It is comparatively easy to find funding for the randomised trials that may or may not show the effectiveness of innovations, but much harder to fund scale-up
Spreading innovation in health systems, including the NHS, is famously hard. Spread takes years, and many innovations that could bring benefit never spread. There have been many attempts to improve spread—including clinical governance, quality improvement, and collaboratives—but all have been disappointing. Now the Health Foundation is trying another method—social franchising and licensing—and launched its programme last week.
Franchising is familiar to everybody as it’s the method used by McDonalds, Starbucks, Subway, and many businesses to expand their business and bring their burgers, coffee, and sandwiches to millions. Social franchising, using the methods of businesses to spread programmes that bring social benefits, has long been used by non-governmental organisations in low and middle income countries and by charities and others in Britain—but it hasn’t been much used in the NHS. Now the Health Foundation is working with Spring Impact, which has worked with some 120 organisations across the globe to spread social impact, to see if social franchising can work within the NHS.
The two organisations have selected four programmes for the experiment, and the leaders of the programmes met together at the launch of the programme. Spring Impact has a five stage process for spreading social impact: prove the effectiveness of the innovation; design a process to spread it; systematise the process; pilot it; and then scale up. The Flow Coaching Academy developed by Sheffield Teaching Hospitals has been working with Spring Impact for some time, and its clinical lead, Professor Tom Downes, a geriatrician, described how they are scaling up their innovation.
Experience of Flow Coaching Academy
Their innovation is a form of quality improvement. Learning from the car company Toyota, they introduced a Big Room (Toyota called it a War Room) in which once a week they gathered together all the people involved in a service—patients, clinicians, managers, and others—and with two quality coaches worked on projects. One of their first projects was a “frailty unit,” and using the PDSA (plan, do, study, act) cycle they planned and improved the unit and cut bed days by 15%. Every meeting begins with a patient story, and a patient suggested a system of “discharge to assess” in the patient’s own home rather than the traditional “assess to discharge” in the hospital. The result was that the hospital cut the average time to discharge from 6.7 days to 0.4 days. In one year they used this with 10 000 patients, saving 40 000 bed days.
The next question was whether the method could work with others, and Downes worked with respiratory medicine within the Sheffield Teaching Hospitals and succeeded in reducing bed occupancy by 30%. Showing a picture of Subway, Downes said “We have a sandwich-improving patient flow.”
They then designed a model of spread and opted for social franchising of a one year training course. Other models could have been wholly owning subsidiaries, accreditation, or simply dissemination of the methods and results. They had to systematise the training, and they opted for training pairs of quality coaches, one a clinician and one a manager; 18 days of face to face training; and working with the coaches over three months to set up a “Big Room” and then work on an improvement project.
After piloting the training in Sheffield they set up the Flow Coaching Academy and trained 24 coaches from Bath and South Warwickshire, who between them set up 12 Big Rooms. A team in South Warwickshire managed in one year to achieve a reduction of length of stay that had taken three years in Sheffield. A surgical team in Bath introduced cholecystectomy in patients with acute cholecystitis after a few days of antibiotic treatment, rather than the traditional discharge and readmission, and reduced the time from presentation to operation from 100 to four days.
In the second wave of franchising Bath introduced the training course with the support of Sheffield, and between the two centres 46 people were trained as quality coaches. Bath taught exactly the same course as Sheffield. In the third wave Imperial from London and Nottingham joined, and Northern Ireland, Devon, and Birmingham are now joining. By 2022 the programme should have trained 1000 quality coaches.
Funding is always one of the hardest parts of social franchising. Sheffield did not charge for the course, and initially through a grant provided backfill for people attending the course. That doesn’t happen now, and Bath and Imperial charge for the course and pass so much money per trained coach back to Sheffield to maintain the hub.
People across the NHS have heard of Sheffield’s work and come to see what has been achieved. Downes emphasises, however, that the “solution” cannot be simply copied and that the “journey” is important.
