Article Text
PDF
Abstract
Introduction Evidence suggests that diabetes burden can be reduced by implementing early lifestyle intervention programmes in population with pre-diabetes in high-income countries. However, little is known in developing nations like Nepal. This study aims to assess effectiveness of community-based Diabetes Prevention Education Program (DiPEP) on haemoglobin A1c (HbA1c) level, proportion of pre-diabetes reverting to normoglycaemia, diet, physical activity, weight reduction, diabetes knowledge and health literacy after 6 months of follow-up. Furthermore, we will also conduct qualitative studies to explore experiences of participants of intervention sessions and perception of healthcare workers/volunteers about DiPEP.
Methods and analyses This is a community-based two-arm, open-label, cluster randomised controlled trial. We will randomise 14 clusters into intervention arm and control arm. Estimated total sample size is 448. We will screen individuals without diabetes, aged 18–64 years, and permanent residents of study sites. HbA1c test will be only performed if both Indian Diabetes Risk Score and random blood sugar value are ≥60 and 140–250 mg/dL, respectively. At baseline, participants in intervention arm will receive DiPEP package (including intensive intervention classes, diabetes prevention brochure, exercise calendar and food record booklet), and participants in control arm will be provided only with diabetes prevention brochure. The change in outcome measures will be compared between intervention to control arm after 6 months of follow-up by linear mixed models. Also, we will conduct individual interviews among participants and healthcare workers as part of a qualitative study. We will use thematic analysis to analyse qualitative data.
Ethics and dissemination Regional Committee for Medical and Health Research Ethics, Norway; Nepal Health Research Council, Nepal and Institutional Review Committee, Kathmandu University School of Medical Sciences have approved the study. The DiPEP package can be implemented in other communities of Nepal if it is effective in preventing diabetes.
Trial registration number NCT04074148, 2019/783.
- preventive medicine
- public health
- diabetes & endocrinology
- general diabetes
- education & training (see medical education & training)
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
- preventive medicine
- public health
- diabetes & endocrinology
- general diabetes
- education & training (see medical education & training)
Strengths and limitations of this study
Use of cluster randomisation allows identification of effects when randomisation at the individual level is not suitable.
Cluster randomisation helps to prevent contamination due to sharing of information between participants in intervention and control arm.
In-depth perception of both participants with pre-diabetes and community healthcare workers/volunteers about this intervention programme will be obtained.
Participation of interested candidates in the screening camps might lead to selection bias.
A limitation concerning the qualitative part is the purposive selection of participants based on availability for an online interview.
Introduction
The prevalence of diabetes is increasing globally 1 and is expected to increase from 463 million to 700 million by the year 2045.2 Among the existing number, 79% occurs in low and middle-income countries (LMIC) while the South East Asia (SEA) had 87.6 million of them in 2019.2 The national prevalence of diabetes in India, Sri Lanka, Bhutan, Bangladesh was 8.9%, 8.7%, 8.7% and 8.1%, respectively, whereas it was 4% in Nepal according to the International Diabetes Federation (IDF) atlas 2019.2 However, a meta-analysis and systematic review showed that the pooled prevalence of diabetes in Nepal from 2000 to 2014 was 8.5%,3 which is not very different from the national prevalence of diabetes in other countries of SEA.2 Diabetes is associated with an increased risk of myocardial infarction, stroke, peripheral arterial disease, diabetic nephropathy, diabetic retinopathy4 and all-cause mortality.5–7 Diabetes causes 4.2 million deaths8 and 760 billion dollars expenditure in 2019, which is 10% of global health expenditure.9
