Research BY General Paediatricians FOR General Paediatricians (and their patients)

On the 22 June a group of individuals met to launch GAPR-UKI: General and Adolescent Paediatric Research Collaborative - UK and Ireland.

'Why yet another research outfit?', you may ask. There is the Medicines for Children Network, PERUKI, the Scottish Paediatric Reasearch Network etc. etc.

Well, this time it's a network run by people who are working general paediatric themselves. Why is that important? Take this example: You may well have contributed to research projects like PREDNOS (prednisolone treatment for nephrotic syndrome) because the outcome is important for the management of these patients, the vast majority of which is overseen by general paediatricians. Have you had a say in the design of the study (it's run by a tertiary nephrologist), have you had any support in getting it up and running, help with getting your R & D department interested, showing you step by step how to overcome the administrative hurdles while your Trust bags the fee and expects you to cram your research time into your CPD hours?

I guess not.

So this is a new approach by having the people (i.e. YOU) who deal with millions of children day in and day out on the ward, in outpatients and in admission units involved in paediatric research from the outset, to hear YOUR clinical questions and translate them into research and to support and guide you every step along the way to get involved in hgh quality research and remain a busy general paediatrican.

So look out for the next phase of recruitment of interested centres and put your unit forward!

Summary of Discussion (divided into Groups/Themes) at the inaugural GAPRUKI meeting June 22^nd 2016 UCL

To be successful a research network we must be:

Collaborative and Inclusive
- Personalise (communication)
- Prioritise
- Publicise and Brand
- (Not reinvent the wheel by using/ modifying {with permissions} existing documents of other networks)

Include mentoring for general paediatricians with little research experience (some paediatricians think that doing research is like going into a wood without a path)

Start small
- Concentrate on ‘low hanging fruit’ in order to complete pieces of work, publish and gain a reputation
- E.G. surveys/ observational studies/ practice variation

Set research priorities
- What are the biggest/ most important research questions to be addressed within general paediatrics?

Don’t repeat work (applies to setting up GAPRUKI/ setting research priorities/ actually undertaking research).

Randomised Controlled Trials Group Discussion

GAPRUKI should begin considering ideas for RCTs from the start. They take a lot of time to set up, and this can be started in parallel with conducting quicker observational studies. Also, it was felt to be important to have / be developing ideas for RCTs when approaching local paediatricians to get involved in GAPRUKI.

a) Broad approach

GAPRUKI needs to keep a broad view when considering ideas for RCTs, beyond just drug trials to trials of behavioural interventions and trials of different models of service delivery (for example trials of Nudge type interventions). It was suggested that these may be less challenging / difficult to set up. And also the need to consider not just acute conditions but also outpatient and chronic conditions (e.g. children with complex needs, and children with medically unexplained symptoms etc).

b) Collaboration

Other groups within paediatrics (e.g. PEM, PICU) when designing trials that involve care spanning both general paediatrics and these areas. The FISH trial was mentioned as an example of the sort of study GAPRUKI would be well placed to collaborate in (currently a pilot study but will go on to a larger RCT). It was also suggested that collaboration with neonatal groups would be beneficial when planning studies in children with complex care needs, some of whom will be ex-preterms.

c) Local Leads and support

The importance of an enthusiastic lead clinician at each local site was emphasised. The challenges for general paediatricians in being a local site lead are considerable such as time demands on already busy clinicians, challenges of dealing with local R&D departments and getting R&D approval, the lack of consistent support available from a local research nurse for most studies (unless the study if an NIHR portfolio study when the trust will receive funding for participation). There was a general consensus that GAPRUKI would need to think hard about how to incentivise local clinician involvement, and how it can support local clinicians as much as possible, for example in navigating R&D processes. It was suggested that GAPRUKI can learn much from PERUKI on how to overcome these difficulties. It was suggested that there should be a regional structure to GAPRUKI, with a lead for each region (e.g. North West). This person would then be well placed to maintain ties with and provide support for lead investigators in each site in the region.

Personal contact

The importance of getting GAPRUKI members / investigators together for face-to-face meetings at least yearly was discussed (to maintain enthusiasm and support), along with the practical challenges of bringing people from across the UK together. It was suggested to tag GAPRUKI meetings on to conferences (e.g. RCPCH) to help facilitate this. It was suggested to have a regular call (e.g. every 6 months or 1 year) for PICO study proposal ideas going out to all GAPRUKI members, which could then be ranked, to help identify new studies. The PICS study group already have such a system.

