Article Text
PDF
Abstract
Background Rheumatoid arthritis (RA) can affect women of childbearing age. The management of patients with RA during pregnancy has evolved over the past decades, especially with the availability of new therapeutic molecules.
Objectives To describe pregnancy in women with RA, to compare pregnancy outcomes with those of women in the general population and to compare pregnancy outcomes in women with active and inactive RA.
Methods Using the French National Health Data System, we identified all pregnancies ending between 2010 and 2020 in patients with and without RA. Characteristics were described. Active RA was defined by conventional synthetic/biological/targeted synthetic disease-modifying antirheumatic drug initiation, systemic or intra-articular corticosteroid administration and/or RA-related hospitalisation. Pregnancy outcomes were compared computing multivariable logistic marginal regression model using generalised estimating equation (GEE).
Results We included 11 792 RA and 10 413 681 non-RA pregnancies. Among RA pregnancies, 74.5% ended in live births and 0.4% in stillbirths. RA pregnancies resulted more frequently in preterm births (adjusted OR (ORa) 1.84; 95% CI 1.69 to 2.00) and very preterm births (ORa 1.43; 95% CI 1.20 to 1.71), low birth weight (ORa 1.65; 95% CI: 1.52 to 1.90), caesarean section (ORa 1.46; 95% CI 1.38 to 1.55) and pregnancy-related hospitalisation (ORa 1.30; 95% CI 1.22 to 1.39). Disease activity decreased during pregnancy. Active RA had higher rates of prematurity (ORa 2.02; 95% CI 1.71 to 2.38), small for gestational age (ORa 1.53; 95% CI 1.28 to 1.83) and caesarean section (ORa 1.25; 95% CI 1.11 to 1.40) than non-active RA.
Conclusion Pregnancies in women with RA were associated with more adverse outcomes, especially if the disease was active. These findings should encourage physicians to closely monitor RA during this crucial period.
- Arthritis, Rheumatoid
- Antirheumatic Agents
- Epidemiology
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information. All relevant data are reported in the article. Additional details can be provided by the corresponding author upon reasonable request.
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
Rheumatoid arthritis (RA), the most common chronic inflammatory rheumatic disease, can affect women of childbearing age.
The management of patients with RA during pregnancy has evolved over the past decades, especially with the availability of new therapeutic molecules.
WHAT THIS STUDY ADDS
RA-related pregnancies more frequently end in preterm births, low birth weight, caesarean sections and pregnancy-related hospitalisation.
Active RA appeared to be associated with prematurity, small for gestational age and caesarean section.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These results underscore the importance of preconception counselling, pregnancy planning and RA activity monitoring throughout pregnancy in women with RA.
Introduction
Rheumatoid arthritis (RA) is the most common chronic inflammatory rheumatic disease. It is characterised by painful and swollen joints and is a major public health problem. Data on the prevalence of RA in adults vary considerably between studies and countries, ranging from 0.00% to 2.70%.1 In France, its prevalence was recently estimated at 0.47%.2 There is a clear female predominance, with women being 2.6 times more affected than men. Although the peak age of onset is around 50 years, RA can also affect women of childbearing age.3
Pregnancy is a particular period, which may be associated with a spontaneous but transient improvement in rheumatic activity despite reduced rates of medication use compared with before conception. A recent meta-analysis showed that disease activity improved in 60% of patients with RA in pregnancy and flared in 47% post partum.4 On the other hand, RA can have a negative influence on pregnancy and may lead to adverse birth outcomes, such as prematurity, caesarean sections and low birth weight.5 This would appear to be particularly significant in active RA.6
The management of these patients has evolved over the last decade, especially with the development of new therapeutic molecules. Recent European Alliance of Associations for Rheumatology recommendations on the use of antirheumatic drugs during pregnancy emphasise the need to adjust treatment in patients planning a pregnancy to promote disease control in the mother while limiting risks to the fetus. Indeed, only a limited number of drugs available in the RA therapeutic arsenal meet these conditions. Compatibility with pregnancy has been found for sulfasalazine and glucocorticoids, but not for methotrexate, which should be discontinued before conception because of its proven teratogenicity. Among biologics, tumour necrosis factor inhibitors (TNFi) are best studied and appear reasonably safe during pregnancy.7 Insufficient documentation regarding fetal safety implies the discontinuation of other therapeutic classes (interleukin-6 inhibitors (IL-6i), cytotoxic T lymphocyte-associated protein 4 immunoglobulin (CTLA4-Ig), anti-cluster of differentiation 20 (CD20) antibody, Janus kinase inhibitors (JAKi)) before a planned pregnancy.8
Despite various international efforts, knowledge regarding pregnancy in patients with RA is still limited. A few studies have been conducted on pregnancy outcomes in women with RA, but they were restricted by the small number of patients included and most were conducted before the era of widespread use of targeted therapies.6 9–13
Our main objectives were to describe pregnancy in women with RA from a recent large national cohort, and to compare pregnancy outcomes with those of women in the general population. Our secondary objective was to compare pregnancy outcomes in women with active RA with those with inactive RA.