The population with pre-diabetes is a high risk group for developing diabetes. Around 5%–10% and 15%–30% of individuals with pre-diabetes develop diabetes annually10 11 and within 5 years,12 respectively. According to the American Diabetes Association (ADA), a fasting blood glucose level of 100–125 mg/dL, oral glucose tolerance test 2-hour plasma glucose level of 140–199 mg/dL and/or glycated haemoglobin (HbA1c) levels of 5.7%–6.4% suggests pre-diabetes—an intermediate state of hyperlycaemia.5 The prevalence of pre-diabetes is also increasing rapidly.13 About 347 million adults between the ages of 20 and 79 (7.5% of the population) have pre-diabetes according to the IDF atlas 2019.2 In Nepal, the prevalence of pre-diabetes was 10.3%.3 Pre-diabetes is associated with complications like cardiovascular disease, coronary artery disease and stroke.14 15 Therefore, early identification of individuals with pre-diabetes, thereby allowing an earlier cost-effective intervention, may potentially prevent diabetes manifestation either by reverting into normoglycaemia or stabilising blood glucose levels.16 Evidence suggests that pre-diabetes can be reversed by the implementation of lifestyle modification programmes by increasing the level of physical activity, following a healthy diet, and reducing weight, which is efficacious, safe and cost-effective.11 17
Most interventions related to ‘lifestyle modification programmes’ have been conducted and implemented in Western nations.11 To our knowledge, so far, no community-based intervention study to prevent diabetes has been conducted in Nepal among individuals with pre-diabetes. Therefore, in our study we developed a lifestyle modification-based educational package-Diabetes Prevention Education Program (DiPEP) in order to prevent diabetes. If the DiPEP study turns out to be successful in reducing the incidence of diabetes, it could be implemented within a community-based healthcare system throughout the country. The trial results may potentially be extrapolated to fill a knowledge gap in other LMICs. Therefore, the primary objective of the study is to evaluate the effect of a Diabetes Prevention Education Program (DiPEP) on HbA1c level, and secondary objectives are to assess the effectiveness of DiPEP on reversion to normoglycaemia, weight reduction, improvement in diet, physical activity among the Nepalese population with pre-diabetes. In addition, we will assess diabetes knowledge and health literacy. A qualitative approach will be used to explore the acceptability and usability of the DiPEP among the participants with pre-diabetes and also to explore the perception of healthcare workers/volunteers about DiPEP.
Methods/design
Study design and setting
This is a community-based open-label two-armed cluster randomised controlled trial (RCT), where the intervention group will follow the lifestyle modification-based educational package—DiPEP and the control group will just receive an educational brochure on diabetes prevention. Two populations in urban regions of Nepal will be screened for participation; Dhulikhel municipality, located 30 km southeast from Kathmandu with a population of 32 026 and a literacy rate of 75.26%,18 and Patan, which is the core part of Lalitpur Metropolitan City (LMC, 5 km southeast of Kathmandu with a population of 284 922 and a literacy rate of 80.35%).19 Both study sites have similar healthcare services as found in general urban areas of Nepal. Four out of 12 administrative units, also known as wards in Dhulikhel,18 and 10 out of 29 administrative units in Patan19 will be operationalised as clusters for data collection. Details are presented in online supplemental tables S1 and S2.
Supplemental material
Study population
The inclusion and exclusion criteria of the study population are presented in table 1.
Inclusion and exclusion criteria of the study
Outcomes
The primary outcome is the HbA1c level (%) after 6 months of intervention, and secondary outcomes are proportion of pre-diabetes reverting to normoglycaemia; change in weight, diet, physical activity, improvement in diabetes knowledge and health literacy after 6 months of follow-up.
Sample size
It has been estimated that a total of 448 pre-diabetes individuals (224 in each arm; 32 per cluster) are needed to detect a minimum detectable difference in HbA1c of 0.12%, as shown in a previous study20 between the intervention and control arm with 80% power, 5% significance level, assuming an intracluster correlation coefficient of 0.01,21 SD=0.3620 and 30% loss to follow-up. This study is powered to detect the minimum detectable difference in HbA1c of 0.12%, which was considered from a local worksite study in the Nepalese population of pre-diabetes.20 Powering the study for a small difference ensures a very high probability of detecting larger differences. We admit that 0.12% may seem small, however, the study will have high power to detect larger differences between groups with the same sample size. We consider that the change in 0.12% in 6 months might be relevant if the change can be sustained over time with the behavioural intervention. Whatever the actual benefit of the intervention, any clinical relevance will be assessed by the estimated difference and its 95% CI once the study is finalised.
