Capacity building at an Australian regional clinical trials unit: determinants of growth and correlation with a capability framework Authors: Lisa A Bell, Rebecca Gurnett, Carmel E Goss, Geoffrey Chong, Ethan L Pitman, Michael Tuohill, Ashleigh Clarke, Virginia Murphy, Wasek Faisal Published: 2026 ABSTRACT AbstractBackgroundExpanding the capacity for clinical trials in regional/nonmetropolitan locations will increase accessibility for patients who live remotely, however there are limited studies on the growth of regional clinical trials units over time. This study aims to explore the status of the growth of a regional Australian clinical trial unit between 2011 and 2022, utilising a recently published clinical trial capability framework.MethodsAn evaluation of a regional clinical trial unit in Victoria, Australia, between 2011 and 2022 was conducted to objectively describe growth, including the number of trials, number of patients and number of principal investigators (PIs). We used an objective capability framework published by Woollett et al. (2023). The framework uses a maturity model applied across several domains. Descriptive statistics are reported as totals and means. Scatter plots, pairwise correlations and Bland–Altman adjustments for trends were computed to show correlations and agreements between the framework average score and objective measures of the clinical trials between 2011 and 2022.ResultsThe number of clinical trials active in 2011 and 2022 were n = 4 and n = 70, respectively. Across all domains, the Grampians Health Clinical Trials Unit was assessed as ‘established’ in 2011, whereas the trial unit was considered ‘high performing’ in 2022. The framework capability score correlated well with the number of PIs per year (r (10) = 0.9, p < 0.001), the number of open trials per year (r (10) = 0.9, p < 0.001), and the number of patients (r (10) = 0.8, p < 0.001). The key contributors to growth identified were organisational support and infrastructure. Growth was largely driven by oncology trials.ConclusionSignificant growth was observed for a regional clinical trial unit across a decade. The framework score strongly correlated with objective clinical trial data. Understanding the growth of clinical trial units in nonmetropolitan areas is essential for developing the capacity of clinical trials in regional and rural areas.Keywords: Clinical Trials, Non-metropolitan, Capability, Capacity, Capability framework Background Clinical trials are considered the gold standard of evidence-based healthcare [1]. Participation in randomised clinical trials tends to be associated with beneficial health outcomes and not associated with harm compared with those who do not participate [2]. According to the Australian Statistical Geography Standard, regional and rural areas encompass all areas outside metropolitan cities [3]. The prevalence of clinical trial participation among regional (nonmetropolitan populated region) and rural patients is lower than that among their metropolitan counterparts [4]. Moreover, those living in regional and rural areas, including cancer patients, experience poorer health outcomes [5, 6]. Malignant neoplasms are the most prevalent health subcategory of clinical trials [7]. Improving access to clinical trials can be expected to increase treatment options and improve health outcomes [8, 9] in potential clinical trials’ participants that disproportionately affect regional and rural Australians. To offer clinical trial options to patients closer to home, the Grampians Health Clinical Trials Unit (GHCTU) in Ballarat, a nonmetropolitan/regional city in the central highlands of Victoria, Australia, has progressively increased in capacity and capability since its initiation in 2004. Travel distance, cost, and inconvenience are commonly reported as barriers to remote patients accessing clinical trials [10]. The establishment and growth of regional and rural clinical trial units are important strategies for increasing the accessibility of clinical trials. However, there is a lack of publications that objectively describe the growth trajectory of regional or rural clinical trial units over time. To assess clinical trial growth and capacity, there are two published frameworks: the Canadian Remote Access Framework for Clinical Trials (CRAFT) and an Australian maturity model developed by the Alfred Health TrialHub [11, 12]. CRAFT provides a risk-based approach to assist in the delegation of activities within a hub and spoke clinical trial model [9]. Inspired by the CRAFT framework, a clinical trial unit capability framework was published to assess and inform the fundamental requirements for development, growth, and longevity in Australian regional and rural areas [12]. The framework developed by TrialHub was based on data collected in Victoria from 2020 to 21 and is the first publication to provide a maturity model in the clinical trial capability setting [12]. This analysis aims to (i) explore the status of the growth of a regional clinical trial unit between 2011 and 2022, (ii) evaluate the capacity of a clinical trial unit based on of a published framework for regional sites, and (iii) compare objective measures of clinical trial unit growth with the average framework score to characterise organisational development over time. Methods To collect objective descriptive data on the growth of the GHCTU, paper and digital annual reports submitted to Grampians’ health governance between 2011 and 2022 for clinical trials and registries were examined by three authors (LB, EP and MT). Data was obtained from Grampians Health Ethical Review Manager Software (Infonetica), and scanned paper documents submitted to the governance office. Data was extracted to excel spreadsheet by authors EP and MT, and LB double checked data extracted, helped find and reconcile missing documents, and in the event of any disagreements between EP and MT helped reach a consensus. Clinical trials were defined