Article Text
PDF
Abstract
Objectives Little prior research investigated how to promote physical activity (PA), which can reduce cancer risk, to immigrant women. The overall aim of the current study was to pilot test education session feasibility. The objectives were to assess participation in, satisfaction with and potential impact of the education session, knowledge needed to refine the education session prior to a future trial.
Design Pre–post multiple-methods comparative cohort,
Setting Canadian immigrant settlement agencies recruited intervention and control women.
Participants Intervention: 60 baseline, 53 education (49 virtual group, 4 video only), 1 month (43 virtual, 4 video), 6 months (38 virtual, 4 video), 37% African black; control: 41, 32% African black.
Results Among intervention women, PA knowledge increased significantly from baseline at 1 (p<0.001) and 6 (p=0.01) months, as did PA confidence at both time points (p<0.001). PA behaviour increased significantly from baseline at 1 and 6 months for (p<0.001), moderate (p=0.02) and mild (p=0.05) intensity PA. Total PA metabolic equivalent units (METs) also increased significantly from baseline at both time points (p=0.01). PA confidence (p=0.002) and behaviour assessed by weekly minutes of vigorous (p=0.04, n2=0.05) and moderate (p=0.005) intensity PA, and total PA METs (p=0.01) were significantly greater among intervention women compared with control women. PA knowledge was greater among intervention women compared with control women but not significantly (p=0.8).
Conclusions The findings underscore an important health promotion role for community agencies, which may interest policy-makers, healthcare leaders and health promotion specialists.
- Sexual and Gender Minorities
- Public Health
- Primary Prevention
- Oncology
- Health Education
- Preventive Medicine
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information. All data generated and analysed during the current study are available in the manuscripts and their supplemental files.
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
- Sexual and Gender Minorities
- Public Health
- Primary Prevention
- Oncology
- Health Education
- Preventive Medicine
STRENGTHS AND LIMITATIONS OF THIS STUDY
We employed rigorous methods, complied with standards for conducting and reporting research and supplemented quantitative data with survey and interview data.
We actively engaged immigrant women and community agency advisors in planning and executing this study.
Participating women varied by age, ethnocultural group and years since immigration.
As volunteers, the participants may have a particular interest in physical activity, thus their views may be biased.
Because we focused on ethnoculturally diverse Canadian immigrant women, the findings may not be relevant to those from other disadvantaged groups or to immigrant women in settings outside of Canada with differing health systems.
Introduction
Cancer remains a leading cause of mortality; thus, prevention is an important consideration.1 Physical activity (PA) can reduce the risk of 13 types of cancer: bladder, breast, blood, colon, endometrial, oesophageal, head-neck, kidney, liver, lung, rectal and stomach.2 Modelling suggests that nearly 4 in 10 cancer cases could be prevented if people followed recommended PA levels.3 PA is defined as any bodily movement via skeletal muscles4 and can be achieved in various ways (eg, daily tasks, exercise), and in different environments (eg, home, work, community) and contexts (eg, leisure, household, transportation).5 While PA recommendations vary globally, guidelines commonly recommend a minimum number of weekly minutes of aerobic activity of differing intensity and resistance exercise.5–7
Considerable research has examined how to promote PA at the population level but generated limited insight on how best to do so. A synthesis of 57 reviews found that PA-specific policies (eg, urban design, school or workplace programmes) positively influenced short-term and long-term PA outcomes but authors noted that evidence was sparse.8 Another synthesis of 10 reviews reported that primary care PA promotion interventions (INT) had a small to moderate effect on PA levels.9 Another review of nine studies showed that social marketing increased PA participation but did not identify which approaches suited diverse groups.10 Little research has examined how to promote PA to equity-deserving groups who may be at high risk of cancer due to low PA rates. Cancer risk increases among immigrants from a range of ethnocultural groups when they move to higher-income countries.11 12 Immigrant women may be particularly at risk because they have lower PA rates compared with immigrant men or non-immigrants due to social, economic and cultural factors.13–16
Community-based health promotion may be an ideal approach to reach equity-deserving groups such as ethnoculturally diverse immigrant women, who often lack access to mainstream healthcare services.17 A review of 264 studies found that community-based education about reducing the risk of dementia, which includes PA, was more effective than mass media or the Internet to reach immigrants.18 A 24-week mosque-based initiative increased PA knowledge among Muslim women in Canada.19 Another review of 13 studies of community-based PA promotion for immigrant women from 13 countries found that all but one study significantly improved PA knowledge, levels and anthropomorphic measures.20 Most studies evaluated culturally tailored (delivered in first language, offered solutions to PA barriers) in-person group education reinforced with take-home educational material.20
In prior research, we codesigned an education session on the role of PA in reducing cancer risk with immigrant women and managers of the immigrant settlement agencies that would host sessions.21 The overall aim of the current study was to pilot test the feasibility of education sessions delivered via community agencies. The objectives were to assess participation in, satisfaction with and potential impact of the education session, knowledge needed to refine the education session prior to a future trial.
