Article Text

Physiotherapy exercise rehabilitation with tailored exercise adherence support for people with osteoporosis and vertebral fractures: protocol for a randomised controlled trial – the OsteoPorosis Tailored exercise adherence INtervention (OPTIN) study
  1. Karen L Barker1,
  2. Jonathan Room2,
  3. Ruth Knight1,
  4. Erin Hannink2,
  5. Meredith Newman2
  1. 1Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK
  2. 2Physiotherapy Research Unit, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
  1. Correspondence to Professor Karen L Barker; Karen.Barker{at}


Introduction Vertebral fragility fractures affect at least 20% of the older population in the UK. Best practice guidelines recommend the use of exercise to slow the rate of bone loss, to maintain muscle strength and physical function, and to prevent falls and further fractures. However, treatment effects are often small and difficult to sustain and adherence, or the extent to which patients engage in treatment, has been identified as an important issue by many studies. Our hypothesis is that integrating adherence intervention strategies with an exercise intervention will be beneficial. We will compare physiotherapy exercise rehabilitation with adherence support versus physiotherapy exercise rehabilitation alone in terms of effects on (A) physical function, quality of life and fear of falling and (B) exercise self-efficacy and adherence.

Methods and analysis A multicentre, two-arm, parallel group, superiority randomised controlled trial with blinded assessments at baseline (0) and 4, 8 and 12 months, with a nested qualitative study and health economic analysis. 116 participants will be allocated to either (1) outpatient physiotherapy which will include a musculoskeletal assessment and treatment including balance, posture, strength training and low impact weight-bearing exercises over 16 weeks or (2) OsteoPorosis Tailored exercise adherence INtervention intervention. This includes standard physiotherapy as above plus an additional, integrated assessment interview (30 min) and 60 min of adherence support spread over the subsequent 16 weeks.

Ethics and dissemination The study protocol was approved by West of Scotland Research Ethics Committee 4 (21/WS/0071). Trial registration number ISRCTN 14465704. The paper is based on Protocol V.4.

Trial registration number ISRCTN 14465704.

  • Back pain

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • The physiotherapy exercise rehabilitation with tailored exercise adherence support for people with osteoporosis and vertebral fractures (OsteoPorosis Tailored exercise adherence INtervention) study is a multicentre randomised controlled trial with an embedded qualitative study and economic evaluation.

  • It will recruit from at least six National Health Service hospitals.

  • The intervention addresses adherence which is an important confounder in many trials of physiotherapy.

  • The intervention was developed using current research evidence, input from expert clinicians, researchers and patient/public representatives.

  • Due to the nature of the interventions, the physiotherapists delivering the treatments and the participants cannot be blinded.


Vertebral fragility fractures (VFFs) affect at least 20% of the older population in the UK and present a significant health and economic burden.1 2 They are associated with back pain, fatigue, low mood, restrictions in physical function and activities of daily living and marked, persistent reductions in quality of life (QoL).1 2 Without treatment, progression and functional decline are expected. Conservative treatment for osteoporosis includes bone protective medications and lifestyle adaptations. Guidelines recommend people with osteoporosis keep active and exercise to slow the rate of bone loss, to maintain muscle strength and physical function and to prevent falls and further fractures.3 4 Exercise prescription with multicomponent exercise programmes that include postural, balance, aerobic weight-bearing and strength exercises are recommended.1–3

Trials evaluating exercise in people with VFFs have reported benefits across a range of outcomes,1 2 5–10 with a recent Cochrane review concluding there is moderate-quality evidence that exercise improves physical function.1 However, treatment effects are often small and difficult to sustain.1 2 5–10 For example, in the PROVE trial, significant, clinically relevant benefits to back muscle endurance, balance, walking capacity and physical function following physiotherapy exercise at 4 months postrandomisation did not persist at 12 months.2 Adherence, or the extent to which patients engage in treatment, has been identified as an important issue by many studies.1 2 5 6 9 Partial adherence or non-adherence is associated with worse outcomes and conversely, higher adherence with better outcomes.1 2 5–10

