Article Text
PDF
Abstract
Background This study aims to evaluate the long-term trend of prevalence and DALY (disability-adjusted life-year) rate on the age, period and cohort (APC) of the BRICS (Brazil, Russia, India, China and South Africa) country for autoimmune diseases (rheumatoid arthritis (RA), inflammatory bowel disease (IBD), multiple sclerosis (MS) and psoriasis).
Methods The data are sourced from the Global Burden of Disease Study 2019, and it uses the Joinpoint regression model to estimate the time trends of autoimmune diseases from 1990 to 2019. Additionally, it employs the Age-Period-Cohort (APC) model to estimate the age, period, and cohort effects from 1990 to 2019.
Results For 1990 to 2019, the ASPR (age-standardised prevalence rate) of IBD increased significantly for China and South Africa, and decreased significantly for Brazil, India, Russian. The Russian ASPR of MS demonstrated a significantly decreasing trend (average annual percent change=−0.5%, 95% CI −0.6 to −0.5), with the most increased occurring in Brazil at 2009–2014. The cohort effect on DALY rates for Psoriasis displayed an ongoing decreasing trend from the 1929–1933 birth cohort to the 1999–2003 birth cohort. Specifically, the five countries relative risk values (RRs) of DALYs due to RA increased significantly by 7.98, 16.07, 5.98, 3.19, 9.13 times, from 20 to 24 age group to 65 to 69 age group.
Conclusions The population of the BRICS countries accounts for more than 40% of the global population. And we found that the age effect of various autoimmune diseases is heavily influenced by population ageing.
- Autoimmune diseases
- Joinpoint regression analysis
- Prevalence
- Age-period-cohort model
- DALY
Data availability statement
Data are available in a public, open access repository. The list of data sources used is publicly available at the Global Health Data Exchange website (http://ghdx.healthdata.org/gbd-results-tool). No additional data are available.
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
WHAT IS ALREADY KNOWN ON THIS TOPIC
This study aims to evaluate the long-term trend of prevalence and disability-adjusted life-year rate on the age, period and cohort of the Brazil, Russia, India, China and South Africa (BRICS) country for autoimmune diseases (ADs) (rheumatoid arthritis (RA), inflammatory bowel disease (IBD), multiple sclerosis (MS) and psoriasis).
WHAT THIS STUDY ADDS
While these studies provide valuable data, further research on the comparison of ADs prevalence among the BRICS countries is limited due to the heterogeneity in the methods used for disease assessment across different studies, as well as the lack of comparative data on ADs prevalence in the BRICS countries with varying levels of economic development. Therefore, our study can explore the burden of four ADs of different BRICS countries (China, India, Russian, Brazil and South Africa). Four of ADs are RA, IBD, MS and psoriasis.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Immune diseases have received extensive attention in recent years. In recent years, numerous studies have been published on topics such as epidemiology, underlying mechanisms, early diagnosis and intervention measures. The prevalence of ADs may vary significantly in different regions and countries, and this could be attributed to factors such as disease specificity, complex genetic backgrounds and environmental influences.
