[Skip to Content]
[Skip to Content Landing]
October 25, 2000

Effects and Ethics of Sanctions on Childhood Immunization Rates

Author Affiliations

Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorPhil B.FontanarosaMD, Executive Deputy EditorIndividualAuthor

JAMA. 2000;284(16):2056-2057. doi:10.1001/jama.284.16.2053

To the Editor: In their Commentary, Drs Davis and Lantos1 make a common mistake when they state that, "the PIP improved immunization rates" may have produced "untoward effects." They attribute these effects and "ethical problems involved in policy evaluation" to the evaluation rather than to the policy itself. In Georgia, it was the coercive statewide policy that improved immunization rates and may have had adverse effects. By adding a small temporarily exempt control group, the PIP2 merely increased the ethical quality of an otherwise uncontrolled, potentially unethical statewide policy experiment. When an agency laudably uses a randomized delayed control group to make a policy's effectiveness (or ineffectiveness) more measurable, it also makes other aspects of the policy, such as coercion and adverse effects, more evident. The challenge for ethicists, researchers, clinicians, and policy makers is to find ways to increase the linkage of policies to evidence, including ethics-related evidence. This can be better achieved by praising randomized policy trials for increasing visibility of ethical issues than by criticizing them for ethical aspects of the policies they are evaluating.

Our views are based on experience with our randomized policy trial. British Columbia's publicly funded drug benefit program, Pharmacare, introduced a cost-saving policy in 1999 that restricted coverage of nebulized medications. As usual, there was no informed consent and some theoretical risk of potential adverse effects. Ten percent of physicians were randomly allocated to an optional 6-month exemption. This is an inversion of Zelen's randomized consent design3,4 in which consent is sought from subjects in the treatment arm after randomization. In our trial, only the 474 privileged controls were given the option of withholding their consent. Only 1 physician with patients receiving nebulized medication asked to withdraw from the control group. A university ethics committee approved our delayed controls but had no authority over the policy itself. Yet, as advisors who helped implement the policy, 2 of us (M.M., B.C.) were aware that the control group increased scrutiny of the policy from an ethical perspective. We found that this increased the sensitivity to ethical issues and helped produce a better policy. We believe randomized delayed controls can often increase the ethical quality of restrictive or coercive policies.

Davis  MMLantos  JD Ethical considerations in the public policy laboratory.  JAMA. 2000;284:85-87.Google Scholar
Kerpelman  LCConnell  DBGunn  WJ Effect of a monetary sanction on immunization rates of recipients of Aid to Families With Dependent Children.  JAMA. 2000;284:53-59.Google Scholar
Zelen  M A new design for randomized clinical trials.  N Engl J Med. 1979;300:1242-1245.Google Scholar
Zelen  M Randomised consent trials [comment].  Lancet. 1992;340:375.Google Scholar