Article Text

Original research
Association between physician characteristics and practice-level uptake of paediatric virtual mental healthcare: a population-based study
  1. Natasha Ruth Saunders1,2,3,4,
  2. Therese A Stukel3,4,
  3. Rachel Strauss3,
  4. Longdi Fu3,
  5. Jun Guan3,
  6. Eyal Cohen1,2,3,4,
  7. Simone Vigod4,5,6,
  8. Astrid Guttmann1,2,3,4,
  9. Paul Kurdyak4,5,7,
  10. Alene Toulany1,3,4,8
  1. 1Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada
  2. 2Division of Pediatric Medicine, The Hospital for Sick Children, Toronto, Ontario, Canada
  3. 3ICES, Toronto, Ontario, Canada
  4. 4Insitute for Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada
  5. 5Department of Psychiatry, Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada
  6. 6Women's College Hospital and Women's College Research Institute, Toronto, Ontario, Canada
  7. 7Centre for Addiction and Mental Health, Toronto, Ontario, Canada
  8. 8Division of Adolescent Medicine, The Hospital for Sick Children, Toronto, Ontario, Canada
  1. Correspondence to Dr Natasha Ruth Saunders; natasha.saunders{at}sickkids.ca

Abstract

Objective To examine physician factors associated with practice-level uptake of virtual mental healthcare for children and adolescents.

Design, setting and participants A population-based data linkage study of a cohort of all physicians (n=12 054) providing outpatient mental healthcare to children and adolescents (aged 3–17 years, n=303 185) in a single-payer provincial health system in Ontario, Canada from 1 July 2020 to 31 July 2021.

Exposures Physician characteristics including gender, age, specialty, location of training, practice region, practice size and overall and mental health practice size.

Main outcomes Practice-level proportion of outpatient virtual care provided: (1) mostly in-person (<25% virtual care), (2) hybrid (25%–99% virtual care) or (3) exclusively virtual (100% virtual care). Multinomial logistic regression models tested the association between practice-level virtual care provided and physician characteristics.

Results Among physicians, 1589 (13.2%) provided mostly in-person mental healthcare with 8714 (67.8%) providing hybrid care, and 2291 (19.0%) providing exclusively virtual care. The provision of exclusive virtual care (vs mostly in-person) was associated with female sex (adjusted OR (aOR) 1.97, 95% CI 1.70 to 2.27 (ref: male)), foreign training (aOR 1.27, 95% CI 1.07 to 1.50 (ref: Canadian-trained)), family physicians (aOR 2.05, 95% CI 1.56 to 2.69 (ref: psychiatrist)) and reversely associated with large practice size (aOR 0.32, 95% CI 0.25 to 0.40 (ref smallest quintile)). Mostly in-person care was associated with older age physicians (71+ years) and practice outside the Toronto region.

Conclusions and relevance In a single-payer universal healthcare system that remunerates physicians using the same fee structure for in-person and virtual outpatient care, there is heterogeneity in utilisation of virtual care that is associated with provider factors. This practice variation, with limited evidence on effectiveness and appropriate contexts for virtual care use, suggests there may be opportunity for further outcomes research and guidance on appropriate context for paediatric virtual mental healthcare delivery.

  • mental health
  • telemedicine
  • COVID-19

Data availability statement

The datasets from this study are held securely in coded form at ICES. Data-sharing agreements prohibit ICES from making the datasets publicly available, but access may be granted to those who meet prespecified criteria for confidential access, available at www.ices.on.ca/DAS. The complete dataset creation plan, and underlying analytic code are available from the authors on request, understanding that the programmes may rely on coding templates or macros unique to ICES.

http://creativecommons.org/licenses/by-nc/4.0/

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: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

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.

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • This study used several linked datasets, inclusive of all physicians who provided outpatient mental healthcare during the pandemic, in a single-payer universal healthcare system.

  • The methodology involved the use of validated fee codes to identify mental health diagnoses and outpatient visit modality in the paediatric population.

