Aims: This paper reports on the methods used in two studies to obtain access to subjects to comply with the common law duty of confidence laid out in the Data Protection Act (1998) and discusses the researchers' problems in interpreting the procedures.
Rationale: The amendments to the United Kingdom (UK) Data Protection Act (1998) are causing confusion within the health service and academic institutions. There is a need to balance patient confidentiality with the requirement to conduct vital, unbiased research in which health service professionals are not subject to ethical dilemmas. This paper examines the recruitment methods used in two studies in which the researchers' attempts to adhere to the requirements lengthened the study costs and may have produced less reliable results.
Methods: The methodological difficulties in two studies are presented. In Study 1, the difficulties encountered when the Multicentre Research Ethics Committee refused permission for researchers to recruit patients directly to a multicentre randomized controlled trial are discussed. In Study 2, the method used to compile a sampling frame for a national questionnaire survey following the eight principles of the Act are described.
Findings: Our experience has shown that health care professionals are increasingly required to recruit patients to intervention trials, and that researchers are not allowed access to the names of patients or other subjects to ask them for consent to participate in a study. The requirement for researchers to use "intermediaries" to obtain consent from and recruit subjects to studies increases the risk of selection bias, may expose the practitioner to ethical difficulties and may compromise the external validity of trial results. There is also a danger that research costs will soar when the Data Protection Act (1998) is fully realized.
Conclusion: The Data Protection Act (1998) is currently being interpreted in a number of different ways. We conclude there is an urgent need for consensus within the health service and academic communities.