Methods
Data source and study design
This nationwide cohort study was based on data of the French National Health Data System (Système National des Données de Santé (SNDS)).14 This database contains individualised anonymous health data and covers 98.8% of the French population (over 67 million individuals). Each person is identified by a unique, anonymous number. In France, all citizens have free, equal, universal access to healthcare for costly chronic diseases. The SNDS contains exhaustive data for all reimbursements for health-related expenditures and outpatient medical care and nursing care prescribed or performed by healthcare professionals, together with demographic data, such as year of birth, sex, area of residence, degree of social deprivation in the geographical area15 and vital status. The data on all pharmacy-dispensed medications include the date of prescription delivery, the type of formulation and the quantity delivered. The SNDS also contains information on patient eligibility for fully reimbursed care related to severe, costly chronic diseases (long-term diseases (LTDs)), such as RA, encoded according to the International Classification of Diseases, 10th Revision (ICD-10).16 The SNDS provides detailed medical information concerning all admissions to French public-sector and private-sector hospitals (French National Hospital Discharge Database: Programme de Médicalisation des Systèmes d’Information), including the dates of hospital admission and discharge, the ICD-10 code at discharge, the medical procedures performed in hospital and costly drugs (such as biological/targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs)) administered in hospital. This large database has been used to conduct several pharmaco-epidemiological studies.17–20
This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline.
RA and pregnancy identification
All patients identified with RA between 1 January 2010 and 31 December 2020 were detected. RA patients included: (1) all inpatients (primary, related, associated diagnosis) admitted with the ICD-10 diagnostic codes for RA (M05 and M06, except M06.1, which corresponds to adult Still’s disease) and (2) all patients with an LTD status for RA. The identification period for hospitalisation corresponds to the maximum data storage period available in the databases, that is, extending back to 2006.
We used a previously developed and used algorithm to identify pregnancies ending between 1 April 2010 and 31 December 2020 among women aged 15–49 years in the SNDS (online supplemental table 1).21 22 Briefly, the pregnancy outcome date was the date of delivery or termination of pregnancy, or, when missing, the date of admission for pregnancy completion. The date of conception was calculated using the pregnancy outcome date and the gestational age, or, when missing, the date of the last menstrual period entered by the physician at the end of the pregnancy. Information on pregnancy duration was available for all pregnancies ending in live births, stillbirths and therapeutic abortions after 22 weeks of amenorrhoea. For other abortions and ectopic pregnancies, the pregnancy duration was unknown for 27.1% of pregnancies, and was therefore imputed by the median pregnancy duration for pregnancies with the same outcome and known duration.23 Neonatal data were available in the SNDS. Linkage between maternal and neonatal data has been possible by means of a common identifier shared by the mother and her child and present in both the delivery stay and the birth stay.21
Supplemental material
Women with RA diagnosed before or during pregnancy were included in the RA pregnancy cohort, and women without RA before or during pregnancy were included in the non-RA pregnancy cohort. Pregnant women with no healthcare record in the SNDS database during the 2 years preceding pregnancy were excluded (less than 1% of the population). The non-RA pregnancy cohort was composed of all the 10 413 681 pregnancies of women without RA, aged over 15 and under 49 years, who had at least one healthcare utilisation during the 2 years before the pregnancy.
