Impact case study database

In the UK alone, there are an estimated two million people with eating disorders. The prevalence of eating disorders varies between countries, but estimates place the total prevalence, across the different forms, as around 4% of those over 16 years of age. Most cases of eating disorders are self-perpetuating in the absence of effective treatment.

Taken together, the findings from clinical trials provided a substantial body of work demonstrating the effectiveness of CBT-E can treat all forms of eating disorder and that it is as effective, or more effective, than the main alternative treatments. For example, one of the systematic reviews of the effects of CBT-E, conducted independently of the Oxford team, concluded that there is robust evidence to support its effectiveness, ending their abstract with the statement that “ There is robust evidence that CBT-E is an effective treatment for patients with an eating disorder” [A].

In terms of specific benefit to patients, the Oxford studies showed that around a half to two-thirds of patients who received CBT-E recovered from the eating disorder. Almost half of the patients who received CBT-E reported no binge eating, self-induced vomiting or laxative misuse at the end of treatment [3]. These studies also demonstrated that the benefits of CBT-E are well maintained one year later, despite the patients in these studies having had an eating disorder for eight years on average [2,3]. Importantly, a significantly greater proportion of patients receiving CBT-E recovered by the end of the treatment (66%) compared to those receiving an alternative leading psychological treatment, Interpersonal Psychotherapy (33%) [3]. The findings also indicated that improvements following CBT-E extended meaningfully beyond the eating disorder, including important reductions in depressive features, that were maintained beyond 1 year [3].

The clinical and research findings on CBT-E led the Chief Medical Officer in her 2014 annual report to recommend its wide use:

“*I recommend that Clinical Commissioning Groups ensure prompt access to evidence-based enhanced cognitive behaviour therapy (CBT-E) and family-based therapy for eating disorders. This should be available in all areas, as in the NICE guidance, and not restricted to specific age groups.*” [B, page 12].

In 2015, NHS England published access and waiting time standards for the treatment of younger patients with an eating disorder. It too specified that there should be prompt access to CBT-E as well as to a specific form of family therapy [C, pages 24 and 45].

In 2017 the National Institute for Health and Care Excellence (NICE) published new clinical guidelines on the treatment of eating disorders [D], strongly endorsing CBT-E. The guidelines state that CBT-E (referred to as CBT-ED) is the leading specialist treatment for bulimia nervosa [D, p.20] and binge eating disorder [D, p.19] and one of three leading specialist treatments for anorexia nervosa [D, p.11-12] in adults. For younger patients, CBT-E was one of the two leading recommended treatments for all three eating disorders [D p.15, 20 and 22] (the other treatment being a specific form of family therapy).

3,500 therapists worldwide, across 40 countries, have received the free web-based CBT-E training programme that CREDO have shown to be effective [E, F], including 1,200 therapists within the NHS. The last annual survey of users’ experience of this training has shown that 98% would recommend it to others and over 90% have found it to be very helpful [G]. Testimonials from clinicians include:

“We have found your online training to be extremely valuable for our clinicians. It is very reassuring to know that my clinicians have received world class training in CBT-E and I have observed first-hand the value of this during our clinic meetings and supervision sessions.” (Clinical Director, Western Australia, 2020) [Hi]

“I continue to use the treatment myself with patients and also supervise others in use of the CBT-E model. I often consult the online training resources for myself and use them additionally as a supervision tool. The scalability of this mode of delivery has meant that therapists who are relatively novice in eating disorders can have flexible access to modern training techniques to supplement the delivery of their treatment.”

(Clinical Psychologist, describing work in a private service) [Hii]

The additional face-to-face workshops held by CREDO have supplemented the web-based programme, training a further 750 therapists (who have not taken part in the web-based training), including 300 NHS therapists. CREDO also provided two advanced workshops for 300 therapists already trained in CBT-E from a diverse range of countries.

In total, around 1,500 NHS professionals have received training in CBT-E from the CREDO research group using both forms of training since 2013 Following consultation with trainees, it is estimated that the average therapist who has participated in this training will have treated at least 20 cases using CBT-E. This suggests that at least 30,000 patients have received CBT-E. In addition to the research studies supporting the effectiveness of CBT-E there are numerous positive reports from clinicians using CBT-E in the real-world. For example, an outpatient service described that:

“As a team, we implemented CBT-E across the outpatient service in around 2014 [as] we felt close adherence to the CBT-E model would help more of our patients get better. ... I believe that the service we provided to the patients was of exceptional quality and this was supported in large part by the work of the CREDO team and their materials.”

(Clinical Psychologist, for work in an NHS eating disorders service) [Hii]

and an inpatient unit that introduced CBT-E in 2016 reported that:

“... we have experienced a transformation of the clinical service, not only in the practice, but also the philosophy of collaboratively working with patients.. We have been able to discharge 5 to 6 patients completely from services to the care of a [GP] with full remission of eating disorder symptoms and restoration of weight. In the previous 26 years of practice I have not been able to discharge patients in a similar manner.”

(Consultant Adolescent Psychiatrist, NHS inpatient unit) [Hiii]

In response to high levels of demand for information about CBT-E during the REF period, in 2020 CREDO, together with Italian colleagues, launched a dedicated website [I]. This website provides up to date information on CBT-E for members of the public (whether they are patients or family members) and health professionals. A recent analysis of its use over a ten-month period showed that over 30,000 users from 144 countries accessed this website [J].

The research described above has had a major impact on the diagnosis and treatment outcomes of PTSD in the NHS and overseas, and on the delivery of more effective resilience training to emergency workers and other people exposed to high levels of stress.

The National Institute for Health and Care Excellence (NICE) has issued revised guidelines on the optimal treatment of PTSD (NG116, 2018) on the basis of a meta-analysis of RCTs. Based on the RCTs conducted at Oxford (e.g. [ 5, 6]), CT-PTSD received the highest level of endorsement as a first-choice treatment for acute and chronic PTSD. The guidelines state that cognitive therapy for PTSD should be used as a psychological intervention for the prevention of PTSD in adults (1.6.15) and as a treatment for adults with a diagnosis of PTSD or clinically important symptoms of PTSD (1.6.16) [A(i)]. Similarly, the clinical guidelines of the American Psychological Association (2017) state *“For adult patients with PTSD, the panel strongly recommends that clinicians offer … cognitive therapy (CT) ...*” [A(ii), Table 1]. The International Society for Traumatic Stress Studies (2019) also give a ‘Strong Recommendation’ to CT-PTSD as a first-line intervention on the basis of their meta-analysis [A(iii)].

CT- PTSD [ 1, 5, 6] is now widely taught in the UK and is included in the national training curriculum for Improving Access to Psychological Therapies (IAPT) high intensity therapists [B] and is taught on the majority of IAPT high intensity therapy courses. Since August 2013, around 4,800 IAPT therapists have learned the treatment approach and are delivering it in over 180 local services (over 74,000 patients treated since 2013, currently over 10,000 per year) [C], in addition to patients treated in secondary and tertiary care. The treatment is taught on other post-graduate diploma courses in psychological therapies and in specialist CBT diploma/ MSc course at Queens University, Belfast (180 therapists trained in the treatment approach on these courses since August 2013) and clinical psychology courses. In the early stages of the COVID pandemic, the Oxford team were asked to give webinars for NHS England on how to remotely deliver CT for PTSD (viewed by 3,270 clinicians until 31/07/20) and traumatic bereavement (2,779 clinicians). The Oxford team was also commissioned by Health Education England to offer a national top-up training for high intensity IAPT therapists delivering CT-PTSD, starting in November 2020.

The PTSD prevention programme developed by the Oxford group on the basis of the prospective studies [ 3, 4] is a recommended intervention for emergency workers and is being disseminated across England and Wales through the Blue Light Programme [D(i)]. By December 2020, 32 emergency services across England and Wales have benefited to date with an additional 10 services currently undergoing training. The benefits have been twofold: access to evidence-based training, and significant improvements in wellbeing as measured by the Warwick Edinburgh Mental Wellbeing Scale [D(i)]. Newly recruited emergency workers (223 to Dec 2020) have benefited from the programme with significant improvements in confidence managing their mental health once they start full-time work, and another 570 student paramedics at 15 universities across England have completed the programme [E(i)]. During 2020 the programme was also being delivered to paramedics in Singapore during the COVID pandemic.