In the qualitative study, 10–16 interviews with participants with pre-diabetes of the intervention group and 10–11 interviews with community healthcare workers/volunteers appointed for each cluster in the intervention arm will be conducted. The participants attending at least one of four intervention classes will be selected for the interview. Participants who do not attend any class are not eligible for the interview, which is meant to explore experiences of participants taking part in DiPEP intervention. There will be a purposive selection of participants considering variation in gender, cluster, ethnic group and attendance of the class. Consent of the participants is important to enrol the participants in the interview. The interviews will be conducted until we reach data saturation. Data saturation is defined as the point where no new information or themes are observed in the data.22 The consort diagram with estimated numbers of participants is shown in figure 1.
Consort diagram. This figure illustrates the study design (RCT). A total of 448 participants with pre-diabetes will be randomised to an intervention arm and a control arm. RCT, randomised controlled trial.
Randomisation
Out of four clusters in Dhulikhel Municipality, two will be randomised to the intervention arm and other two to the control arm. Similarly, out of 10 clusters in Patan, five will be randomised to the intervention arm and five to the control arm. Hence, there will be 14 clusters (seven in the intervention arm and seven in the control arm), and for each cluster, 32 participants will be enrolled. The randomisation will be performed using STATA software (runiform).
Screening and recruitment process
Participants will be recruited by organising screening campaigns in the allocated clusters of Dhulikhel and Patan. For this, a temporary place will be set up with all required equipment in the open visible space in the community, where the interested eligible participants could come to have their blood sugar test for the screening purpose. Banners and speakers will be used to make public announcements in the Nepali language to provide awareness about conducting the screening campaigns. The recruitment process will continue until the required sample size in each cluster is attained. Participation in the screening will be voluntary. However, the COVID-19 pandemic will complicate the recruitment process, and the final number of participants is expected to be smaller than originally planned.
The purpose of the screening will be explained to the interested participants in their local language. After obtaining verbal consent, the screening process will be initiated following Indian Diabetes Risk Score (IDRS),23 which takes into account parameters such as age, abdominal obesity, self-reported physical activity and family history of diabetes.23 Random blood sugar (RBS) tests will also be performed via glucometer for the screening process. If a participant meets both the criteria of IDRS ≥60 and RBS between 140 mg/dL and 250 mg/dL, then HbA1c via point-of-care test (POCT) will be performed. The use of different methods and reagents in POCT and lab tests makes some variability in the result.24 However, we will use only one HbA1c analyser called DCA Vantage 2000 throughout the study to minimise the variability in the results. DCA vantage is certified by National Glycohemoglobin Standardization Program/Diabetes Control and Complications Trial and International Federation of Clinical Chemistry and Laboratory Medicine) with coefficient of variance consistently below 3% in internal quality evaluation.25 If the percentage of HbA1c is 5.7%–6.4%,26 participants will be operationalised as pre-diabetes and detailed information about the research project will be provided. Written informed consent will be obtained from all individuals with pre-diabetes and who want to participate in this study. A thumb impression will be obtained in case a participant who is illiterate.
Data collection and assessment
At inclusion as well as at the 6-month follow-up, data on sociodemographic factors such as age (years), gender (male/female/other), marital status (married/unmarried/widow), residency (rural/urban), level of education (highest education), occupation (business/housewife/agriculture/office/others), income (annual household per capita income) will be registered using a standardised questionnaire. Furthermore, data on smoking (cigarettes per day and duration of smoking), alcohol intake (drinks per week), assessment of the quality of diet using two 24-hour dietary recalls, physical activity (metabolic equivalents (METs) mins per week) assessment using a Global Physical Activity Questionnaire,27 information on family history of diabetes and hypertension will be collected. Diabetes knowledge will be assessed by using a standardised questionnaire28, and literacy in diabetes will be assessed by measuring functional, communicative and critical health literacy among patients with diabetes.29 Details are shown in online supplemental table S3. Telephone calls will be used in case physical presence is not possible due to the COVID-19 pandemic.
Standardised techniques and calibrated equipment will be used for anthropometric measurements. Weight (in kg) will be taken in light indoor clothing without shoes using a digital weighing scale. Height (in cm) will be measured using a height scale using a metal scale; hip circumference will be measured using a plastic tape around the widest portion of the buttocks, with the tape parallel to the floor; and waist circumference (in cm) will be measured by placing a plastic tape horizontally, passing the umbilicus (midway between the 12th rib and the iliac crest on the midaxillary line).
