Abstract
Aim: Our aim was to assess the quality of ethnicity data in the Health New Zealand - Te Whatu Ora human resources (HR) databases.
Methods: This project involved two components. 1) Staff with "not stated" or missing ethnicity in their HR records were identified from Auckland and Waitematā district HR databases on 30 April 2018 and 2 August 2018. They were asked their ethnicity using the Standard questions via an online survey. 2) Staff data were extracted in June 2017 and linked to the National Health Index (NHI) ethnicity data. The concordance of ethnicity data between the two datasets was assessed in three categories: exact match, partial match or total mismatch.
Results: 1) Of the 17,539 staff, the proportions with "not stated" ethnicity were 15.1% at Auckland district and 6.4% at Waitematā district. Among those, 727 Auckland staff and 122 Waitematā staff responded to the survey to update their ethnicity. These respondents most identified as European (64%), followed by Asian (15%) and Pacific and Māori (5% each). 2) Of the 17,539 staff, 86% had matched ethnicity between the HR dataset and the NHI dataset (kappa 0.77, p<0.0001), with the highest agreement level being Asian (93%), followed by European (86%), Pacific (84%) and Māori (83%).
Conclusions: This project assessed the extent of "not stated" staff ethnicity data and misclassification in two large health districts. Staff with "not stated" on their records were willing to provide their ethnicity data when asked the Standard question. This project suggests the need for quality improvement activities in recording HR ethnicity data to support planning and monitoring workforce diversity.