The prevention programme also forms the basis of current resilience and wellbeing interventions, training and outreach programmes that the mental health charity, Mind, is delivering across England and Wales [E]. For example, the intervention is now being routinely offered across 16 Local Minds in Wales (80% of the Local Minds available in Wales) to support a range of people exposed to stress, including 350 older adults, 1,600 university students and families of military veterans (120 people) by December 2020. These recipients have benefited with significant improvements in wellbeing as measured by the short Warwick Edinburgh Mental Wellbeing Scale [D(ii)].

The Oxford group’s research provided evidence for changes in diagnostic criteria for PTSD in the DSM-5 (American Psychiatric Association, 2013) and ICD-11 (World Health Organization, 2018) [F]. Ehlers’ research showing that ruminative thoughts about trauma are functionally different from intrusive memories (as developed in [ 1] and [ 3]) led to a revision of DSM criterion B1 to “ restrict this criterion to involuntary and intrusive distressing memories” [F(i). Similarly the findings on appraisals in Ehlers’ research led to new symptoms in cluster D (“ *Negative alterations in cognitions and mood that are associated with the traumatic event(s)*”) [F(i)]. Ehlers’ research on the perceived ‘nowness’ of trauma memories and the model’s emphasis of a sense of current threat [ 1, 2] was included to define the ICD-11 (2019) symptom clusters “ re-experiencing in the present” and “ persistent perceptions of heightened current threat” [F(ii), citing [1]; and F(iii), disorder 6B40].

The outstanding results obtained with CT-PTSD have led clinicians and health service commissioners from many countries to request training. Since August 2013, Ehlers and her team have provided 112 workshops on the treatment in England, Scotland, Wales, Northern Ireland, The Netherlands, Germany, Norway, Sweden, Finland, Iceland, Switzerland, Italy, Slovenia, Ukraine, Greece, Poland, Hungary, Israel, Sudan, China, Australia and the USA, reaching a combined total of 6,756 attendees, the majority of whom were clinical psychologists, CBT practitioners, psychiatrists and nurse practitioners. A treatment manual and therapist training videos are available free of charge at the Oxford Centre for Anxiety Disorders and Trauma website [G]. Between 1^st January 2019 and 28^th October 2020, 15,340 therapists from 143 countries registered for this training website. A German therapist manual has also been published (Hogrefe, over 5,000 sales since August 2013). A Japanese manual is also available. The treatment was successfully implemented in several Scandinavian countries. It was delivered in a primary care intervention study in Stockholm, with similar effect sizes to the ones in Ehlers et al.’s randomised controlled trials [H]. It is now being rolled out into secondary care services in Norway, in a large government funded project [H]. The success of intensive, 7 day version of CT-PTSD has stimulated the implementation of intensive treatment programmes in several countries including the Netherlands (e.g. [I]) and the USA.

The results for the network meta-analyses and systematic reviews for depression, bipolar disorder and schizophrenia that Cipriani and Geddes with their collaborators developed in Oxford between 2009 and 2018 have had a significant impact on the treatment of these conditions in the NHS and overseas. The team’s 2009 Lancet publication [1] recommended sertraline as first-line treatment for depression, and in 2015 sertraline became the most commonly prescribed antidepressant in the US, with a 22.6% increase in prescriptions, the largest increase in prescribing among all antidepressants (from less than 2,100,000 to more than 2,700,000 prescriptions/year) [A]. It has also been described as the most prescribed antidepressant in Japan [B(ii)].

In China, by 2019 more than 30,000 psychiatrists had followed recommendations for treatment of bipolar disorder [C(i)] and schizophrenia [C(ii)] that were based on findings from the University of Oxford research, directly influencing the treatment of nearly a hundred million patients in China. The Vice-Chairman of the Chinese Neurology Society wrote citing [2, 3, 4], amongst others, and confirmed that the University of Oxford reviews

“... have had a great impact on clinical practice in China: they have changed the prescription patterns of psychotropic drugs in our country and have improved the clinical outcome of our patients, reducing significantly the number of hospital admissions for patient with schizophrenia and reducing the length of stay for inpatients and the number of suicides and deliberate self harm for patients with bipolar disorder.” [B(iii)]

The Professor of Medicine, of Health Research and Policy, and Director, Meta-Research Innovation Center at Stanford University, cited papers including [2,3,4,6] in summarising the influence of this body of work:

“ *This powerful work has had tremendous implications for setting the guidelines of clinical practice... it has been extremely valuable also for offering tangible information on crucial questions to tens of millions of patients worldwide.*” [B(i)]

The University of Oxford work in evidence synthesis [1, 2] allowed for the first time the evidence-based ranking, in order of effectiveness, of all most commonly prescribed pharmacological treatments for depression and bipolar disorder. By including in the analysis drugs that are in the WHO Model List of Essential Medicines, these rankings have greatly facilitated the real-world decision making between clinicians and patients in both primary and secondary care across the globe, and also in low and middle income countries. The Dean of Medical School, University of Toronto, wrote that [1] was “the single most influential paper in shaping treatment guidelines and practice” and went on to explain that, “The more recent network meta-analyses [on depression and bipolar disorder] are another game-changer, as they allow extrapolation from randomized clinical trials to real world clinical settings which are compelling for many clinicians including those outside of academic settings”. [B(iv)]

The team’s research-based recommendations have been included in official clinical guidelines in a number of countries, including:

• Canada - The Canadian Network for Mood and Anxiety Treatments (CANMAT) offers clinicians clear use recommendations for first, second, and third-line treatments, while considering the robustness of evidence and clinical relevance. The 2018 CANMAT guidelines recommend lithium as the best treatment for acute mania, bipolar depression and maintenance treatment for bipolar disorder and suicide prevention [D], based on the results from the network meta-analyses, citing papers including [3] and corroborated by letter [B(iv)].

• United Kingdom: National Institute for Health and Care Excellence (NICE) guidelines on the optimal treatment of bipolar disorder (NCG 185, first released 2014) state that the network meta-analysis [2] “... found robust evidence that several pharmacological interventions are efficacious. Furthermore, there was evidence of differential effectiveness among medications, which is a unique strength of network meta-analysis” [E, section 6.2.4]. The results were then used as the evidence base to support specific recommendations also for the cost-effectiveness analysis: “The interventions assessed in this economic analysis were determined by the availability of data reported in the network meta-analysis by Cipriani and colleagues, 2011” [E, section 6.25, p125].

• United Kingdom: The British Association for Psychopharmacology, the largest such association in Europe, issues guidelines based on the best available evidence to aid clinical decision-making and provide information for patients and carers. Its 2016 guidelines on bipolar disorder draw on the network meta-analysis [2] in the recommendations for treating mania: “The network was highly coherent, and so strongly supports the validity of the overall recommendation to use dopamine antagonist/partial agonists in mania” [F, p523].

• Australia and New Zealand: The Royal Australian and New Zealand College of Psychiatrists Mood Disorders Clinical Practice Guidelines for treating bipolar disorder first drew on Cipriani and Geddes’ work [2] in 2015, stating that, “The choice of medications is determined by the availability of medications, the required speed of onset of action and the proposed future choice of medications (Cipriani et al., 2011)” [G(i), p67]. Their update in 2020 cited [1,2,4,6] both in recommending ‘Choice’ antidepressants [G(ii), section 7.2] and those to discontinue for children and adolescents [I, box 18]. While the 2020 guidelines were in preparation, the Chair of the working group for these guidelines highlighted the role of the most recent work [4,6]:

“Presently, the 2015 guidelines are being updated and will be published in 2020, and once again Cipriani’s articles feature strongly.... In particular his most recent network meta-analyses comparing the efficacy and tolerability of antidepressants for major depression in children and adolescents and the acceptability of 21 antidepressant drugs for the acute treatment of depression in adults (2016 and 2018).” [B(v)]

The 2017 meta-analysis [5] has been used to accurately quantify, for the first time, the magnitude of the effect of antipsychotics versus placebo in the acute treatment of schizophrenia, informing guidelines in the USA and UK.