Blood pressure will be measured in resting position using the ‘Omron’ digital blood pressure measurement instrument. Three blood pressure measurements will be taken at least 5 min apart30 and the average for the first three measurements will be used for the study. RBS test via glucometer and POCT HbA1c26 via DCA Vantage 2000 HbA1c analyser will be measured. All quantitative data will be collected in an electronic version in a tablet with an Android Operating System using Commcare software.
Intervention arm-DiPEP
Theoretical framework
DiPEP incorporates a behavioural framework that includes components of the health belief model,31 social cognitive theory32 and the transtheoretical model33 to tailor the intervention to each participant. We will consider perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action and self-efficacy as described by the health belief model31 and will emphasise social influence and its emphasis on external and internal social reinforcement as described by social cognitive theory.32 We will recognise different levels of willingness of participants to change behaviour and different barriers to behaviour change, as described by the transtheoretical model.33
Operational framework
The intervention will be conducted following all the components of DiPEP as described in box 1 considering the cultural values of the study population. A four-session DiPEP curriculum will be derived from the Diabetes Prevention Program.34
Components of Diabetes Prevention Education Program (DiPEP) intervention
DiPEP curriculum
It includes four sessions
Introduction to diabetes and pre-diabetes.
Healthy eating and physical activity.
Stress management.
Management of social cues.
Each session will last for 1 hour (30 min theory and 30 min practical).
It is tailored to the Nepalese population-based on US-based National Diabetes Education Program and Diabetes Prevention Program.
Community-based or digital-based media will be used for delivery.
Collaboration with healthcare centres
The main aim of the collaboration with healthcare centres is to work efficiently and to sustain the programme.
Community healthcare workers/volunteers will be appointed by healthcare centres for this programme.
Training and orientation
Training to study nurses
The 1-day training will be provided to study nurses followed by 1 week of practice in the delivery of the contents.
The training will cover the purpose, objectives and implications of the research, ethical issues of the intervention study and core contents of the curriculum.
Training to community healthcare workers/volunteers
One-day workshop on the DiPEP curriculum will be provided to community healthcare workers/volunteers.
Two community healthcare workers/volunteers will be appointed from each cluster.
Terms of reference will be provided to each trainee at the beginning of the training.
The main responsibilities of community healthcare workers/volunteers are to coordinate to conduct DiPEP in the community setup and to be in contact with participants via running weekly maintenance classes or biweekly phone calls.
Orientation to peer leaders
Two peer leaders will be appointed from each intervention group.
They will be oriented to help trained community healthcare workers/volunteers to run maintenance classes and to encourage fellow participants to change their lifestyle.
Orientation to local stakeholders
Orientation about DiPEP will be provided to local stakeholders (ward chairman, ward officials, ward health coordinators, community leaders, community health volunteers) for their support.
Group education session
It will be a group-based education session comprising 60 min.
The first half will be interactive theory and the next half will be supervised exercise training.
The goal of group session is to encourage each other to change their lifestyle to prevent diabetes.
Empathetic listening will be implemented if participants want to share their queries and experiences.
Involvement of family members
Interested family members of the participants with pre-diabetes will also be invited in a group education session.
The aim of inviting family members is to support participants with pre-diabetes to adopt changes in their lifestyle to prevent diabetes.
Brochure on diabetes prevention, exercise calendar and food logbook
The pictorial brochure on diabetes prevention education, exercise calendar and food logbook will be provided in the Nepali language to all enrolled participants with pre-diabetes in the intervention arm.
The pictorial brochure is a summary of the education session.
The exercise calendar consists of a box for each date, where participants write the duration of exercise they perform each day. There is an eighth column in the calendar, which is meant to write total minutes of exercise in a week.
A food logbook is meant to keep records of food intake on a daily basis.
The aim of providing these materials is to track them for their lifestyle and motivate them to change accordingly to prevent diabetes.
Maintenance classes
Trained community healthcare workers/volunteers will conduct maintenance classes weekly to their respective groups with help of the peer group leaders.
They have the following responsibilities:
Arranging premises for the class.
Calling and motivating participants to attend the sessions.
Conducting a 45 min group session.
