We conducted a study on research culture and capacity of all health personnel (medical, nursing and allied health) located at the central river city district of the Commission’s region. 11 The study area included two upcoming health and education innovation precincts at Westmead and Blacktown12, along with Auburn, Mt Druitt and Cumberland hospitals and community health centres. Together, they formed part of the Western Sydney Local Health District, which is one of the 15 health administrative districts of NSW Health.13 Geographically, it also includes four local government areas of Parramatta Council, Blacktown Council, Cumberland Council and the Hills Shire. Figure 1 provides an illustration of the study area; locating the study area within the Greater Sydney Region and identifying the upcoming health and education innovation precincts.
The Universities of Sydney and Western Sydney are the main higher education universities in the study area and have established bases in the region. Both Universities have made considerable investments and envision the upcoming health and education precincts at Westmead and Blacktown as global centres for excellence and multidisciplinary innovation14.
Approximately 7150 clinical personnel (2079 medical; 4100 nursing and 968 allied health) were employed in the public health sector (i.e. Western Sydney Local Health District) at the time of the study. Overall, the public health sector serves over one million people in the region. Over 43 percent of the population in the region were born overseas, and about 45 percent speak a language other than English at home.15 The region is also the home to the highest urban aboriginal population in NSW (about 13,400 aboriginal people based on 2016 census).15
All health personnel (medical, nursing and allied health) employed at the study area were invited to participate in a survey through an online questionnaire via a Research Electronic Data Capture (REDCap) system . The survey was administered between November 2016 to January 2017. A range of data items, including demographic, work status and research biodata were collected. Research culture (outcome variable) was assessed using a battery of 51 items, classified across three domains: individual level (14 items); team level (19 items) and organisational level (18 items). Health professionals (medical, nursing and allied health); gender (male, female); age groups (less than 35 yrs, 35-45 yrs and 55+ yrs); educational qualification (undergraduate, graduate, higher degree research); team role (clinician, management, teaching/research); experience years (5 yrs or less, 6 to 15 yrs, and 15+ yrs); enrolled in a research program and having research in role description - were used as covariates.
Respondents were asked to rank their level of skill/ confidence to each item/statement of the research culture domains ranging from 1 to 10, with 1 being in least agreement and 10 being in strong agreement to that item/statement. Responses were later dichotomised into two groups using mean scores of < 6 as cut-off point (scores < 6 were considered to be a lower level of skill/confidence in the concerned scale/domain and coded as 0; scores of ≥ 6 were considered have a higher level of skill/confidence with the concerned scale/domain, and were coded as 1). These dichotomised scales were first examined by sample characteristics using chi-square tests, and a level of significance set at p<0.05. Characteristics found significant at any research culture domain were included as covariates in the multivariate log-binomial regression models. Adjusted odds ratios were generated. Ethics approval for the study was obtained from an approved Human Research Ethics Committee in Australia. In this viewpoint, we have presented results of the logistic regression analysis; a prior publication has examined the descriptive nature of the findings from the study.11
Table 1 presents the findings from the regression analysis of the dichotomised research culture domains with selected sample characteristics. Adjusted log-binomial models are presented for each domain. Respondents with a higher degree by research qualification had a consistently higher odds ratio of having a higher level of skill/confidence the three research culture domains compared with respondents with an undergraduate or graduate qualification. Respondents engaged in teaching and research, also had a higher odds ratio of identifying higher skill level / confidence for the team and organisational research culture domains, compared with respondents mainly involved in clinical and management/executive tasks. A further finding was respondents not having research within their role description were less likely to identify higher skills / confidence for each of the research culture domains, and this was significant at the individual and team domains.
Only health personnel employed by the public health system (Western Sydney Local Health District) were surveyed. Private hospitals in the region were not included. However, the public sector is the largest employer of health personnel in the region. Further, the study did not evaluate the research culture of the health and education precincts per se, but chose to examine only an aspect of the precincts - the health sector. This decision was conscious as the purpose of the study was to build an inclusive research culture for evidence-based practice among the health professionals.
 The Greater Sydney Commissions plan document classifies the region into five districts: eastern harbour city district, central river city district, western parkland city district, north district and south district.
 REDCap is being widely deployed in NSW Health, and used by researchers across the local health districts. The functionality is similar to other online surveys such as Qualtrics/Survey Monkey but provides secure storage functions.