2.1 Stage 1: Evidence review and stakeholder consultation
The overall aim of Stage 1 was to both learn from what had been reported in the literature and to gather the perspectives of multiple stakeholders about the issues related to establishing a research registry.
2.1.1 Evidence review
A non-systematic literature search was conducted to address the following two research questions; In healthcare settings, what initiatives have been used for research participant recruitment? and; In healthcare settings, how have research registries been implemented and evaluated? Proquest and PubMED databases were searched. Key terms were as follows: Registr*, System*, Opt-in, Research consent system, Patient* consent, clinical registr* and paed*. Included were full text, peer-reviewed publications in English between 2013-2018. Additional strategies included searching end-text reference lists of identified articles and searching of key websites. Results of the search identified 256 articles from Canada, United States (US), Europe, United Kingdom (UK) and Australia. Articles included descriptive reports, qualitative and quantitative studies, and systematic reviews. Subsequent searches followed an iterative process and included the identification of relevant CCI literature.
2.1.2 Review findings
There are many challenges to recruiting research participants, some unique to the acute pediatric healthcare setting [33, 34]. Difficulty recruiting participants is a common reason for discontinuation of trials. For example, of 559 paediatric randomised control trials, recruitment difficulties were cited in 37% of the 104 discontinued trials [33]. Further, of 3428 United States (US) closed studies, 152 were terminated before completion, with 83 of these reporting termination was due to insufficient recruitment [35]. A UK review of child health research found less than 5% of registered studies involved children, and less than 2.5% of two million National Health Service paediatric patients were recruited into research studies [1]. A plethora of strategies have been used to improve research recruitment. A systematic review identified 72 strategies, with three that demonstrated good levels of evidence [36]. Only two strategies effectively improved recruitment: conducting open trials rather than blinded placebo-controlled trials and following up postal invitations with telephone reminders [36]. In an attempt to further understand recruitment challenges, interviews with researchers identified four factors that were positive influences; an infrastructure supporting researchers’ access to potential participants, study design, if the treating doctor mentioned the study to potential participants, and participants being motivated by altruism [4]. Recruiting research participants is complex, even more so in the paediatric setting, and successful recruitment requires a systematic approach to be taken.
A consumer research register is one systematic approach [33, 5]. The feasibility of an Australian ‘consumer registry’ to facilitate direct patient recruitment from hospital populations was assessed in New South Wales hospitals [5]. A survey was used in an outpatient population to measure consent rates, preferred methods and frequency of contact, and the feasibility of establishing the register. The concept of a register was found to be feasible, with most participants willing to be contacted multiple times utilising methods such as email [5].
We explored Australian paediatric hospitals’ websites for research registry personnel contact details. At the Children’s Hospital in Victoria, a component of the Electronic Medical Record includes a patient portal for families to register interest for research contact [37]. We were unable to verify how well utilised this capability is, or to find information about other paediatric hospital research registries. We did not identify any existing registries to model for our context.
2.1.3 Consultation
We used a variety of methods for stakeholder consultation ranging from regular meetings with stakeholders to one-off unstructured consultations. A steering group of key stakeholders was established with a smaller working party to progress initiatives and provide representation from key areas of the health service. Stakeholders included health professionals, researchers, information technology professionals, legal services professionals, the health service communications team, and health service clerical staff.
Initially a targeted approach was used to identify stakeholders, and engage with other relevant professionals, individuals and groups. In collaboration with the steering group, we created an initial list of people and departments who would be impacted by or would likely wish to provide input about creating a research registry. We contacted them by email or telephone to arrange to meet in person or consult by telephone. As further stakeholders were identified and recommended by others we expanded the number and range of consultations. The aim was to gather multiple perspectives about possible solutions and identify barriers for a feasible process to create a registry. These unstructured consultations were undertaken as individual interviews or small group meetings where the topic of creating a research registry was broadly introduced and stakeholders had the opportunity to provide their input. For some consultations eg. with information technology professionals, questions were more targeted to understand the capacity to integrate a research registry into the health service’s existing systems. The consultations were verbal with field notes taken. A written record of key points from each consultation was returned to stakeholders to confirm. Some stakeholders provided written input by email. The key points were collated to identify similarities and differences across stakeholder perspectives and to inform subsequent consultations.
We consulted a range of researchers at local private and public hospitals, including research directors who shared their insights, experiences, views and ideas. We found one private hospital had an existing research registry using an opt-out process, administered by clerical staff during admission of patients, however, the completion of the consent field in the database was not a mandatory field and consequently sub-optimally captured. We also sought input from researchers at one public hospital who had previously held a consultation with consumers and community members to explore their perspectives on an opt-out approach to using routinely collected health information for low risk research [38].