• USA: The American Psychiatric Association guidelines recommend that “ patients with schizophrenia be treated with an antipsychotic medication and monitored for effectiveness and side effects.” [H(i), Statement 4] *. * They cite [5] as one of two “high-quality meta-analyses that examined findings from RCTs of antipsychotic medications in schizophrenia” and conclude that “the strength of the research evidence is rated as high in demonstrating that the benefits of treatment with an antipsychotic medication outweigh the harms...” [H(ii) pp210-11].

• United Kingdom: The same meta-analysis [5] was used to inform the updated British Association of Psychopharmacology guidelines recommending pharmacological treatment of schizophrenia: “A meta-analysis of placebo-controlled acute treatment studies in established schizophrenia found that just over 50% of participants responded to antipsychotic treatment, compared with a figure of 30% for those receiving placebo. The respective proportions for ... ‘much improved’ were 23% and 13%”. [I, p11]

The extensive press coverage of the 2018 Cipriani et al. study [6] led to the spontaneous #MedsWorkedForMe hashtag, used to share personal stories about how antidepressants helped them manage their mental health. In just a few hours, the 5 posts in the initial thread were retweeted 335 times, liked 1,448 times, and generated 156 replies and then retweeted 730 times. This social media activity is reflected in the study having the highest ever Altmetric score of any psychiatry paper (4,543), and in the top 5% of all research outputs. The hashtag was a ‘UK Trending Topic’ in February 2018 [J(i)], and the hashtag originator went on to be interviewed by the British Medical Journal [J(ii)]. Patients sharing their positive stories were prominently featured in the extensive media coverage [K]. Example tweets in response include:

“I think seeing people you look up to publicly admitting to their depression helps you realise you’re not alone and sharing real advice on what works for them - drugs, … - is probably one of the most powerful things someone can do.”

“The stigma which surrounds taking medication for mental illness needs to end. Mental illnesses are valid illnesses. No one should ever be made to feel shamed for taking medication.”

“Mental illnesses must be the only life threatening illnesses that medication is seen as a luxury and not a need. Antidepressants aren’t ‘happy pills’; they don’t cure you, they just make it a bit easier to function. Stop stigmatising, these drugs can save lives.”

Chronic insomnia affects people through lack of energy, poor concentration and productivity, increased accident rates, heightened stress and low mood and longer term, increases the risks of diabetes, heart disease, infectious diseases and early mortality. It is a barrier to effective self-care and leads to greater healthcare costs. CBT for insomnia (CBTi) is lastingly effective and is the recommended treatment of first choice for chronic insomnia in European and American guideline documents. Sleepio has a considerable evidence base including the University of Oxford-based trials [3-5], the findings from which have influenced uptake of digital CBTi as an intervention to improve mental health and wellbeing in particular. Two of these trials [3,4] are highlighted in the 2019 consensus statement for evidence-based treatment of insomnia from the British Association for Psychopharmacology as demonstrating a causal relationship between improvements in sleep and improvements in mental health symptoms, wellbeing and quality of life [A]. The considerable evidence base for Sleepio contributes to the statement in this document that digital CBTi has the potential to offer a choice amongst evidence-based alternatives (CBT or drugs) in routine clinical care. Additionally, the University of Oxford studies on students and employees [3,5] are cited in a RAND Europe report reviewing promising workplace wellbeing interventions to aid organisations in developing their workplace wellbeing offer and contribute to Sleepio being given the highest evidence-based rating in the report [B].

The trials conducted by University of Oxford researchers [3-5] have been cited in evaluations of Sleepio, which have subsequently influenced uptake. For example, the first systematic evaluation by NICE of a digital therapeutic intervention was a MedTech Innovation Briefing (MIB) on Sleepio in November 2017 and summarised evidence ‘based on 5 well-designed and well-reported randomised controlled trials’ that included the Freeman et. al. [3] and Bostock et. al. [5] trials [C].

The same trials [3,5] were referenced in the successful application, ‘Enabling better health and self-care at scale with digital sleep medicine’ submitted by Big Health and the Oxford Academic Health Science Network (AHSN) to Innovate UK for its Digital Health Technology Catalyst competition in 2017. The project aimed to explore different routes to enable immediate access to CBT and was one of only 9 funded from several hundred in total [D]. In this first large-scale NHS rollout of direct access digital medicine, Sleepio was made freely available across the Thames Valley (an adult population of 2,700,000) and individuals were able to access the programme without needing a GP referral or prescription. Launched in October 2018, the project has been adopted in local primary care, by local employers and by the Improving Access to Psychological Therapies (IAPT) programme (a national programme which uses talking therapies to help people overcome depression and anxiety and better manage their mental health). From October 2018 to June 2020, more than 16,000 people completed a Sleepio sleep score test, with 7,078 of these going on to start the personalised CBT programme. There were 3,600 referrals from 19 GP practices over this period and there was positive feedback on the ease of prescribing, integration into GP workflow and reduction of the burden on GPs. In addition, results from nine GP surgeries projected savings of around GBP100 for each Sleepio user over 3 years. A joint report from the Oxford AHSN and Big Health published in August 2020 [E] highlights these benefits.

Big Health partnered with the NHS to provide free access to Sleepio for all NHS and social care staff between April 2020 and December 2020 (3,000,000 individuals) as part of a package of health and wellbeing support for keyworkers during the COVID-19 pandemic [F]. Clinical evidence from the University of Oxford trials [3-5] was important in the decision by the NHS to choose Sleepio and Daylight as two of the four mental health and wellbeing apps made available [G,H(i)]. In England and Scotland, 18,000 sleep tests for Sleepio have been accessed through this provision with 10,000 individuals going on to start the CBT programme (up to the end of November 2020) [H(ii)].

Following the success of the Thames Valley project, five health boards in Scotland covering a population of 1,700,000 approved roll-out of Sleepio in evaluation projects in December 2020 to tackle insomnia and anxiety. The University of Oxford efficacy studies [3,4] and impact of the Thames Valley study were said by the National Advisor for Digital Mental Health for the Scottish Government to be crucial in securing the roll-out [I].

Since 2019, there has been a major increase in the uptake of Sleepio in the US where it is being adopted to improve workplace health, wellbeing and productivity. In June 2019, Sleepio announced a landmark partnership with US insurers CVS Health which has 94,000,000 medical benefit members, offering the prospect of mainstream adoption of digital therapeutics for the first time within the US healthcare system [J]. [Text removed for publication]. Uptake of Sleepio in the US has been facilitated by the strong evidence base from University of Oxford research [3,4] and recognition that the product provides a scalable approach to tackling poor sleep amongst the workforce and improving well-being and productivity [L].

Both the 8 and 2-question versions of the SCI have been routinely adopted for use in Sleepio. It is the metric used to assess the benefit of Sleepio and provides a robust tool in place of the sleep diary that individuals previously completed. Users go to the Sleepio website and complete an initial 5-minute sleep score test which includes the SCI. This simple and clinically meaningful entry into Sleepio is believed to have facilitated Sleepio’s adoption by individuals, clinical services and workplace referrers, based on numbers who have enrolled in the initial sleep score test: 16,000 and almost 18,000 for the Thames Valley and NHS/social care staff roll-outs respectively.

Recognised as a robust means of measuring sleep, the SCI has also been adopted as a clinical tool for insomnia outside Sleepio. It was promoted for use in primary care by GPs in 2019 via the Red Whale GP CPD update [M], a recognised and reliable source of clinical knowledge and guidance that is sent to ~30,000 of the UK GPs. The update recommends the 2-item questionnaire, leading to the 8-item questionnaire if required, as a short screening tool to use in some chronic disease reviews and cites Espie and colleagues’ work [1].

The digital therapeutic intervention Daylight, evaluated in the RCT carried out at the University of Oxford [6], has been made freely available alongside Sleepio between April 2020 and December 2020 to support NHS and social care staff during the COVID-19 pandemic [G]. Up to the end of November 2020, almost 9,000 staff in England and Scotland had signed up for Daylight with around 4,600 going on to start the CBT programme [H(ii)]. The roll-out of Daylight in clinical services across 5 health boards in Scotland as part of a national computerised CBT programme, covering 1.7m people, began in December 2020 [I]. Daylight has also been added to the CVS Health formulary in the US making it available to around 2,800,000 individuals there [L].

