The selection process for Indigenous-Australian trials is presented in Fig. 1. We identified 139 trials from 9206 All-Australian trials that were focused on Indigenous health (ANZCTR: 135; CTgov: 4). The remaining 9067 trials were termed General Australian (ANZCTR: 8131; CTgov: 936). An overview of results is provided in Table 2.
Selection process for Indigenous-Australian trials
Table 2 Characteristics of Indigenous-Australian trials compared to general Australian trials
Over the ten-year study period, the absolute number of Indigenous-Australian trials increased from 12 in 2008–09 to 25 in 2017–18 (see Fig. 2). There was no significant increase in the proportion of Indigenous-Australian trials per year when compared against General Australian trials (χ2 (df) = 8.11 (9), p = 0.52) (Fig. 2). The total participant sample size for Indigenous-Australian trials was 155,694, which constituted 5.73% of the recruitment to the corresponding All-Australian trials (2,717,031) in the ten-year period. There was also no significant increase in the participant sample size of Indigenous trials when examined in proportion to All-Australian trials (see Supplementary Fig. 1). The median participant sample size of Indigenous-Australian trials (n = 250, IQR 100–535) was considerably larger than for General Australian trials (n = 60, IQR30–140) (Mann-Whitney U = 297,250, p < 0.001). There was no clear trend in median sample size over time for Indigenous-Australian or General Australian trials (post-hoc analysis, Supplementary Table 1). Indigenous-Australian trials were more likely to list public health as an area of study (59/139, 42%) compared to other Australian trials (958/9067, 11%) (OR 6.24, 95% CI 4.41–8.78).
Percentage of Indigenous-Australian trials as a proportion of All-Australian trials (left) and absolute number of Indigenous-Australian trials (right) per registration year from 2008 to 2018
Allowing up to 14 registered conditions, the median and average number of health conditions registered per Indigenous-Australian trial were 1 (IQR1.00–2.00) and 1.62 (SD0.81) respectively, similar to General Australian trials where median was 1 (SD0.80) and mean was 1.60 (IQR1.00–2.00). The most frequently listed health conditions studied in Indigenous-Australian trials were mental health (28%), cardiovascular disease (20%) and infection (16%), compared to other Australian trials which were mental health (24%), cancer (16%) and cardiovascular disease (11%). The top 14 most frequently studied conditions for Indigenous and General Australian trials are shown in Fig. 3. Between 2008 and 2018, Indigenous-Australian trials were more likely than General Australian trials to study ear conditions (OR 20.26, 95% CI 10.32–37.02, p < 0.001), infection (OR 3.11, 95% CI 1.88–4.85, p < 0.001) and reproductive health (OR 2.59, 95% CI 1.50–4.15, p < 0.001). They were less likely to focus on musculoskeletal conditions (OR 0.09, 95% CI 0.00–0.37, p < 0.001), anaesthesiology (OR 0.16, 95% CI 0.01–0.69, p = 0.021) and surgery (OR0.17, 95% CI 0.01–0.73, p = 0.027). Health conditions that were most and least commonly studied by Indigenous-Australian trials compared to other trials are displayed in Fig. 4.
Top 14 conditions studied in Indigenous-Australian trials compared to General- Australian trials registered 2008–2018. Numbers within the bars are the percentage of trials in that category. Note that multiple conditions may be selected per trial therefore the percentages do not add to 100
Odds ratios of conditions studied in Indigenous-Australian trials, compared to General Australian trials, 2008–2018
Regarding the use of randomisation (Table 2), Indigenous-Australian and General Australian trials did not significantly differ: 70% (97/139) and 74% (6666/9067) were randomised respectively (OR 0.76, 95% CI 0.53–1.11, p = 0.148). Blinding was less common in Indigenous-Australian trials (35%, 48/124) compared to General Australian trials (47%, 4273/9067) (OR 0.58, 95% CI 0.40–0.83, p = 0.003).
The most common categories of interventions studied in Indigenous-Australian trials were prevention (48/139, 36%) and behavioural interventions (41/139, 30%) (Table 2). Compared to General Australian trials, Indigenous-Australian trials were significantly more likely to focus on screening (OR 3.57, 95% CI 2.10–5.70, p < 0.001), prevention (OR 2.24, 95% CI 1.61–3.08, p < 0.001) and behavioural interventions (OR 1.58, 95%CI 1.11–2.20, p = 0.009) and less likely to focus on rehabilitation (OR 0.38, 95% CI 0.13–0.83, p = 0.021) and treatment (OR 0.40, 95% CI 0.30–0.52, p < 0.001).
The most common sponsors of Indigenous-Australian trials were universities (n = 72, 53%), individuals (n = 16, 12%) and government bodies (n = 14, 10%) shown in Table 2. For funding, Indigenous-Australian trials had higher rate of government (OR 2.90, 95% CI 1.57–4.93, p < 0.001) and universities (OR 2.30, 95% CI 1.63–3.24, p < 0.001) support and less funding by hospitals (OR 0.10 95% CI 0.02–0.25, p < 0.001) and industry (OR 0.20, 95% CI 0.05–0.53, p = 0.002). Additionally, only 11.5% (n = 16) of Indigenous-Australian trials had some form of industry involvement (as either a sponsor, collaborator or funder) compared to 24.9% (n = 2255) of General Australian trials (OR 2.52, 95% CI 1.54–4.43, p < 0.001).
The AIHW data highlighted the conditions that contributed most to the burden of disease for Indigenous Australians (see left side of Fig. 5). Our analysis of the frequency that various conditions were studied in Indigenous-Australian trials shows that studied conditions do not necessarily align with research priorities. For example, whilst cardiovascular and mental health conditions were studied with high frequency in Indigenous-Australian trials between 2008 and 2018, which broadly reflects their contribution to the burden of disease, other conditions such as injuries and musculoskeletal disorders were studied less frequently than would be expected relative to their burden of disease. In terms of funding for research in the top ten burden of disease groups, government bodies were the most common funding sources, as shown in Supplementary Table 2 which divides funding into conditions from priority and non-priority areas. In comparison, industry funding for Indigenous-Australian trials was less common, and this affected conditions listed as priority areas (n = 12/162, 7.4% conditions funded from industry) and non-priority areas (n = 3/83, 3.6% conditions funded from industry). On examining the type of interventions used to address top ten burden of disease areas (see Supplementary Table 4), a high proportion of trials studying mental health conditions evaluated behavioural interventions (n = 22), whereas drug-related interventions were scarce, and studied mostly in cardiovascular (n = 5) research.
Comparison of the percentage of total burden of disease measured in DALY as a proportion of total from AIHW Burden of Disease study for Indigenous Australians (10) to percentage of Indigenous Australian trials studying various health conditions registered on ANZCTR and ClinicalTrials.gov