Reporting Indigenous status, ethnicity, language and country of birth to build equity in international paediatric clinical trials with Australian sites: a scoping review

Introduction

Considered by many the gold standard in clinical research, clinical trials utilise rigorous methodologies to explore the therapeutic effectiveness and safety of interventions, which subsequently inform evidence-based practice.^1–4 However, it must be recognised that the safety and efficacy of therapeutic interventions evaluated through clinical trials are subject to limitations imposed by the methodological approaches.^5,6

Indigenous status, race and ethnicity are formed by social, cultural, spiritual, geographical and generational histories that are well documented to influence access to health care.⁷ Constructs that categorise human differences may generate marginalisation, drive discrimination and perpetuate racism; however, in research, understanding these constructs is ethically responsible and valuable.⁸ Failure to capture and report these attributes inadvertently disregards the potential variances observed between racial and ethnic groups regarding incidence, pathophysiology, social determinants and subsequent interventional effectiveness and safety.^9–12 For example, the effectiveness of pulse oximetry in racially diverse populations has recently been questioned, with studies demonstrating poor inclusion of participants from diverse backgrounds in initial validation studies.^11,13

It is essential for equity and safety that clinical trials include diverse participants.^{6,10,12,14–18} Globally, there have been increasing efforts to improve accessibility to clinical trials for children from diverse backgrounds and ensure clinical trial demographic information is transparent.^19,20 However, the context of demographic reporting and diversity within paediatric clinical trials with Australian sites, a multi-cultural nation with highly increasing numbers of migrants, remains unclear. Therefore, this scoping review aims to review the reporting of Indigenous status, ethnicity, language and country of birth in paediatric clinical trials with Australian sites.

Methods

This scoping review is congruent with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis protocol for Scoping Reviews (PRISMA-ScR).²¹

Results

The search yielded 2717 studies, upon removal of 592 duplicates in Endnote V.20.2.0.17373 and 45 duplicates in Covidence, 2080 studies progressed to screening, as illustrated in Supplementary material Fig. S1. A total of 830 studies progressed to full text screening, and overall, 262 studies were eligible for inclusion and proceeded to data extraction in Air Table.

As demonstrated in Table 1, studies included in this review primarily consisted of RCTs (n = 158; 60.3%). Most studies only had Australian sites (n = 209; 80.2%), but of those that had international sites (n = 53; 19.8%), sites were predominately within Australia and New Zealand (n = 53; 15.9%), Europe (n = 28; 8.4%) and North America (n = 27; 8.1%). There was significant heterogeneity in the diseases and interventions examined in the included studies, however, most studies considered diseases from medicine and clinical science research disciplines (n = 211; 80.5%), encompassing medical or health interventions in clinical settings focused on therapeutic outcomes.

Overall, 154 (58.8%) clinical trials did not report Indigenous status or any of the diversity indicators of interest. For studies reporting these variables, ‘Indigenous status’ was most reported (n = 40, 15.3%), followed by ‘ethnicity’ (n = 30, 11.5%), ‘language’ (n = 27, 10.3%), ‘country of birth’ (n = 25, 9.5%), ‘race’ (n = 19, 7.3%), ‘race and ethnicity’ (n = 9, 3.4%) and ‘natural skin colour’ (n = 1, 0.4%). Table 2 demonstrates the reporting of Indigenous status, race, ethnicity, language used at home and country of birth as per author definitions. No study reported a triple-phased descriptor (i.e. reported 1: a cultural indicator such as Indigenous status/race/ethnicity; 2: language and 3: country of birth).

Table 2 also illustrates the differences observed in capturing the variables of interest for Australian only clinical trials and international clinical trials with Australian sites. While Indigenous status (n = 37, 17.7%), language (n = 27, 12.9%) and country of birth (n = 23, 11.0%) were the three most commonly reported variables in Australian only clinical trials, it was apparent that Indigenous status (n = 40, 15.3%), ethnicity (n = 30, 11.5%) and language (n = 27, 10.3%) were the most commonly reported variables in international clinical trials with Australian sites. Race was also reported in higher rates within the international clinical trials with Australian sites (n = 19, 7.3%), in comparison to Australian clinical trials (n = 1, 0.5%).

As illustrated in Table 3, the reporting of ethnicity and/or cultural background increased over the 5 years captured. No trends in the reporting of Indigenous status, language used at home or country of birth were observed over time.

Methods used to capture the variables of interest were predominately self-reported for Indigenous status (n = 14, 35.0%), ethnicity (n = 14, 46.7%), language (n = 10, 37.0%), race (n = 13, 68.4%) and race and ethnicity combined (n = 4, 44.4%). Other common methods of determining the variables of interest included parents and other methods, such as school-identified, carer-identified, educator-identified, Elder-identified or identification based on school or medical records. Failure to disclose the methods used to determine the variables of interest was common as demonstrated in Table 4.

Of the studies that reported Indigenous status (n = 40), 14 (35.0%) included only Indigenous participants. Two studies (5.0%) completed a formal subgroup analysis specific to Indigenous participants, with results specific to the Indigenous participants and intervention under investigation. In these studies (n = 2), the mean percentage of Indigenous participants were 8.3% and 8.85% respectively. Two studies (5.0%) made comments about the results of interventions in relation to the included Indigenous participants but did not report a formal subgroup analysis in the publication. In these studies (n = 2), the mean proportion of Indigenous participants were 7.4% and 14.0%. Twenty-two (55%) studies that reported Indigenous status did not complete a subgroup analysis specific to Indigenous participants and the intervention under investigation. In these studies, the percentage of Indigenous participants ranged from 0.39% to 66.7%.

