Commentary: Why we did not seek informed consent before testing patients for HIV

Commentary

We agree completely with the Nuremberg code and the Helsinki declaration that informed consent is an essential prerequisite for medical research. However, we believe that there may be extraordinary circumstances when this right may be waived. We identify four crucial requirements that must be fulfilled before research without informed consent may be permitted.

Requirements that must be satisfied before research without consent

1. It is impossible to obtain informed consent

Eighty five per cent of admissions to our unit are emergency cases. These patients cannot give informed consent because they are critically ill. A second option is to obtain consent from a relative. In our study this would have resulted in two possible scenarios. Firstly, if the patient survived he or she could choose to be informed about the result of HIV testing and maintain the right to limited disclosure. But if the patient died the relative would have the right to know the result. This would be a serious breach of patient autonomy. Furthermore, such disclosure of results obtained in the course of research when there was no risk of infection to the relative would represent an unacceptable breach of patient confidentiality.1 It was therefore not appropriate to seek consent from relatives. The third option was to obtain consent on discharge. This would have excluded all patients who died, which would have profoundly limited the value of the study.

2. The research is of sufficient importance that patients' right to informed consent may be waived

The problem of HIV and AIDS in South Africa has reached epidemic proportions.2 3 4 By the end of l992 over 300 000 people were infected.5 In 1994 the figure was estimated to be 1.2 million.6 Seroprevalence in the antenatal clinic at our hospital was 12% in 1992 and 23% in 1996 (A N Smith, personal communication). If the worst case scenario materialises, by 2010 it is estimated that 28-52% of all deaths will be related to HIV infection.7 The impact of the epidemic on scarce intensive care resources is likely to be profound. Our 2000 bed hospital is served by 25 intensive care beds (16 in the surgical unit). Furthermore, our unit is the primary referral intensive care unit for the province of Kwazulu-Natal. As a consequence of excessive demand and our limited resources one fifth of all patients referred to our unit are denied admission. Hence given the extent of the HIV epidemic it was essential that any decisions regarding allocation of resources should be based on objective data and not subjective impression (the ethical principle of social justice). The study was therefore deemed to be of sufficient importance to waive patients' right to informed consent.

3. There must be unanimous agreement among appropriate individuals and groups that the aforementioned conclusions are valid

The exhaustive procedure followed in verifying the suitability of the protocol shows that we satisfied the third prerequisite–namely, that there must be unanimous agreement among appropriate individuals and groups about the importance of the research and the impracticability of obtaining consent. In order to pre-empt prejudice against HIV positive patients (and therefore prevent breach of two other principles of medical ethics–namely, beneficence and non-maleficence) and in view of the above considerations it was deemed essential that a prospective blinded trial should be conducted. We consulted three clinical departments, two laboratory departments, and two international AIDS experts. The institutional ethics committee appointed a subcommittee comprising Dr E M Barker (bioethicist and principal author of the Medical Association of South Africa guidelines), Professor D J Pudifin (clinician and AIDS expert), and one of us (SB) to investigate the most suitable approach. Eighteen months after initiation and deliberation among the various parties concerned the protocol was finally approved by the ethics committee.

4. Every attempt must be made to protect patients' interests after enrolment

Every effort was made to protect patients after enrolment. Their HIV status was not disclosed to staff members lest disclosure might result in discrimination. Patient care was never influenced by knowledge of HIV status. HIV status of patients was not disclosed to relatives, and the results were used exclusively for the study. On completion of the study patients' HIV test results were permanently removed from the hospital records.

Conclusion

HIV and AIDS raise unique ethical considerations, which are not limited to patient autonomy but encompass the three other principles of medical ethics (beneficence, non-maleficence, and social justice). In adhering to these three principles we breached the first principle. Our decision to embark on this study was not taken lightly. On the contrary, every attempt was made to ensure that the decision was correct in the light of our unique circumstances.