For over 30 years I’ve been living with the awful inevitability that I was going blind but now, as a result of the operation, there’s a real prospect that I will continue to be able to see and that’s just absolutely fantastic.’ Trial participant [A]

The advance in gene therapy technology developed at the University of Oxford is transforming the treatment of two degenerative retinal disorders. Choroideremia is a degenerative retinal disorder affecting one in every 50,000 people. In Phase 1/2 [1, 2], the Oxford researchers conducted a retinal gene therapy clinical trial in 14 patients with choroideremia, treating one eye in each patient (ClinicalTrials.gov reference NCT01461213, 2011-2017). Visual acuity in treated eyes improved relative to untreated eyes over the two-year trial period, commencing October 2012. Furthermore, testing the single treated eye 12 months after gene therapy showed that three patients could read three further lines on a standard optometric eye test. Longer term follow-up with a mean of 3.6 years for 12 participants confirmed that visual acuity gains were sustained, indicating that retinal gene therapy can improve and sustain visual acuity in a cohort of late-stage choroideremia patients in whom rapid visual acuity loss would ordinarily be expected [3].

One patient in the clinical trial found the improvement very noticeable, reporting:

“Immediately afterwards the visual acuity in my left eye was improved enough to see another couple of lines on the sight screen.”

Another patient’s testimonial described their expectations being exceeded:

“I have had a substantial improvement in the treated eye. This was more than I expected and a better result than the Professor [MacLaren] had contemplated. I have little doubt that without the benefit of this trial my left eye would now be useless.” [A]

X-linked Retinitis Pigmentosa is an inherited disease affecting one in 40,000 people. In the Phase 1 clinical trial in 18 patients with x-linked retinitis pigmentosa, visual field improvements began at one month and were maintained to the last point of follow-up (six months) in six patients. All patients described subjective improvement in visual clarity and increase in field of vision in the treated eye by one month of follow-up [5].

These remarkable outcomes attracted national media coverage in the UK [B, C].

The charity Fight for Sight referred to the capabilities of gene therapy in their 2020 report, ‘Time to Focus’ [D], examining the personal and economic cost of sight loss. The report draws the conclusion that “ sight loss is not inevitable” and highlights the choroideremia trial [1] as an example of success. Its recommendations include a focus on prevention, not only care; and greater investment in sight loss research, including investment “ to harness the potential of new approaches”. It also reports that the World Health Organisation and the United Nations highlighted in 2019 that more investment was needed in sight loss research globally.

Following positive outcomes in patients participating in the original Phase 1/2 study, it was clear that there was a need to expedite clinical evaluation of the gene therapy for choroideremia, and to advance the gene therapies (all developed at Oxford) for X-linked retinitis pigmentosa and other inherited retinal disorders in to human clinical trials.

Professor Robert MacLaren first met with representatives of Syncona (then the investment arm of the Wellcome Trust) to discuss this in November 2012. They had already identified retinal gene therapy as a key area in which a company could be built, and begun sector-wide research, meeting with leaders across the field. The University of Oxford and Syncona founded spin-out company NightstaRx Ltd, trading as Nightstar, in early 2014. [E]

The ABCA4 [6], AAV-CNTF and the AAV-melanopsin programmes were subsequently licenced to Nightstar, thereby adding considerable value by diversifying the retinal gene therapy portfolio to include many causes of inherited sight loss.

Over 2014-2019, Syncona invested funds to develop the company, and licensed further programmes from MacLaren’s research from the University of Oxford. Nightstar developed two key products: NSR-REP1, potentially the first treatment for choroideremia; and NSR-RPGR which is seeking to treat retinitis pigmentosa. By December 2018, Nightstar had grown to comprise 47 employees based in London and Boston offices. [F]

Nightstar was one of Syncona’s first gene therapy companies, and helped establish their portfolio in this area. Nightstar was acquired by Biogen for USD877,000,000 in early 2019. The deal ranks as the third most valuable British biotech exit in the last two decades [G], and, delivering for Syncona a 4.5x return on their original investment of GBP56,400,000. [E]

This was the first time Syncona had sold a portfolio company.

The Chief Executive of Syncona stated at the time:

“In the six years since Syncona founded Nightstar, the company has established itself as one of the global leaders in retinal gene therapies. It is a strong example of our differentiated approach of founding, building and funding global leaders in life science.’ [H]

The acquisition of Nightstar expanded Biogen’s focus in ophthalmology, one of its core growth areas, as well as strengthening Biogen’s foothold in gene therapy. [I] At the announcement of completion of the sale in June 2019, Biogen’s Chief Executive Officer, said

“Today marks a significant achievement for Biogen… The acquisition of Nightstar further bolsters our pipeline and is an important step forward toward our goal of a multi-franchise portfolio across complementary modalities.” [J]

Biogen have made substantial investments in Nightstar, demonstrating that gene therapy is a very active workstream in its portfolio. In December 2020 they completed the Phase 3 STAR study (NCT03496012, 170 participants) for the potential treatment of choroideremia [K]. The study was designed to investigate the safety and efficacy of a single subretinal injection of the gene therapy. Nightstar has also commenced a preclinical development programme for the dual vector technology for Stargardt Disease, based on [6].

In 2014, OANG tested 8,944 samples for autoantibodies against 8 targets. Of these, more than 95% were UK patients. By 2019, the OANG testing had expanded to 24,552 samples on 12 targets, including approximately 10,000 for the Nationally Commissioned Service for patients with neuromyelitis optica (NMO) which focuses on aquaporin-4 antibodies [A], and is the only UK provider of this optimised live-cell based assay. OANG’s test identifies 80 new UK patients/year with aquaporin-4 antibodies. Given that aquaporin-4 antibodies have an overall incidence of 1.3/million person-years (equivalent of 85 new cases/year in the UK), these accurate tests are capturing the whole of the relevant UK patient population. In addition, this aquaporin-4 antibody test is run on 30,000 samples/year at the Mayo Clinic (USA), covering about 60% of USA test requests [B].

OANG’s assay to detect MOG antibodies is the most specific worldwide in comparison to four international testing centres with an interest in MOG antibody testing [5]. OANG is the only provider of MOG-antibody IgG1 testing in the UK. UK test requests for MOG autoantibodies have increased by 61%/year over the last 5 years and this test received statutory approval in the USA in 2018. It is now run on 25,000 samples/year, approximately 55% of USA requests [B].

“The Oxford Autoimmune Neurology Diagnostic Laboratory has had a significant impact, not only on the field of autoimmune neurology but also on the development and optimization of platforms for neural antibody testing, many of which have been incorporated into the current Mayo Clinic Neuroimmunology Laboratory testing platforms.”

Director of the Neuroimmunology Laboratory, Mayo Clinic. [B]

Since the OANG’s original report in 2008, a 2018 literature review identified descriptions of 186 patients with glycine receptor antibodies who were almost universally responsive to immunotherapies. Although rare, the consistent clinical findings and response to immunotherapy in patients from multiple countries has led to the establishment of glycine receptor antibody testing in major testing centres in Germany (Jan 2019, [Ci]), and the USA (Aug 2020; [A], [Cii]).

These autoantibody findings dramatically alter patient care. Three clinically similar diseases: multiple sclerosis, aquaporin-4 antibody and MOG antibody disorders all respond to different therapies and require different durations of administration. Patients with MOG-antibodies typically require short-term immunotherapies. Patients with aquaporin-4 antibodies have lifelong disease relapses which respond to long-term immunotherapies but worsen with treatments that are effective in multiple sclerosis (MS). In particular, misdiagnoses of NMO patients as MS means many patients suffer preventable disability or death and can respond to costly MS treatments such as interferons with catastrophic worsening [D]. OANG’s improved live cell-based assay for aquaporin-4 antibody detection has led to 50% more patients identified with NMO in the USA each year - this equates to 1,658 extra patients captured between 2014 and 2020 using live cell-based assays [4; E].

Importantly for patient benefit, the time to diagnosis was reduced (1 versus 18 months). In 2019, class I evidence led to FDA approval for three effective immunotherapies which treat NMO, highlighting the impact of the new diagnostic criteria. A neurology clinician at the Massachusetts General Hospital summarised the benefits to patients being that, “ without the MOG assay, these patients would be left without a diagnosis or treatment, and would have a poor prognosis”, including hundreds of his patients alone [F(i)].