  • In this study, the type of virtual visit could not be discerned due to the lack of distinguishing codes for video and telephone visits.

  • There was limited data pertaining to provider-level characteristics such as ethnicity, language and feasibility of virtual care use, therefore, these factors could not be studied.

Introduction

Virtual mental healthcare, either via telephone or video, has been widely adopted in recent years as a care delivery modality for children and youth.1–5 Mental healthcare, in contrast to physical healthcare, often does not rely on in-person physical examination for a diagnostic formulation, and, therefore is well suited for virtual care.6 Furthermore, families and school age children and adolescents are now more accustomed to using technology, including for healthcare,7–9 facilitating widespread implementation of virtual modalities.

After years of slow adoption,10 11 virtual mental healthcare was broadly implemented and embraced by patients and physicians out of necessity as a way to continue to provide care while controlling the spread of SARS-CoV2 infection. It continues to be promoted as a potentially more patient-centred and family centred healthcare modality that facilitates more timely and accessible care, mitigates disparities, reduces costs and improves communication.6 12 Broadening of reimbursement models for telephone and video visits has also supported uptake by healthcare practitioners.13 14 The optimal balance of in-person and virtual care needed to achieve high patient and provider satisfaction and, importantly, positive health outcomes in diverse clinical populations has yet to be determined.

The extent of uptake of virtual care for mental healthcare has been variable by jurisdiction, underlying health conditions and type of healthcare providers.1 6 15–17 Patient-level factors associated with use of virtual care include familiarity and trust in the modality and accessibility of supporting technologies, low income status and non-white race.6 Factors contributing to uptake by healthcare providers have been less studied,6 with some early evidence of lower uptake in younger physicians and those practising in rural settings16 17 health system factors, including restrictions on reimbursement and licensing may also contribute to virtual care utilisation. For example, in jurisdictions with little or no physician remuneration for virtual care, uptake is poor.18 In contrast, in jurisdictions where physicians are compensated equally for in-person and virtual care (eg, Ontario, Manitoba), approximately 70% of physician-delivered mental healthcare visits occur virtually.1

As providers, young people and families gain experience with the technologies available for virtual mental healthcare, understanding factors associated with uptake and utilisation is important. Our previous work identified uptake of virtual mental healthcare by physicians was similar across patient-level factors including diagnostic group, patient sex and patient age group.1 In this study, we build on this body of work and evaluate the association between practice-level virtual care use for physician-based paediatric mental healthcare and physician characteristics in the first 15 months since the widespread adoption of virtual care during the COVID-19 pandemic.

Methods

We conducted a population-based cohort study in Ontario, Canada, using data from ICES (formerly the Institute of Clinical Evaluative Sciences), a research institute that collects and uses health administrative data for health system evaluation and improvement. Ontario is Canada’s largest provinces (population ~14 million) and has a single-payer universal healthcare system that provides physician and hospitals services at no direct cost to residents.

Data sources

We used several datasets linked through unique encoded identifiers and available at ICES: Ontario Health Insurance Plan (OHIP) database for outpatient physician visits, Registered Persons Database for patient sex and age and Corporate Providers Database for data on physician characteristics. Data available at ICES are complete and valid for primary discharge diagnoses, physician billing claims and sociodemographics.19

Study population

Our study population included all Ontario family physicians, paediatricians and psychiatrists with an outpatient mental health-related OHIP visit claim for children and adolescents aged 3–17 years between 1 July 2020 and 31 July 2021. Data from the first 3 months of the COVID-19 pandemic were excluded to allow for a washout period as care providers pivoted to set up their virtual practices. We excluded any physicians with missing gender information, those not living in Ontario and those who practised for <1 year prior to the onset of the COVID-19 pandemic (14 March 2020).