Outcomes
Pregnancy outcomes were categorised into six groups: deliveries including live births or stillbirths (delivery of a dead fetus with a gestational age ≥22 weeks), therapeutic abortions after 22 weeks, elective/therapeutic abortions before 22 weeks, spontaneous abortions (death of a fetus with a gestational age <22 weeks), ectopic pregnancies and other pregnancies (hydatidiform mole and other abnormal products of conception). Before 22 weeks, it is impossible to distinguish between elective and therapeutic abortions on the basis of the information available in the database. Other outcomes of interest included prematurity (births occurring before 37 weeks were considered to be preterm and those occurring before 32 weeks were considered to be very preterm), birth weight (a weight <2.5 kg was considered small birth weight and >4.0 kg was considered high), birth weight for gestational age (newborns below the 10th percentile computed in the national pregnancy cohort were considered small for gestational age and those above the 90th percentile as large for gestational age), type of delivery (vaginal or caesarean) and pregnancy-related hospitalisation (ie, hospitalisation during pregnancy for a pregnancy complication, except for hospitalisation for termination of pregnancy).
As disease activity was not directly available in the database, we used a proxy based on healthcare consumption to define it. Active RA was defined by initiation of conventional synthetic DMARD (csDMARD) and/or initiation of b/tsDMARDs and/or at least one specific systemic or intra-articular corticosteroid dispensation, and/or hospitalisation related to an RA flare (hospitalisation with a main diagnosis code related to RA). The initiation of a csDMARD was defined as the delivery of one of the study treatments after a period of at least 6 months without reimbursement of any of these molecules. An equivalent definition was used for the initiation of a b/tsDMARD. The systemic corticosteroids studied were those commonly prescribed in RA in France: prednisone, betamethasone per os, methylprednisolone and prednisolone. The csDMARDs studied were methotrexate, leflunomide and sulfasalazine. The bDMARDs studied (biological originator and biosimilar) included adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab for the TNFi class, tocilizumab and sarilumab for the IL-6i class, abatacept as a CTLA4-Ig, and rituximab as a monoclonal anti-CD20 antibody. The tsDMARDs studied included baricitinib and tofacitinib for the JAKi class. Upadacitinib and filgotinib, JAKi that received marketing authorisation for RA in April and October 2020, respectively, were not studied. Drugs were identified by Anatomical Therapeutic Chemical codes in outpatient databases and Unité Commune de Dispensation codes in hospital discharge databases. RA activity was approximated during seven trimesters: two trimesters before the beginning of pregnancy (T-2: day −182 to day −92; T-1: day −91 to day −1), each trimester of pregnancy (T1: day 0, that is, fertilisation, to day 90; T2: day 91–day 181; T3: day 182 to delivery −1) and two trimesters after the end of pregnancy (T+1: delivery to delivery +91; T+2: delivery +92 to delivery +182). The day of delivery was included in the post-pregnancy trimester (T+1).
RA and pregnancy characteristics
Baseline maternal characteristics included: age, income, complementary universal health insurance status (free access to healthcare for people with low income) and deprivation index.15 Income was calculated from the woman’s salary during the 3 months before maternity leave and was only available for pregnancies for which women received maternity benefits (ie, pregnancies ending in a delivery among general health insurance scheme beneficiaries). Gravidity, defined as the number of times the woman had been pregnant during the study period, was collected. Assisted reproduction was defined as any of the following procedures performed during the 2 months preceding pregnancy: ovarian stimulation, oocyte retrieval, artificial insemination or in vitro fertilisation. The clinical setting of termination of pregnancy was recorded: university hospital, general hospital, private hospital or outpatient procedure. The time since first RA-related LTD or first hospitalisation diagnosis was used as a proxy for RA duration. The number of RA-related hospitalisation in the 6 months prior to pregnancy, and the type and number of RA treatment dispensed during pregnancy (ie, between fertilisation and delivery) were also reported.