OANG’s work has revealed that of all patients with VGKC antibodies, only those with either LGI1- or CASPR2-antibodies benefit from immunotherapies [1, 3, 6], and that treating the remaining 80-95% with VGKC-antibodies is harmful, or at least non-beneficial [G]. Consequently, VGKC testing was ceased by the Mayo Clinic in 2019 [H] and replaced by the OANG tests for LGI1 and CASPR2 antibodies, now performed on approximately 50,000 samples per year [B]. Many other centres worldwide have followed suit, including the Oxford NHS Trust immunology laboratory in 2020 [H]. Hence, the approximately 30,000 and approximately 100,000 VGKC-antibody tests performed in the UK and USA, respectively, are now preferentially tested for OANG’s discovery of LGI1/CASPR2 antibodies. The new tests help avoid unnecessary immunotherapies in approximately 20,000 individuals/year in the USA, Europe and the UK [G].

The OANG’s development and evaluation of the most accurate aquaporin-4 and MOG autoantibody assays [4, 5] led to their uptake during the current REF Impact period in South Korea [Fiii], USA (Mayo Clinic, Minnesota [B], Harvard, Boston [Fi]) and Canada [Fii]. In South Korea, this has led to an increase in identification of patients with NMO (by 18.5%) and a 10% increase in the disease diagnosis, patients who would previously been left undiagnosed and untreated. The OANG’s aquaporin-4-antibody test has been established at the British Columbia Neuroimmunology laboratory in Vancouver, Canada, now providing 100 tests/month. Cell-based assays are now recommended as gold standard in the 2015 international guidelines for diagnosis of NMO spectrum disorders [I, 2,090 citations]. These highly cited criteria have improved the diagnosis of NMO globally, with validation studies from every continent demonstrating improved patient identification: e.g. France (97% accuracy vs. 82%), in a single centre in one year in India (91 vs. 30 patients), and in four countries in Latin America (104 vs. 64 patients).

a. Commercial clinical trials: Identification of LGI1 and CASPR2 antibodies [1, 6] has been used as an inclusion criterion for an industry-funded clinical trial in the field [G]. Patients benefited from a reduced frequency of seizures on the test immunotherapy (intravenous immunoglobulins), which confirmed the previous OANG data [6]. This result now permits approximately 250 US patients per year access to immunotherapies otherwise denied by insurers [B].

[Text removed for publication]

b. New technologies to support drug discovery: OANG’s LGI1- and CASPR2-antibody tests have been commercialised and sold in 57 countries worldwide [K]. The OANG-pioneered membrane-tethering technology, used in the development of the LGI1-antibody assay [1], has been taken up by Retrogenix Ltd, to detect binding partners of soluble molecules and off-target binding of biotherapeutics [K]. Since 2017 this technology has been integrated into nearly every client project run by Retrogenix including screening 1,000 drug molecules in more than 400 projects with more than 200 clients worldwide. The data generated supported 75 successful commercial clinical trial applications in the last 2 years. Retrogenix describe the benefit to their clients who “ have been able to make more informed decisions as to which molecules to progress within their pipelines, and into the clinic. This reduces their failure rates, and allows safer drugs to enter clinical trials, meaning an economic benefit, and even more importantly, less likelihood that patients will encounter unexpected toxicities.” [L].

Correct patient diagnosis and treatment leads to better patient outcomes, often with a return to activities of normal daily living. Accurate assays have also provided savings to the NHS since 2019 by treating patients who are CASPR2 or LGI1 antibody positive, and not treating (80-95%) those with VGKC antibodies (without CASPR2 or LGI1 reactivities) where there is no evidence of benefit on immunotherapy.

Williams’ new meditations, a translation of his research into practice, were integrated into the second edition of the widely-used MBCT therapist manual he co-authored, ‘ Mindfulness-Based Cognitive Therapy for Depression’, published in 2012 (Segal, Williams and Teasdale, [A]). This second edition also drew on his wider body of work for a new chapter, ‘ *How Does MBCT Achieve Its Effects?*’, citing 7 of his papers from Oxford work (published 2007 to 2010). The manual includes detailed guidelines on therapist training; all the resources required for disseminating MBCT with high fidelity; and recordings of his new meditation practices. Other changes drawing on research since the first edition included allowing additional time in the pre-class interview for participants more likely to drop out, citing the determination of the relevant factors in Williams’ research [2]. This manual defines MCBT practice internationally, and in what follows, ‘MBCT’ denotes its formulation in this second edition unless otherwise stated.

In 2016, Kuyken was appointed by NHS England to develop a national curriculum to train NHS staff working in Improving Access to Psychological Therapies (IAPT) to deliver MBCT. This curriculum [B] led to Health Education England directly commissioning training. In its first two years (2018-19), the programme trained 65 NHS staff members to deliver MBCT, with a further 86 places commissioned in 2020. 57% of NHS IAPT services had MBCT trained staff by 2019 [C].

MBCT is now widely accessible to those in need through the Improving Access to Psychological Therapy (IAPT) programme, with care to maintain its potency and fidelity. IAPT services were able to expand MBCT provision significantly (treating 1,807 patients when first recorded separately in 2016-17, growing to 3,957 patients in 2018-19) [Di]. From analysis of IAPT data, Kuyken and team have demonstrated in a study published in 2019 (n=1,554) that outcomes of MBCT in routine clinical practice match those observed in research settings, with 45% of those entering treatment with current depression recovering, and 96% of those entering treatment in remission sustaining their recovery [Dii].

MBCT is recommended in many international treatment guidelines, including the 2019 North American APA guidelines [Ei] and the UK 2018 draft updated NHS NICE Guidelines [Eii]. These guidelines are the basis for mental health care in these nations. For example, the APA guideline recommends either second generation antidepressants or psychotherapy, including MBCT in a list of five approaches with comparable effects on long-term outcome. The UK NICE draft updated guideline reaffirms MBCT for people “ *who have recovered from more severe depression when treated with antidepressant medication (alone or in combination with a psychological therapy) but are assessed as having a higher risk of relapse.*” The Canadian guidelines [Eiii] recommends MBCT as a first-line maintenance treatment adjunctive to medication. The Oxford-based trial [3] and/or 2016 meta-analysis [5] are cited in the UK, Belgian [Evi], Canadian [Eiii] and Australian/New Zealand guidelines [Ev].

The positive results of multiple MBCT clinical trials, including [3], and Kuyken’s patient-level meta-analysis [5] have resulted in clinicians and health commissioners from many countries implementing MBCT. From 2014 onwards Kuyken, Williams and their Oxford team have delivered much of this international training to build capacity in countries where no established MBCT services existed, including China, Hong Kong, Brazil, Taiwan, Singapore, Japan, Czech Republic, India, Hungary, Uruguay and Poland, following guidelines based on the therapists’ manual. In Hong Kong, Taiwan, Brazil, Uruguay and Singapore there are now established self-sustaining centres. Together these initiatives have trained over 1,070 MBCT therapists (including, 456 in mainland China; 158 in Hong Kong,150 in Spain, and 84 in Brazil) [F].

Williams’ work has shown that MBCT prevents future episodes in those who are most vulnerable, including patients with suicidality [4]. Its evidence-based nature was cited as one of the reasons that led prison healthcare staff in Pentonville Prison, UK (a remand prison housing 1,300 male prisoners where suicide is high) to invite Williams to initiate a pilot project in their Wellbeing Unit [G]. The project included running MBCT groups for vulnerable prisoners (2017-18) as well as training prison staff to deliver MBCT (2018-19) to ensure sustainability. The initial pilot project had 39 prisoners and with significant positive change for those involved, with a mean self-reported score of 8.8 out of 10 for importance [G]. Resulting changes in behaviour included reduction in self-harm, fewer conflicts with prison staff, and better sleep [G]. It also reached 26 prison staff including staff from other prisons, of whom 16 trained to be MBCT therapists. As a result, MBCT is now taught in prisons not only at Pentonville, but also at Brixton, Wormwood Scrubs, and Springhill (Bucks) [G]. The same factors prompted a prison project in Taiwan where four prisons have introduced MBCT. To date, MBCT has been taught to over 400 prisoners in Taiwan and benefits to them have been summarised as “ more positive thinking, self-confidence, and hopes for their future life” [G].