Multiple factors affect exercise adherence in older people with chronic health conditions including osteoporosis.11 12 These include low exercise self-efficacy, low motivation, depression, insufficient exercise knowledge or skill, physical ability, negative views about treatment and exercise programme design.11 12 Considering this complexity, interventions to support adherence that recognise personal barriers and facilitators to exercise and that can draw on multiple adherence techniques are recommended.12 13

Behavioural approaches can include interventions that support exercise through providing additional monitoring, interventions that aim to alter thinking patterns that contribute to non-adherence and ones that strengthen behaviours that support adherence.12 Motivational interviewing is a collaborative process that explores potential ambivalence, obstacles and facilitators surrounding behaviour change.6 There is evidence that additional monitoring, prompts and feedback can benefit adherence in older adults, for example, via telephone call/ text messages, wearable activity monitors or by enriching environmental cues.1 Incorporating exercise into everyday routines can make it easier to initiate and sustain and creating ‘Exercise Action Plans’ that specify when, where and how exercises are undertaken can support this process.1 13 Using an intervention mapping approach, we developed an exercise adherence intervention underpinned theoretically by the Capability, Opportunity, Motivation-Behaviour (COM-B) behaviour change model.14 15


The aims of this study are:

  • To compare physiotherapy exercise rehabilitation with adherence support with physiotherapy exercise rehabilitation alone in terms of effects on: (A) physical function, QoL and fear of falling and (B) exercise self-efficacy and adherence.

  • To explore patient and physiotherapist views of the intervention and of adhering to exercise.

  • To understand if physiotherapy exercise rehabilitation with adherence support is cost-effective.

Methods and analyses

Study design

A multicentre, two-arm, parallel group, superiority randomised controlled trial (RCT) with blinded assessments at baseline (0) and 4, 8 and 12 months following randomisation, with a nested qualitative study and health economic analysis. Participants will be allocated to either (1) outpatient physiotherapy which will include a musculoskeletal assessment and six treatment sessions over 16 weeks based on the current best practice guidance from the Royal Osteoporosis Society or (2) the OsteoPorosis Tailored exercise adherence INtervention (Opt-In) intervention. This includes outpatient physiotherapy as described above, plus an additional, integrated assessment interview (30 min) and 60 min of adherence support spread over the subsequent treatment period of 16 weeks as prescribed by the physiotherapist in collaboration with the participant. Sessions in both arms can be in-person or virtually via video-call/telephone as agreed between participant and therapist (figure 1).

The trial started recruitment of patients in August 2021 and will continue recruiting until June 2023.


At least six National Health Service (NHS) hospitals and their related physiotherapy services.

Study participants

Adults aged 55 years or over who have a diagnosis of at least one previous osteoporotic vertebral fracture and back pain.


Inclusion criteria

Participants may enter the study if they meet ALL the following criteria:

  • Men and women ≥55 years: all women must be at least 1-year postmenopausal.

  • One or more VFFs confirmed by radiography, X-ray, MRI, CT or DEXA scan, people with VFF of any severity and at any time point postfracture are eligible.

  • They must have had an episode of back pain in the previous 12 months.

  • All must be able to walk at least 10 metres independently with or without a walking aid.

Exclusion criteria

Participants may not enter the study if any of the following apply:

  • Current conditions that would make participating in physiotherapy or exercise unsafe or confound results. This includes those with significant neurological and psychiatric conditions, severe unstable cardiovascular or pulmonary disease.

  • Bone loss secondary to other metabolic disorders, diseases or medication for example, rheumatoid arthritis, anorexia, cancer, coeliac disease, steroid use.

  • Individuals whose primary problem is back pain that involves pain radiating into the lower limbs.

  • Vertebroplasty, facet joint injection or physiotherapy within past 12 weeks.


A member of the patient’s direct care team will identify potential participants with VFFs via clinic lists and electronic medical records from relevant metabolic bone clinics, radiology clinics (DEXA), physiotherapy referral lists and from rheumatology clinics.

Screening and eligibility assessment

Potential participants, who respond to an invitation letter will be contacted by telephone to discuss the study further, to check eligibility and to answer any questions. Patients who do not meet the eligibility criteria or who do not wish to participate will receive standard NHS treatment. We will record the age and gender of these patients to assess the generalisability of those recruited.