Methods
Approach
This study employed a pre–post multiple-methods comparative cohort study design and compared outcomes (knowledge, confidence, PA, satisfaction, qualitative views) between women exposed and not exposed to the education session.22–24 Specific community agencies recruited INT women while others recruited CON women, thus delineating the two groups. We followed the STROBE (observational studies), CROSS (surveys) and SRQR (qualitative) research reporting criteria.23–25 The Toronto General Hospital Research Ethics Board granted study approval. All participants provided written informed consent and received a US$100 e-gift card following participation. The research team included five women advisors representing African black, and East and South Asian immigrants, six investigators with expertise in qualitative research, immigrant health, INT design, cancer prevention, PA and PA promotion; and the Canadian Society for Exercise Physiology, which led development of the 24-hour Movement Guidelines.5 There was no prior relationship between the research team members and participants.
Sampling and recruitment
Eligible women were adults (aged 18+) of immigrant groups most common in Canada: African and Caribbean Black, Chinese, Filipino, Indian and Pakistani persons11 who could understand and communicate in the English language. Based on a moderate effect size in the outcome of confidence to perform PA,26 alpha 0.05, power 80%, we aimed to recruit 32 INT and 32 control (CON) women. However, we opted to recruit a larger number to ensure representation across immigrant groups and accommodate potential loss to follow-up. We recruited women through 15 community agencies from across Canada that provided immigrant settlement services including 6 agencies that agreed to offer sessions and recruit INT women, and 9 agencies that agreed to recruit CON women. Agencies shared study information with women clients instructing those interested to contact the study coordinator, who established informed consent, and worked with INT agencies and women to schedule sessions. Although our prior research found that immigrant women preferred in-person sessions,21 the COVID-19 pandemic necessitated virtual sessions live-streamed through an online platform.
Intervention
The INT consisted of an education session and handout. The 1-hour group session included a didactic lecture and interactive discussion delivered in English language, preferred by women we consulted in prior research because they said it would help them practice speaking English.21 However, in advance of sessions, we emailed participants the lecture slides and handout translated into Hindi, Urdu, Tagalog or Chinese. Session content included the benefits of PA, PA recommendations (amount, types, examples), how to overcome barriers of PA and additional ways to reduce cancer risk (eg, use sunscreen, avoid smoking). The handout provided links to tools that support PA, and information about cancer prevention and screening. Based on preferences of each INT host agency and women they recruited, women participated from their own home or in person at the agency. A registered kinesiologist (self-identifying South Asian woman) with expertise in pretreatment and post-treatment PA for patients with cancer facilitated sessions. INT participants unable to attend scheduled sessions received an 11 min voice-over narrated video to review on their own, plus the slides and handout. While this differed from group session, as a pilot-test, we wanted to assess the feasibility of this alternative means of sharing the information. CON women received the video and handout on study conclusion.
