Informing a national rare disease registry strategy in Australia: a mixed methods study

Abstract

Background

Rare disease registries (RDRs) facilitate monitoring of rare diseases by pooling small datasets to increase clinical and epidemiological knowledge of rare diseases and promote patient centred best practice. The aim of this study was to understand the current state of RDRs in Australia, data captured, impact on patient outcomes, funding models, and barriers and enablers regarding their establishment and maintenance.

Methods

An exploratory sequential mixed methods study design was adopted. First, a list of Australian RDRs, primary contacts and data custodians was generated through online and consumer group (Rare Voices Australia (RVA)) contacts. A cross-sectional, anonymous online survey was distributed to registry custodians, managers, or principal investigators of 74 identified Australian RDRs, 88 RVA Partners, 17 pharmaceutical organizations and 12 RVA Scientific and Medical Advisory Committee members. Next, managers and coordinators of RDRs and databases who participated in the survey were invited to participate in semi-structured interviews. Quantitative and qualitative data were analysed using basic descriptive statistics and content analysis, respectively.

Results

Forty RDRs responded to the survey; nine were national, five were based in Australia and New Zealand, and the remaining were global. Of the 40 survey respondents, eight were interviewed. Most of the RDRs captured similar information regarding patient characteristics, comorbidities and clinical features, diagnosis, family history, genetic testing, procedures or treatment types, response to treatments and complications of treatments. Better treatment outcomes, changes in process of care and changes in quality of care were the most frequently reported benefits of the RDRs. The main challenges proved to be cost/funding of data collection, data completeness, and patient consent. When asked, the participants identified opportunities and challenges regarding potential options to streamline RDRs in Australia in the future.

Conclusion

Findings from this study highlighted significant dataset heterogeneity based on the individual disease, and current lack of interoperability and coordination between different existing RDRs in Australia. Nevertheless, a nationally coordinated approach to RDRs should be investigated given the particular benefits RDRs offer, such as access to research and the monitoring of new disease-modifying treatments.

Background

A rare disease (RD) is defined as one with a population prevalence of fewer than five in 10,000 people [1, 2]. While estimates of the number of RDs may vary between countries, due to varying definitions and challenges with data collection, it is prominently cited that there are more than 7,000 different RDs [3]. Approximately two million people in Australia live with a RD [4]. The impact of RD on a person’s life and their family can be devasting. Diagnosis can often take time because these conditions are complex. For many RDs, there are no effective treatments or cures.

Rare Voices Australia (RVA) in consultation with stakeholders across Australia developed the National Strategic Action Plan for Rare Diseases (the Action Plan) [5]. The Action Plan demonstrated the need for a national, coordinated and systematic approach to the collection and use of RD data, including registries. Clinical registries systematically monitor the quality of healthcare within specific clinical domains by routinely collecting, analysing and reporting health-related information [6]. Clinical registries have been developed for various health conditions and are powerful tools for reducing variation in clinical practice and improving healthcare quality, informing care management, research and health system planning [7].

In Australia, there are no national datasets for RDs, meaning that often estimates of the number and prevalence of RDs are based on international figures. There is also a lack of data on RD treatments and health outcomes. Without this data it is impossible to measure the overall burden and cost of RD in Australia. Registries can fill these data gaps and inform systems changes for better outcomes for people living with a RD [8]. Rare disease registries (RDRs) can enable recruitment to clinical trials. Patient reported outcome measures (PROMs) collected by some RDRs can inform a patient-centred approach to clinical care, by measuring meaningful outcomes for patients and their families [9].

Some RDs, such as cystic fibrosis [10], have benefited from long-standing Australian registries; however, standardised data collection sets or minimum data elements are not available for most disorders. RVA advocates for better health, disability and other systems for people living with a RD [4]. RVA is leading the collaborative implementation of the Action Plan, including the development of a national approach to person-centred RDRs to support national standards, best practice and minimum data sets [5].