Participants who are eligible and willing to proceed will be approached for informed consent; they and the researcher will sign and date a consent form. For participants who are recruited to the additional nested qualitative study and interviewed on-line or via telephone, informed consent will be obtained verbally before the interview. The researcher taking consent will read, and fill out, the consent form on behalf of the participant and then sign the form (online supplemental file).


Consented participants will be randomised 1:1 using a computer-generated randomisation schedule prepared by the trial statistician (RK). Individual randomisation will be stratified by recruitment centre and permuted blocks of varying undisclosed sizes will be used. The randomisation schedule will be concealed in sequentially numbered, opaque, sealed envelopes for each site. A study administrator who has no interaction with blinded study staff will manage these envelopes. The administrator will open the randomisation envelope, and then communicate with the local site who will make the participant aware of their allocated group and refer for physiotherapy; making sure that participants are allocated to physiotherapists delivering the treatment for their allocated arm.


Physiotherapists delivering the interventions and participants will be told the treatment allocation. Initial baseline assessment will occur prior to randomisation and the researcher undertaking assessments will not be involved in any part of the randomisation procedure to ensure that they are not able to bias the group allocation. The researcher conducting follow-up measures and the research team personnel entering data will also not be informed of allocated group and participants will be asked and reminded not to disclose their treatment group to the researcher at follow-up appointments.


Training and monitoring

Sessions in both arms can be delivered in-person or virtually via video-call/telephone as agreed between participant and therapist; to allow flexibility and resilience as COVID-19 restrictions on physical attendance vary. Treatments were standardised and manualised and the study team provided training in the multicomponent exercise therapy treatments (delivered to all participants) to all treating physiotherapists. Training in the behavioural adherence support intervention which comprised assessment and a set of nine behavioural interventions (the Opt-In toolkit) was delivered separately to the therapists in the Opt-In arm. Treating physiotherapists will record the delivery and content of each treatment session in adherence logs for each participant. Regular site visits will be carried out to monitor intervention fidelity.

Standard care

Participants will be offered a 1-hour physiotherapy assessment and six individual outpatient physiotherapy sessions spread over 16 weeks.2 The physiotherapy will include a musculoskeletal assessment and treatment including a multicomponent, progressed balance, posture, strength training and low impact weight-bearing exercise.3 Exercise intensity will be assessed using the 10-point Rating of Perceived Exertion scale (CR10-RPE), so participants work at a moderately hard to hard (RPE 4–6) intensity. Although current practice may vary, the package agreed as the standard care is based on consensus, best practice guidelines and successful delivery in the PROVE trial exercise arm2 and aims to be a credible representation of current best practice treatment across the NHS. Treating therapists will receive prior training on prescription of the exercises.2

OsteoPorosis Tailored exercise adherence INtervention

Participants allocated to Opt-In will receive the standard package described above, plus an additional, integrated assessment interview (30 min) and 60 min of adherence support spread over the subsequent treatment period of 16 weeks in an individualised pattern as required by the participant.

Participants in the Opt-In arm will complete the Personalised Exercise Questionnaire (PEQ). The PEQ was developed in Canada to support patient-centred exercise prescription for people with osteoporosis and covers topics such as barriers to exercise and goals of treatment.16 Treating physiotherapists will have a collaborative discussion with the participant using a motivational interviewing approach drawing on PEQ responses and considering goals, motivators, facilitators and barriers surrounding exercise. It aims to provide physiotherapists with a deeper understanding of patient motivations and circumstances, to strengthen the therapeutic alliance and the patient’s own motivations for adopting exercise.12 Using their assessment findings, the questionnaire and collaborative interview the physiotherapist will assess a participant’s exercise capability (C), opportunity (O) and motivation (M) to carry out exercise behaviour (B) (COM-B) and select an adherence technique from the Opt-In toolkit in response.14 Techniques can include education about osteoporosis and exercises, education about and practice of fall prevention strategies, Exercise Action Plans, a contact telephone call and self-monitoring and feedback strategies such as, exercise confidence rating scales or using an exercise diary. Techniques are linked to COM-B domains to facilitate physiotherapist decision-making for example, Education improves capability and motivation (C, M) and a diagrammatic decision aid was developed to facilitate rapid decision-making during treatment. Each Opt-In arm treating physiotherapist received a toolkit and training by the study team about techniques and how to use them. Physiotherapists were asked to prescribe at least three adherence techniques from the Opt-In toolkit over 16 weeks but could use more. The exact techniques selected were personalised to the patient as was the pattern and spread of the 60 min adherence support time. Participants in the intervention arm were given a folder that included their exercises and selected adherence materials, for example, exercise diary, education leaflet, action plan record. Figure 2 summarises the intervention in a logic model.