Encouraging participants in discussion related to the benefits and challenges of lifestyle modification.
Keeping records on attendance, track of weight, diet and physical activity.
A biweekly phone call to participants if there are no physical sessions.
Monthly meeting with study nurses and participants.
The development of a four-session intervention was based on the researcher’s experience in diabetes counselling where Nepali people focus more on symptoms and treatment from the doctor and less on holistic healthcare with preventive approach. Multiple classes and follow-up sessions are challenging in Nepal in terms of both interest of the participants and also the resource for intensive interventions, and the number of sessions was, therefore, restricted to four. Furthermore, the curriculum will be tailored to the Nepalese population by considering the availability and acceptability of foods and the feasibility and acceptability of exercises. Collaboration with local stakeholders; training to study nurses and community healthcare workers/volunteers; orientation to peer group leaders; conduction of group sessions in a community using Nepali language; provision of motivational materials in the Nepali language; involvement of family members and maintenance sessions by community healthcare workers/volunteers will be essential components of this DiPEP intervention. The digital intervention will be performed if physical presence is not possible due to the COVID-19 pandemic.
Control arm
After randomisation, participants allocated to the control group will receive pictorial brochures in Nepali illustrating instructions for physical activity and a healthy diet to prevent diabetes. Thereafter, follow-up on the same participants will take place after 6 months. Later, all sessions demonstrated to the intervention group will also be conducted for the control group once the study is completed.
Qualitative study
To explore the acceptability and usability of DiPEP intervention, participants from the intervention arm will be invited for interviews after 3 months of DiPEP intervention. Participants belonging to different age groups, with varied educational backgrounds, will be invited for an interview to obtain diverse data. Similarly, all trained community healthcare workers/volunteers will also be invited for interviews. Both interview themes are presented in table 2. There will be one interviewer and one-note-taker for each interview. Each interview session is expected to last approximately 60–90 min, and only verbal informed consent will be taken from the potential participants before conducting the interview if the interview will be taken on a digital platform due to the COVID-19 pandemic. The interview will be audio-recorded and significant points will be noted. Recorded interviews will be transcribed (verbatim) and translated into English.
Themes of a qualitative interview with participants with pre-diabetes and community healthcare workers/volunteers
Statistical analysis
Data will be described by mean and SD for normally distributed variables, and median and IQRs for variables with skewed distribution. Frequencies and percentages will be applied for the description of categorical variables. The primary analyses will be performed based on the intention-to-treat principle. Linear mixed models will be used to compare the change in HbA1c in percentage, diet in Kcal, physical activity in terms of change in METs min/week, weight reduction in kilogram, over time between intervention to control groups. Similarly, change in diabetes knowledge and health literacy score will be analysed by linear mixed models.
The effect of DiPEP on proportion reverting to normoglycaemia, quality of diet (poor=0, good=1); physical activity status (inactive=0, active=1, based on a cut-off value of METs minutes/week recommended by WHO); weight loss status (weight loss by less than 5% from baseline=0, weight loss by more than or equal to 5% from baseline=1); on diabetes knowledge and health literacy (not adequate=0, adequate=1), respectively, will be analysed by generalised linear regression with logit and exchangeable correlation coefficient and presented as ORs and 95% CI. Per-protocol analysis will be performed as sensitivity analysis including participants of intervention clusters who have attended all four intervention sessions. In all analyses, standard errors will be estimated by a clustered sandwich estimator. All statistical analyses will be performed using STATA V.16 (IBM, USA).
Thematic analysis
English transcripts of the interviews will be read several times to become familiar with the discussions and obtain manifesting meanings. There will be at least two team members who will review the data. We will consider an 80% consensus process for identification of themes based on intercode reliability and intercode agreement. The meaning units and codes will be condensed to create subcategories and categories depending on their cohering findings. Categories capture the fullness of the experiences and actions studied.35 Mindjet MindManager (software application for mind mapping) will be used to facilitate organising the meaning units into codes and merge into subcategories and categories.