A collaboration between the RVA Scientific and Medical Advisory Committee (SMAC) and Monash University researchers was established to understand the current landscape of RDRs both in Australia and internationally. A series of studies were conducted to describe the current state of the RDRs capturing Australian data. The first study was a scoping review of the literature in which we identified seventy-four RDRs [11]. The present study aims to explore data captured by the RDRs identified in the scoping review, to describe registry impact, barriers and enablers of data collection, registry funding, and the future potential of RDRs to improve patient outcomes, facilitate research and clinical trials and provide ongoing surveillance.

Methods

Mixed method design

In this study we adopted a mixed methods exploratory sequential study design [12]. A quantitative component using a cross-sectional survey was conducted first which was followed by qualitative interviews. Data were analysed separately and integrated at the final stage of the study in order to answer the research question.

Survey

The survey was developed by Monash University researchers in collaboration with the RVA. It comprised four sections, with a total of 37 questions. The first section of the survey collected registry information, such as the number and type of RDs captured by RDRs, jurisdictions covered, data sharing, year established, participating sites, population of interest and size, ethics details, consent model, governance structure and funding source. The second section gathered information on RDR data collection, including parties responsible for data entry, methods of data collection, types and categories of data captured, outcomes collected, frequency and timing of data collection. The third section focused on research and reporting, and gathered information of stakeholders, frequency of data reporting, use of data and publication history. The final section collected qualitative data on the impact that registry has made, its challenges, barriers and enablers.

A list of Australian RDRs and RD organisations, contact details of their managers, investigators and data custodians was obtained through a search of online registry webpages and RVA contacts. The survey was distributed, via email, to registry custodians, managers, or principal investigators of 74 Australian RDRs, 88 RVA Partner organizations, 17 pharmaceutical organizations from RVA’s roundtable of companies and 12 RVA SMAC members.

The link to the survey included an invitation to participate, which explained the aims of the survey, its voluntary nature, and the requirements for participation. An implied consent process was utilised. A survey was administered online through Qualtrics Survey Software [13] from 19 to 2021 to 25 January 2022.

Interviews

Registry custodians, managers, or principal investigators who responded to the survey were invited to participate in the second part of the study. We developed a semi-structured topic guide with open-ended questions (see Additional file 1). Interview topics covered objectives and research priorities for the RDRs, their functions in terms of improving health outcomes, resources required to maintain a registry, the coverage, attributes and feedback mechanisms to the different stakeholders such as clinicians, managers, policymakers and researchers, and barriers and enablers in achieving goals of the registry. Follow-up questions and prompts were used to obtain rich data. Critical discussions held throughout the analysis facilitated both self-reflection and common discussion for distinguishing between participant meaning and research interpretation.

Each potential participant who expressed their interest in the study was provided with an invitation letter, an explanatory statement and a consent form before the interview. All interviews were audio-recorded, subject to the participant’s consent. Consent was given verbally before the interview. It was not possible to establish how many participants saw the invitation to participate but decided not to volunteer. No participants dropped out of the research.

Interviews with the study participants were conducted on the phone by the experienced qualitative researcher (RR). On average, the interviews lasted 29 min (range 21–46). The interviews were conducted from 18 January to 12 February 2022.

Data analysis

Quantitative data were statistically analysed in two stages using SPSS V26. Firstly, descriptive statistics were calculated for appropriate variables and responses were reported as both whole numbers and proportions. Secondly, sub-analyses by participant characteristics were undertaken for questions where the participant responses were varied.

The voice files of interviews were transcribed by a paid transcription service. To ensure data quality, the research team checked all transcriptions against the voice files. All participants were offered the opportunity to review transcripts. The process of analysing qualitative data involved coding and categorising the data from interview transcripts using NVivo software (Version 12, QSR, Australia). Transcripts were reviewed, and RR and MC thematically analysed the transcripts identifying quotes and words and grouping them according to themes and sub-themes as they emerged from the interviews [14]. Data saturation was determined when no new information was generated from successive interviews.

Results

Managers and coordinators of 40 Australian RDRs responded to the survey (Table 1). Subsequently, the managers of these RDRs were invited to participate in the interviews. Eight people expressed their interest and were interviewed.