Figure 2

Logic model for intervention. NHS, National Health Service.

Concomitant care

Other aspects of health and social care will continue as usual. Analgesia and other medication use will be collected by self-report diary. Additional treatments sourced outside of the trial including contact with general practitioners (GPs) and other healthcare professionals will be recorded in self-report health utilisation diaries in which participants will asked to record their use of health and social care services across the study, for example, GP, nurse, other physio, hospital admissions, home carer visits in standardised study diaries. Diaries will be from 0 to 4, 4 to 8 and 8 to 12 months.17

Outcome measures

The primary outcome measure will be the Timed Up and Go (TUG) at 12 months. The TUG is a test of balance, lower limb strength and walking ability with established reliability and validity. It records the time a person takes to stand up from a chair, walk 3 m at a self-selected speed, turn, walk back and sit down.18

Secondary outcome measures are:

  • QUALEFFO 41: a disease-specific measure of health-related QoL applicable to patients with osteoporosis and vertebral fractures. It is a self-administered questionnaire that provides scores on five domains: pain, physical function, social function, general health perception, mental performance and a total score.19

  • Timed Loaded Standing (TLS): an assessment of shoulder and back muscle endurance for people with VFF(s).20

  • Thoracic kyphosis angle: measured non-radiographically using a flexicurve ruler, allowing an angle of kyphosis to be calculated which is approximated to radiological measures of kyphosis (Cobb angle) using a standardised formula.21

  • Back pain: measured with a 10-point Numeric Pain Rating Scale (NPRS).22

  • Functional Reach test: a measure of dynamic standing balance developed for older adults. T, the test has been used in people with VFF and performance is predictive of falls risk.23

  • Six-minute walk test: a measure of functional walking capacity and aerobic cardiorespiratory fitness.24

  • Falls Efficacy Scale International (FES-I): a 16-item (3 min) self-report measure of fear or concern about falling during activities.25

  • Grip Strength: is the maximum force the hand and forearm muscles can generate measured with an isometric hand dynamometer in kilograms; maximum strength is the mean of three trials (3 s each) and measured for both hands.26

  • Self-efficacy for exercise (SEE) scale: a brief (<5 min) nine-item scale that asks participants to rate how confident they would be that they would engage in exercise on a 10-point scale (not confident to very confident) under different situations, for example, if they were tired. SEE ratings are predictive of exercise behaviour.27

  • Adherence. This will be measured in two ways:

    1. Attendance records via clinician completed treatment logs, including a checkbox to log whether adherence techniques have been prescribed (intervention group only).

    2. Exercise Adherence Rating Scale (EARS): a brief six-item scale that asks participants to describe how they do their recommended exercises on a five-point scale.28

  • Falls: documented on the case report form (CRF) and prospectively using participant completed event diaries. These will be collected in blocks from 0 to 4 months, 4 to 8 months and 8 to 12 months during the study. Incidence and severity formation will be recorded, for example, nature of the fall, its outcome (no-harm, fracture, etc) and any treatment required.2

  • EuroQol - version EQ-5D-5L is a short, generic measure of health related QoL and will be completed to assist assessment of health economics.29

A summary of outcome measures and time points is shown in table 1.

