Patients rarely detect breaches of confidence.

Abstract

OBJECTIVE: To measure the prevalence of patient reported breaches of confidence emanating from health services in South Australia. METHODS: A community household survey with follow up telephone interviews of those reporting unauthorised disclosures. Recruitment interviews were conducted during October and November 2001 with 3037 South Australian residents over 15 years of age. South Australian households were selected randomly within a representative selection of census collection districts. Follow up telephone interviews were conducted with 18 respondents who reported unauthorised information releases in the twelve months prior to the initial survey. RESULTS: The one year prevalence of breaches of confidence reported by patients was 0.23 +/- 0.17% (1.96 x SE). This prevalence was significantly lower than a previously published estimate (Mulligan, 2001) of 1.1 +/- 0.4% (difference between proportions p= 0.003, z= 4.415) which relied upon reports of information disclosed without the patients permission as a proxy for breaches of confidence. An incidental finding was that 0.20+/- 0.16% (1.96x SE) of respondents reported exchanges of information amongst members of treating teams which had not been authorised by the patient. CONCLUSION: True breaches of confidence arising from health care interactions are rarely detected by patients. Ordinary information sharing within treating teams may be a more important source of misunderstandings between patients and health care providers. This issue can be addressed at the point of care by informing patients what will be done with their health information and who may later gain access to it.