation, we have created the conditions for fraud, so our first problem is one of policing. Where payment depends on physical consumption, supervised administration or testing is needed. Such systems are costly and cumbersome to administer. Arguably they also infringe personal privacy. Watching someone produce a urine sample to monitor adherence may not be as invasive as taking blood, but it may be even less appealing to health professionals.
There is another group of patients to consider. Where very ill people are not in a position to understand and make positive choices to take their drugs, and the medical benefits are very clear, is there an ethical argument for payment, especially if relatively small financial incentives can make a real difference? Absolutely not.
Paying for adherence undermines the basis for informed agreement about the best treatment for the individual, which should be at the heart of health care. Paying people to take medicines sends a signal that they need to be compensated for doing something that is not inherently in their own interests. It is coercion by carrot rather than stick, but coercion none the less. Can we imagine paying people to have lobotomies? Paying people to take drugs occupies the same ethical territory and is no more acceptable just because we don't use the same explicit written consent processes for drugs as we do for some other medical interventions. Convincing people to accept treatment when they are reluctant to do so is a genuine problem, but paying for adherence, however seductive it may appear, will never be the way to solve it.
Competing interests: None declared.
Where do you stand on the issue? Vote in our poll at www.surveymk.com//s.aspx?sm=vqkSSGRBEiE7dQ9ns8x16g_3d_3d