The need to use the unlinked anonymous technique for surveillance of HIV was evident once it was appreciated that data from diagnostic testing, or obtained after explicit consent for unlinked testing, inevitably produced biased estimates of the prevalence of HIV.² The technique was adopted nationally in 1990 only after extensive consultation and general agreement that, with safeguards, it was legal and ethical.³ The surveys, which are overseen by the Department of Health, provide information that would otherwise not be available and is essential for planning and monitoring the control of the spread of HIV.⁴ They use blood that would eventually be discarded, which is left over after the completion of screening tests in genitourinary medicine and antenatal clinics. Before testing, every specimen is irreversibly unlinked from information that would identify the source individual. Hence infection status can never be traced back to a person. Essentially, the results represent groups in the community and not individuals.^2,4

Unlinked anonymous surveys in the United Kingdom started only after approval by local ethics committees, and refusals have been rare. Recently, these committees approved extension of the technique for surveillance of hepatitis A, B, and C. Internationally, most countries have followed the United Nations Programme on AIDS and the World Health Organization's recommendation to use the technique; only two countries have decided against it.^1,5

A door to door survey found that only a third of the general public seemed to be aware of unlinked anonymous surveys, and that only a quarter disagreed with them.¹ People mainly become aware of the unlinked surveys when they attend participating clinics where patients are informed of the surveys by posters and leaflets. Patients who have concerns can discuss them with staff (an option not available in the door to door survey), and the proportion who finally object to their specimens being included is low.⁴ The percentage reported to be aware of the surveys is considerably higher than expected. It is higher for adults aged 25 to 54 years—those most likely to have attended a participating antenatal or genitourinary medicine clinic since the surveys began.¹

There is broad professional and public acceptance of the unlinked anonymous technique. In 1996 the chief medical officer's expert advisory group on AIDS reviewed the surveys and concluded again that they were ethical and should continue as they provide essential information on public health. ⁴

Foëx, Bernard A (Hope Hospital, Salford M6 8HD bfoex@zen.co.uk).

Editor—Kessel et al found that most (69%) respondents to their questions were not aware of the anonymous testing of blood for HIV and other diseases. They also found that 26% of respondents disagreed with the practice of testing.^1-1

The authors said that the policy should be reconsidered by the government in the interests of rebuilding confidence in the NHS. They also said that the ethical debate had changed because the HIV epidemic had not materialised. Have the ethical issues changed with time? In 1966 Henry Beecher wrote, “An experiment is ethical or not at its inception; it does not become ethical post hoc—ends do not justify means. There is no ethical distinction between ends and means.”^1-2 For him neither the passage of time nor the results obtained changed the ethics of a study.

Is anonymous testing of blood for HIV and other diseases unethical? Does such a policy breach the four principles of biomedical ethics—namely, autonomy, non-maleficence, beneficence, and justice?^1-3 It may be argued that anonymous testing breaches patient autonomy. However, the blood was taken, with consent in the first instance, for a particular purpose. It was used for that original purpose. There is no direct maleficence to the patient from its secondary use. There is no direct benefit either. The policy does not breach the principle of justice: anyone's blood might be used, as would have been the case even if there had been an HIV epidemic. The use of stored tissues for research remains controversial. At least one authority has argued that consent is not needed to use stored tissue, and by implication blood, from anonymous donors for research.^1-4

The anonymous testing of blood for HIV and other diseases seems hardly comparable to the Tuskegee “bad blood” study. As only a minority of respondents disagreed with the policy, is there really a case to suggest that the testing should stop?

Kessel, Anthony (Epidemiology Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT anthony.kessel@lshtm.ac.uk).

Editor—As well as explaining how unlinked anonymous testing is done, Nicoll et al reiterate that the results of the HIV seroprevalence surveys are of value to public health and that the methodology has been widely approved. None of this is contested in the original paper.^2-1 Nicoll et al also point out that refusals have been rare, but this tells us only about those people who are aware of the testing. The true number of objections may be much higher.

There is clearly a need to widen the debate. Important questions remain. Under what conditions is it acceptable not to let people (explicitly) know what is happening with their tissue samples? Does the ethical debate change as circumstances change? What are the best ways to engage the public in research and development and what are the consequences of not doing so? What are the best means of improving trust in the NHS? We have argued for a stronger emphasis on education and a need for further research. In time this may encourage embracing of some of these important issues.

Foëx makes a number of misplaced claims. Firstly, the ethical validity of the surveys at their inception is not questioned, only the validity now. The original moral evaluation does not change with what we have learned subsequently, but undoubtedly the ethical debate shifts with time. Foëx adds that there is no moral distinction between ends and means—a statement that will sound strange to utilitarians or Kantians.^2-1

Next, Foëx's statement that there is no direct maleficence from unlinked anonymous testing of blood takes a narrow view of harm. After all, patients may feel psychologically harmed from blood being used without their agreement, and lack of openness may (indirectly) damage relationships between patients and healthcare workers and adversely affect trust in organisations such as the NHS.^2-2 The simplistic “four principles” approach of biomedical ethics has been extensively criticised.^2-3,2-4

Lastly, Foëx points out that our research is not comparable to the Tuskegee “bad blood” study. We have made no attempt to draw a comparison, and a literature search of PubMed showed many other medicoscientific papers with “bad blood” in the title, covering a range of topics.