Statistical analyses
Quantitative variables are reported as means±SDs or median and IQR and categorical variables as frequencies (percentages). The unit of analysis was a pregnancy. To compare pregnancy outcomes among pregnancies resulting in delivery in the RA and non-RA cohorts and given that the observations were not independent for women who delivered more than once during the study period, we performed multivariable logistic marginal regression models using generalised estimating equation for each pregnancy outcome to estimate adjusted ORs (ORa) and their 95% CIs. Variables entered into the multivariable models were age at the beginning of pregnancy, income (none, <€2000/month, ≥€2000/month or missing data), complementary universal health insurance status, gravidity, place of end of pregnancy, assisted reproduction and study period (categorised into four periods according to the date of the end of pregnancy: 2010–2012, 2013–2015, 2016–2018 and 2019–2020). Similar models were also performed among patients with active RA during the 6 months before pregnancy and during pregnancy, compared with patients with inactive RA over the same period.
To assess the robustness of the estimated ORa, we conducted the following sensitivity analyses: (1) considering only the first pregnancy per woman over the study period; (2) excluding multiple pregnancies; (3) excluding pregnancies with extreme maternal age (considering only pregnancies of women aged 18–35 years). Finally, we performed prespecified subgroup analyses according to the mother’s age (15–24 years, 25–34 years and 35–49 years) and by year of pregnancy end (2010–2012, 2013–2015, 2016–2018 and 2019–2020).
All tests were two tailed with a limit of significance of p<0.05. All analyses were performed with SAS software V.9.4 (SAS Institute).
Results
Study population
A total of 10 735 583 pregnancies were identified in the database between 1 April 2010 and 31 December 2020. After applying our exclusion criteria, 10 438 004 pregnancies were detected, including 24 323 in women with RA. After excluding those with a diagnosis of RA after delivery, 11 792 pregnancies were included in the RA cohort and 10 413 681 in the non-RA cohort (figure 1).
Flow chart. RA, rheumatoid arthritis.
Maternal and pregnancy characteristics
Maternal characteristics of pregnancies of the RA cohort are presented in table 1. The median age at the beginning of pregnancy was 32 years (IQR 28–36). During pregnancy, 1077 (9.1%) received at least one dispensation of a TNFi and 343 (2.9%) received at least three dispensations (ie, sustained exposure). Details of the number of dispensations per molecule during pregnancy are shown in online supplemental table 2. The most frequently prescribed long-term molecule during pregnancy was certolizumab (in 43.7% of cases), followed by etanercept (22.2%) and adalimumab (16.5%). Sulfasalazine was prescribed in 501 pregnancies (4.2%) and methotrexate in 332 pregnancies (2.8%) in women with RA. In the latter group, 51.5% of pregnancies resulted in live birth, 36.1% in elective/therapeutic abortion before 22 weeks and 6.3% in spontaneous abortion.
Maternal characteristics for each pregnancy in the rheumatoid arthritis (RA) cohort (n=11 792)
Pregnancy characteristics in the non-RA and RA pregnancy cohorts (raw data, unmatched) are presented in table 2. Assisted reproduction concerned 420 (3.6%) pregnancies in women with RA, compared with 2.4% in the non-RA pregnancy cohort. RA pregnancies resulted in 8788 (74.5%) live births, 52 (0.4%) stillbirths, 204 (1.7%) therapeutic abortions after 22 weeks, 1836 (15.6%) elective/therapeutic abortions before 22 weeks, 562 (4.8%) spontaneous abortions and 154 (1.3%) ectopic pregnancies. Among pregnancies resulting in delivery, 1049 (11.9%) were preterm/very preterm births and 2376 (26.9%) required caesarean section in the RA cohort. Among newborns from mothers with RA, the median birth weight was 3190 g (IQR 2830–3510 g); 791 (9.8%) were small for gestational age and 797 (9.8%) were large for gestational age.