Table 1

Time points at which outcomes will be assessed

Adverse events

Adverse events (AEs) occurring because of the trial interventions will be recorded. Participants will receive information on potential AEs resulting from the exercises and what they should do if they experience an AE, as would be part of standard NHS procedure. Adverse symptoms in response to treatment and any AEs will be monitored by clinicians regularly and in line with local departmental procedures and captured on AE forms, and via questions on the CRF.

A serious AE (SAE) is any untoward medical occurrence related to the trial interventions that results in death, is life-threatening, requires inpatient hospitalisation or prolongation of existing hospitalisation, or results in persistent or significant disability/incapacity. SAEs are likely to be rare and are unlikely to occur as a result of the exercise programmes delivered in this study.

Any reports of SAE will be reported to the trial office within 24 hours of the local research team becoming aware of the event. They will be reviewed by an independent medically qualified assessor within 3 days.

Statistics and analysis

Sample size

The primary outcome is the TUG test. This is the most widely used physical function measure in RCTs of exercise for people with VFF.4 The minimal clinically important difference (MCID) for the TUG has not been established in people with VFF(s), but a MCID of 1.4 s is reported for similar older populations with chronic musculoskeletal disorders.30 The study requires 104 participants (52 per arm) to be 80% powered to detect a 1.4 s difference in TUG score between groups at a 5% significance level (two sided) assuming that the SD is 2.5 s. Similar trials have had loss to follow-up rates of 10% at 12 months.2 To account for this the sample size has been inflated to 116 participants (58 per arm).

Statistical analysis

The study will be reported according to the Consolidated Standards of Reporting Trials (CONSORT) 2010 statement utilising the non-pharmacological and patient-reported outcome extensions.31 32 Standard descriptive statistics will be used to describe the characteristics of the two groups at baseline. Means and SDs or medians and IQRs as appropriate will be used for continuous variables, and numbers and percentages will be used for binary and categorical variables.

Compliance with the intervention will be defined as participating in the extended interview and prescription of at least three adherence support techniques. This will be recorded on treatment logs. Details of the number of physiotherapy sessions attended will also be summarised by treatment group. The number and proportion of participants who withdraw will be summarised along with reasons for these. Deaths are not anticipated in this study, but details of any that do occur will also be summarised by treatment arm.

Summary statistics will be presented for all comparative outcomes, and effect estimates will be reported together with 95% CIs with all tests carried out at a 5% two-sided significance level.

At 12 months postrandomisation the two treatment groups will be compared on the TUG measure using a multivariate linear regression model adjusting for recruiting centre (stratification factor), age and baseline TUG score. An unadjusted t-test will also be undertaken. The TUG is also recorded at 4 and 8 months after randomisation, and an additional analysis utilising all time points, using multilevel modelling and including a treatment by time interaction if appropriate will be undertaken. For each of these models, the assumption of approximate normality will be assessed by examining the residuals. If this assumption is not met the first approach will be to consider a transformation to achieve normality. If this is not possible, the two groups will be compared using non-parametric methods (eg, Mann-Whitney U test). This analysis will be unadjusted and will consider each time point separately.

Similar analyses will be performed for secondary outcomes which can be considered approximately continuous (QUALEFFO-41, FES-1, NPRS, TLS, Grip strength, 6 MW, FRT, Thoracic kyphosis, SEE and EARS) at 4-months, 8-months and 12 months postrandomisation. The appropriateness of the assumption of approximate normality will also be considered and transformation to normality or non-parametric methods used as appropriate. It is not anticipated that the number of falls will be approximately normal, therefore, this will be summarised by treatment group using medians and IQRs and compared using non-parametric methods. The number and proportion of participants experiencing an AE during the follow-up will be summarised by treatment group and a logistic regression model adjusted for recruiting centre will be used to compare the rates in the two groups. Severity of AEs will also be summarised by treatment group.

In addition, since previous work suggests that change in thoracic kyphosis at follow-up is closely related to baseline values, a subgroup analysis of thoracic kyphosis at follow-up will be completed dependent on whether the participant was kyphotic at baseline.33

All analyses will be performed for the intention-to-treat population. This will include all randomised participants with available data who will be analysed according to their allocated intervention regardless of the treatment they received.