Practice-level virtual care use for mental healthcare

To compare practice-level virtual mental healthcare use, we first identified all outpatient mental health visits between 1 July 2020 and 31 July 2021 to individual active physicians in the Corporate Providers Database using the OHIP physician billings claims with fee codes that have been validated to identify mental health diagnoses or counselling and modified for paediatric populations (online supplemental eTable 1).20 We then determined if visits were conducted virtually (telephone or video) or in-person using visit modality fee codes introduced by the Ontario Ministry of Health in March 2020.21 We determined the proportion of practice-level virtual care by measuring the total number of paediatric mental healthcare visits conducted virtually divided by the total number of paediatric mental healthcare visits for each physician during the study period. We stratified physicians into three groups based on the proportion of virtual care provided: (1) mostly in-person (<25% virtual care), (2) hybrid (25%–99% virtual care) and (3) exclusively virtual (100% virtual care) and used this as our main outcome measure. We determined these groups based on clinical relevance and prior work by our team that described the proportion of paediatric virtual mental health and primary care delivery in the pandemic.1 22

Physician and patient characteristics

We considered the following physician-level characteristics as factors potentially associated with virtual care use: physician age, gender, specialty (family physician, paediatrician, psychiatrist), location of graduation from medical school (Canada vs foreign), number of years in practice, region of practice within Ontario, prepandemic practice mental health practice size (number of unique individuals with mental health visits to a physician for mental health concerns), prepandemic proportion of practice focused on mental health (proportion of all visits in year prior to the pandemic that were for mental health concerns). To provide practice-level context, we measured patient-level characteristics (case-mix) including patient age, patient sex and patient mental health diagnosis grouping using the International Statistical Classification of Diseases and Related Health Problems, Eight Revision diagnoses (psychotic disorders, mood and anxiety disorders, substance use disorders, social problems, neurodevelopmental and other concerns).

Statistical analysis

We described the physician demographic, training and practice characteristics, for all physicians combined and then separately by specialty (family physician, paediatrician, psychiatrist), and also described these characteristics by practice-level proportion of outpatient virtual care provided (grouped as mostly in-person, hybrid and exclusively virtual).

The analysis focused on comparing characteristics of physicians and their practices with hybrid or exclusively virtual versus mostly in-person (the reference category) outpatient physician-based mental healthcare. We performed multinomial logistic regression, modelling practice-level proportion of outpatient virtual care provided as a nominal outcome on the physician and practice factors described above. This form of the multinomial logistic regression model fits simultaneous logistic regressions on each outcome category (hybrid or exclusively virtual) with respect to the reference category (mostly in-person) in a saturated model where every covariate has a separate effect on each pair of outcomes, and estimates the corresponding adjusted ORs (aORs) and 95% CIs. All analyses were performed using SAS V.9.4 LOGISTIC procedure.

ICES is a prescribed entity under section 45 of Ontario’s Personal Health Information Protection Act. Projects conducted under section 45 do not require research ethics board review and this study was approved by the ICES Privacy Office.Cell sizes less that six were not reported in accordance with Ontario privacy regulations. This study followed the Reporting of studies Conducted using Observational Routinely collected health Data guideline.23

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Results

Physician and patient characteristics

Between 1 July 2020 and 31 July 2021, 12 054 physicians (10 015 (83.1%) family physicians, 1139 (9.4%) paediatricians and 903 (7.5%) psychiatrists) had at least one claim for outpatient mental healthcare for 303 185 unique paediatric patients and met inclusion criteria for this study (table 1). The overall mean±SD physician age was 50.8±12.1 years with psychiatrists slightly older (mean age 53.5±11.9 years) compared with family physicians (mean age 50.2±12.2 years) and paediatricians (50.8±11.7 years). Half (50.2%) of all physicians were female, although paediatricians were disproportionately female (61.4%) and slightly more than half (55.7%) of psychiatrists were male. Physicians were in practice for a mean of 19.2±13.3 years. Prior to the pandemic, the mean mental health-specific practice size was 73.7±172.6 with an overall patient practice size of 958.9±1318.5 individuals (810.1±969.0 family physicians, 2863.8±2402.2 paediatricians and 208.3±371.5 psychiatrists).