When reported, terms used by study authors varied extensively with authors primarily categorising cultural indicators as ‘Indigenous status’, ‘race’, ‘race and ethnicity’, ‘ethnicity’ or ‘natural skin colour’. As shown in Fig. 1, while these terms formed the main categories used for analysis, there was significant heterogeneity in the terms reported by authors. Overall, more than 60 terms were used to categorise study participants in relation to ‘Indigenous status’, ‘race’, ‘race and ethnicity’, ‘ethnicity’ or ‘natural skin colour’.

Fig. 1.

Dendrogram of author defined categorisation and terminology.

Discussion

This scoping review has demonstrated low levels of reporting of Indigenous status, race, ethnicity, language and country of birth for children participating in international clinical trials with Australian sites. It is well established that the effectiveness and safety of clinical interventions may be influenced by ethnicity and Indigeneity indicators. Subsequently, many of these clinical trials are not generalisable and do not adequately reflect the appropriateness of interventions for an entire population.^{10,12,14,15,25} This review has provided a comprehensive and contemporaneous aggregation of clinical trial diversity reporting. The level of reporting captured in the clinical trial is lower than previously reported in prior studies evaluating other regions of the world.¹⁵ Our review demonstrated that 41.2% of clinical trials captured reported at least one diversity indicator. Marginally better reporting was observed when results considered international clinical trials with Australian study sites (n = 108/262; 41.2%), in comparison to clinical trials with only Australian sites (n = 78/209; 37.3%). This may indicate that Australia’s diversity reporting is lagging behind global benchmarks.

Clinical trial participant diversity is a socio-political, moral and ethical issue that requires prompt rectification to minimise harm and improve patient outcomes.^10,26 The low level of reporting demonstrated in this review raises questions as to the diversity of participants in those trials that did not report any diversity indicators. Thus, concerns remain surrounding access to these trials for children from Indigenous, racially and ethnically diverse backgrounds. This review demonstrated that the frequency of reporting diversity indicators was 41.2% in international clinical trials with Australian sites in paediatric or mixed paediatric/adult populations over the preceding 5 years. Similar levels of reporting were observed by Turner et al.¹⁵ who reported that overall, 43% of United States (US) clinical trials conducted between 2000 and 2020 reported race/ethnicity data, with 45% of trials reporting race/ethnicity data since 2007.¹⁵ Turner et al.¹⁵ reported that the reporting of any race/ethnicity enrolment data had increased from 26% in 2008 to 91% in 2018. Turner et al.¹⁵ also reported an increase in reporting over time, with an annual growth rate of 13.5%, which was not observed in this review. This may suggest that clinical trials with Australian sites are not accelerating in the reporting domain at the same pace as international counterparts. Despite similarities between this review and this US study, reviews conducted on United Kingdom (UK) RCTs from 2016 to 2021 have yielded far higher rates of reporting for ethnicity at 67.3%.²⁷ Currently, global interventions including the United Nation’s Sustainable Development Goals and Leave No One Behind Agenda are leading the charge to ensure adequate access to health care for everyone.^19,20 The increasing global movement to address racism has expedited reporting requirements in some countries, resulting in a global urgency to improve the transparency in the diversity of data collection and reporting for clinical trials, particularly as it pertains to Indigenous status and ethnicity.^1,6,18,26 Subsequently, the inclusivity and reporting of clinical trial population characteristics have modestly improved over the past decade internationally, such as in the UK and the US.^15,28 In particular, the Children’s Oncology Group supported by the National Cancer Institute has placed significant emphasis on ensuring diversity in paediatric cancer clinical trials, resulting in improved diversity reporting and improved access to clinical trials for diverse populations.¹⁷

Improvements to diversity reporting in clinical trials are multifaceted. Ensuring that Indigenous and ethnicity measures are appropriate to the study and study population and reported transparently with justification for the decisions made as per international guidelines should be employed by all clinical trialists. Indigenous status, race, ethnicity and cultural background are self-defined attributes and constitute social rather than biological constructs.^8,29 There is no one size fits all. It is common for a person to identify with more than one ethnicity or cultural background and subsequently it is paramount that categories are not viewed in isolation with a single absolute option.⁸ Subsequently, it is essential that clinical trialists ensure participants or parents are identifying their Indigenous status and ethnicity indicators, rather than researchers using other methods to determine these factors. The heterogeneity in terminology associated with the diversity attributes of interest reported in this scoping review reflects the diversity of clinical trial participants and forms a basis to advocate for self-identification. Furthermore, adequate reporting includes capturing multiple variables.⁸ Through capturing multiple demographic variables, participants or parents can ensure that their or their child’s representation is an authentic reflection. Engaging with consumers across all geographic clinical trial sites to co-create clinical trial diversity data collection tools may offer a simple solution to ensure diversity information is adequately captured and reported.³⁰ Ensuring transparency in research encapsulates a clear notation of how demographic variables are determined. When reporting clinical trials, methods used to capture diversity variables must be adequately defined.³¹ Clearer reporting guidelines, with consideration for socio-geographical differences across clinical trial sites and mandating the capturing of participant diversity information, are urgently required.

Conclusion

This review has highlighted low rates of reporting of diversity variables in international paediatric clinical trials with Australian sites. Clearer reporting standards, partnering with consumers to co-design trials and self-identification during data collection are required. Ensuring adequate access to clinical trials for Indigenous children, and children from ethnically, linguistically and geographically diverse backgrounds, is essential in building health equity and ensuring patient safety.

Supplementary material

Supplementary material is available online.