In addition, analysis of the primary outcome (TUG at 12 months) will be repeated for the per protocol population which will include only those participants who received their allocated treatment. Participants with other major protocol deviations (eg, recruited and later found to be ineligible) will also be excluded from this population.

Health economic analysis

The relative efficiency of the intervention will be assessed by within-trial cost–utility and cost–consequences analyses.17 The evaluation will take an NHS and personal social services perspective. Resource use for the delivery of Opt-In and at participant level will be combined with unit cost from standard national sources to estimate average total costs. We will estimate the incremental cost per quality-adjusted life-year (from EQ-5D-5L) and present the different cost components and multiple benefits of Opt-In in a ‘balance sheet’ in the cost–consequences analysis.

Embedded qualitative study

As part of the main study a nested qualitative study will take place. The qualitative element of this study will involve a subset (12–15) of patients who undertake the Opt-In intervention who will be invited to take part in 4 short (15–20 min) interviews about their experiences and views about exercise adherence and the adherence intervention at the following time points: (1) following assessment, (2) during treatment (after three sessions), (3) post-treatment (after 4 months) and (4) at 12 months. The interviews may occur online via video-call or face to face in the person’s home or at a local clinic, depending on participant preference. The interviews will be audiorecorded and transcribed verbatim. Participants will be given an opportunity to check the interview transcript.

Focus groups will be conducted with physiotherapists who undertake the Opt-In intervention, asking them to share their views about promoting exercise adherence and the Opt-In intervention, these will be audiorecorded with a Dictaphone.

Purposive sampling will be used to achieve a sample which includes men and women, patients of varying activity levels and patients of different ages and disease severity (pain/number of fractures). These factors may influence the ability to engage with an exercise programme. Since most research regarding adherence in osteoporosis has previously been undertaken with women, it also considered important to capture the views of men within the current study and to capture the experiences of people with differing physical activity levels prior to the programme. The quality of a qualitative study is not dependent on its sample size; however, the sample size needs to be sufficiently large to enable relevant data to be obtained, without being so overly large that detailed analysis is subsequently prevented.34 Information about physiotherapists views of delivering the adherence interventions will also be sought. All those who deliver the adherence techniques will be invited to participate in a focus group.

Audiorecordings will be listened to, and transcripts read until they become familiar. Data from the interviews with physiotherapists and participants will be analysed separately to understand the perspectives of each group. We will use collaborative methods to ensure a strong voice from patient and public involvement (PPI) members and research rigour. We will use thematic analysis, using the six steps proposed by Braun and Clarke.35

Patient and public involvement

The study funding application, intervention development and study materials preparation were supported by our PPI members who will be involved across the course of the study.

Ethics and dissemination

The study protocol was approved by West of Scotland Research Ethics Committee 4 (Reference 21/WS/0071). The University of Oxford is the sponsor. The trial is registered with the International Standard Randomised Controlled Trials database ISRCTN reference number 14465704.

The protocol has been reported following the Standard Protocol Items: Recommendations for Interventional Trials statement.36 Results will be published reported following the CONSORT guidelines.37 The Template for Intervention Description and Replication statement will be used to report the intervention ensuring replication is possible.38 Results will be published in a peer reviewed journal with authorship eligibility according to International Committee of Medical Journal Editors criteria. Participants will be asked if they wish to have the results shared with them prior to publication and we will share with those who request this. We plan to publish results in an international peer-reviewed journal and at international rehabilitation and bone health focused conferences.

Ethics statements

Patient consent for publication


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Twitter @jonroom

  • Contributors KLB: chief investigator, conceived and designed the study, was awarded the funding and had overall responsibility for the study design and delivery and drafted the manuscript. She is the guarantor. JR: contributed to study design and provided specific content and edited manuscript. Qualitative lead. EH: contributed to study design and provided specific content and edited manuscript. RK (statistical coapplicant) performed the sample size calculation, prepared randomisation schedules, planned statistical analysis methods. MN (trial manager), trial design, intervention development, manualisation, training, supervision, writing and reviewing report.

  • Funding The study is supported by an independent grant from the Chartered Society of Physiotherapy Charitable Trust (RP/19/01).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.