A confidential con job

Andrew Leigh Brown fears the implications of a court's use of personal details and blood samples given in confidence for research.

In March, a man who had admitted having lied about his HIV-positive status before having unprotected sex was convicted of a serious common-law offence and imprisoned for five years by the High Court in Glasgow. This established a unique legal status for HIV infection that returned it to its pariah state of the 1980s. As a consequence, we can expect a big drop in uptake of HIV testing, with significant harm to HIV prevention programmes. Vital efforts to convince at-risk individuals to have a test have been undermined because, while you might be imprisoned for lying about your status, you cannot be prosecuted for not knowing it.

The implications of the decision for medical practice relating to HIV are manifold, but of wider concern is the way in which the prosecution was brought and critical evidence gained. This case of heterosexual transmission does not stand out in any significant aspect. Two individuals had casual sex. One lied about his sexual health. The other chose to ignore safe-sex advice and agreed to unprotected sex. A relationship developed that, from evidence given at the trial, included high-risk sexual practices. By consenting, the complainer must have known she increased her risk.

A second issue, of wider significance, arises from the the way in which evidence was collected. Although the basic facts were admitted, there remained a legal requirement for corroboration. This was addressed by the seizure of confidential medical and research records. The individuals concerned had contributed, in confidence, information and blood samples important for a research study performed during 1993-95 with a remit to determine the rate and extent of spread of HIV infection in a Scottish prison. The study, published in the British Medical Journal in 1997, established the scale of the prison outbreak and, by including the sample from the complainer, showed that the viral strain associated with the outbreak could be spread by sexual contact as well as by needles.

The samples and information obtained in the study were coded in a way that only the physicians directly responsible for the patients knew their identity. Outside the clinic, following normal practice, only codes were used to identify the participants. None of those involved - patients or research workers - could have imagined that confidentiality arrangements so carefully established could be completely ignored, such that those responsible for the research study became the unwilling contributors to the defendant's conviction.

When the prosecution was initiated in 1997, a warrant was issued permitting the seizure of materials considered by the detective sergeant to be relevant. The investigation started by identifying the doctors who saw the individuals concerned, and from their medical records obtained the codes used in the study. Then the scientific evidence was seized, allowing the information that established the similarity of the virus in the two individuals to be decoded.

It was convenient for the prosecution that the defendant had taken part in the study and that standard research practice involves the storage of such data for several years after the results are published, but this case has revealed that confidentiality agreements provide no legal protection in Scotland. What are the likely consequences for other medical studies? Preventive efforts in HIV, like those used in other sexually transmitted diseases that involve contact tracing, clearly involve the possibility of self-incrimination, which must be rendered virtually impossible.

In this case, although the data were already available, the police were dissuaded from seizing the blood samples themselves only when their hazardous nature was explained. Now, it would seem to be unethical to perform similar studies of variation among HIV strains without explaining to the patients explicitly that if they participate and evidence that they have transmitted the infection is obtained, they are potentially liable for prosecution. This is not likely to encourage recruitment.

Beyond studies of viral infections, there are much wider implications because every human blood sample carries the genetic signature of the individual donating it. Huge comprehensive studies of genetic variation in the human population planned by the Medical Research Council and the Wellcome Trust will establish databases of large numbers of genetic variants using single nucleotide polymorphisms. Seizure of such information will provide a detailed genetic signature of an individual and could be used as forensic evidence.

These databases will have an important role to play in developing our understanding of human genetic variation and disease, but what will protect them from seizure by legal authorities? Confidentiality and consent agreements are becoming more formal, but they are still much less so in the United Kingdom than in the United States. Evidence obtained the way it was in this case would be inadmissible in a US court, and the US institutions responsible for the research also bear responsibility for preventing violation of confidentiality agreements, providing appropriate legal advice and action in court if necessary.

UK medical research is bound by the Helsinki declaration, in which the principle of confidentiality and "the right of research subjects to safeguard their integrity" are explicitly incorporated, as well as by the Data Protection Act. The law in both Scotland and England is also subject to the European Court, which might take a different view on confidentiality, although it does not appear likely that this will be taken forward as a test case. Thus, until we know why one case of HIV transmission out of many was chosen to establish this precedent, we cannot know how much of a threat this breach of confidentiality poses to current and proposed medical research studies. In view of the scale of existing and proposed studies, the need for clarification of the legal status of consent agreements and confidentiality in medical research is urgent.

Andrew J. Leigh Brown is professor at the University of Edinburgh and director of its Centre for HIV Research. He was responsible for part of the study referred to and was an expert witness at the trial.