Given the recognised benefits and value of consumer and community involvement in the development of research, consumer and community consultation was also undertaken from the outset of the project. A range of strategies were employed, including engaging with Nursing Research Consumer Advisory Panels (consisting of parents and young people who had utilised the health service), the health service Consumer Advisory Committee and holding “a Community Conversation”; a consumer and community event involving young people and parent users of the health service. See Table 1 further detailing consultation.
Table 1 Consultation
Groups
|
n
|
Health professionals
|
|
Steering group
|
25
|
Working party
|
6
|
Other Divisions of Child Health Service including Community and Mental Health groups
|
40
|
Health Service Research and Research Governance
|
12
|
Allied Health professionals
|
3
|
Other professions
|
|
Patient management systems
|
35
|
Health Information management system
|
3
|
Information and Communication Technology
|
2
|
Legal services team
|
2
|
Communications team
|
5
|
Business Intelligence professional
|
1
|
Consumer and Family Engagement team
|
3
|
Consumers
|
|
Community Conversation
|
26
|
Youth Advisory Panel
|
7
|
Consumer Advisory Panel
|
6
|
Individuals
|
|
Researchers and healthcare professionals within health service
|
14
|
Health and relevant professionals from other organisations
|
13
|
We worked together with the CCI Program team [27] to host the ‘Community Conversation’ using a World Café format [39]. The Nursing Research Consumer Advisory Panels also advised on the planning and promotion of the event, and the wording of the discussion questions. Attendees were accessed through social media platforms and the CCI Program consumer database. Seated in small groups in a welcoming “café”-style environment, 16 parents/carers and 10 young people discussed the following three questions: what are your thoughts about a research registry?; how and when would you prefer to be approached to join a registry, and who should approach you to join the registry?; and what information should be collected and how should it be managed? To avoid impeding the flow of conversation, consumers did not move tables in between questions, as is the standard process for the World Café methodology, and most of the young people were seated at one table. After the three rounds of questions, a spokesperson from each table shared insights and summarised the conversations. Facilitators and note-takers were situated at each of the tables and later compared notes, synthesised and agreed on the key findings.
2.1.4 Consultation findings
In total, more than 80 meetings were held with individuals and groups, of which over a quarter were representative of the patient and carer population. We found strong consumer and health service support for establishing a systematic process to contact families about research and learnt that families are generally motivated to participate in research by altruism, and a desire to improve future health outcomes. The feedback indicated that families would want to manage their own registry preferences using online technology. Importantly, we found we had incorrectly assumed that consumers already understand research and what research participation involves. In addition, consumers identified that the term ‘registry’ was unclear and that this may be a barrier for some. Following the Community Conversation, the Nursing Research Consumer Advisory Panels recommended using the term ‘contact list’ instead. It was clear there was a gap in communicating with consumers and the community about the research conducted by the health service and about research participation. Barriers to creating a research registry were mostly raised by health service staff. These included views that consumers would be too concerned about their privacy to wish to join a contact list, would be overburdened, were unlikely to join a contact list if their own doctor did not suggest it, or would be concerned about how to update changes. Other health service staff concerns were about using existing service-based information technology systems, uncertainty about governance and legal requirements for patient privacy, using families’ contact details for secondary purposes other than clinical care, and the need for a research registry to be sustainable (not dependent on specific project funding or individuals).
The Community Conversation identified seven key issues relevant to establishing a research registry. These were; foreseen concerns, benefits, consent methods, raising awareness, approaching participants, information collection and information management. Logistical issues foreseen by consumers included how participant updates would be managed, such as ‘unsubscribing’ if a person no longer wanted to be part of the registry, managing data once a child turns 18 years, and knowing who would have access to their data. Registry benefits included facilitating research to increase knowledge to benefit others in the future, despite no direct benefits for participants. Most consumers supported an opt-in approach to consent. Importantly the group highlighted that adequate information was key to raise consumer awareness about research and a research registry at the health service. Consumer input regarding how best and, who would be the most appropriate person to approach families to join the registry, revealed a range of potentially acceptable options requiring further consideration. Last but not least, consumers emphasised the need for consistent transparency about the collection and management of personal information. Although there was broad agreement supporting the use of health information for research, consumers felt that the general public may be reluctant to adopt an opt-out approach, unless this was preceded by a concerted consumer engagement and public awareness campaign.
2.2 Stage 2: Co-production
The outcomes of the Stage 1 consultation led us to first address the need for communication of research information to health service consumers. The aim of Stage 2 was to co-produce these solutions with the health service communications team, a project working party (members of the working party were purposively selected to represent key stakeholder areas), and consumers (members of the Nursing Research Youth and Consumer Advisory Panels). A research communications plan was developed which included three key strategies, namely, a social media campaign, messaging throughout the health service and developing a prototype research information brochure.