as studies which have focused on novel tests and treatments and their effects on human health outcomes [13]. A registry study is defined as any observational study to evaluate outcomes for a specific disease, condition, or exposure that does not include a novel treatment intervention [14]. If a study was commercially sponsored i.e. by a for-profit company it was defined as “industry sponsored”. Studies sponsored by an individual investigator or collaborative group of investigators were defined as “investigator-initiated”. All studies conducted at GHCTU fall within these two categories. The data within these documents included the open date, close date, type and phase of the clinical trial, Principal Investigator (PI), Study Coordinators (SC), and medical specialty of the clinical trial (i.e., oncology). The study open date was defined as the date opened listed on the annual report, and the study close date was the date on which the letter close to the study was submitted to governance. The number of PIs and SCs was the sum of individuals listed as PIs and primary SCs. Sub-Investigators and backup study coordinators were not captured in the annual reports, and the working hours of staff were not collected. Where a progress report was unable to be located, the data from the previous year were used, i.e., in the first year of a trial being open, one participant was recruited to the trial, but two participants were withdrawn, and the number of patients in the trial was recorded as two, not one. It is likely that a participant was recruited and withdrawn within the year of that progress report. Where no additional source data could be located, the data were coded as missing. If a formal annual report was unavailable or incomplete, onsite records were used as source documentation. Surveys of GHCTU resources, processes and practices were evaluated via a capability framework for outer metropolitan and rural areas proposed by Woollett et al., 2023. The framework was used retrospectively assess trials unit capacity, not prospectively to inform organisational decisions or implementation interventions. The Australian framework was selected over CRAFT to assess capacity for the GHCTU, as it is regionally appropriate and not specific to hub and spoke model clinical trials. The framework uses a maturity model that allows the assessment of the capacity of a clinical trial unit by assigning a score (0–4) across several domains, including trial infrastructure, leadership and culture, technology, skilled people, professional networks and collaborations, and organisational support. A survey of the GHCTU was conducted to assign a score for each domain and the average sum of the scores was used for an overall assessment: formative (0), developing (1), established (2), high performing (3), and leading (4). Some survey items within the framework did not provide individual scores but instead combined them. For example, for survey items for office and clinic spaces, the available options are scored as 0, 1–3 or 4. There is no guidance on how to compute combined scores, so for consistency in the analysis, the highest score in a given range was recorded for analysis, i.e., if the score was 1–3, then 3 was recorded for analysis. Using the highest score was selected as an appropriate objective method to address this issue, although it may overestimate scores. If the lowest score was assigned, the bias would be equal but inverse. If reviewers gave their best estimate within the range without any predefined criteria, the score would be neither objective nor reproducible. The higher range of the score rather than the lowest was selected, as points should be awarded if the capability of the clinical trial unit met the criteria. The framework was completed independently by four GHCTU staff (AC, CG, RG and VM), including the GHCTU manager, who has overseen the unit for the duration of the study period. All four reviewers compared scores and reconciled differences to form a final agreed score used in this analysis. Statistical analysis The objective measures of the GHCTU included the number of clinical trials, number of registry studies, PI, SC, medical specialties (i.e., oncology, cardiology, etc.), and patients per year. Total studies per year combined registry and clinical trials. Descriptive statistics for all objective measures are reported as totals and means. Framework scores were reported as the mean score per category and the mean total score for each year. To compare objective measures of the GHCTU score and framework score scatter plots, pairwise correlations and Bland–Altman adjusted for trends were computed to determine the correlation and agreement between the framework score and the number of clinical trials per year and the number of clinical trial patients per year [15]. Proportional bias was evaluated by conducting regression analysis between the difference on the mean, regression-based limits of agreement were used when a significant trend was detected. A strong correlation between objective measures and the framework score was considered r = 0.8 and p < 0.05, and acceptable variance in Bland–Altman analysis was set at ≤ 5% outside limits of agreement a priori. Three sensitivity analyses were conducted; (i) repeat correlation analyses using the lower bound of framework score ranges i.e. (if the item scoring range is 0–3, 0 is input rather than 3), (ii) Spearman vs. Pearson’s correlation analysis, (iii) leave one year out correlation analysis for each year. Results Data were available for 122 studies across 12 years. For number of participants per year 5/122 studies were missing data for one or more years (4%), no other variables had data coded as missing. The number of active trials increased across all objective measures of the GHCTU between 2011 and 2022 (Fig. 1). In 2011, four research studies were active, all of which were clinical trials; there were no active registry studies. In 2022, there were 82 active