Pregnancy characteristics in the non-rheumatoid arthritis (RA) and RA cohorts
Comparison of pregnancy outcomes in RA and non-RA cohorts
In multivariate analyses, compared with non-RA pregnancies, those of women with RA more frequently resulted in preterm birth (ORa 1.84; 95% CI 1.69 to 2.00) and very preterm birth (ORa 1.43; 95% CI 1.20 to 1.71), low birth weight (ORa 1.65; 95% CI 1.52 to 1.90), caesarean section (ORa 1.46; 95% CI 1.38 to 1.55) and pregnancy-related hospitalisation (ORa 1.30; 95% CI 1.22 to 1.39). In contrast, RA pregnancy less often resulted in large birth weight (ORa 0.72; 95% CI 0.64 to 0.81). There was no difference in stillbirth between the two cohorts (ORa 1.04; 95% CI 0.76 to 1.41) (table 3).
Comparison of pregnancy outcome among live births and stillbirths in rheumatoid arthritis and non-rheumatoid arthritis cohorts (multivariate analyses)
Similar results were obtained in sensitivity analyses (online supplemental table 3). Analyses stratified by maternal age and year of pregnancy end are provided in online supplemental tables 4 and 5 and yielded similar conclusions.
RA activity and pregnancy outcomes
RA activity appeared to be lower during pregnancy, especially in the third trimester (lower proportion of active RA), than before pregnancy. There was a rebound in activity after delivery (higher proportion of active RA) (figure 2). More specifically, only a very small proportion of pregnancies were considered to have active RA due to the initiation of a DMARD (0.9% of pregnancies in the first trimester, 0.5% in the second trimester and 0.9% in the third trimester). In most cases, RA was considered active due to the delivery of a specific corticosteroid (22.4% of pregnancies in the first trimester, 22.7% in the second trimester and 18.0% in the third trimester) (table 1). In multivariate analyses, compared with pregnancies in women with inactive RA, there were significantly more preterm births, low birth weights, small for gestational age and caesarean sections in women with active RA during pregnancy but also in the 6 months before pregnancy. Nevertheless, ORa were generally higher among pregnancies in women with active RA during pregnancy (ORa of 2.02, 1.51, 1.53 and 1.25, respectively) than in the 6 months before pregnancy (ORa of 1.56, 1.52, 1.52 and 1.16, respectively) (table 4).
Proportion of active rheumatoid arthritis before, during and after pregnancy.
Pregnancy outcome among live births and stillbirths by rheumatoid arthritis (RA) activity before and during pregnancy (multivariate analyses)
Discussion
In this nationwide study involving 11 792 RA pregnancies between 2010 and 2020, we were able to provide a detailed description of pregnancy in the targeted therapy era and compare pregnancy outcomes with 10 413 681 non-RA pregnancies from an unselected population. We report further evidence that, compared with pregnancies in women without RA, RA-related pregnancies more frequently end in preterm births and very preterm births, low birth weight, caesarean sections and pregnancy-related hospitalisation. In addition, active RA before and during pregnancy appeared to be associated with prematurity, small for gestational age and caesarean section.