Introduction
Autoimmune diseases (ADs) are related to significant incidence and mortality, and can affect many organs and systems, with significant clinical heterogeneity.1 It is estimated that the proportion of population affected by ADs globally is about 10%,2 of which in developed countries, the prevalence of ADs accounts for 4%–5% of the total population.3 In general, ADs affect approximately 20 million Americans, especially women. Moreover, the incidence of ADs is on the rise in industrialised nations.4 However, in the past few decades, the influence of BRICS (Brazil, Russia, India, China and South Africa) countries has greatly increased on the international stage5. The population of these five countries accounts for more than 40% of the global population, and the number of diseases accounts for about 40% of the burden on global diseases.6 With the rapid expansion of the economies of BRICS countries, there has been a consequent increase in population, environmental pollution, unequal access to healthcare resources and a shift in disease burden towards non-communicable diseases.7 The development of ADs in low-income and middle-income countries has drawn global attention.8 Brazil, China and India have emerged as leaders in the production of low-cost drugs and vaccines, and the BRICS countries have accumulated significant experience in achieving universal health coverage and providing affordable drugs and vaccines.9 10 Studies have shown that in China, the prevalence of rheumatoid arthritis (RA) is approximately 0.28%, with the peak of onset occurring in median or older age groups.11 In recent years, numerous studies have been published on topics such as epidemiology, underlying mechanisms, early diagnosis and intervention measures.12 The prevalence of ADs may vary significantly in different regions and countries, and this could be attributed to factors such as disease specificity, complex genetic backgrounds13 14 and environmental influences.15
Although these four ADs have higher prevalence rates in Europe and the Americas from a global perspective, their incidence levels are lower in Africa and Asia, and they tend to cluster in relatively developed regions.16 However, studying the prevalence of ADs in emerging economies such as the BRICS countries is also crucial. While these studies provide valuable data, further research on the comparison of ADs prevalence among the BRICS countries is limited due to the heterogeneity in the methods used for disease assessment across different studies, as well as the lack of comparative data on ADs prevalence in the BRICS countries with varying levels of economic development. Therefore, our study can explore the burden of four ADs of different BRICS countries. Four of ADs are RA, inflammatory bowel disease (IBD), multiple sclerosis (MS) and psoriasis.
Methods
Data sources
The Institute for Health Metrics and Evaluation, an independent global health research centre at the University of Washington, provided the data for the Global Burden of Disease 2019 (GBD 2019), which includes national ADs incidence and prevalence from 1990 to 2019 (http://ghdx.healthdata.org/gbd-results-tool). The GBD study aims to establish comprehensive and comparable global health metrics, reporting various indicators such as incidence, prevalence, death, years of life lost, years lived with disability and disability-adjusted life-years (DALYs) for each disease and injury, categorised by year, location, age group and sex. In this study, we obtained estimates along with their 95% uncertainty intervals (UI) from the GBD 2019 database for various metrics, including incidence, prevalence, DALYs and age-standardised rates (ASRs). These metrics served as measures of the burden of ADs. ASRs were age-standardised using the GBD standard and reported per 100 000 population. Statistical descriptions and analyses of the incident cases were conducted using the R programme (V.3.6.0, R Core Team).
Statistical analysis
Joinpoint regression analysis
The Joinpoint regression model is a statistical method that analyses the trend of the incidence of disease or mortality over time. The model starts by fitting the original data with the minimum number of joinpoints (eg, 0 joinpoints, representing a straight line). It then tests whether additional joinpoints are statistically significant and must be added to the model. The maximum number of connection points is related to the input data. The number of observation values in this study is greater than equal to 29. Therefore, the maximum number of connection points by default is 4, divided into five trend groups, ensuring the results were credible. This study used the Joinpoint regression analysis model to perform trend analysis on the prevalence data of ADs in the BRICS countries. The analysis was based on annual per cent change and average annual per cent change (AAPC) to assess the trends in incidence and prevalence of ADs in BRICS countries. If the annual per cent change value is0, it indicates an increasing trend in ADs prevalence, whereas0 indicates a decreasing trend. The AAPC value represents the geometric weighted AAPC values.17 The statistical significance of each trend segment’s annual per cent change values and the overall AAPC value was validated using the Monte-Carlo permutation method. All the above procedures were implemented using the Joinpoint regression program V.4.7.0.0 provided by the National Cancer Institute (NCI) of the USA.
Age–period–cohort analysis
Collinearity is a common issue in the application of the age–period–cohort (APC) model, as it involves a relationship between cohort=period age. In this study, we employed the intrinsic estimator (IE) algorithm, whose parameters have been proven to be estimable, unbiased, efficient and asymptotically normal.18 The model is represented as follows: Y=log(M)=μ+αage i+βperiod j+γ cohort k+ε. Here, α, β and γ are the coefficients for the three dimensions (α represents the age effect, indicating the mortality or incidence risk for a specific age group; β represents the period effect, indicating the mortality or incidence risk for a specific time period; γ represents the cohort effect, indicating the mortality or incidence risk for all individuals in the same birth cohort); μ and ε are defined as the intercept and random error, respectively.
The age, period, and cohort coefficients estimated by the APC model are not directly interpretable.