Identity and branding

The importance of GAPRUKI having a very clear identity and priority was discussed. It should be clear that GAPRUKI is focused on high-quality research and facilitating multi-centre collaborative studies.

Generalisable and representative

The importance of ensuring that GAPRUKI sites are representative was discussed – i.e. to include DGHs and teaching hospitals.

g) Organisational structure

It was suggested that it is important for GAPRUKI to have an efficient organiser / secretariat to drive the organisation.

Initial Ideas

Some potential RCT ideas mentioned included:

Head to head studies of the comparative efficacy of new drugs coming through development for bronchiolitis.

Antibiotic prophylaxis for chest infections in children with cerebral palsy
Antibiotic prophylaxis for UTIs
Early allergen exposure for allergens other than peanuts (i.e. replicating LEAP trial for different allergens)
Medical vs psychosocial model for management of headaches

2. Minimum Core Datasets Group Discussion

a) Barriers

General paediatrics is a large specialty covering ground in numerous subspecialties, and that there is a risk that these subspecialties may also be collecting routine data- need to avoid overlap.

NICU and PICU both have their own research-oriented routine dataset (PICANet http://www.picanet.org.uk/) applied to every inpatient, though does not believe that their model is scalable to fit our purposes.

b) Other models

The registry model by which data on chronic conditions (e.g. CF) is collected prospectively for each patient from their first diagnosis but this method is of little use for one-off acute illnesses which make up a large proportion of paediatric care.

c) Function

What the purpose of this dataset will be - is it for research, for monitoring quality, for making comparisons? If for research, it would be easier if it was designed with a specific trial in mind than trying to make a general set to suit all possible purposes. Could it be tailored to specific conditions and settings or whether it will be more generalised- must find balance between useful data which is better than HES while making sure that it is not so extensive that clinicians do not bother completing it properly. PERUKI tried to tackle this issue for ED presentations but found that it was too difficult and were forced to abandon the idea. They have recently begun to look at it again.

d) Outpatients

Perhaps outpatients should be a priority here- HES ( Hospital Episode Statistics) provides some useful data for inpatient admissions, though there are questions how reliably collected it is beyond dates of admission/discharge, but virtually no routine data are collected for outpatient appointments.

e) Non specific presentations in outpatients-

Headaches, abdominal pain etc. ;lots of guidelines exist for managing these but since no data is collected we aren’t really sure what the prognosis is for these patients after their appointments- are these guidelines correct?

f) Data governance

regulations on use of routine data for research exist . Perhaps a notice in the waiting room is sufficient as long as only the routine data is used, and no additional information is collected.

g) What is happening now?

Many centres do their own routine collection, and all use their own systems for classification and coding- how do we overcome this? The minimum core dataset could actually be a solution to that problem, by encouraging unification of different systems and consensus on which ones to use.

Should a core data set be a a priority for GAPRUKI- clearly very problematic and even PERUKI failed to solve it, major challenge to take on so early.

Starting small may help – PERUKI are currently working on a minimum core dataset purely for cardiac arrest. Many different specialties and professions will be affected, so it will be important to collaborate with and involve many of these people when making these decisions.

3. NUDGE

a) How to use NUDGE

Consenting into studies and increasing awareness of research to the public.

To improve the rate of consent for inclusion of paediatric patients in clinical research by their parents.; they are often confused by the information given to them about studies. Most parents are enthusiastic about participating in research and believe that it is important, but then do not go on to get involved in it due to poor communication of important information.

Social media may be an answer- says that the FiSh study had more success engaging parents through these avenues than with letters.

Engaging adolescent patients in research- approaches must be very different to those used in communicating with parents.

b) Using NUDGE to encourage other clinicians to join GAPRUKI and participate in its projects.

These kinds of approaches were very useful in attracting people to PERUKI, The key is to make clinicians feel safe and involved, and to add a personal touch- face-to-face engagement very important. Branding, is important example of the value of the owl brand to PERUKI’s success. A concern that engaging general paediatricians in this network will be more difficult because few are involved in research. The long and tedious governance process do cause problems. Providing incentives is crucial

Personalise all communication

Named emails
BCC on all group emails
Phone calls

Include face to face meetings

Free lunches!
Meetings in places where people already are (for example at conferences)

Make members of the network feel valued!