Large scale replication
The meeting also heard from Kate Tilley, the director of business development for Pause, a programme for women who have had two or more children taken into care, which has been quickly and extensively replicated. A study in Hackney showed that some 49 women had between them had 205 children taken into care. The suffering and the cost are clearly huge, and this problem is seen across the country. A programme was developed in Hackney of encouraging women to pause, use contraception, and reflect on what they wanted from life. A pilot found positive results. The programme is centred on the women and depends primarily on skilled staff; it cannot, in the jargon of social enterprise, be “manualised,” which makes it harder to replicate.
Nevertheless, with government funding that tapers off Pause has been replicated, and will be in 18 sites by the end of May. Pause has developed a model for replicating its programme that has three phases: feasibility and scoping; set up; and ongoing support. They accept that the programme may not be right for everywhere. Pause provides help selecting staff, training, and continuing support. Getting the right staff and the right governance are both essential for success. The sites pay a membership fee to Pause for the support. One thing that Pause has discovered is that a replication model that works for six sites won’t work for 20—for example, in the length of training and the frequency of support visits.
Tilley summarised what Pause had learnt, and the biggest learning has been the need to balance flexibility in the programme, letting sites do their own thing, and fidelity, keeping to the core of the programme so that it continues to work. From the beginning they have “thought big,” trying to devise a programme that will work eventually for 60-100 sites. “Keep hold,” concluded Tilley, “of what is integral to the model, and focus on what you are trying to achieve.”
Four programmes to try social franchising
Those from the four programmes selected by the Health Foundation and Spring Impact to try and scale up using social franchising and licensing seemed both inspired and somewhat intimidated by these accounts of success in scaling up. They each described their programme and their progress so far. All have evidence of effectiveness, based on randomised trials in some cases, and all have made some progress with scaling up. But all hope that by following the Spring Impact model, with support from Spring Impact, they can be more rigorous and successful.
IRISi is a programme of helping GPs identify and refer women who have experienced domestic violence. They have found much demand for the programme, but their challenges have been to find the resource to support sites replicating the model, being sure that the sites are using the programme accurately, and funding.
Pathway is a programme of incorporating into hospitals multidisciplinary teams led by primary care physicians with experience of working with homeless people to improve their management. It started at University College Hospital and has spread to 12 hospitals. The need is large and growing, and Pathway would like to spread to many more acute hospitals and try working in mental health hospitals.
Pincer was developed in Nottingham University and is a programme of getting pharmacists to identify in GP lists patients at high risk of clinically important prescribing errors. It has shown benefit, including economic benefit, in a randomised trial and is now used in 2000 practices in 199 clinical commissioning groups. Their ambitious aim is to reach 60% of practices within five years.
Developed at Southmead Hospital in Bristol, Prompt is a programme of multiprofessional training to improve care during childbirth. Suboptimal care accounts for half of maternal deaths, and litigation related to problems at birth has cost the NHS £3.1 billion in the past 10 years—now more than £600 for every birth. Prompt has developed a “train the trainer” model and a set of tools. They have reached 70-80% of units in England and have extended into Scotland and Wales, but they worry about quality assurance and long term sustainability.
All of the groups worry about finance, and a panel discussion gave them ideas on finance. I find it ironic that it is comparatively easy to find funding for the randomised trials that may or may not show the effectiveness of innovations but much harder to fund scale-up. I’ve also come to recognise that doing randomised trials is easier than scaling up, and yet academics are well rewarded both financially and professionally for conducting randomised trials but not at all for scaling up. The result is thousands of innovations, most of them having little or no impact, and few successful replications of innovations.
The four programmes will continue to meet and learn together as they continue their journeys of trying to introduce across the NHS innovations known to have both health and economic benefit. I will be watching their progress, or lack of it, and reporting back.
Richard Smith was the editor of The BMJ until 2004.
Competing interest: RS has spoken with Spring Impact several times and helped introduce the organisation to the Health Foundation, but he has received no payment from either organisation. He did eat a few wraps at the launch.