Background

Increasing physical activity (PA) is a global public health priority.1 For decades, exercise referral schemes (ERS) have been a popular way for healthcare professionals in primary and secondary care to help patients increase their PA.

Delivery of ERS varies widely, with the construction of an evidence base informing ‘what works best’ limited by a lack of understanding about what individual schemes deliver and how.2 Between-scheme analyses are extremely challenging due to varying quality of reporting (eg, of scheme delivery components and processes) and evaluations.3 4 As a consequence, overviews of ERS evidence5 6 are flawed by combining heterogeneous interventions (eg, falls prevention via physiotherapist referral and hypertension management via general practitioner (GP) referral) and datasets.

Collectively, the underwhelming findings of such overviews lead to concerns over commissioning ERS, and the inability of national policy and best practice guidelines to recommend a ‘gold standard’ structure, or even comment on ‘what good looks like’.5 We do not know whether local tailoring of ERS is more effective and efficient than a standardised approach.

To advance knowledge and practice about ERS, we therefore propose a universal classification taxonomy, grounded in practice-based experience and theory. We believe that this will help appropriately identify meaningfully different ERS classifications, leading to improvements in the interpretation and understanding of the evidence base for policy-makers and practitioners.

Is the term ERS outdated?

As traditionally defined, ERS contain four essential components: (i) an assessment involving a healthcare professional to determine that someone who has a health condition or other factors that put them at risk of ill-health is sedentary or inactive, (ii) referral by this professional to a PA specialist or service, (iii) a personal needs assessment by the specialist or service and (iv) an opportunity to participate in a PA programme.2 7 8

This definition now fails to represent a myriad of innovations in both evidence-informed models and contemporary practices, which support PA uptake. We suggest that - ‘PA referral scheme’ more appropriately describes the range of interventions offered. Specific examples include entry routes via self-referral or from other professionals (eg, health trainers) and group-based needs assessments. Our taxonomy therefore encompasses all PA schemes that: (1) have the primary aim of increasing PA; (2) have a formalised referral process; and (3) are provided for individuals who are inactive/sedentary, and/or have or are at risk of a health condition.

These inclusion criteria enable us to usefully classify and compare traditional ERS alongside rapidly emerging innovations as described above. We acknowledge that inactive but otherwise healthy individuals attend schemes, despite current recommendations that ERS are only for those with, or at risk of, health conditions.2 We exclude therapeutic ERS provided by health practitioners in a clinical environment (eg, physiotherapy-based rehabilitation in hospitals), general signposting to PA opportunities or -interventions where increasing PA is not the direct service aim.

Thus, the first distinction the taxonomy requires the user to make is between ‘traditional ERS’ (think classic assessment-based referral from a GP to a supervised gym session) and ‘non-traditional PA referral’ (think new trends for social prescribing, self-referral and digital interventions).

The PA referral scheme taxonomy

Our proposed taxonomy operates at three levels:

Level 1: classification

This high-level classification allows for the identification of scheme subcategories for study and comparison (figure 1). It details whether a scheme is traditional or non-traditional, who the provider is, whom it is for and activities offered.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

PA referral scheme classification framework. ACS, acute coronary syndrome; COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; HF, heart failure; OA, osteoarthritis.

How to use the PA referral scheme taxonomy and next steps

We propose using the taxonomy as a reporting checklist in the practice-based and academic literature, a classification system for evidence reviews of delivery and effectiveness, and an audit and monitoring tool for commissioners and providers to capture service delivery.

In terms of the next steps, we invite comments, critique and engagement from the policy, practice and academic sectors. To this end, we are delivering practitioner and expert consensus events for late 2019. We are presenting this as an idea, not the finished product, and are keen to seek consensus on what factors are appropriate, what needs changing and what needs adding. We believe that an agreed framework will benefit the implementation of PA referral schemes internationally and ultimately benefit population health. The next step is to test the utility of the taxonomy to meaningfully classify reach, uptake and efficacy (or effectiveness) of the different scheme types.