Therefore, we calculated the relative risk (RR) values to help explain the independent effects of APC on the DALY rates of ADs. In this work, age-specific DALY rates of ADs were classified by consecutive age groups (20–24, 25–29, 65–69, 5-year periods (1994, 1999, 2004, 2009, 2014, 2019), and correspondingly 5-year birth cohort groups (1929–1933, 1934–1938, 1999–2003). To calculate the relative risk values, we used the 20–24 age group, the 1994–1999 period and the 1929–1933 birth cohort as the reference group. For each other group, we computed the difference in estimated parameters compared with the reference group and then took the exponential of this difference to obtain the corresponding relative risk value. The analysis was performed by using Stata 16.0 software, developed by StataCorp.
Results
As far as China is concerned, Psoriasis was the most common ADs, of the other ADs, RA showed the next largest proportion, followed by IBD, with MS representing a small incident case. Moreover, Psoriasis with incident cases ranging from 388 948.9 (95% UI 375 495, 403 586) in 1990 to 635 345 (95% UI 613 080, 657 983) in 2019 in India. While incident cases of MS decreased slightly from 1763 (95% UI 1486, 2025) to 1339 (95% UI 1130, 1547) in Russian from 1990 to 2019. Therefore, except for incident cases of ADs of Russian, the number of ADs in other BRICS countries has increased significantly In South Africa, psoriasis was the most common ADs, with an incident cases ranging from 12 277.80 in 1990 to 17 912.04 in 2019. Of the other ADs, RA showed the next largest proportion, followed by IBD and MS (figure 1).
The incident cases of inflammatory bowel disease, multiple sclerosis, psoriasis and rheumatoid arthritis by BRICS countries. BRICS, Brazil, Russia, India, China and South Africa.
The age-standardised incidence rate (ASIR) for RA has an upward trend from 1990 to 2019 in China. The ASIR, age-standardised prevalence rate (ASPR) and DALY for MS have an upward trend from 1990 to 2019 India. The Russia ASPR of MS decreased from 22.86 (18.85, 27.05) in 1990 to 19.60 (16.04, 23.35) in 2019 per 100 000 persons. The Brazil DALY rate of RA in 2019 was 5.3% higher than that of 1990s (40.33 cases per 100 000 population (29.30,51.97) vs 42.60 (31.08, 54.68)). The ASIR, ASPR and DALY for IBD have a decreasing trend from 1990 to 2019 India (table 1).
Age-standardised incidence, prevalence and DALY rate of autoimmune diseases in the BRICS countries, 1990–2019
We described the trend changes and annual change percentage in Prevalence of IBD, MS, psoriasis and RA of the BRICS five countries through the Join-point model. For 1990 to 2019, the ASPR of IBD increased significantly for China (AAPC 2.5%, 95% CI 2.4% to 2.7%; p<0.001) and decreased significantly for Brazil (AAPC −1.1%, 95% CI −1.3% to −0.8%; p<0.001), India (AAPC −0.1%, 95% CI −0.1% to −0.0%; p<0.001), Russian (AAPC −0.3%, 95% CI −0.4% to−0.3%; p<0.001) and South Africa(AAPC −0.4%, 95% CI 0.3% to 0.5%; p<0.001). From 1990 to 2019, the Russian ASPR of MS demonstrated a significantly decreasing trend (AAPC −0.5%, 95% CI −0.6% to −0.5%; p<0.001), with the most substantial increased occurring in Brazil at 2009–2014 (APC 2.0 %, 95% CI 1.3% to 2.8%; p<0.001). The ASPR of psoriasis varied significantly across BRICS countries (figure 2 and table 2), with the highest ASPR in Brazil and lowest ASPR in South Africa. In addition, the ASPR of RA in China and South Africa showed a trend of rising first and then decline.
Trends in autoimmune diseases prevalence in the BRICS countries, 1990–2019
Age-standardised prevalence rate per 100 000 for inflammatory bowel disease (A), multiple sclerosis, (B) psoriasis (C) and rheumatoid arthritis (D) in China, India, Russian, Brazil and South Africa from 1990 and 2019. APC, age–period–cohort.