Table 1 Survey Results. Characteristics of Rare Disease Registries and Databases

Survey results

General registry characteristics

Eighteen (45%) registries recorded data relating to a single RD. The remaining registries collected data from numerous RDs or conditions (e.g. the Australian Mitochondrial Disease Foundation Patient Registry collects data from more than 50 different RDs). Ten (25%) RDRs provided biobank facilities. Eleven (28%) RDRs were patient-initiated.

The number of patients in each registry varied, with the mean (SD) of 1249 (1826) reported by 28 (70%) RDRs. For example, the Rare Genetic Lipid Disorder Registry recorded data from 25 patients; however, the Sanofi Genzyme Rare Disease Registries captured data from 7,725 patients with Fabry disease. The Morquio A Registry Study stated that their population coverage was nearly 100% of Australian patients who were under the Life Saving Drugs Program. The Australian National Creutzfeldt-Jakob Disease Registry (ANCJDR) reported similar coverage. Thirty-five (87.5%) registries required a consent and ethics approval. Twenty-two (55%) RDRs were funded primarily via the public or private sector, a charity (18, 45%) or via a research grant (9, 23%). Most registries (29, 73%) were managed by a steering/governing committee or a management group (27, 68%), with established terms of reference.

Data collection methods

In most registries (34, 85%), data were entered online by clinicians or staff members (Table 2), but there were a few exceptions.

Table 2 Survey Results. Data collected by Rare Disease Registries and Databases

These included the Australian Autoinflammatory Disease Registry collecting data through telephone calls only, and the Australian Idiopathic Pulmonary Fibrosis Registry (AIPFR) and Children with Interstitial Lung Disease Research Australia and New Zealand collected data via postal/paper methods. Five (12.5%) registries collected data via electronic medical record.

Demographic information

Thirty-eight (95%) registries answered the survey questions regarding type of data collected by their registry. Of these, thirty-seven (97%) registries captured patient name, surname, date of birth, gender, postal and email address and country of birth. Additionally, some registries (5, 13%) captured information about ethnicity, race, religion, occupation, education, marital status, employment status and income.

Diagnostic/clinical variables

Thirty-seven registries (92.5%) collected patient diagnostic data. These included age and year of diagnosis, diagnostic tests, clinical visits, pregnancy/childbirth, comorbidities, family history of disease and genetic variations. Some registries captured additional diagnostic and clinical data specific to their registries. For example, Sanofi Genzyme RDRs collected details of immunogenicity, and the International Fragile X Premutation Registry records neurological, autoimmune problems, fertility/menstruation information, age of menopause, as well as information on other specific health issues such as hypertension, diabetes and kidney disease.

Treatment/Procedure and adverse events data

Clinical assessment and treatment data were captured by thirty-two registries (80%). This information included any/all of clinical management, imaging, therapy, laboratory results, medication type, frequency of treatment dosages and lifestyle interventions.

Information on procedure types (e.g. surgical procedures, radiotherapy and chemotherapy) was entered by twenty-six (65%) RDRs. In addition, the Australasian Interstitial Lung Disease Registry collected information on lung transplants and lung biopsies; the Australian Haemoglobinopathy Registry (HbR) and the Neonatal Alloimmune Thrombocytopenia (NAIT) Registry both captured transfusion data. The Sanofi Genzyme RDR recorded information on audiology, pulmonary function tests, biopsies, stress tests, ECG, cardiac events and cerebrovascular events; and the Thrombotic Microangiopathy Registry collected data on plasma exchange.

Only twenty-one registries (52.5%) captured data of adverse events and complications. This information included details of medical devices, cause investigations, adverse health effects, and clinical outcomes for mother and baby.