research studies, of which 14 were registries and the remaining were clinical trials. The number of participants increased 61-fold from n = 4 clinical trial patients and no registry participants in 2011 to n = 366 clinical trial patients and n = 664 registry study patients in 2022. Concurrent with the increase in the number of active research studies, the number of PIs and SCs also increased from n = 2 PIs and n = 2 SCs in 2011 to n = 17 and n = 13, respectively, in 2022. Open clinical trials and clinical trial patients per year between 2011 and 2022. The number of open clinical trials (red) and clinical trial patients (blue) was determined through annual progress reports submitted to Grampians Health governance based on the open and closure dates of a clinical trial and the number of patients reported to be enrolled in the trial by year In 2011, the four clinical trials conducted were limited to the field of oncology and were predominantly investigator-initiated (75%). In 2022, investigator-initiated trials composed 44% of the 82 clinical trials conducted, with the majority being industry-sponsored clinical trials (66%). The fields of research were also expanded beyond oncology (79%) to include gastroenterology (9%), cardiology (6%), infectious diseases (4%), neurology (1%) and anaesthetics (1%). The number of trials by department is shown in Fig. 2. Number of open research studies by department between 2011 and 2022 The number of open research studies each year was documented by open and closure dates in annual progress reports submitted to Grampians’ health governance and categorised by the treating department. The overall framework average score increased between 2011 and 2022. In 2011, GHCTU was considered ‘established’, which increased to ‘high performing’ (average score 3.0) in 2016, demonstrating the largest growth in framework score compared with years prior to 2015 (average of 2.6). The lowest scoring domain in 2022 was leadership and culture, which had an average score of 2.6 and was the only domain to score an average of less than 3. The framework score correlated well with the number of PIs per year (r (10) = 0.9, p < 0.001, 95%CI 0.8-1.0), the number of open trials per year (r (10) = 0.9, p < 0.001, p < 0.001, 95%CI 0.6-1.0), and the number of patients (r (10) = 0.9, 95%CI 0.6-1.0). The score continued to correlate well when the lower bound of framework score was input number of PIs per year (r (10) = 0.9, p < 0.001), the number of open trials per year (r (10) = 0.9, p < 0.001), and the number of patients (r (10) = 0.9, p < 0.001). There was no change in interpretation when pairwise Spearman correlation was conducted (vs. Pearson) for framework score and PIs per year (ρ = 1.0, p < 0.001, n = 12), number of trials open per year (ρ = 1.0, p < 0.001, n = 12), and number of patients per year (ρ = 0.9, p < 0.001, n = 12),. The framework score was likewise unchanged when one year was left out for each year for number of PIs per year (all r(9) = 0.9–1.0, p < 0.001), number of trials open per year (all r (9) = 0.9, p < 0.001), and number of patients per year (all, r(9) = 0.9, p < 0.001). Correlation analysis revealed the similarities between objective variables and the average framework score per year over 12 years, and Bland–Altman adjusted for trend was used to show the degree of difference. The Bland‒Altman plots show 0% outside limits of agreement for overall framework score open trials per year and studies open per year (Fig. 3). Bland–Altman plot of the average framework score and number of studies at GHCTU per year between 2011 and 2022. Bland–Altman analysis was used to quantify the percent difference between the average framework score per year and n=studies opened per year plotted against their averages. The analysis was adjusted for trend by fitting a regression model to the data to account for a linear relationship. The grey shading shows that a range within 95% of the two measurement methods is expected to fall. All data points (red dots) are within the grey shaded area, showing high agreement between the average framework score per year and n=studies opened per year between 2011 and 2022 at the GHCTU Bland–Altman analysis was used to quantify the percent difference between the average framework score per year and n=clinical trial patients per year plotted against their averages. The analysis was adjusted for trend by fitting a regression model to the data to account for a linear relationship. All data points fell within the 95% interval, meaning there was less than 5% difference between the variables. This shows good agreement between the average framework score per year and n=patients per year between 2011 and 2022 at the GHCTU. Discussion This study describes the growth of a regional Australian clinical trial unit between 2011 and 2022 using measures of growth from annual reports and correlating these data with scores from a published capability framework [12]. When growth derived from annual reports were compared to the GHCTU score using the published capability framework, the scores correlated well. The average score from the framework capability score was positively correlated with objective measures of clinical trial unit growth, demonstrating that the framework was appropriate for assessing the capacity of a regional clinical trials unit. The capability framework has not been previously validated, and this study provides some evidence of construct validity given that the average score computed via the framework correlates well with multiple objective measures of clinical trial unit growth, including the number of trials, number of PIs, and number of patients recruited. Further research is required to show other forms of validity such as inter-rater reliability and predictive validity. Our results reflect overall growth in clinical trials compared with Australian data from clinical trials between 2015 and 2020 [16, 17]. In Australia the growth of clinical trials