Table 1

Characteristics of physicians and their paediatric patients with outpatient mental health visits in Ontario, overall and stratified by physician specialty, 1 July 2020 to 31 July 2021

The mean patient age in the practice was 10.0±4.3 years and ranged from 9.4±4.1 (paediatricians) and 13.0±3.5 years (psychiatrists). Patients seen by paediatricians were less frequently female (37.4%, n=57 713) compared with family physicians (52.4%, n=81 314) and psychiatrists (52.6%, n=25 162). A higher proportion of patients seen by psychiatrists had visits related to a psychotic disorder (6.5%, n=3108), compared with those seen by family physicians (0.7%, n=1098) or paediatricians (0.1%, n=192). Paediatricians had the largest proportion (77.8%, n=119 965) of visits for neurodevelopmental and other disorders (eg, autism, attention deficit hyperactivity disorder, etc) compared with family physicians (33.9%, n=55 531) and psychiatrists (41.2%, n=19 738).

Proportion of virtual care by practice-level characteristics

Among all physicians, 1589 (13.2%) provided mostly in-person care with 8714 (67.8%) provided hybrid mental healthcare, and 2291 (19.0%) provided exclusively virtual care (table 2). Physicians ≥61 years and ≤40 years had the largest proportion providing exclusively virtual care. Among female physicians, 19.7% provided exclusively virtual care with 9.1% providing mostly in-person care; males provided 18.3% exclusively virtual care and 17.3% provided mostly in-person care. There were large differences in the proportion of exclusively virtual care by specialty with 6.0% of paediatricians providing exclusively virtual care and 28.5% of psychiatrists and 19.6% of family physicians providing only virtual care (online supplemental eTable 2). In the overall cohort, we observed no major differences in provision of virtual care by years in practice, but within specialties, 38.8% of psychiatrists who had been in practice for >20 years provided exclusively virtual care in contrast to only 8.6% of paediatricians who had been in practice for the same amount of time (online supplemental eTable 2). There were major differences in exclusive virtual care delivery by region with 31.1% of Toronto physicians vs 10.2% of Northern region physicians providing exclusively virtual care (table 2).

Table 2

Characteristics of physicians and their paediatric patients with outpatient mental health visits in Ontario, stratified by proportion virtual, 1 July 2020 to 31 July 2021

The adjusted odds of hybrid care versus mostly in-person and exclusively virtual versus mostly in-person care by provider characteristics are shown in figures 1 and 2. In both models, female physicians, compared with males, were less likely to provide in-person care. More specifically, female physicians were 73% more likely than males to use hybrid versus mostly in-person mental healthcare (aOR 1.73, 95% CI 1.53 to 1.96) and 97% more likely to use exclusively virtual versus mostly in-person care (aOR 1.97, 95% CI 1.70 to 2.27). As a function of specialty, family physicians were over four times more likely (aOR 4.07, 95% CI 3.20 to 5.18) to provide hybrid versus mostly in-person mental healthcare and two times more likely (aOR 2.05, 95% CI 1.56 to 2.69) to provide exclusively virtual versus mostly in-person mental healthcare compared with psychiatrists. In adjusted models, there were no major differences by physician age except that physicians ≥71 years were less likely than those ≤40 years to provide hybrid (aOR 0.40, 95% CI 0.31 to 0.53) mental healthcare. Physicians who had the largest pre-COVID-19 practice panel sizes (quintile 5) were almost three times more likely (aOR 2.89, 95% CI 2.32 to 3.60) to use hybrid versus mostly in-person care, although were almost 70% less likely (aOR 0.32, 95% CI 0.25 to 0.41) to use exclusively virtual versus mostly in-person mental healthcare, compared with physicians who had the smallest pre-COVID-19 panel sizes. Physicians practising in east, west and north regions of the province had lower odds of using either hybrid versus mostly in-person or exclusively virtual versus mostly in-person, when compared with physicians practising in Toronto.