Ethics and dissemination
This study involves human participants and wasapproved by an Ethics Committee(s) or Institutional Board(s). The study is approved by the Regional Committee for Medical and Health Research Ethics, Norway (registration number 2019/783); Nepal Health Research Council (registration number 324/2019), Nepal and Institutional Review Committee, Kathmandu University School of Medical Sciences (registration number 196/19). Also, written permission from Dhulikhel Municipality and LMC have been obtained to conduct the study. A key output of this project will be to facilitate the implementation of this DiPEP package in other communities of Nepal if this intervention programme is effective in preventing diabetes.
Patient and public involvement
General public were involved during the development stage of the diabetes prevention brochure and were asked to give us the feedback in order to improvise and finalise it. The public was not directly involved in developing research questions, study design, intervention designs and outcome measures. The results of this study will be disseminated to the associated organisations, local committees, etc.
Discussion
Diabetes is a serious global health concern and imposes an economic burden on the healthcare system particularly in resource-limited countries like Nepal. Evidence suggests that individuals originating from Asian countries develop type 2 diabetes at a higher rate than Caucasians counterparts.36 The prevalence of individuals with pre-diabetes is also increasing worldwide, and those with pre-diabetes are at an increased risk of developing diabetes and its associated complications. Therefore, the selection of an appropriate target population and cost-effective diabetes prevention lifestyle intervention strategies are much needed.
This community-based cluster RCT (DiPEP) is remarkable in terms of its novelty and potential impact on the country’s policy. Moreover, as this trial is conducted in a resource-constrained country like Nepal, the results of this study might influence activities in other LMIC. Nepal has a federal government, and each municipality has its independence to plan and conduct programmes required for its population. Hence, these kinds of community-based prevention programmes would be resourceful in preventing diabetes.
The main strength of this study is that it is a randomised trial making it possible to evaluate the effectiveness of the lifestyle intervention programme DiPEP to prevent diabetes. The use of cluster randomisation allows identification of effects when randomisation at the individual level is not suitable. Second, as all cluster participants will be offered the same intervention, this may ease the recruitment process, simplify administrative tasks and enhance compliance due to interaction among cluster participants in the intervention arm.37 ,38 Moreover, it provides an environment to motivate each other following social cognitive behaviour theory. Third, cluster randomisation may help to prevent contamination due to sharing of information between intervention and control participants.39 Furthermore, we will obtain an in-depth perception of both participants of pre-diabetes and community healthcare workers/volunteers about this intervention programme from qualitative part of this study.
There are some limitations in this study. First, in general, a cluster RCT design may require a larger sample size than a trial with randomisation at the individual level because of intracluster correlation.40 ,41 We assumed a very small intra-cluster correlation for sample size calculation for this study. Second, there might be a possibility of potential spillover among participants in intervention and control groups due to short geographical distances between clusters and potential contact with participants in other clusters. However, to reduce this risk, we will provide intensive education sessions with follow-up to the specific group participants of the intervention arm. Third, participation of only interested candidates in the screening camps might also lead to selection bias. Fourth, the variability in HbA1c result with POCT and lab test should be considered if the test results from POCT method in this study are to be compared with results from lab test because of the use of different methods and agents in these two methods and also because of discrepancies between capillary, venous and arterial blood sample.42 Finally, there are some limitations in regards to the qualitative parts such as purposive selection of participants based on availability for online interview and degradation of the meaning of the statements due to translation from Nepali language to English language.
Ethics statements
Patient consent for publication
Acknowledgments
Authors would like to thank physiotherapist (Nistha Shrestha), research officer (Seema Das), research assistants (Smriti Shrestha, Suruchi Lama, Krishnaa Barun Shrestha, Supriya Kharel), intern (Ishwori Byanju Shrestha) who contributed to the development this project. Authors would like to remember the contribution of the Late Dr. Kedar Manandhar, who guided in the initial phase of project design.
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 PS, ArS, AbSh, AbSe, BMK, BEK conceived the study, designed the final study protocol and provided the domain knowledge expertise. PS and ArS sampled the study participants. ArS and ES contributed to the technical design and provided biostatistical and epidemiological support. PS and AbSe drafted the manuscript. All authors critically revised the manuscript for important intellectual content and approved the final version.
Funding The grant was obtained by BEK. This work was supported by the Department of Public Health and Nursing, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway (project number: 81850087)
Competing interests None declared.
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.