The Co-production process involved an action research cycle of a series of meetings to develop an information brochure, that was titled ‘What is child health research?’, and a survey. There were iterative rounds of discussion, rewording and redrafting using multiple mediums (in-person meetings, email communication, and voting) until agreement was reached. Careful consideration was given to balancing the need for clear messaging to a mixed audience of varying levels of health literacy and providing sufficient detail in the brochure to achieve the health service duty of care regarding full disclosure of information. The brochure contained information to address issues raised at the Community Conversation such as explaining child health research, the goal and planned process of creating a research contact list, the information to be collected and how patient privacy would be managed. It was agreed that the term ‘research contact list” be used to reflect the collection of consumer contact details only at this point (not linked to health information). The initial survey items were created and further developed in Stage 3. The brochure content and survey item refinements were made, presented and discussed with the different groups until agreement was reached about the content and presentation. This stage resulted in the development of a prototype colour brochure [40] and a draft survey presented in English language.
2.3 Stage 3: Prototyping
Using the survey developed in Stage 2, the aim of Stage 3 was to assess health service users’ views about:
- The prototype brochure
- Being contacted about research and being part of a possible future research contact list
- A number of uncertainties regarding potential consumer participation in the research contact list, as identified in stage 1 of the project
The quality improvement activity involved requesting consumers to complete an online survey and agree to receiving a telephone call.Health service governance approval was obtained (Reg No 29335). Participation was voluntary, and the survey did not collect any identifying information unless participants agreed to be contacted for telephone follow-up and provided their name and telephone number.
The objectives for Stage 3 (and the intervention) were to:
- Assess appropriateness of approaching families to discuss research in the hospital settings of outpatient clinics and Emergency Department
- Assess content, clarity and presentation of the prototype brochure via a survey
- Measure families’ recall and satisfaction with their responses and obtain feedback and suggestions to refine the brochure.
2.3.1 Setting
The setting of the quality improvement activity was a 250 bed specialist children’s hospital serving a population of 500 000 children and young people. During 2018/19 there were 227 337 outpatient clinic attendances and 67 592 Emergency Department attendances [41]. Participants were recruited while attending a wide range of outpatient clinics including medical, surgical, and dental clinics. Emergency Department participants were recruited from areas in the triage waiting room catering for low-acuity patients. Participants who could speak, read and understand English were included. Data collectors were experienced in communicating with families in healthcare settings. Parents or carers with acutely ill children or who appeared to be anxious or distressed were not approached. Data collection was coordinated with clinical and clerical staff to minimise disruption to workflow.
2.3.2 Planning the intervention
A 20 item survey was initially tested with 40 families who attended the children’s hospital and a community health clinic during an annual week-long patient engagement event. The testing resulted in minor wording changes to improve item clarity. The iterative testing process resulted in an additional question to understand families’ preferences about being contacted about research by health service researchers only, or by researchers from other organisations. This was to answer another uncertainty about whether families’ preference may differ depending on the research affiliation. The first 11 survey items captured participant characteristics. There were six items addressing key issues raised during consultations including willingness of families to be part of a contact list for future research studies, preferences for how to be contacted, preferred frequency of contact, the preferred person to inform families about the contact list and feasibility of the proposed process. The final four items requested participants to rate (using 5-point Likert type scale from strongly agree to strongly disagree) their level of agreement about the brochure and the online survey content and presentation.
The telephone follow up consisted of four closed questions capturing participants’ recall about receiving the brochure, whether they would have liked additional information, whether they remained satisfied with the response given to being contacted about research, and their knowledge of how to update a change of preference about being contacted for research. The final question was open ended and invited suggestions to improve the delivery of information in future.
2.3.3 The intervention
One of two data collectors approached families as they waited for their outpatient clinic appointment or in the Emergency Department triage waiting room. Participants who agreed to complete the survey were invited to participate in a follow up phone call. These two settings were selected to access a large volume of families who may not have previously been exposed to research and to assess the appropriateness of approaching families in these settings.
2.3.4 Methods of evaluation
Data collection was completed in 2 parts between July and August 2019.
Part 1. Each participant was provided the brochure to read and invited to complete an online survey. The brochure was available to be taken away and some families did take a brochure. Survey responses were entered by participants on a tablet using Survey Monkey© software, There was no time limit to complete the survey.
Part 2. Two to three weeks later a telephone call was made to evaluate the experience of participants in Part 1. A maximum of three attempts to contact participants were made over a seven-day period. The purpose of the phone call was to capture participant recall of receiving the brochure, completing the survey, and to obtain feedback to improve the brochure.
2.3.5 Analysis
For the survey and telephone call responses, descriptive data analyses were undertaken using frequencies and examined for distribution and percentages. Categorical data are presented using proportions and frequencies.