Data collection
We assessed PA knowledge with a single question of ‘How much PA is recommended for an adult per week?’, scored as correct if women report between 100 and 210 min/week. While Canada’s 24-hour Movement Guidelines recommend at least 150 min per week of moderate-to-vigorous activity, the key message is to ‘move more’ so the broad range of acceptable responses reflects this message.5 We assessed PA confidence using the nine-item Self-efficacy for Exercise Scale with item responses ranging from 0 being not confident to 10 being very confident, and a higher mean score representing higher confidence.26 We assessed PA behaviour using the six-item International Physical Activity Questionnaire Short Form that elicits self-reported minutes of vigorous, moderate and mild PA, and number of days for each intensity over the last week.27 We multiplied the number of days and number of minutes for a total number of minutes engaged at each intensity, and multiplied this (light×3.3, moderate×4 and vigorous×8) to convert PA scores to metabolic equivalent units (METs).28
The study coordinator and research assistant administered questions and instruments by telephone to INT women 1 month before (baseline), and at 1 and 6 months after the education session and recorded their responses. To avoid temporal effects of the cold Canadian winter climate on PA, and because the CON group was not exposed to any INT, the same staff similarly surveyed CON women a single time concurrent to the 1-month postsession survey of INT women. When surveying INT women, we assessed satisfaction with session content, format, facilitator and usefulness on a 5-point Likert scale (1=strongly disagree, 5=strongly agree) using a 16-item survey that we developed (online supplemental additional file 1). The 12-item satisfaction survey sent to video-only women did not include these questions: I had no problem joining the session, 1 hour was enough time to review and discuss the information, I was asked to share ideas and ask questions, and the trainer did a good job of answering questions. We also asked qualitative questions about satisfaction and potential impact: what did you like and not like about the education session; how did the education session influence PA knowledge and behaviour and how could we improve the education session? We derived survey and interview questions based on the Workgroup for Intervention Development and Evaluation Research framework to elicit views about session content, format, delivery, timing and facilitation.29
Supplemental material
Data analysis
We used summary statistics to describe participant characteristics (age, ethnocultural group, number of years in Canada), compare the characteristics of CON and INT women and assess PA knowledge, confidence and behaviour, and education session satisfaction. We tested the normality of the data sets from both groups using the Shapiro-Wilk normality test. All group and time differences for dichotomous data were explored using χ2 analysis. We compared continuous PA knowledge, confidence and behaviour data (total PA score and each intensity) between INT and CON groups at the 1-month post-INT time-point using a MANOVA model with follow-up univariate effects. We also compared INT group PA knowledge, confidence and behaviour over time using a repeated-measures MANOVA model with follow-up univariate effects. We applied a threshold for statistical significance of p<0.05. Despite the fact that there were only two tests, we ran the Bonferroni correction, which did not impact the findings. We used thematic analysis to analyse qualitative data.30
Patient and public involvement
Five ethnoculturally diverse immigrant women contributed to study planning and execution.
Results
Participants
Table 1 summarises the characteristics of participating women. In the INT group, we surveyed 60 women presession (ie, baseline), 53 received the INT (49 virtual group, 4 video only) and we surveyed 43 (4 video only) and 38 (4 video only) at 1 and 6 months postsession, respectively. In the CON group, we surveyed 41 women. Women in the baseline INT group and CON group were largely 35–50 years of age (60.0% and 51.2%, respectively), African black (36.6% and 31.7%, respectively) and in Canada from 1 to 6 years (46.6% and 39.0%, respectively). Among video only women, 2 were <35 years and 2 between 35 and 50 years of age; 2 were South Asian, 1 East Asian and 1 African black and all 4 resided in Canada from 1 to 6 years. There were no statistically significant differences between CON and INT women at baseline in age (t=1.44, p=0.15), ethnocultural group (χ2=3.37, p=0.64) and years in Canada (t=1.42, p=0.16) despite having been recruited from different agencies.
Characteristics of study participants
Satisfaction
Online supplemental additional file 1 reports satisfaction data from 43 (81.0%) responding INT group women. Most were satisfied or very satisfied (98.0%). Over 90% of respondents agreed or strongly agreed that they learnt about the types of cancer that PA can prevent, recommended weekly levels of PA, different ways to achieve PA levels and how to overcome barriers to PA. All respondents had no problem joining the virtual session or accessing the video. Most (95.0%) respondents agreed or strongly agreed that the lecture slides were easy to understand. Over 95% agreed or strongly agreed that the trainer had good knowledge, spoke clearly and answered participants’ questions well. All respondents planned to do more PA.