Engagement with wider public: Williams has co-developed a number of more accessible lower-intensity and self-help programmes for people who experience sub-threshold symptoms of depression, or have major depression but are not accessing health services [G]. Williams collaborated with Penman in Mindfulness: Finding Peace in a Frantic World (2011). Translated into 26 languages, since 2013 it has sold more than 743,000 copies worldwide since 2014 [H], is rated as 4.6/5 on Amazon with over 5,000 reviews, and has been in the top ten for lifestyle/mental health on Amazon throughout the REF period. Since 2014 it has been used in schools, universities, workplaces, parliaments [K] and corporate settings, and is the basis for training those who wish to deliver MBCT to staff in these settings (>700 trained in Oxford). Four research trials have demonstrated its acceptability and effectiveness in these contexts [I], an evidence-base that is unusual for a self-help guide. The earlier Mindful Way Through Depression (2007) has sold over 103,000 copies since 2014 and The Mindful Way Workbook (2014) provided a new, highly accessible, self-help format of MBCT for depression and has sold over 104,000 copies [H]. Finally, the Oxford Mindfulness Centre has developed a series of digital resources designed for direct access by people with an interest in mindfulness and MBCT. Its website has been accessed by over 1,500,000 people since 2018 [J].

Engagement with policy makers and public service provision: Following a successful pilot by Williams and Oxford colleagues in spring 2013, a termly series of mindfulness courses to members of parliament, based on Williams & Penman (2011) was established and have continued throughout the impact period. These courses have now reached a total of 286 parliamentarians in Lords and Commons, and been extended to 480 of their staff, with benefits reported in both work and personal contexts [K]. In 2014 parliamentarians who had been through these groups formed the Mindfulness All-Party Parliamentary Group (MAPPG). The MAPPG President spearheaded international political engagement and dissemination of the model of MBCT used in the UK parliament to 39 other countries. He has assisted Parliaments in The Netherlands, Ireland, Denmark, France, Estonia, Wales, Scotland and Iceland to establish similar mindfulness initiatives [K]. In 2015 the MAPPG commissioned the first UK Mindful Nation Report [L] to review evidence and best practice. The report cited Oxford research [3] and the MBCT manual and its recommendations have led to increased work to bring mindfulness into education, the criminal justice system and workplaces [G,K]. More broadly, the MAPPG President credited the “ meticulous work Mark Williams, Willem Kuyken and his colleagues have undertaken in scientific research on preventing depression” as the biggest reason for mindfulness practice being ‘normalised’ [K].

The English Improving Access to Psychological Therapies (IAPT) has greatly increased public access to effectively delivered, NICE recommended psychological therapies for anxiety disorders and depression [A]. From small beginnings in 2008 (when less than 80,000 patients were seen) it now sees around 1,200,000 people per year (NHS Digital, 2020) [B] and the NHS Long-Term Plan (2019) [C(iii)] commits to a further expansion with the programme seeing 1.9 million people per year by 2024. During the impact period the IAPT programme has tripled in size (from 434,247 in 2012/13 to 1,165,653 in 2019/20) [B]. The England-wide clinical recovery rate for individuals who receive a course of psychological treatment has improved from 43% to 52% [B] on the same timescale and is now in line with expectation from randomised controlled trials. Furthermore, nearly 7 in every 10 treated patients (68%) show reliable improvement [B]. New integrated IAPT services that bring together mental health and physical health care have been created and are achieving substantial savings in physical health care costs [D]. Clark has been the National Clinical and Informatics Advisor for IAPT since the programme started and is widely considered to be its principal architect. The work of his research group at the University of Oxford has underpinned substantial parts of the impact of IAPT, as outlined below.

Collecting Outcome data on everyone who has a course of treatment (two or more sessions before discharge). When Alan Johnson MP, announced the IAPT programme on 10^th October 2007, he stated that, once it was established, 50% of treated patients will recover “as expected by NICE guidelines”. In order to assess whether this ambitious target was met, it was necessary to devise a system for collecting outcome data from everyone who was treated. The existing NHS outcome monitoring system, which involved giving patients a symptom questionnaire to compete at the first and last session of treatment, was not up to the job. A national audit of counselling services showed that this system only collected outcome data on 38% of treated patients [A]. To get around this problem, IAPT adopted the session-by-session monitoring system pioneered by Clark and Hackmann [1] in their Omagh community treatment project [E]. On advice from Clark, the programme also decided that the nationally reported outcome metrics for IAPT (recovery rate and reliable improvement rate) would remove any incentive for services to fail to collect data as they would assume that anyone without a post-treatment (last available session) score would have failed to recover/improve. Together, these two decisions have enabled IAPT to achieve an unprecedented pre-post treatment data completeness rate of 99% for everyone who is seen at least twice (NHS Digital) [B]. This is better than in most randomized controlled trials and means that IAPT can accurately assess the outcomes that it achieves.

Improving clinical outcomes. The initial clinical outcomes achieved in IAPT (recovery = 40%) were well-below target. Clark & colleagues’ patient level analysis [2] of outcomes during the first year of the programme showed that better outcomes were associated with giving the NICE recommended treatment, and with services that had higher average numbers of therapy sessions and higher step-up rates. Their subsequent service level analysis [3] of data from 2014 confirmed the importance of an adequate number of sessions and higher step-up rates. It also showed the services that had shorter waiting times, lower DNA rates, and consistently identified the problems they were treating using ICD-10 codes, had better outcomes. Clark fed these findings back to services in an extensive series of regional and national workshops between 2013 and 2018 and they were incorporated into NHS England’s IAPT Manual (2018) [C(iv)], advance draft copies of which were available from late 2016. IAPT activity and outcomes have improved dramatically (recovery from 43% to 52%), during the REF2021 Impact period, meaning that 230,000 more people a year recover and 314,000 more a year improve than in 2013. Regional variation in outcomes has also halved. Longitudinal analyses show that improvements in services’ outcomes are strongly related to adoption of the quality standards identified in the Oxford analyses. Using the full national dataset, Clark et al [3] showed that the between-year improvements they identified in the service organization features (2014/15 to 2015/16) predicted service-level improvement in outcome. Subsequently, Clark reported that services continued to improve on these variables and that the degree of these improvements between 2015/16 and 2018/19 predicted over half the variance in service improvements in the proportion of people who have recovered by the end of treatment (R²=.54) [F(i)]. These findings were independently confirmed using data collected from London IAPT services between 2013 and 2019 [F(ii)].

Providing the combined clinical and economic arguments for further expansion and development of IAPT . On 14^th July 2014 the arguments outlined in Layard and Clark’s 2014 book entitled Thrive: the power of psychological therapies were presented at a major public event [G] in London, chaired by Andrew Marr and attended by senior politicians of all persuasions (including Alan Johnson MP for Labour and Norman Lamb MP for the Coalition). This event, and multiple private meetings with politicians and treasury officials, helped ensure that Labour, the Liberal Democrats and the Conservatives all included a commitment to expand IAPT in their manifestos for the 2015 General election [H]. The government and the NHS subsequently confirmed major expansions of IAPT in the NHS Five Year Forward View (February 2016, page 15) [C(ii)] and the NHS Long-Term Plan (January 2019, page 68) [C(iii)]. Thrive (page 209) also argued for the creation of integrated IAPT services that provide co-located mental and physical healthcare for people with anxiety/depression and a long-term physical health condition. In March 2015, NHS England asked David Clark and the President of the Royal College of Psychiatrists to co-chair a working group to define the operational principles that would underpin integrated IAPT [H]. The working group’s recommendations were confirmed by the British Psychological Society, and the Royal Colleges of Psychiatry, General Practice and Physicians in a joint statement issued in November 2015 [C(i)]. Clark was then asked by NHS England to Chair the Education and Training Group that would develop curricula for training the psychological therapists who would work in the new integrated IAPT services and to also join the group that elaborated the clinical model, selected CCGs to receive early adopter funding, and set-up national and local evaluations [H]. 37 CCGs were funded to develop integrated IAPT services in 2016-2018. The official NHS England Blog (10^th July 2019) [C(vi)] confirmed that these services were achieving excellent clinical outcomes while savings physical healthcare costs as predicted in Thrive. NHS England has confirmed that 75% of all CCGs in England are now developing integrated IAPT services.

