Feasibility of using self-reported ethnicity in pregnancy according to the gestation-related optimal weight...

To evaluate the feasibility of self-reported ethnicity using the gestation-related optimal growth (GROW) classification in a contemporary multicultural antenatal population.

Cross-sectional study.

Tertiary obstetric hospital in Melbourne, Australia.

Pregnant women attending the antenatal clinic.

We surveyed pregnant women during April-June 2016 regarding their understanding of the term 'ethnicity', and how they would classify the ethnicity of themselves, their partner, and family members according to the Australian GROW classification.

Two hundred and thirty-five women completed the survey. When describing 'ethnicity', most women (103, 44%) chose multiple descriptors, most frequently country of birth (54%) and region of ancestry (47%). Interpretation of 'ethnicity' varied significantly between ethnic groups: those choosing 'country of birth' were more likely to identify as Indian (odds ratio, OR 3.5, P = 0.03), whereas those choosing 'physical appearance' were more likely to identify as Chinese (OR 3.0, P = 0.047). Thirty participants (13%) were unable to describe their ethnicity from the available GROW options. Sixty-one (26%) respondents' ethnicity was inconsistent with that of their parents' heritage. A further 35% had a partner of different ethnicity. The agreement between country of birth and self-reported ethnicity was only fair (kappa 0.73, 95% confidence interval, 95% CI 0.64-0.82).

This study confirms the complexity of defining ethnicity in contemporary multicultural settings. Self-reported ethnicity is often inaccurate, concepts of ethnicity vary by ethnic group, and country of birth is a poor descriptive surrogate. Adjustment for maternal ethnicity should be undertaken with caution in the customised assessment of fetal growth.

Is self-reported maternal ethnicity reliable? We think not.