Impact
Table 2 provides descriptive statistics for the INT group over time and the CON group at post-INT.
Impact of the education session among intervention (INT) women compared with control (CON) women
Group differences: The MANOVA model testing differences between the INT and CON at post-INT were significant (F(5,78)=3.85, p=0.004, η²ₚ=0.20). Univariate models for confidence (F(1,82)=9.83, p=0.002, η²ₚ=0.11) and weekly minutes of vigorous (F(1,82)=4.34, p=0.04, η²ₚ=0.05) and moderate (F(1,82)=8.27, p=0.005, η²ₚ=0.09) intensity PA were significant, as well as total PA METS (F(1,82)=6.32, p=0.01, η²ₚ=0.07). Minutes of weekly mild PA were not significantly different between groups.
Time effects: The models testing effects over time were limited to the 38 women who completed all three time points. Among the INT group, compared with baseline, the percentage of women reporting PA knowledge increased significantly at post-INT (χ2 (1)=12.5, p<0.001) and 6-month follow-up (χ2(1)=7.6, p=0.01). These results are significant at a Bonferroni corrected p value of 0.025 taking into account the two comparisons. The repeated measures MANOVA model exploring changes in the main study variables for the INT group was significant, F (10,28)=7.29, p<0.00, η²ₚ=0.72. In univariate effects, significant changes were observed for confidence in PA (F(1,37)=10.91, p<0.001, η²ₚ=0.23) and knowledge of PA (F (1,37)=8.90, p<0.001, η²ₚ=0.19) with significant (p<0.02) linear and quadratic effects demonstrating an increase from baseline to post-INT, and a slight decrease at 6-month follow-up that remained significantly higher than baseline. For the univariate effects specific to behaviour, vigorous-intensity (F (1,37)=7.80, p<0.001, η²ₚ=0.17) and moderate-intensity (F (1,37)=3.91, p=0.02, η²ₚ=0.10) PA changed significantly as well as total PA METS (F (1,37)=6.17, p=0.003, η²ₚ=0.07). The time contrasts revealed significant (p<0.01) quadratic effects for weekly minutes of vigorous and moderate PA, and total PA METS with increases post-INT followed by decreases in scores at 6 months. As seen in table 2, the 6-month follow-up data for weekly minutes of vigorous and mild intensity PA and total PA METS remained higher than baseline scores.
Among four video-only INT women, this self-directed version of the INT had mixed results but increased PA knowledge and behaviour for some. Knowledge increased from baseline for three at 1 month and was sustained for two at 6 months. Confidence increased from baseline for four at 1 month and was sustained for 1 at 6 months. Behaviour decreased from baseline for 4 at 1 month but increased from both baseline and 1 month for 3 of those women at 6 months.
Qualitative findings
Online supplemental additional file 2 includes all themes and quotes arising from qualitative survey questions collected from 42 INT women including 3 videos only. Each quote shows participant ID, ethnocultural group and age. Table 3 summarises themes with exemplar quotes. Video-only women expressed similar views to virtual group session participants. Here, we discuss key themes, which correspond to and supplement quantitative findings.
Supplemental material
Summary of themes and quotes from qualitative interviews
When asked what they liked about the session, some participants commented on content, noting that they appreciated learning about PA options and benefits, the difference between intensities, and tips to motivate PA. Participants said that the information was easy to understand and empowering.
Let me know how to do the PA and what kind of PA and make us body healthy…There are a lot of types of activities that can prevent cancer (11 Chinese)
I know the difference between the medium and high intensity workout…These are the information I learn newly (42 Indian video)
We all different type of women from different backgrounds…but I felt all of us were able to understand (34 African)
Others commented on format, noting that they appreciated the inclusion of women only, translated materials and interactivity (sharing experiences, time for questions). They said it was easy to participate because the session was virtual and scheduled in the evening.