Figure 1

Adjusted odds of hybrid versus mostly in-person outpatient physician-based mental healthcare for paediatric patients by physician characteristics in Ontario, 1 July 2020 to 31 July 2021.

Figure 2

Adjusted odds of exclusively virtual versus mostly in-person outpatient physician-based mental healthcare for paediatric patients by physician characteristics in Ontario, 1 July 2020 to 31 July 2021.

Discussion

In this population-based study of physicians providing mental healthcare to children and adolescents in a single-payer universally funded health system, 19% provided exclusively virtual mental healthcare, 13% provided mostly in-person care and 68% practised a hybrid model with at least one-quarter of their mental healthcare delivered virtually. The provision of exclusive virtual care was associated with female gender, foreign training, family medicine specialty, large overall practice size and small mental health practice size. Providing mostly in-person care was associated with older physician age (71+ years) and practice outside of the Toronto region. These findings suggest that there is heterogeneity in utilisation of virtual care that are associated with provider factors.

In the current study, we observed that most physicians used virtual modalities to deliver at least some mental healthcare. Virtual modalities have the potential to facilitate more timely and accessible mental healthcare, mitigate disparities and address workforce shortages.24 They may also reduce the amount of time required for each visit and improve clinic efficiency.25 In our previous work, we have shown that approximately 70% of physician-delivered paediatric mental healthcare in Ontario is delivered virtually, with few differences in virtual care uptake by diagnostic grouping, age group and sex.1 These findings contrast with other jurisdictions and health systems where 30%6 physicians used virtual care. Specifically, in the USA, only 50% of psychiatrists used telemedicine after the onset of the COVID-19 pandemic.6 Since its widespread introduction in March 2020, Ontario has a funding mechanism to support equivalent remuneration for physicians for in-person and virtual visits.13 Other jurisdictions that do not remunerate equally or remunerate at a lower fee than in-person have low utilisation of virtual modalities, particularly for non-mental health conditions.16 Insufficient remuneration for telehealth services have been reported to be the largest barrier to uptake in the USA.3 26 While Medicaid programmes have payment provisions for live video telehealth services, few pay for such services when patients are in their home and even fewer pay the same rate as in-person visits.14 27 Early in the pandemic in paediatric primary practices in Western Pennsylvania, significant practice-level variation in telemedicine use was observed with increased uptake associated with more varied diagnoses and increased overall primary care encounter volume, although this study only examined the first 2 months of the pandemic and did not restrict to mental health diagnoses.24 Others have focused on patient-level factors and reported those of races other than white non-Hispanic were less likely than other races to have a telemedicine visit.24 It is not known if these patient-level race factors extend to utilisation by physicians or their practices. In Ontario, while population-level literacy, education, race and ethnicity data are not available in linked datasets and may be important for virtual care utilisation, we have shown neighbourhood-level income and rural residence are not associated with uptake of virtual care but refugees used a lower proportion of virtual mental healthcare compared with other paediatric groups including non-refugee immigrants.28

There are several possible explanations for our findings. We showed substantial gender-based differences in provision of mental healthcare services with female physicians providing more exclusively virtual and less mostly in-person care compared with males. During the pandemic, female physicians may have had greater ‘dual caregiver’ roles, providing care for both patients and children at home during school closures, thereby making virtual care a more attractive and feasible option.29–31 The age findings of younger physicians (<40 years) having the highest odds of exclusively virtual mental healthcare may also reflect the dual caregiving roles and work responsibilities.29 31 In contrast, the high proportion of older (71+ years) physicians providing mostly in-person care may reflect less familiarity with technology for some, or potentially fewer dual caregiving responsibilities.32 Our findings that the Toronto region had the largest proportion of exclusively virtual mental healthcare while Western and Northern Ontario had the highest proportion of in-person visits may be reflective of the relatively high SARS-CoV2 infection rates in the Toronto region during this period with greater mobility restrictions.33 Family physicians had the highest odds of providing exclusively virtual mental healthcare. In contrast, only a small proportion (6%) of paediatricians provided exclusively virtual mental healthcare. One might expect variation in in-person visits based on the need to perform a physical examination to guide diagnosis and management. In mental healthcare however, in-person physical examinations are not as critical for diagnosis and monitoring, except perhaps for neurodevelopmental disorders. Findings may be related to personal protective equipment availability or many paediatricians already keeping offices open to deliver immunisations to the relatively large proportion of young children in their practices for vaccine-preventable diseases, or the large proportion of children with neurodevelopmental disorders managed by paediatricians where a physical examination may be warranted.17 This is supported by the case-mix differences observed by the practice-level proportion of virtual care. More specifically, in exclusively virtual practices, social problems and neurodevelopmental disorders were under-represented.

