OBJECTIVES: To determine the accuracy of routine identification of Aboriginal and Torres Strait Islander women confining at King George V (KGV) Hospital, located in Sydney, Australia. DESIGN: Interviewer-administered survey. PARTICIPANTS: Consecutive sample of women who delivered live, well infants from May to July 1999. MAIN OUTCOME MEASURE: Comparison of hospital documentation compared with confidential self-disclosure of Aboriginal or Torres Strait Islander status to a female Aboriginal health professional. RESULTS: Of 536 women in our sample, 29 (5%) self-disclosed as being Aboriginal or Torres Strait Islander. Only 10 of these were identified as Aboriginal or Torres Strait Islander in hospital records (p<0.001). While specificity as determined by us was 100%, sensitivity was low (34.5%). Those Aboriginal and Torres Strait Islander women referred by another organisation were significantly more likely than those who self-referred to the hospital to be correctly identified (p=0.011). Only 1% of non-Aboriginal women indicated they would have objected to an explicit question by staff about their Aboriginal or Torres Strait Islander status. CONCLUSIONS: Routine identification significantly under-represents Aboriginal or Torres Strait Islander women giving birth at an urban obstetric hospital. We recommend the development and use of a sensitive but also specific series of questions to ensure women always are given the opportunity to disclose their status, especially as few women appear to mind such questions.
OBJECTIVES: To determine the accuracy of routine identification of Aboriginal and Torres Strait Islander women confining at King George V (KGV) Hospital, located in Sydney, Australia. DESIGN: Interviewer-administered survey. PARTICIPANTS: Consecutive sample of women who delivered live, well infants from May to July 1999. MAIN OUTCOME MEASURE: Comparison of hospital documentation compared with confidential self-disclosure of Aboriginal or Torres Strait Islander status to a female Aboriginal health professional. RESULTS: Of 536 women in our sample, 29 (5%) self-disclosed as being Aboriginal or Torres Strait Islander. Only 10 of these were identified as Aboriginal or Torres Strait Islander in hospital records (p<0.001). While specificity as determined by us was 100%, sensitivity was low (34.5%). Those Aboriginal and Torres Strait Islander women referred by another organisation were significantly more likely than those who self-referred to the hospital to be correctly identified (p=0.011). Only 1% of non-Aboriginal women indicated they would have objected to an explicit question by staff about their Aboriginal or Torres Strait Islander status. CONCLUSIONS: Routine identification significantly under-represents Aboriginal or Torres Strait Islander women giving birth at an urban obstetric hospital. We recommend the development and use of a sensitive but also specific series of questions to ensure women always are given the opportunity to disclose their status, especially as few women appear to mind such questions.