This site is intended for health professionals only

How we cut direct referrals for back pain to nothing


The problem

It is well recognised that spinal pain, with or without nerve root irritation, is one of the most common conditions encountered in primary care. Most people recover spontaneously by six weeks. However, a significant proportion (3-10% depending on the study) develop chronic problematic long-term pain. The main risk factors for developing chronic back pain are psychosocial factors, such as fear or avoidance of movement, depression and a ‘catastrophising attitude' towards the pain. Back pain accounts for significant costs for both healthcare and society – estimated to be around £12.3bn a year in the UK.1

As GPSIs in musculoskeletal medicine, we were asked by clinical commissioners locally – then Central Sheffield Consortium, now Sheffield GP Commissioning Consortia – to look at this problem. Access and provision of service was inequitable across the city for this condition, with variable waiting times and quality of physiotherapy. We were also influenced by a musculoskeletal service framework released in 2006,2 and the Arthritis and Musculoskeletal Alliance document, Standards of care for back pain.3

It was clear from the evidence base that early intervention for musculoskeletal conditions could prevent the progression to persistent problematic pain. We felt applying best practice could significantly improve the quality of patient care, and reduce progression to chronic back pain.

Figures for referrals to secondary care surgical spinal teams before our initiative were around 100 a month with a conversion to surgery rate of 20%. It was felt we could achieve savings in this area, as spinal surgery is only indicated for persisting nerve root compression.

What we did

The initial work between 2008 and 2009 came down to two main approaches.

First was the production of a local pathway. We convened a stakeholder group from primary care, physiotherapy and secondary care, and we drew up a patient-centred pathway that took into account best evidence and local consensus. We did this with a mix of a couple of large group workshops, with smaller email and face-to-face groups to debate areas of contention.

We made a few key changes:

• We created a specification for a physiotherapy-led spinal team that would act as the first point of referral from primary care. These would be led by a specialist physiotherapist in spinal pain, and allow access to a variety of skills including acupuncture, manipulation and pain management approaches. We requested that all non-urgent spinal referrals pass through these teams, as opposed to direct referral to hospital surgical spinal teams.

• We specified standard maximum timescales for patients to be seen once they are referred – one week for severe pain, two weeks for standard pain levels.

• We recommended that GPs use the STarT back screening tool to help stratify risk of chronicity (an evidence-based, simple nine-question tool designed for primary care).

• We organised access to MRI for these spinal teams.

• We specified that all cases being considered for surgical intervention (with nerve root pain) be discussed at a surgical multidisciplinary team meeting where case history and MRI are considered by a team including a surgeon, physiotherapist and pain specialist.

We then went on to set up a website, The process of developing the site created a common goal for the group, and helped to focus our advice on written and video formats for patient information leaflets, assessment tools and education materials for professionals and patients.

We launched the new website and pathway as part of a learning event for GPs on musculoskeletal issues. We printed flyers, posters and cards for practices, and followed up with letter and email updates to remind people about use of the pathway.

In 2011, as a result of local audit and the report of the STarT trial,4 we made some significant updates to the pathway and referral processes. We realised that despite clear approval of the pathway by primary care, secondary care was getting the same number of referrals. In two out of three cases for all referrals to any specialty, GPs were not highlighting psychosocial risk factors for chronic pain and made very little use of the STarT back tool. The physiotherapy team also requested we include neck pain.

We amended the pathway so that all non-urgent referrals passed through spinal physiotherapy teams, and all referrals included a STarT score. Physiotherapy teams would only accept patients with a medium or high risk on the STarT score, unless problematic pain persisted beyond six weeks despite a low score.

The secondary care and physiotherapy-led spinal teams agreed to bounce back referrals that fell outside the agreed pathways and also agreed to stratify the care that patients received according to the risk score on the STarT tool (high-risk patients receiving ‘psychologically informed physiotherapy' that addressed barriers and fears).

We designed a concise referral form, with mandatory sections in bold, to facilitate collection of our minimum data set.

Lessons learned

Despite achieving good buy-in from GPs, we learned that in this case it required mandated use of the STarT tool to ensure all patients had a bio-psychosocial assessment communicated in referral information. We also learned we needed misdirected referrals bounced back to achieve full compliance with agreed pathways. However, a certain level of trust is required before the local health community accepts such directive strategies.

The physiotherapy teams have concerns that only 70% of all referrals bounced back to referrers are returned. However, this may reflect the natural history of spinal pain, where many cases settle spontaneously.


Our pathways now ensure that all patients (>95%) are referred having received a bio-psychosocial assessment. This allows easier triage to appropriate treatment from first physiotherapy appointment. It also allows GPs to reassure patients unlikely to need further treatment. Patients have reported high levels of satisfaction at this approach, and physiotherapy teams find it efficient.

Physiotherapists report a skew to more complex cases and a trend towards reduced overall referrals (although six months in it may be too early to tell). This suggests using the STarT tool is helping to select patients with higher need, and ensuring that those who are likely to recover spontaneously remain managed in primary care.

Direct secondary care referrals have reduced to zero – saving on average 80 appointments a month. Using the multidisciplinary team approach generally halves the number of patients being referred to surgical teams from physiotherapy services, and has increased surgical conversion rates from 20% to 50%. It has also allowed faster access to surgical treatment for patients in severe pain. It will require re-audit of the whole system to assess the change in numbers of referred patients, but initial estimates suggest this approach has at least halved the number of referrals to surgical outpatients – saving more than 40 new appointments a month.

The website has attracted significant use – averaging 100 visitors a day with more than 5,000 regular users (patients who have visited more than five times). The PCT has funded expansion of the site to include problems with the hip and knee, foot and ankle, hand and elbow, and shoulder.

The future

The next steps are to work with local academic institutions to study the effects of self-care websites on patient outcomes. While we have anecdotal evidence of patient benefit, it would be important to quantify the value of such approaches, so as to assess their importance in a commissioned model of care. Research suggests these benefits to patients, of website information and tools, may be enhanced by peer support while accessing sites. We intend to include research questions that assess this.

Dr Ollie Hart and Dr Andy Hilton are both GPSIs in musculoskeletal medicine in Sheffield


1 Dagenais S, Caro J and Haldeman S. A systematic review of low back pain cost of illness studies in the US and internationally. Spine 2008;8:8-20

2 Department of Health. The Musculoskeletal Services Framework, 2006.

3 Arthritis and Musculoskeletal Alliance. Standards of Care for People with Back Pain 2004.

4 Hill JC, Whitehurst D, Lewis M et al. Comparison of stratified primary care management for low back pain with current best practice (STarT Back): a randomised controlled trial. Lancet 2011;378:1560-71.


Visit Pulse Reference for details on 140 symptoms, including easily searchable symptoms and categories, offering you a free platform to check symptoms and receive potential diagnoses during consultations.