Take-Up of NICE recommended treatment for social anxiety disorder. The specialised cognitive therapy for social anxiety disorder (CT-SAD) that Clark and colleagues developed and initially tested in Oxford [5] had been recommended as the first line treatment for social anxiety disorder in the NICE Clinical Guideline (CG159, May 2013) that was published just before the start of the Impact period. This recommendation has since facilitated its widespread adoption. CT-SAD is in the national training curriculum for IAPT CBT therapists (latest edition January 2019 [C(v)]), has been taught to over 4,800 English IAPT therapists and has been used to treat 64,000 patients during the Impact period. During COVID all the IAPT therapists moved to delivering treatment remotely (by video link). NHS England asked Clark and colleagues to provide detailed guidance on how to do this and the resulting webinar was viewed by 2,613 NHS therapists up to 22/07/20. The Canadian Province of Ontario has recently started its own IAPT programme (called “Structured Psychotherapy”) and has adopted CT-SAD as the social anxiety treatment to be included in its therapist training programme [I]. To assist with such the training initiatives, the Oxford team created a free therapist resources website [J] that includes the therapist manual, specialized measures to guide treatment, and stream-able videos of a full-day clinical workshop by Clark plus 45 short videos illustrating key techniques. As well as being extensively used by IAPT training courses, the website had over 16,500 registered clinical users in 144 countries. Together they visited the site 268,459 times during 2020.

International Impact of IAPT. The success of IAPT has attracted considerable international attention during the REF impact period. As the NHS’s Clinical and Informatics Advisor for IAPT and the author of Thrive, Clark has been invited to advise health commissioners and ministers in multiple countries [I]. Many of these countries have subsequently started IAPT-like services and others are considering doing so. The most advanced is Norway, which now has over 50 IAPT services (termed “Prompt Mental Health Care”) that use the same outcome monitoring system to England and report similarly good outcomes [I(i)]. The Ontario [I(ii),(iii)], Israel [I(iv)] and Australia initiatives are also all adopting the outcome monitoring system and other IAPT features that derive from Clark’s research. International press coverage includes a feature length article in the New York Times (24^th July, 2017) [K] described it as “ the world’s most ambitious effort to treat depression, anxiety and other common mental illnesses”, and a December 2018 article that discussed IAPT in the Canadian Globe and Mail [K] was simply entitled, “ For better mental health care in Canada: look to Britain”. An independent report in 2018 [L], commissioned by the Ontario (Canada) Department of Health, concluded that IAPT was the world’s most successful attempt to increase public access to evidence-based psychological therapies for mental health problems. Finally, IAPT was selected for presentation as an outstanding example of UK innovation in mental health at the First Global Ministerial Mental Health Summit Government in London on 10^th October 2018 and in the main plenary session of the World Government Summit in Dubai on 10^th February 2019.

Following a stroke, it is common for a patient to experience cognitive deficits and failure to detect these can have a significant impact on post-stroke recovery. Cognitive screening is vital to facilitate rehabilitation, ensure appropriate support and enhance the quality of life of stroke survivors. The psychological consequences of stroke are some of the most disabling and are also the complications most feared by people living with stroke. In clinical practice, guidelines recommend that stroke survivors are assessed for cognitive issues, but until recently, the brief tools used for assessing post stroke cognitive problems were ‘borrowed’ from other conditions such as Alzheimer’s disease.

The OCS was the first tool to provide a short, neuropsychology based cognitive assessment, in line with the NICE guidelines on domain specific screening. The current Chair of the Organisation for Psychological Research in Stroke (OPSYRIS), a consultant stroke physician, substantiated the role of the Oxford Cognitive Screen in improving standards in post-stroke assessment, writing that

“... the stroke specific Oxford Cognitive Screen, which is free to use and has accompanying comprehensive, easy access training materials exactly matches the clinical needs. Indeed, the Oxford Cognitive Screen (OCS) has rapidly gained visibility and traction within the clinical and research fields since its publication only a few years ago.

I am now using OCS in preference to other tools in our clinical stroke setting, and I know that other Principal Investigators and lead clinicians are following suit. In a recent International consensus statement (Stroke Recovery and Rehabilitation Roundtable) we collated the essential and desirable characteristics of a cognitive assessment tool for stroke research. It was not our remit to name a preferred tool but OCS maps to all the criteria we specified”. [A(i)]

The OCS is licensed free of charge for publicly funded clinical use through Oxford University Innovation (OUI), with 1,214 users licenced from 31 March 2014 to 31 Dec 2020 [B]. It has been widely adopted for clinical use in the NHS: 169 out of 217 NHS Trusts in England and Wales had taken a licence by July 2020. Outside the UK, other major clinical licensees include: New South Wales, Australia, where the OCS will be the standard screen adopted throughout the care pathway in all of NSW; Policlinico Gemelli IRCCS (a large hospital in Rome); all large hospitals in Belgium (Gent, Leuven, Antwerp, Hasselt) and the Veterans hospital in the USA covering Medicare/Medicaid patients.

In order to estimate the frequency of use by each licencee, a survey was conducted of OCS licensees and training attendees [C(i)], returning responses from a subset of 164 users. The reported use of OCS ranged from occasional and regular use to OCS being the standard screen on the unit, where everyone is screened with OCS (reported by 21 out of 164 survey responses including Oxford University Hospitals and Oxford Health NHS Foundation Trust, Somerset Partnership NHS Foundation Trust, Gloucestershire Health, and Care NHS Foundation Trust Early Supported Discharge). A subset of respondents reported using OCS for specialist screening particularly with patients who have language deficits (21% of respondents). On average, survey responders reported screening 10 patients per month with OCS (responses ranging from 0 to 150). Extrapolating to all licensees suggests that over 100,000 stroke survivors are being screened with OCS every year.

The high frequency of use and familiarity with OCS ‘on the ground’ is further corroborated by an independent study in which the 21 occupational therapists interviewed named both OCS and MoCA as the main tools being used in post stroke cognitive screening [D].

So far, 13 linguistically validated versions of OCS have been published in peer reviewed publications (including Spanish, Russian, Chinese, Brazilian-Portuguese, Italian, Danish, Dutch). These include new normative data and new validation testing against local standard clinical practice. In each case, the Oxford team work closely with the translating teams, providing advice and guidance, and supporting cultural adaptations (e.g. ensuring no meaning is lost, and in adapting stimuli to be appropriate for the country and culture). All translations are done under licence from Oxford University Innovation. A further 22 translations are under development including in German, Greek, Polish, Korean, Arabic, Urdu, Lithuanian and Norwegian. The eagerness of international clinician researchers to translate this tool, along with labour-intensive normative and psychometric data collection demonstrates the perceived gap in domain-specific screening throughout the world, which is being filled by OCS. These adaptations and translations have had widespread uptake: licences have been taken out across the world, including (number of licences in brackets): Australia (56), Belgium (79), Canada (23), China (11), Denmark (34), Hong Kong (7), Ireland (56), New Zealand (17), South Africa (18) and USA (36) [B].

Between 2016 and 2020, 11 in-person training sessions have taken place at local and regional stroke units on request (approx 20 allied health professionals each), and a ‘train the trainer’ event organised by Oxford University Innovation (March 2019). Example training feedback from a session organised by the College of Occupational Therapists, on Specialist Section Neurological Practice (February 2020), attended by 36 specialist OTs, found that 57% would ‘definitely’ change their practice and 43% were ‘quite likely’ to change their practice having learnt about OCS. Larger externally organised events in which OCS training and dissemination was included include two Stroke Association Masterclasses in 2016 (approx 50 clinicians); and clinical audiences at the regional South-West Conference for Stroke in 2018 and the 2019 Welsh Stroke Conference (approx 200 each). A course is offered by the Somerset Partnership as part of the Stroke-Specific Education Framework [E(i)].

Yearly platforms and invited presentations at the UK Stroke Forum have formed the most direct, national dissemination to clinicians. The UK Stroke Forum is a coalition of over 30 organisations all committed to improving stroke care in the UK, underwritten by the Stroke Association and the British Association of Stroke Physicians. In 2019, the Royal College of Occupational Therapists organised two back-to-back training sessions at the UK Stroke Forum (1,200 attendees). This OCS training was attended by over 200 delegates. The most popular and accessible training consists of a demonstration video on administering the OCS has been widely viewed (over 12,000 views on YouTube) [E(ii)].