Our study has several limitations. We examined provider factors related to uptake of virtual care during the first 15 months following the introduction of fee codes for widespread availability of virtual care. It is unknown whether, after pandemic restrictions have been completely lifted and there is resumption of ‘normal’ activities, there will continue to be heterogeneity in practice and whether the same association will be observed. This may be particularly relevant as guidance for appropriate use of virtual versus in-person care evolves. We could not differentiate between video and telephone visits as distinguishing codes were not introduced to Ontario until October 2021 at the end of the study period. Data were not available on feasibility or digital access factors that may affect a provider’s ability to provide mental health visits virtually. We have data on billing code diagnoses rather than clinical diagnoses, limiting our understanding of the richer clinical context in which these encounters occurred. Data are not available and linked for provider-level ethnicity or languages, which may also affect virtual care uptake.

Conclusions

We showed wide variation in uptake of virtual mental healthcare at the physician practice-level. Several physician factors, including age, gender, foreign training, region and specialty were associated with virtual care utilisation. This variation in practice, with little evidence on the quality of mental healthcare or mental health outcomes with widespread adoption of virtual mental healthcare, is concerning and suggests the need for greater guidance for providers on appropriate contexts for paediatric virtual mental healthcare delivery. This will be important as different funding models and guidance on appropriate use of the provision of virtual care are developed.

Data availability statement

The datasets from this study are held securely in coded form at ICES. Data-sharing agreements prohibit ICES from making the datasets publicly available, but access may be granted to those who meet prespecified criteria for confidential access, available at www.ices.on.ca/DAS. The complete dataset creation plan, and underlying analytic code are available from the authors on request, understanding that the programmes may rely on coding templates or macros unique to ICES.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by the ICES Privacy Office and and this project was conducted under section 45 of Ontario’s Personal Health Information Protection Act does not require research ethics board review.

References

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.

Footnotes

  • Contributors NRS conceptualised and designed the study, interpreted the results, drafted the initial manuscript and revised the manuscript and as the study guarantor, accepts full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish. TAS, RS, and AT conceptualised and designed the study, interpreted the results and revised the manuscript. EC, SV, AG and PK interpreted the results and revised the manuscript. LF and JG designed the study, had access to and analysed the data, interpreted the results and revised the manuscript. All authors reviewed and approved the final manuscript as submitted and agreed to be accountable for all aspects of the work.

  • Funding This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). This study also received funding from: MOH Grant 710 and the Canadian Institutes of Health Research Grant MS1-173069, both awarded to Drs NRS and AT. This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under license from ©Canada Post Corporation and Statistics Canada. Parts of this material are based on data and/or information compiled and provided by: CIHI and the Ontario Ministry of Health. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.

  • Competing interests NRS reported receiving personal fees from the Archives of Diseases in Childhood, grants from the Canadian Institutes for Health Research and the Ontario Ministry of Health, grants from the Centre for Addiction and Mental Health and grants from The Hospital for Sick Children outside the submitted work. SV reported receiving royalties from UpToDate for authorship of materials related to depression and pregnancy outside the submitted work. No other disclosures were reported.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • 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.