APC analysis on DALYs from BRICS in AD
Figure 3 presents the impacts of age (as rates per 100 000) and cohort and period effects (as relative risk, rate ratio) on the DALYs trends of IBD in all BRICS countries. The APC study findings indicated that the entire APC model was the best-fitting model for IBD DALYs in these countries. This model effectively captures the influences of APC on the disease burden of IBD. In China, India, Russian and Brazil, the age rate ratio of DALYs for (IBD) shows an exponential increase in the age group of 20–69 years. Additionally, in Russia, the rate ratio values for IBD DALY rates show a slight declining trend in the age group of 55–69 years. On the other hand, in South Africa, the relative risk values for IBD DALY rates exhibit a fluctuating upward trend across all age groups (p<0.001 for all age groups).
Disability-adjusted life-years (DALYs) relative risks of inflammatory bowel disease due to age, period and cohort effects for BRICS countries. BRICS, Brazil, Russia, India, China and South Africa.
For period effects, In Brazil and Russia, the period effect rate ratio (relative risk) for IBD DALY rates shows a similar pattern of a slight decline followed by a slight increase. Between 1994 and 2004, South Africa experienced a substantial increase in period relative risk for DALY rates, followed by a downward trend from 2004 to 2019.When comparing the period relative risk for DALY rates in 2019 to the same period cohort in 1994, China and India exhibit opposite trends for IBD. China’s period relative risk increased to 1.19, indicating a higher burden, while India’s period relative risk decreased to 0.83, suggesting a lower burden in 2019 compared with 1994.
The cohort relative risk (RR) values for IBD DALY rates in each cohort show an overall declining trend, with India, Russia and Brazil’s cohort relative risk values almost overlapping. China exhibits the lowest relative risk value, indicating a relatively lower burden of IBD in the cohort analysis. However, in South Africa, the relative risk values show an upward trend in the age group of 1939–1958, indicating an increased burden for that specific cohort. Despite varying peak risks in different countries, the curves for all groups still exhibit convergence (online supplemental table 1).
Supplemental material
The age relative risks of DALYs for MS in BRICS countries are shown in figure 4. The peak values for MS in different countries occur at different age groups. In Russia, the peak value for MS DALYs is in the age group of 45–49 years, with an relative risk value of 5.13. On the other hand, the peak values for the other four countries (China, India, Brazil and South Africa) occur in the age group of 55–59 years, with Brazil having the highest relative risk value among them.
Disability-adjusted life-years (DALYs) relative risks of multiple sclerosis due to age, period and cohort effects for BRICS countries. BRICS, Brazil, Russia, India, China and South Africa.
We found that the relative risks of period effects for MS in 2019 (relative risk=0.94 for China, relative risk=1.38 for India, relative risk=0.77 for Russian, relative risk=1.38 for Brazil, relative risk=1.20 for South Africa).
The cohort relative risks in the five BRICS countries for MS DALYs showed an overall downward trend with the birth cohort. However, there was a notable difference in the trend observed in Russia compared with the other countries. In Russia, the cohort relative risks increased in the 1949–1963 birth cohorts. In contrast, the cohort relative risks slowed down in the other birth cohorts in online supplemental table 2.
The coefficients of age effect on Psoriasis DALYs showed different patterns across the BRICS countries. In China, India and South Africa, the relative risks for psoriasis DALYs increased from the 20–24 age group to the 50–59 age group. However, in Russia and Brazil, the age relative risks peaked in the 60–64 age group (figure 5).
Disability-adjusted life-years (DALYs) relative risks of psoriasis due to age, period and cohort effects for BRICS countries. BRICS, Brazil, Russia, India, China and South Africa.
The period effect on DALY rates in BRICS countries displayed varying trends among different countries. In China and South Africa, the period effect showed a slight decline over time. In India and Brazil, the period effect exhibited a slight increase. Meanwhile, in Russia, the period effect first experienced a slight decline from 1994 to 2004 and then showed a slight increase from 2004 to 2019. The cohort effect on DALY rates for psoriasis in all these countries (China, India, South Africa, Brazil and Russia) displayed an ongoing decreasing trend from the 1929–1933 birth cohort to the 1999–2003 birth cohort in (online supplemental table 3).