Other outcomes and quality of life data

Nineteen (47.5%) RDRs reported patient survival or mortality. Further to this, sixteen registries (40%) collected PROMs or Patient Reported Experience Measures. The Australian Bronchiectasis Registry uses the Quality of Life-Bronchiectasis Questionnaire and the Bronchiectasis Health Questionnaire. The Australian Idiopathic Pulmonary Fibrosis Registry collected shortness of breath, cough and wheeze, gastro-oesophageal reflux, St George Respiratory Questionnaire, hospital anxiety and depression, and sleepiness and tiredness instruments. The Australian and New Zealand Vasculitis Quality and Disease Registry used the ANCA-associated Vasculitis Patient-Reported Outcome Questionnaire and the EQ-5D-5 L. The Global Atypical Haemolytic Uremic Syndrome (aHUS) Registry utilised the FACIT Fatigue Scale v4 (Adults), Paediatric FACIT and the Patient Questionnaire (Adults and Paediatrics) to collect quality of life data.

Timing of data collection

Six registries (15%) captured data at a single-time point. Thirty-two registries (80%) collected data on multiple occasions, including at initial diagnosis, at time of procedure, or at 6, 12- and 24-month intervals. Some registries (5, 13%) collected their data more regularly. Specifically, the Australian Cystic Fibrosis Data Registry (ACFDR) collected data at 3-month intervals, and the ANCJDR and the Observational Longitudinal Prospective Long-Term Registry of Patients with Hypophosphatasia collected data at each clinic visit.

Output reporting and data use

Most registries (33, 83%) reported their outputs to multiple stakeholders including clinicians, hospitals/sites, jurisdictions, consumers, funders, industry and government departments. For example, the Australian HbR reported its output only to their funders. The ANCJDR and the Global aHUS Registry reports output to government departments only, and the Lysosomal acid lipase deficiency (LAL-D) Registry reported their output to regulatory agencies only.

Eight (20%) registries published annual reports. Two (5%) RDRs reported their data quarterly (e.g. AILDR and Simons Searchlight) or twice a year (AIPFR and Myeloma and Related Diseases Registry). Six registries (15%) have not produced any reports.

Data use for research purposes/post-marketing surveillance

Thirty-three registries (82.4%) used their data for research and publication purposes, including clinical trials, epidemiological modelling, collaborative projects, secondary use of data and availability of registries for data access requests.

Nine registries (22.5%) used their data for post-marketing surveillance for high-cost medicines. These registries included the ACFDR, Australian HbR, AIPFR, Australian Neuromuscular Disease Registry, MGBase, Morquio A Registry Study, aHUS Registry, LAL-D Registry, and the Observational, Longitudinal, Prospective, Long-Term Registry of Patients with Hypophosphatasia.

Registry impact on patient outcomes

Registry managers and coordinators were asked to describe the impact they have made on patient outcomes. The distribution of responses and exemplary quotes are shown in Table 3. The top three responses included “better treatment outcomes”, “changes in process of care” and “changes in quality of care”.

Table 3 Survey Results. Summary of Impact of Rare Disease Registries and Databases on Patients Outcomes, Barriers and Enablers in Meeting Registry objectives

Six (15%) registries indicated “other” type of impacts made on patient outcomes. For example, the RDR33 “assisted researchers in addressing many of the unanswered questions surrounding the clinical course and outcomes.” The RDR30 increased the “understanding of disease and ability to communicate to physicians” while the RDR18 “supported reimbursed access in Australia”.

Enablers, challenges and barriers towards operating RDRs

Patient participation (57.5%) was recognised as the main enabler towards operating RDRs followed by registry funding (50%), expertise in running a RDR (50%), consent of participants (43%), volunteer participation (30%) and identified data collection (16%).

52.5% of registries listed funding as the main barrier towards successful maintenance of the RDR, followed by the availability of data “due to the low frequency of people living with particular RDs” (32.5%), patient participation (30%), low registry staffing levels (27.5%), issues around identifiable data (15%), a lack of volunteers (15%) and consent issues (12.5%).

Other challengers included data completeness, reported by 28 (70%) of respondents and “cost/funding of data collection,” mentioned by 16 (40%) of participants. Eleven (27.5%) of survey participants listed consent issues, 8 (20%) mentioned data validity while the remaining participants stated data security and time allocated to data entry as major concerns for their registries (Table 3).