I can go back to slides and learn again or remember all the things on slides. I read them before the class, and after the class, I read it again (15 Filipino)
It was in the evening. It made it very easy. I think most importantly the fact that it was virtual (09 African)
It was easier for me to attend in the sense that there will be all the immigrant women (04 Indian)
When asked about impact, many women said that the session increased their knowledge about the importance of PA and its specific role in reducing cancer risk; gave them confidence to do PA; and that they planned to or had already started to do more PA. A few women noted that the benefits extended to family as well.
After the session, I realized that PA is important, and evening, I do some dancing with my child as exercise (29 Indian)
I feel more confident to do more PA after class (20 Chinese)
After I attended that session, I came home and talked to my husband, and we decided to do walking together (03 African)
When asked how to improve the session, a few women suggested choice of in-person or virtual session, delivering the session in multiple languages, including some PA during the session, and offering longer or multiple sessions.
There was nothing that I didn’t like. It was totally informative (41 Pakistani video)
I don’t think one workshop is enough for this topic for me…I might need more instructions about those physical activities (31 Chinese)
However, the majority said they had no critiques, and several recommended offering the programme to more women.
I think we should expand the audience. We should probably have more about this kind of program so that other women can also get the information (21 Pakistani)
Discussion
This feasibility study revealed the potential benefits of a community-based education session for ethnoculturally diverse immigrant women on the role of PA in reducing cancer risk. INT participants were highly satisfied with the session design, planned to do more PA and recommended offering it widely to benefit more women. PA knowledge and confidence increased significantly from baseline at 1 and 6 months among the INT group. PA behaviour also increased over time. Compared with the CON group, PA knowledge was greater among the INT group, but not significantly; and PA confidence and weekly minutes of vigorous-intensity and moderate-intensity PA behaviour were significantly greater among the INT group. A self-directed version of the INT shared with four women who could not attend education sessions increased PA knowledge and behaviour in some, but given the small number of those exposed to this form of the INT, requires future research.
This research contributes to a growing understanding of how to promote PA to equity-deserving groups, in this case, ethnoculturally diverse immigrant women. Prior research focused on exercise programmes, which were not consistently effective because participation and impact waned over time, perhaps because INTs were not culturally tailored.31,32 PA promotion research focused on the impact of policies, primary care counselling or social marketing on PA, but evidence was sparse and those studies did not identify the approaches most suitable for promoting PA among equity-deserving groups.8–10 Other research on community-based health promotion to immigrant groups addressed chronic conditions, not PA, and did not examine which approaches were most suitable to immigrant women.18 A review of in-person community-based PA promotion for immigrant groups included a mere 13 studies, and of those, only 3 included solely women.20 Our study generated preliminary data showing that virtual rather than in-person education sessions can improve PA knowledge, confidence and levels among immigrant women. Therefore, this study was unique in that it examined PA promotion to reduce cancer risk delivered virtually to ethnoculturally diverse immigrant women via community agencies, demonstrating possible benefits.
This research is relevant to health promotion in many countries worldwide due to steadily increasing rates of migration,33 34 and lack of access to or poor mainstream healthcare experiences among immigrants, particularly immigrant women.35–39 Furthermore, the WHO Global Action Plan for Healthy Lives and Well-being for All emphasises the need to improve women’s health.34 In particular, this research considered cultural safety in education session design. Cultural safety has been defined as effective care of a person from another culture as determined by that person, where culture includes but is not limited to age or generation, gender, sexual orientation, occupation, socioeconomic status, ethnic origin, migrant experience, religious/spiritual beliefs or ability.40 41 The concept of cultural safety was developed for research involving Indigenous groups, and there is no established model or framework.41 It is likely that cultural safety must address context-specific determinants of health inequities. This study, including foundational work,21 revealed key components of culturally safe PA promotion education such as women-only, interactive group format, community-based, translated materials, convenient scheduling and strategies to motivate PA.