Our results are important because, to our knowledge, this is the largest study on RA pregnancy in a real-world setting. This study confirms the few previous studies and meta-analysis reporting an increased risk of adverse birth outcomes among women with RA.9–11 24–28 In particular, our study showed a significantly higher risk of preterm/very preterm birth in patients with RA (prevalence of 11.9% in the RA cohort vs 6.6% in the non-RA cohort), at a similar prevalence to a recently published US study,12 but slightly lower than that reported in older studies.10 29 This could reflect both the wider use of highly effective treatments (such as targeted therapies) in recent years but also the improved knowledge about pregnancy, leading to changes in rheumatology and obstetric routines. We also observed that patients with active RA during pregnancy were at higher risk of preterm birth and low birth weights for gestational age. These results are in line with three studies involving 152, 657 and 1739 pregnant women with RA, respectively, which found that disease activity was associated with preterm delivery and lower birth weight, independent of other factors (including corticosteroid therapy).6 25 30 The mechanisms leading to preterm birth and growth restriction are poorly understood, but appear to result from a complex interaction between the mother, fetus and environment.31 One hypothesis is that proinflammatory mediators, released in women with active RA, stimulate prostaglandins, thereby promoting uterine contractions.32 Moreover, lifestyle factors, such as smoking or socioeconomic disadvantage (which may themselves be related to RA activity), may also lead to premature births and small for gestational age.33–35 Furthermore, we found no risk of stillbirth in RA compared with the general population. This result, corroborated by previous studies on smaller samples,11 36 is a reassuring aspect in the current management of pregnant women with RA. Finally, we confirmed a higher frequency of caesarean sections in patients with RA compared with healthy controls (26.9% vs 20.0%).9 10 13 Reasons for caesarean delivery in this population could include fetal distress, high disease activity or elective surgery due to limited range of motion of the hip joints. Nevertheless, as the indication for caesarean section was not available in our database, we were unable to investigate its causes.
This study has several strengths. First, it involved a large number of women from an unselected national cohort, with linkage of maternal and infant discharge records, over a recent study period of 11 years. Furthermore, it used previously developed and validated algorithms to identify pregnancies and RA from a database with a data quality and consistency plan ensuring homogeneous data processing and with information captured during routine medical care and used a previously developed and used algorithm to identify pregnancies.2 14 21 This framework minimises selection bias. This population-based study allowed analysis of rare complications of pregnancy such as assisted reproduction, ectopic pregnancies and stillbirths. Finally, several sensitivity analyses were performed and supported the integrity of our results.
This study also has some limitations. Although we controlled for some possible confounding factors, such as maternal age, income or study period, our analyses are limited by the availability of data on some individual risk factors, such as smoking. Moreover, the database does not contain all the clinical data that would allow an accurate assessment of RA activity. We therefore used proxies to estimate disease activity such as initiation of cs/b/tsDMARDs, specific systemic or intra-articular corticosteroid dispensation, or RA-related hospitalisation. It should be noted that a similar definition of disease activity for another chronic inflammatory disease, inflammatory bowel disease, has already been used from this database.22 However, these proxies are not perfect and may miss flares of RA or reflect other intercurrent events. Furthermore, the rebound in activity after delivery observed in our study may partly reflect the reinitiation of DMARD, since there is a concern to use DMARD during pregnancy. Indeed, 20% of pregnancies initiated a cs/b/tsDMARD in the trimester after delivery, but it is important to note that 25% of pregnancies also received specific corticosteroids and 3% had to be hospitalised due to an RA flare in the trimester following delivery, underlining the resurgence of disease activity. In addition, some spontaneous abortions may have been missed because they did not require hospitalisation or treatments and therefore could not be identified in the database. However, data concerning pregnancies with delivery are comprehensive. Finally, although we have precisely reported the use of DMARDs during pregnancy in France, we did not investigate the impact of these molecules on pregnancy outcomes here, and further studies specifically addressing this issue are needed in the future.
Conclusion
This large-scale population-based study shows that in the era of targeted therapies, pregnancies in women with RA remain associated with more adverse outcomes than pregnancies in women without RA, especially if the disease is active during but also before pregnancy. These findings highlight the importance of preconception counselling and pregnancy planning in these women and should encourage physicians to optimise the monitoring of RA activity throughout pregnancy.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information. All relevant data are reported in the article. Additional details can be provided by the corresponding author upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The French public institution which conducted this study has permanent access to the SNDS database in application of the provisions of Articles R. 1461–12 et seq. of the French Public Health Code and the French data protection authority decision (CNIL-2016-316). No informed consent was therefore required.
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
Twitter @LauraPnVg
Contributors Study idea and concept—AW, RD-S and LPV. Data collection—JD and LPV. Data analysis—LPV. LPV drafted the article and acts as the guarantor of the study. All authors interpreted the results, critically revised the manuscript and approved the final version.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement statement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.