Physical activity: exercise referral schemes

4.1 The PHAC noted that many of those involved in commissioning, developing and delivering exercise referral schemes believe they are an effective use of public money. This is evident in the number of schemes, the popularity of referrals and anecdotal reports of an increase in physical activity levels and other health benefits among participants. However, the economic analyses demonstrated that these schemes are less cost effective than giving brief advice, as recommended in NICE's guideline on physical activity: brief advice for adults in primary care. That is because they have a very small additional effect and are relatively expensive. See sections 4.11 to 4.20 for further details.

4.2 The PHAC acknowledged that people can be grouped together in a variety of ways. These different categories helped clarify exactly who would benefit. People were grouped as follows:

4.3 The PHAC acknowledged that a number of different types of exercise referral schemes have been set up in the UK since the publication of NICE's guideline on four commonly used methods to increase physical activity. There was insufficient evidence to assess the relative cost effectiveness of the different types of schemes. Overall, the new evidence identified does not support exercise referral schemes as a cost-effective means of improving health by increasing levels of physical activity. (See section 4.8 for other benefits of such schemes.)

4.4 The PHAC noted that a number of factors may influence effectiveness. These include: the intensity, length and frequency of the exercise referral scheme; and the experience, skills and knowledge of people who provide or deliver it. However, the evidence on the specific impact of these factors was very limited.

4.5 The PHAC noted that the overall aim of exercise referral schemes is to improve health and that an increase in physical activity is not always the primary outcome. Other outcomes, including an increased sense of belonging and social interaction ('social capital') may be important, but these have not been measured in most studies and were not specifically considered here. Members agreed that these outcomes were potentially important in their own right. But they also agreed that it should be made explicit if they are the primary goal for such a scheme.

4.6 The PHAC was disappointed at the relatively small number of studies identified for this update. Members discussed the details of each study including: participants referred; reasons for referral and exclusion – including their risk factors for ill health and existing diseases or other health conditions; and types of exercise referral intervention (see the evidence).

4.7 The PHAC noted that exercise referral schemes, compared with brief advice, resulted in a 1.08 relative risk of participants meeting the Chief Medical Officers' (CMOs') recommended level of physical activity. So, if 36 people participate in an exercise referral scheme only one of them will achieve the recommended levels of physical activity.

4.8 Members noted that getting people who are sedentary or inactive to be more physically active will lead to health benefits, even if they do not meet the CMOs' recommended levels. They also noted that this is not always captured in the evidence base.

4.9 Members noted that the evidence on the medium- or long-term health benefits associated with exercise referral schemes was very limited.

4.10 Members noted that data collected via self-reporting methods may overestimate how physically active each participant has been compared with more objective measures of physical activity.

4.11 The PHAC noted that exercise referral schemes are only marginally more effective than brief advice and lead to a very small additional gain in quality-adjusted life years (QALYs). However, there were considerable uncertainties about the correct parameters to use for the economic modelling and members noted that the model does not capture all the potential benefits.

4.12 Using the base case assumptions, the incremental intervention cost of £217 led to an incremental cost-effectiveness ratio (ICER) between £72,748 and £113,931 per QALY gained. Even in the best case scenario, the estimated incremental cost effectiveness ratio was £31,009 per QALY gained. NICE normally considers that any interventions over a threshold of £20,000 to £30,000 per QALY are not cost effective. However, because current evidence to inform the assumptions in the model was insufficient, members did not feel they could recommend disinvestment in such schemes. Further, some schemes may be cost effective, or may only be cost effective for some subgroups. Again, however, there was insufficient evidence to make recommendations on this, hence data collection has been made a condition for any exercise referral scheme that is commissioned.

4.13 The PHAC noted that if exercise referral schemes collected more detailed data commissioners would be able to make a more informed decision on future investment. Such a decision would take into account the prevailing local priorities, the nature of the schemes, evidence of effectiveness and the primary aim of the scheme (such as social engagement).

4.14 The PHAC noted that set up costs have not been considered in the economic model and that their inclusion would increase the incremental cost-effectiveness ratio.

4.15 If the relative risk of exercise referral schemes (compared with usual practice) is 1.08, even in the best case scenario schemes costing over £150 per participant would not be cost-effective at a threshold of £20,000 per QALY. However, there was no evidence on how a reduction in costs would affect effectiveness so it was not possible to recommend a cap on the cost of such schemes.

4.16 The PHAC noted that the full cost to participants (including travel and childcare costs) was not considered in the economic model.

4.17 The PHAC noted that any increase in physical activity is associated with positive health benefits. But unless people achieved the CMOs' recommended levels of activity, these benefits were not captured in the economic modelling. This means that the true gains from exercise referral schemes are likely to be underestimated by the model. However, the economic model used is comparable to that used to assess the cost effectiveness of brief advice to increase physical activity. The latter is often used as the comparator in many of the included studies. So the finding that exercise referral schemes cost considerably more per QALYs than brief advice is likely to be valid.