Testimonial letters, online conversations and survey responses [A, C] attest that the OCS has had impact on clinical decision making, in particular regarding therapy plans, discharge planning (care packages), and in changing approaches to patient and family communications regarding cognition. In addition, a key attribute of the OCS reported by users is the ability to conduct the screen in patients with aphasia and neglect. This was clear from comments such as “I really like using the OCS with patients who have language deficits, as it gives them the opportunity to score well with a cognitive (standardised) test”, and “Allows different aspects of cognition to be assessed in the presence of visual inattention”. [C(ii)]

The survey by OUI of OCS licencees found that of the 74 respondents to the question, “Has the OCS impacted on clinical decision making? e.g. regarding therapy plans, discharge planning (care packages) or in changing approaches to patient and family communications regarding cognition?”, 64 (90%) answered affirmatively (3 no responses) [C(i)]. Examples of free comments include:

“It has supported and provided evidence in MDT meetings, and to support discharge plans. It has supported education of family members, and patients. It has been an integral part of assessment and treatment planning of the more complex cognitive patients. Particularly around praxis, visual perception and executive function. It has been particularly helpful for patients with language deficits.”

“Yes, the OCS has helped change how we understand cognition on the ward. Whereas before cognition was considered an all or nothing entity based on using a cut off score on a dementia screening test, now the staff feel more confident discussing the individual domains of cognition, and integrating this information into their formulations, discharge plans etc.” [C(i)]

The 2016 Royal Colleges of Physicians Guidelines included OCS alongside MoCA as a first-line screen, only one year after OCS was published [Fi]. Though the Montreal Cognitive Assessment (MoCA) is still a much used cognitive screen in stroke patients, due to historic embeddedness, the research on its usefulness in stroke, sensitivity for detecting impairments and inclusivity (e.g. [3]), as well as newly introduced costs related to training qualifications, has led to a steady rise of clinicians switching to the OCS. New guidelines in countries where OCS has recently been translated are also emerging. For example, the Danish Health Authority has published guidelines in which OCS is recommended [F(ii)]. Chapter 4 of the guidelines, ‘Tracing Cognitive Function’, recommends that: “ *Cognitive function in adults with acquired brain damage is initially assessed with the tool Oxford Cognitive Screen (OCS)...*”.[F(ii), translated]. Other independent studies include a systematic review in 2019 [G] which reported that, “ The Oxford Cognitive Screen (OCS), a multidomain tool, was found to be a better predictor of PSCI/PSD than the MOCA or MMSE.”

There are over 13,700,000 new strokes worldwide each year, with a lifetime risk of 20%. Prior to the Oxford research, most guidelines had suggested that patients should be assessed and treated within four weeks of a TIA or minor stroke. After the University of Oxford team’s research showed that the early risk of stroke after a TIA had been substantially underestimated, guidelines rapidly changed, markedly increasing the urgency with which assessment, investigation and treatment of minor stroke and TIA is recommended. Prior to the start of this REF period, this recommendation had been adopted in the 2007 Department of Health National Stroke Strategy (NSS), the NICE guidelines for stroke (2008), and by the European Stroke Association (2008) [A(i)], US National Stroke Association (2011) [A(ii)] and UK Royal College of Physicians (2012). Similarly the identification of urgent intervention in prevention of stroke [3] had been included in the 2007 NSS and 2008 NICE guidance. Both research and impact have broadened since the REF2014 case study that presented those outcomes.

During the current REF period, the updated versions of the Royal College of Physicians guideline (2016) [B, section 2.3] and the NICE guideline (2019) [C(i)] reaffirmed this recommendation. In coming to recommendations about rapid recognition and treatment of TIAs, the detailed version of the revised NICE guideline draws heavily on Rothwell’s work (e.g. [6] showing benefits of urgent treatment with aspirin [Cii]), and rates it as high-quality evidence. In addition, the 2014 US Guideline on the prevention of stroke after TIA [D] recommends early carotid surgery on the basis of the Oxford research in [3]. The other prior guidelines [A] remain active.

Repeated UK national audits showing improvements between 2006 and 2014 in provision of acute services for TIA and minor stroke and in urgent carotid imaging were sustained in the current REF period. The median delay from TIA or minor stroke to assessment in a specialist clinic in the UK had fallen from 12 days in 2006 to 2 days in 2012. The 2014 and 2016 audits showed this maintained at 2 days [E]. Rolling the service out UK-wide was shown to prevent about 10,000 strokes per year [5], thus approximately 70,000 over this REF period, continuing to save the NHS up to GBP200,000,000 annually in acute care costs alone. Equivalent benefits in stroke prevention and reduced healthcare costs are considered to have resulted from recommendations outside the UK.

The improved guidelines on management of TIA and minor stroke has also driven further improvements in patient care during the current period. For example, NHS service provision has continued to improve, with the proportion of hospitals that provide a 7-days a week service able to see, investigate and treat ‘high-risk’ patients on the day of referral increasing: for inpatients from 60% in 2014 to 71% in 2016 and similarly 45% increasing to 52% for outpatients [E].

Rothwell’s findings on carotid endarterectomy [3] led to guidelines that stipulated it should be performed within 14 days of a TIA/stroke, with continued inclusion in the 2016 Royal College of Physicians guideline [B, section 5.3] and the 2019 NICE guideline [C(i)]. The ongoing UK National Vascular Registry showed that median delay to carotid surgery (approx. 5,000 operations per year) fell from over 3 months in the early 2000s to 12 days in 2018, with the upper quartile cut falling from 28 days in 2015 (2016 report) to 23 days in 2018 (2019 report) [F]. The UK national stroke audit has also shown that the proportion of hospitals that are able to provide same-day carotid imaging 7-days a week for ‘high-risk’ patients has continued to increase: from 42% in 2014 to 50% in 2016 [E].

Rothwell’s finding of the substantial reduction in early risk and severity of major stroke after TIA and minor stroke with aspirin alone [5] led subsequent UK guidelines to recommend immediate use of aspirin even prior to specialist assessment/investigation (Royal College of Physicians 2016 [B, section 3.1]; NICE 2019 [C(i)]). In particular, the Evidence Review for the 2019 NICE Guideline [C(ii)] recommended that aspirin is offered immediately for suspected TIA, requiring that first-line healthcare professionals (e.g. paramedics, NHS 111, GPs, nurses, ED physicians) advise aspirin, and that general practices have adequate supplies to enable treatment, and cited [3] and [5] as two of its three key papers. The 2017 Australian guidelines made a ‘Strong recommendation’ for the use of antiplatelet therapy in all people with TIA or ischaemic stroke, citing [6]. [G, Chapters 3 & 4].

The recommendation for immediate use of aspirin prior to specialist assessment/ investigation is also now supported in online advice to patients and the public, including the NHS website [H]. which states in a section encouraging prompt action to seek medical advice, ‘ If a TIA is suspected, you should be offered aspirin to take straight away. This helps to prevent a stroke.’ This change was called for by Rothwell and colleagues in their 2016 paper [6] and was shown to be necessary by their further work on continuing delays in patients seeking medical attention. As immediate aspirin reduces the risk of disabling stroke by 80% [6], any increase in aspirin self-medication has benefits by reducing early recurrent stroke.

Use of the Oxford Team’s ABCD2 system [2] to triage patients at high risk of stroke after TIA wass recommended in international guidelines [A,D,G,I]. Use of the score in the UK was recommended until 2019 whilst clinical services were under-resourced and needed to be expanded to meet demand. It was used as a Key Performance Indicator (KPI) to identify ‘high risk’ patients in the 2014 and 2016 UK national stroke audits [E].

The ABCD2 score also underpins the eligibility criteria for TIA (usually ABCD score >4 or 5) in all recent major international randomised trials of acute antiplatelet treatment (CHANCE trial NEJM 2013; SOCRATES trial NEJM 2016; POINT trial – NEJM 2018; THALES trial; NEJM 2020) and hence for treatment recommendations based on the results of these trials. Use of the ABCD2 score to identify high-risk TIA patients for dual antiplatelet treatment is recommended in the most recent Australian [G] and Chinese guidelines [I] and was used by the European Stroke Organization (ESO) in the formulation in 2020 of its updated guidelines for management of TIAs [J]. The latter cites [1,4,6] and was submitted in October 2020 for public release in 2021.