Interview results

Four interview participants were volunteers who worked with clinicians across the country and internationally in their organisations. Other participants had a professional role before taking up the role with their registry. Most of the participants came into the RD area because they either had a family member or friend with a RD. Four of the participating interviewees belonged to global registries with the size of their RDRs ranging from 80 to 500 patients.

Table 4 displays the discussion topics and sample quotes provided by the interview participants. The topics included “Data collected”, “Registry platforms”, “Registry Management and Consumer Involvement”, “Post-marketing Surveillance”, “Funding”, “Registry Impact”, “Challenges” and “Views towards RVA’s National Strategy”.

Table 4 Interview results. Themes and exemplary quotes with the participants of Rare Disease Registries and Databases

In terms of the data collected, registry management, consumer involvement, funding and impact made, the interviews confirmed findings from the survey.

When asked on the impact made, the interview participants said that impact was either not measured or no direct impact has been made. This appeared due to either a small sample size of the registry participants or a slow roll out of the registry.

When asked regarding the challenges faced when maintaining a registry, the respondents agreed that funding was the most pressing issue. Other challenges included fundraising activities, resources required to setup and maintain a registry, ethics and governance, and database issues.

The participants were also asked to provide their feedback regarding the RVA’s national strategy around centralised data set, and what that would it look like with a national registry of RDs. Main comments included feedback regarding the goals and priorities: “We should be focussed on what the real goal is, which is being able to combine data to get an accurate snapshot or accurate information about all the [strains] of rare diseases, and let’s keep an open mind about what the solution looks like” (RDR1). Other responses included benefits of collaboration between the RVA and RD groups: “shared knowledge about running registries could be really useful. The better job that we can do, […] the more that the patients will benefit. […] Feeling like we’re connected to a community of people trying to run the rare disease registries will be really useful” (RDR33).

Comments around challenges of a national registry included concerns for existing registries, education and participation for patients, confidentiality, and on what level would a national registry engage with international RDRs. Participants were cautious regarding the possible impact of the national registry: “Being able to facilitate bringing all that information together confidentially would be a big challenge. Having just seen how difficult getting patients to put their details on the global patient register for their own disease has been, I’m not sure how you would go about collecting that information” (RDR42).

Other comments raised on the discussion around support for a national registry included partnering with international registries, platform, standardised guidelines and challenges around funding: “The criteria for success is that data can be pooled with international data, and I think if we had one Australian registry, I just don’t see how that can happen without […] patients and clinicians having to contribute to multiple (databases)” (RDR1).

Discussion

To our knowledge, this was the first mixed-methods study of RDRs’ capturing Australian data and investigating the impact of these RDRs on patient outcomes, data captured, funding sources, governance, and the barriers and enablers of building and managing these registries. RDRs collecting Australian data vary in their purpose, coverage, treatment types, target population, data collected, governance models and sources of funding. Information collected by most RDRs include patient demographics, diagnostic and laboratory data, as well as data related to therapies and treatments. Common challenges to data collection reported in the survey and interviews were data completeness, errors and other data quality issues. To prevent such errors, harmonization of the data items, as well as identification and correction of causes of poor data quality should be continuously considered [15,16,17]. Using standard coding and classification systems along with the standard definitions for data items should also be considered to enhance comparability [18]. To improve the quality of registry data, RDR managers should develop clear definitions for core data elements, to ensure that procedures for checking the data are specified, and that feedback has been provided [15].

Numbers of patients captured by RDRs and quality of data collected may relate to the impact of registries on patient outcomes. The quality of a registry is closely associated with the validity of information that is published through reports and other forms of dissemination [19]. Thus, publications from registries have great value in guiding clinicians with respect to therapeutic recommendations, treatment expectations and severe complications that may occur during the evolution of a disease. Despite RDRs comprising small patient numbers, the burden of RDs on patients, families, health systems and the economy is proportionally higher than the general population [20]. Therefore, having a coordinated approach to collecting RD data may provide economies of scale and thereby enable a wider positive impact on patient outcomes [5].