4.18 The PHAC discussed the importance of additional, health-related quality of life gains and the 'feel good' factor (process utility) gained from being physically active. Both feature as inputs of the model. However, there was uncertainty around the magnitude of the process utility and how long it would last. This meant that the PHAC was unable to agree or disagree on this key assumption in the cost effectiveness model. This added to the uncertainty about estimates of cost effectiveness.

4.19 The PHAC was aware that the economic modelling to determine the long-term health benefits of exercise referral schemes was based on cohort studies limited to coronary heart disease, stroke and type 2 diabetes. The PHAC noted that the other benefits of physical activity are not captured by the model (for example, alleviation of mental health problems, musculoskeletal conditions and some cancers). Taking these into account could lower the ICER, but the magnitude of this effect was unclear.

4.20 The PHAC noted that the economic model over-simplifies the clinical situation. That is because it does not allow for someone having more than 1 of the 3 health conditions (coronary heart disease, stroke or type 2 diabetes). Members also noted that the model does not consider that having one 'comorbidity' may affect the likelihood of experiencing another. These limitations mean that the cost effectiveness of exercise referral schemes may be underestimated. But the comparison with brief physical activity advice (usual care) is still valid as the same limitations apply.

4.21 The PHAC noted that some exercise referral schemes may or may not be more effective and cost effective than others. Some approaches may be cheaper to deliver (see section 4.22). Others may be more effective for specific subgroups (see section 4.23). Some are better at helping to maintain physical activity levels after the scheme ends (see section 4.24).

4.22 The PHAC noted that cost effectiveness and effectiveness varied according to the type of exercise referral scheme. There was a feeling that self-directed and less resource-intensive activities (such as walking and cycling) may be more acceptable than gym-based activities. Therefore, they may be more effective and more cost effective for certain subgroups. However, no review evidence was identified to verify this assumption.

4.23 The PHAC felt that exercise referral schemes may be cost effective in encouraging physical activity among specific groups. For example, it may help people with multiple disease risk factors such as hypertension, obesity or poor mental health, or those who would not otherwise have access to supervised exercise programmes. Members also noted that the people who appear to benefit most from these programmes may gain similar benefits from brief physical activity advice. However, because of a lack of evidence the PHAC was unable to make specific recommendations on the best ways to increase physical activity among specific groups.

4.24 The PHAC discussed the importance of increasing adherence throughout the duration of an exercise referral scheme and participating in physical activity beyond the end of the programme. For example, it agreed that helping participants to develop the skills they need to be physically active on their own, or providing social support during the intervention, might encourage adherence to the scheme. In turn, this might increase the chances of participants being physically active in the longer term.

4.25 The PHAC noted that poor referral practices affect the overall effectiveness of schemes. This could be due to the initial assessment or the type of activity someone has been referred to. Because the participant may not be interested in a particular type of activity, or may not be able to complete it because of their current fitness level. Or schemes may be less effective because they do not fully take participants' motivation and ability into account. Members noted that better use of triage or a 'stepped approach' that includes brief physical activity advice (see NICE's guideline on Physical activity: brief advice for adults in primary care) may overcome these problems. However, no evidence was identified to substantiate this assumption.

4.26 Stakeholder and fieldwork feedback highlighted that there is a lack of consistent and appropriately delivered brief advice on physical activity in primary care. Members were concerned that removing exercise referral schemes for specific populations and in certain locations, combined with these problems, may reduce the priority given to physical activity in primary care. The impact of this guideline on the wider physical activity agenda was also discussed.

4.27 The PHAC noted that a lack of focus on relapse prevention and sustainability negatively impacts on effectiveness. Alongside improving referral practices (see section 4.25), members discussed the need for improved follow-up to identify why people drop out and how this might inform development of future schemes. Members also discussed the importance of following up participants who have completed a scheme and supporting them to continue to increase or maintain their activity.

4.28 The PHAC considered how staff training affects the effectiveness of exercise referral schemes. Members noted the training outlined in the Department of Health's National Quality Assurance Framework and British Heart Foundation exercise referral tool kit. Members also noted that this training could help alleviate concerns about possible litigation issues. The latter was highlighted as a significant barrier to referral in evidence review 2: the factors that influence referral to, attendance at and successful completion of exercise schemes and longer term participation in physical activity undertaken for this guideline.

4.29 The PHAC noted that the range of physical activities provided is a key factor in whether or not someone adheres to a scheme. Those offering alternatives to gym-based activities, that are less expensive and give a degree of personal choice, seem to improve adherence.