The APC analysis of RA in BRICS countries revealed several important trends in figure 6. First, there was an increasing age trend, indicating that the burden of RA DALY rate tends to be higher in older age groups. Specifically, the five countries (China, India, Russian, Brazil, South Africa) relative risks of DALYs due to RA increased significantly by 7.98, 16.07, 5.98, 3.19, 9.13 times, from the age group of 20–24 years to age group of 65–69 years.
Disability-adjusted life-years (DALYs) relative risks of rheumatoid arthritis due to age, period and cohort effects for BRICS countries. BRICS, Brazil, Russia, India, China and South Africa.
Second, the period effect showed a continuous increase in DALY rates of RA. Over the period from 1994 to 2019, the period relative risks increased over time in these countries.
Lastly, the cohort effect demonstrated a decreasing trend from the 1929–1933 birth cohort to the most recent birth cohorts in BRICS countries in (online supplemental table 4).
Discussion
In this study, we used the global data from GBD 2019 to describe the burden of RA, IBD, MS, psoriasis across different APC group across the BRICS countries. According to the results of GBD 2019, the number of psoriasis incidence is the most in the BRICS countries, followed by the number of incidences cases of RA from 1990 to 2019. It is encouraging that since 1990, the ASPR of psoriasis in the BRICS countries has decreased.
Inflammatory bowel disease
Within the BRICS, there are significant differences in the long-term trends and age of four ADs between five countries. Almost three-quarters of the world (about 350 million) live in developing countries, nearly 270 million people live in India and China.19 Therefore, even if the incidence of chronic diseases with a low rate of mortality such as IBD has increased, and it will cause destructive effects on developing countries in the next few years.20 Studies are consistent with our research, with the development of the economy and changes in lifestyle, the prevalence of IBD has been rapidly increasing since 1995 in China.21 22 Research indicates that the continuous development of industrialisation and urbanisation in our country has made it easier for people to suffer from IBD.23 24 Simultaneously, the imperfect medical security system and uneven medical conditions in China exacerbate the economic pressure and difficulties in seeking medical treatment for IBD patients.25 As a result, the rising incidence of IBD will impact the quality of life and lead to significantly increased hospitalisation expenses, thus substantially increasing the disease burden.15 26 In China, the relative risk values of IBD and RA show an exponential increase with age, whereas psoriasis and MS demonstrate an initial rise followed by a subsequent decline. The relative risks of IBD reaches its highest peak in the age group of 65–69 years, which is consistent with global research. The incidence rate of IBD varies depending on different age groups.27 The relative risks of IBD’s DALY increases with age. In the cohort effect, the DALY relative risks of premature queue are relatively high. With the ageing of the Chinese population, the high mortality rate of the elderly IBD patients leads to the high burden on IBD’s disease.
The incidence and prevalence of IBD are continuously increasing in India. It was only after the 1980s, with the widespread use of colonoscopy, that IBD began to receive significant attention in India.28 The IBD of Western countries is distributed in double peaks, with peaks 20–39 years and 60–79 years. The age distribution of India is similar to other Asian countries. There is no double peak distribution, showing an upward trend.29 Studies have shown that patients with IBD in Russia are closely associated with genetic factors.30 Many studies have indicated that the incidence and prevalence of IBD in Brazil have significantly increased, with a higher number of incident cases in developed regions.31 The rising number of new cases and the prevalence showed a continuous growth in IBD patients in Brazil.32 However, our research results indicate a slight downward trend in the prevalence of IBD in Brazil. Recently, a preliminary analysis of South Africa patients showed an exponential increase in IBD over the past 70 years. Among the Indian community, IBD showed a relatively low prevalence.33
Multiple sclerosis
The age effect of MS in BRICS countries follows a unimodal distribution, and the cohort effect shows improvement in the relative risk values. Research has found that the peak incidence of MS occurs between the ages of 40 and 49. Recently, there has been a shift in the peak onset from the age range of 40–49 to the age range of 30–34.34 Our study also observed a rapid increase in relative risks of MS starting from the age of 40. Consistent with our research results, the number of patients with MS in China seems to have not changed significantly in the past 30 years, and it is significantly lower than Europe and the USA.35 The prevalence of MS in China is 1–5 individuals per 100 000 people.36 According to hospital statistics in India, the prevalence of MS has increased from 1.33 per 100 000 individuals to 8.35 over the past 30 years, its average age of onset is 38.3 years.37 The higher prevalence of MS in Russia compared with other countries may be attributed to the fact that residents in high latitude and high altitude regions are more susceptible to developing MS.38