RDs can lead to a significant reduction in quality of life for patients and their families. Ensuring the patient voice is central to clinical decision making is key to delivering, evaluating and understanding the efficacy of therapeutic interventions [21]. PROMs are highly effective for ensuring the patient voice informs best practice care [22, 23]. The collection of PROMs by RDRs offers the potential to improve patient care and clinical outcomes [21, 24]. Nonetheless, only 40% RDRs reported capturing PROMs in our study. One reason for this may be that administering PROMs for RDs poses unique challenges, including small patient populations, disease heterogeneity and a lack of natural history knowledge [25]. Given the small number of people living with individual RDs, PROMs should be supported by RDRs [21].

Similar surveys have been conducted in other countries. In 2011, the European Commission funded the EPIRARE project (‘Building Consensus and Synergies for the European Registration of Rare Disease Patients’), to inform the development of a European Platform for RDRs [26,27,28]. The EPIRARE survey investigated the minimum data elements of RDRs, to address various methodological, technical, and regulatory issues, and ways to find resources to develop and sustain registries. Of the 272 registries surveyed, 48% did not have a clear strategy for long-term sustainability, 34% did not have a specific management group, 30% did not share data, and 21% were established without any clear funding [26]. The results of the survey demonstrated various issues regarding financial support, data quality, and the need to improve data sharing. Moreover, the registry holders were supportive of a common platform for RD registries.

Ali et al. [15] conducted an international survey of RDR leaders to ascertain level of consensus regarding the quality criteria that should be considered essential features of a RDR. Of 35 respondents representing 40 RDRs, over 95% said that essential quality criteria should include establishment of a good governance system (ethics approval, registry management team, standard operating protocol and long-term sustainability plan), data quality (personnel responsible for data entry and procedures for checking data quality) and construction of an IT infrastructure.

A survey of RDRs in Japan reported similar challenges to RDR’s surveyed as part of this Australian study, including difficulties securing sustainable funding, reducing operational burdens, gaining cooperation from participants and promoting the use of data for research or to improve health outcomes [29]. To address these issues, the Japanese study proposed to: (1) build a common platform with universal facilities, (2) establish collaboration with educational institutions to increase awareness of RDRs, and (3) secure a sufficient budget from collaboration between academia and industry to maintain RDRs.

Our findings are similar to those reported above, supporting the need for a nationally coordinated approach of RD data capture. To achieve this in Australia, well-coordinated efforts should involve all stakeholders [5]. Dedicated funding and incentives should be allocated for RDRs to ensure full coverage of the eligible patient population. Investing in the infrastructure and staffing would assist in streamlining and simplifying the maintenance and data entry demands of RDRs across Australia. Finally, RDRs should be promoted in hospitals and relevant clinics, and collaborations with relevant international registries should be sought [11].

Strengths and limitations

Integration of the quantitative and qualitative study components has strengthened the validity of the findings and provided a richer understanding of participants’ experiences regarding the RVA strategy towards building a national RDR in Australia. A major limitation of our study was that of the 191 survey recipients only a small number participated in the survey and qualitative interviews despite invitations and reminders. Reasons for the low rate of participation are unfortunately unknown, although small staffing of RDRs may be associated. Furthermore, this was a cross-sectional study, therefore responses provided could only give a snapshot experience at a given time.

Conclusions

The results of our study highlighted that while RDRs are feasible and valuable to multiple stakeholders, there are key principles required for success. These include developing a process of low burden data collection; having sustainable funding; contributions of consumer wellbeing information; and ensuring multiple data uses, including reporting, epidemiology, secondary research and access to clinical trials. Opportunities to consolidate existing RDR information and ensure access to RDR expertise to support new RDRs is a starting point to realising the potential of RDRs in Australia. Precision medicine is likely to see more RDRs established to support clinical trials and post-marketing surveillance. Having a co-ordinated approach nationally will enable maximum impact from RDRs in Australia, and maximum benefit for people with RDs.