This research underscores the value of involving community agencies such as immigrant settlement services in culturally safe health promotion. Doing so could address gaps in overburdened, under-resourced healthcare systems and help equity-deserving women to prevent or manage disease; build and expand social networks; interface and self-advocate with the mainstream healthcare system and optimise health and wellness. However, most community agencies are non-profit organisations that rely on volunteers. A case study of chronic disease prevention among community agencies revealed that time and investment are needed to build the capacity for health promotion.42 A systematic review of strategies to build health promotion capacity included only 14 studies and INTs largely targeted individuals only (eg, education, training) rather than a more holistic model that could include organisational infrastructure and partnerships.43 Therefore, further research is needed to establish what constitutes health promotion capacity of community agencies including the role, training and certification requirements of community health workers, who are usually non-health professionals with knowledge of the sociocultural norms, values and behaviours of clients.44–46
This study featured many strengths. We complied with standards for rigorous research methods and reporting criteria.22–25 30 47 48 To inform, plan and execute the study, we employed an integrated knowledge translation approach to engage persons with lived experience including immigrant women and community agency managers as both research advisors and participants.21 We used a multiple methods approach to assess satisfaction with and potential impact of the educational session. Participants varied by age, ethnocultural group and years since immigration to Canada, and their views about the design and benefits of the education session were largely uniform, suggesting the potential relevance to other immigrant women in Canada. We must also acknowledge limitations. The sample size, while sufficiently powered to detect statistically significant differences in outcomes, was small, thus future research must confirm these findings. We recruited women who could understand and communicate in English language, which often correlates with socioeconomic status, and this may have affected study outcomes. Some INT group women did not complete the satisfaction survey; however, those that did expressed high satisfaction with the education session. PA knowledge was not significantly different between INT and CON women. The positive impact in the INT group may have been caused by repeated responses to instruments, a factor requiring consideration in the design of future research. While we did not share study materials with CON women until study conclusion, as part of the informed consent process, we informed CON women that the study aimed to learn about PA among immigrant women, which may have prompted them to search for PA recommendations. Among video-only women, PA knowledge and confidence seemed to degrade by 6 months, but PA behaviour increased by 6 months. While this may signal that women need time to digest and act on information shared with them, this group included only four women, so future research should investigate how to support PA knowledge, confidence and behaviour among women not able to attend group education sessions. While we did recruit beyond the target sample size to accommodate for loss to follow-up, some women were not able to take part in any of the scheduled education sessions, and among those who did, we were not able to collect post-INT data at 1 and 6 months. For those unable to attend, we offered a self-directed version of the INT, which appeared to positive impact PA knowledge and behaviour, but requires future research. To facilitate longitudinal data collection, we may engage immigrant women advisors in collecting pre-INT and post-INT data in future research. As a pilot study focused on assessing INT feasibility and potential impact, we did not collect information about health status; however, a future trial should assess if or how health status influences participation in and impact of the INT. The experiences of immigrant women in Canada may vary from those of immigrant women in other countries, and health promotion infrastructure may differ between Canada and other countries, so the results may not be broadly relevant.
Conclusions
While PA can reduce the risk of cancer, we lacked knowledge of how to promote PA to immigrant women who often have low rates of PA, which increases their risk of cancer. This pilot study suggests that culturally tailored virtual group education delivered via community agencies can increase PA knowledge, confidence and behaviour among ethnoculturally diverse immigrant women. Participants were very satisfied with all aspects of the education session and recommended it be widely offered to immigrant women. While further research is needed to confirm these findings among a larger group of immigrant women, the findings may be of interest to healthcare policy-makers, system leaders and public health agencies as a way to bolster the healthcare system and enhance health among equity-deserving groups.
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information. All data generated and analysed during the current study are available in the manuscripts and their supplemental files.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and University Health Network Research Ethics Board approved the study (REB approval number 22-5434). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank the immigrant women advisors on our research team: KS, DS, TW, VC and LT.
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
X @https://x.com/NazillaKhanlou
Contributors ARG, PB, CMF, NK, CMS, JS-T and JRT conceptualised the study. ARG acquired funding. PB facilitated education sessions. SI, JL and BU assisted with data collection and analysis. CMS conducted statistical analyses. All authors contributed to research design and planning, and interpretation of the findings. All authors contributed to drafting the manuscript. ARG is the guarantor.
Funding This work was supported by the Canadian Cancer Society (grant number 707219) and Canadian Institutes of Health Research (grant number 02270-000).
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.