based on number of trials started was 38.2% (n = 1360 in 2015 and n = 1880 in 2019) [18], whereas growth at GHCTU based on new trials started was 300% (n = 3 in 2015 and n = 12 in 2019). Despite growth of clinical trial availability, participation awareness and participation in clinical trials per capita in Australia remains relatively low based on per capita participation compared to similar countries such as the UK and USA [18, 19]. The observed growth in clinical trials at GHCTU was driven by both investigator-initiated and industry-sponsored oncology trials. This is consistent with nationwide data that oncology streams are the most prevalent indication studied, representing 20% of all trials. The results were inconsistent with nationwide data in that non-commercial sponsors have increased nationwide compared with industry-sponsored sponsors [17]. A key difference in clinical trial growth in Australia compared to GHCTU was during the COVID-19 pandemic in Australia (2019 and 2020), which resulted in an overall decline in trials being registered nationally [16, 17, 18], whereas the number of clinical trials registered at the GHCTU continued to increase (n = 12 and n = 17 of new trials opened in 2019 and 2020, respectively). This may reflect less severe COVID-19-related restrictions in the nonmetropolitan regions of Victoria and the ability of the GHCTU to review new trials on a case-by-case basis, enabling trials to continue to open despite a nationwide decline. Further research comparing the effects of the COVID-19 pandemic on clinical trials in Australian regional areas to those in metropolitan areas is needed. According to the capability framework, technology, infrastructure and organisational support were leading areas of growth between 2011 and 2022. The framework may identify areas of improvement overlooked by site staff; however, it lacks specificity to the region, culture, and unique needs of an individual clinical trial unit. Site expertise, objective data and the capacity framework help to build a holistic perspective of the growth of clinical trials. Emerging regional clinical trials units may find investment in infrastructure and organisation support similarly drive increases in commercial and non-commercial research studies. Further research from the perspective of non-metropolitan clinical trials units is warranted to replicate findings. A key strength of this study is the use of longitudinal objective data combined with a capacity framework assessment of a regional clinical trial unit. This is the first study to show that the Australian maturity model developed by Wollert et al. correlates well with objective retrospective data from a regional clinical trial unit. Limitations of this study include an intentional omission of data for the formative years of the GHCTU, as paper documents were exclusively used prior to 2011. In addition, the correlation analysis demonstrates association and not causation. As both objective measures of GHCTU and framework score increased over time, the correlation may be partly attributable to shared temporal trends. Whether using the framework to inform organisational decisions would result in increased growth in capacity was not investigated. The capability framework did not specify how to compute survey items with a score range, i.e., 0–3. Without criteria to assign a score within the range given, the highest value was used, which may overestimate scores for these survey items. This method was considered the best objective way to address this issue compared with assigning the lowest score (equal but opposite bias) or assessors making a subjective, unreproducible judgement. This study did not examine other forms of validity of the framework such as inter-rater reliability, known-groups comparisons and predictive validity. Finally, demographic information, such as age, sex, and ethnicity, was not available for inclusion in this study. Additional research examining the equity of clinical trials across regions is warranted. This study considered all open studies equally rather than putting additional weight on studies opened each year, the number of new studies being opened clearly increased in successive years. Conclusion The growth of clinical trial units in these regions provides necessary care and treatment opportunities for underserved populations. An objective survey of GHCTU annual reports demonstrated substantial growth across a decade, particularly driven by oncology trials and enabled by improvements in infrastructure and organisational support. This was a novel study that showed that a capacity framework for outer metropolitan and regional clinical trial units correlated well with objective measures of growth and, in doing so, supported the construct validity of the capacity framework tool. More generally, the study provides insights into the determinants of capacity building and growth for a clinical trial unit, which ultimately can aid in service delivery design for a health service seeking to integrate trials into everyday clinical practice. Acknowledgements We thank John Hand and the Research and Ethics Governance Office at Grampians Health for their valuable assistance. Our gratitude also goes to those affiliated with the Grampians’ Health Clinical Trials Unit and the patients who trusted the team with their care. Abbreviations Principal Investigator Grampians Health Clinical Trials Unit Canadian Remote Access Framework for Clinical Trials Study Coordinator Authors' contributions WF and LB designed the research study. LB, MT and EP collated governance data for objective data. AC, CG, RG and VM completed framework assessments for all years. Statistical analysis was conducted by LB. All authors were involved in the writing, editing and final approval of the manuscript. Funding None. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Not Applicable. Consent for publication Not Applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.