Psoriasis
The age effect of psoriasis in BRICS countries exhibits an exponential increase followed by a decline. The period effect and cohort effect show a decrease in the relative risks. Studies have indicated that psoriasis can occur at any age, with peak incidences typically observed around 20–30 years old and 50–60 years old.39 However, our results show a peak incidence at the age range of 5–60 years. Furthermore, due to major risk factors such as air pollution and tobacco, the Chinese government has implemented a series of control measures targeting ADs, including psoriasis .40 Since 1990, environmental air pollution has gradually improved in most parts of China, and there has been a reduction in exposure to environmental and indoor air pollution in rural areas.41 As a result, the prevalence of psoriasis in China is also showing a downward trend. In India, the prevalence of psoriasis is 0.44%–2.8%, typically affecting individuals in their thirties or forties.42 In Russia, the prevalence of severe psoriasis is higher among patients aged 45 and above. Having severe disease or comorbidities further worsens the quality of life for these patients.43
Rheumatoid arthritis
Exposed to high concentration PM 2.5 and NO2 are related to the increase in admission risks of China RA.44 In India, the age relative risks value of RA has increased rapidly, and it may be in the world’s second-largest population (1.2 billion). As the population increases and ages, the age of RA in India has also risen.45 In Brazil, there are socioeconomic disparities in the incidence of RA, which are related to environmental factors, lifestyle and disease development risks.46 A meta-analysis revealed that the overall prevalence of RA in South Africa is 2.5%.47 The prevalence of RA in South Africa is significantly higher than that in other BRICS countries. Due to the low early diagnosis rates of RA in South Africa, and with increasing attention to population health, the burden of non-communicable diseases is gradually rising. Regional disparities and socioeconomic status differences also present challenges for the prevention and treatment of RA in South Africa. There is a need to further focus on patients with ADs in South Africa.48
Different countries exhibit varying trends in ADs over different periods. These disparities in the temporal trends of ADs among BRICS countries may be attributed to differences in the underlying inflammatory mechanisms and complex interactions between exposure to specific environmental factors and the development of ADs.14 49
There are several limitations in this study. The research results based on GBD depend on modelling data to a large extent, so the quality of raw data collected from different countries will be uneven. Second, the geographical areas of various BRICS countries are vast and there are many ethnic groups. The amount of data that can be used to evaluate the frequency differences between different provinces and nations of various countries is limited.
Conclusion
In conclusion, this study described the current status of four ADs, namely RA, IBD, MS and psoriasis, in five BRICS countries. It also analysed the APC trends in disease burden measured by DALYs. Our findings indicate that, with the exception of Russia, the incidence of ADs in the remaining four BRICS countries has been on the rise. There is notable variation in the prevalence of different ADs, the prevalence of psoriasis has decreased across all five countries. Additionally, we found that the age effect of various ADs is heavily influenced by population ageing. Moreover, each country exhibited potential disparities in the prevention, management and treatment of ADs. These findings underscore the importance of further prioritising the management of autoimmune ADs in BRICS countries and other developing nations.
Data availability statement
Data are available in a public, open access repository. The list of data sources used is publicly available at the Global Health Data Exchange website (http://ghdx.healthdata.org/gbd-results-tool). No additional data are available.
Ethics statements
Patient consent for publication
Acknowledgments
We appreciate the works by the 2019 Global Burden of Disease study collaborators.
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
FZ and YC contributed equally.
Contributors YC and FZ conceptualised and designed the study, drafted the initial manuscript and critically reviewed and revised the manuscript. XG reviewed and revised the manuscript. XG and YC reviewed and revised the manuscript. Both authors approve the submitted version.
Funding This work was funded by the National Science Foundation of Hunan Province (2021JJ40367), the National Science Foundation of Hunan Provincial Health Commission (202212055647), the open project of Hunan Normal University School of Medicine (KF2021016).
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.