Provide rewards which are agreed in advance

Named on publications/ CEA points

Break down barriers to research

Particularly where governance is involved
Provide advice and expertise regarding ethics/ R&D etc

Consider making data relative open access once collected

c) How could GAPRUKI research ideas use NUDGE?

i) the national childhood measurement programme could be amenable to NUDGE within the UK part of GAPRUKI – obesity work- this work will be developed pre GAPRUKI for GAPRUKI by KC,VS and AS and circulated in a more advanced stage later this year

ii) A Was Not Brought (WNB) children’s parents letter can be subject to nudging, effecting the choice /decision architecture and we can trial that across the GAPRUKI network as an early win for the group this work will be developed pre GAPRUKI for GAPRUKI by KC,VS and AS and circulated in a more advanced stage later this year

iii) Medication adherence in Asthma and Epilepsy could likewise be amenable and this may well be one to go for external funding.

This will require further reading as there is an extensive research portfolio on adherence.

iv) reducing unnecessary ED presentations could be the number one priority for use of NUDGE- best chance of altering patient/parent behaviour with a potential big positive impact.

The principle of ‘under promise and over deliver’ may boost the reputation of the network and making people more motivated to work with it. It is important to achieve some good results quickly/early will provide a good foundation to build on.

4. Patient and Public Involvement and James Lind

JLA can provide an infrastructure and process by which research questions in a specific area can be collated and prioritised.

a) Process

The JLA process involves patient/ carer involvement

Often data on which questions are important come directly from patients/ parents
Patients/ carers are then involved in the prioritisation stage too (50:50 with clinicians)
Researchers are not involved in the prioritisation process

Topics

What areas do we want to gather research priorities about?

Broad and open- collecting info about any general paediatric question? or
Narrow and focussed- collecting information about 3-5 areas
- Either most important few conditions or most important few processes/ area of care
Consensus was that collecting research priorities on the most important few topics in general paediatrics would probably be the most effective way to move forward initially.

c) Who?

Who do we want to gather research priorities from?

Patients/ Carers
- If so where would they be identified from?
  - Existing patient support groups (condition specific)
  - Contact-a-Family?
  - Directly from clinical contact?
General Paediatricians or General practitioners

d) How?

How are data collected?

Most often this is done by survey with very open ended questions

e) Funding

How are JLA PSPs funded?

There are a variety of models
Total costs including everything is £40K. However there are ways to reduce the costs
Fixed costs
- JLA advisor costs- about 10 days worth of time - £5-6K
- Workshop costs- 2-3 JLA advisors for 1 day- £1-2K
Other work such as admin, data collection, data processing, checking against existing evidence could be done by existing staff (for example it could be a PHD or other research project). However, there is quite a lot of work involved and it may take about 2 days per week worth of time for someone over the duration of the process (which could be several months).

5. Health Economics

There was a general consensus that individual research projects generated by the GAPRUKI collaborative group should incorporate economic considerations at the design stage where this is merited. A health economist should be invited to collaborate at an early stage of the development of new research proposals where this is merited (Professor Stavros Petrou one of the delegates has kindly agreed to act as our Health Economics Advisor).

Initial areas for research

Bronchiolitis where the clinical and economic burden is considerable and where new drugs in the pipeline await evaluation; and
Children with complex needs who tend to be under-represented in randomised controlled trials.
Development of individual-level of observational datasets should aim, where possible, to incorporate relevant economic variables that would permit estimation of economic outcomes associated with clinical events of interest to clinical and policy decision-makers.
The potential for subsequent linkage to HES and CPRD data should be explored. These observational datasets might also form a basis for exploring the relationship between organisational and individual-level variables and long-term clinical and economic outcomes.
There was a recognition that several relevant economic questions, e.g. estimating the economic burden of specific clinical conditions, can only be addressed using data from large networks of clinical centres.
Primary data collection through GAPRUKI could provide a basis for methodological research, e.g. development of new methods for utility measurement in childhood, development of trials methodology, etc.

6. Systematic review group

Yes it is of course possible to do systematic reviews and Dr Bob Phillips will chair this part of the GAPRUKI activities going forwards.

Dear GAPR-UKI delegate,

On behalf of myself and team and the co-chairman of GAP-RUKI, Dr Colin Powell, I am writing to thank you for attending our inaugural meeting of GAP-RUKI. The written records of the various group discussions are starting to come in and I would urge you to send anything that you may find helpful. Dr Powell and I will then send a written report. I am going to get the logo finally edited when I return next week. There are lots of potential actions for GAPR-UKI, but at this stage I would like to remind the paediatrician delegates that we wish to attract further collaborating units throughout UK and Ireland. This is best done by yourselves according to your regional knowledge. Any interested unit will then be sent a formal invitation. We will also develop a website and a standardised operating procedure for the group.

Thank you very much for being a part of this collaboration and sharing this vision.

Best wishes,