Article Text

Original research
Explore the effect of apparent temperature and air pollutants on the admission rate of acute myocardial infarction in Chongqing, China: a time-series study
  1. Xiuyuan Bai1,
  2. Xin Ming2,3,4,
  3. Mingming Zhao4,
  4. Li Zhou4
  1. 1 School of Public Health, Chongqing Medical University, Chongqing, China
  2. 2 Chongqing Health Center for Women and Children, Chongqing, China
  3. 3 Department of quality management section, Women and Children's Hospital of Chongqing Medical University, Chongqing, China
  4. 4 Department of Epidemiology, School of Public Health, Chongqing Medical University, Chongqing, China
  1. Correspondence to Dr Li Zhou; zhouli{at}


Objective Limited research has been conducted on the correlation between apparent temperature and acute myocardial infarction (AMI), as well as the potential impact of air pollutants in modifying this relationship. The objective of this study is to investigate the lagged effect of apparent temperature on AMI and assess the effect modification of environmental pollutants on this association.

Design A time-series study.

Setting and participants The data for this study were obtained from the Academy of Medical Data Science at Chongqing Medical University, covering daily hospitalisations for AMI between 1 January 2015 and 31 December 2016. Meteorological and air pollutant data were provided by China’s National Meteorological Information Centre.

Outcome measures We used a combined approach of quasi-Poisson generalised linear model and distributed lag non-linear model to thoroughly analyse the relationships. Additionally, we employed a generalised additive model to investigate the interaction between air pollutants and apparent temperature on the effect of AMI.

Result A total of 872 patients admitted to hospital with AMI were studied based on the median apparent temperature (20.43°C) in Chongqing. Low apparent temperature (10th, 7.19℃) has obvious lagged effect on acute myocardial infarction, first appearing on the 8th day (risk ratio (RR) 1.081, 95% CI 1.010 to 1.158) and the greatest risk on the 11th day (RR 1.094, 95% CI 1.037 to 1.153). No lagged effect was observed at high apparent temperature. In subgroup analysis, women and individuals aged 75 and above were at high risk. The interaction analysis indicates that there exist significant interactions between PM2.5 and high apparent temperature, as well as nitrogen dioxide (NO2) and low apparent temperature.

Conclusion The occurrence of decreased apparent temperature levels was discovered to be linked with a heightened relative risk of hospitalisations for AMI. PM2.5 and NO2 have an effect modification on the association between apparent temperature and admission rate of AMI.

  • myocardial infarction
  • public health
  • epidemiology

Data availability statement

Data are available on reasonable request. Data are available on reasonable request. Researchers can apply for deidentified data and biomaterial by submitting a proposal to the author LZ.

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:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • This study used the index of apparent temperature to comprehensively analyse the effects of ambient temperature, relative humidity and wind speed on acute myocardial infarction.

  • This study explores the effect modification of air pollutants on the association between apparent temperature and acute myocardial infarction (AMI).

  • Subgroup analysis was used to explore the lagged effect of apparent temperature on AMI in different age and gender participants.

  • This study did not include individual health factors in the analysis.

  • The research has regional limitations.


Acute myocardial infarction (AMI) is commonly acknowledged as one of the most perilous ailments confronting the population. According to previous studies, AMI is a disease in which the coronary artery is blocked and the blood flow is interrupted, causing partial necrosis of the myocardium due to severe and persistent ischaemia.1 2 The relationship between individual factors such as alcohol, specific drugs3 and cigarettes4 5 and the occurrence of AMI has been extensively investigated in recent years.

With the increasing incidence of extreme weather, the incidence of AMI is also increasing, which has prompted researchers to focus on the effect of extreme weather on AMI. Claeys et al found the occurrence of AMI is primarily influenced by the surrounding temperature.6 A study of AMI patients in Beijing observed that low temperatures, especially extreme low temperatures, were responsible for the increased risk of AMI hospitalisations and their lagged effects could last more than 2 weeks.7 According to research using data from the Myocardial Ischemia National Audit Project database in England and Wales, there was a 1.9% rise (0.5% –3.3%, p=0.009) in the likelihood of experiencing a myocardial infarction for every 1℃ increase above a certain temperature threshold.8 Research has shown that both extreme heat and extreme cold lead to a higher occurrence of sudden heart attack.9 The underlying mechanism may be that the increase and decrease of temperature may affect platelet production and blood pressure changes. The core reason is that hypertension damages the coronary artery and forms coronary atherosclerosis, while platelets repair when the atherosclerotic plaque cracks and aggregate to form platelet thrombosis, resulting in myocardial ischaemia.10 11

Furthermore, the impacts caused by air pollutants on AMI have undergone extensive research.12–14 Air pollution comes in various forms, but the most significant ones are fine particulate matter (PM₂.₅), particulate matter 10 (PM₁₀), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), carbon monoxide (CO) and ozone (O₃). The biological mechanisms between air pollutants and AMI have already been expounded by the American Heart Association and European Society of Cardiology,15 16 which include inflammatory response, the activation of sympathetic nerve and direct impacts of atmospheric contaminants on the cardiovascular system.

However, there are also problems with the previous research. It is vital to note that relying solely on environmental temperatures can be misleading when trying to understand how people feel about the weather. Because the apparent temperature takes into account the influence of relative humidity and average wind speed on the human body’s perceived temperature, apparent temperature is a better indicator of how humans perceive temperature and should be considered as a significant factor for AMI.17 18 Additionally, numerous studies have investigated the combined impacts of air pollutants and temperature on human health19–21, but none have delved into the effect modification of air pollutants on the lagged effects of temperature and AMI.

This study aims to investigate the lagged effect of apparent temperature on the admission rate of AMI under the synergistic effect of air pollutants, explore whether the lagged effect is affected by age and gender, and investigate whether air pollutants modify this effect.

Material and method

Study area

Chongqing is situated in the southwestern region of China, specifically within the central and upper sections of the Yangtze River, longitude 105°11'–110°11' east, latitude 28°10'–32°13' north. The climate of Chongqing is a typical subtropical monsoon humid climate, hot and dry in summer and less snow and fog in winter. Chongqing is renowned for being the city with the highest population in China, with a population of 34.15 million in 2021.

Data collection

The data used in this study were obtained from the Medical Data Science Academy Platform at Chongqing Medical University. These data were daily hospital admissions for AMI, covering the period from 1 January 2015 to 31 December 2016. The data were provided by the Second Affiliated Hospital of Chongqing Medical University, the University-Town Hospital of Chongqing Medical University and the Southeast Hospital. In our research, we considered hospitalisation due to AMI (coded as I21 according to the International Classification of Diseases-10) as the primary outcome. We analysed various factors including admission timing, age and gender for each participant. The data used in this study were provided by the Yidu Cloud data platform.

Meteorological data, including daily average temperature, daily average wind speed, daily average relative humidity, daily rainfall and sunshine duration, were provided by the National Meteorological Information Center of China. Air pollutant data are provided by 14 monitoring points under the China National Urban Air Quality Real-time Publishing Platform, which include PM2.5, PM10, SO₂, NO₂, CO and O₃_8h.

Calculation of apparent temperature

Apparent temperature was calculated from the ambient temperature, wind speed and relative humidity. The formula is as follows:

Embedded Image

Embedded Image

T represents the ambient temperature, e denotes the water vapour pressure derived from the combination of relative humidity and wind speed, WS indicates the wind speed and Rh stands for the relative humidity.

Refer to the analysis method of previous literature,7 22 the low temperature was established at the 10th percentile of apparent temperature (P10, 7.18°C), while the high temperature limit was set at the 90th percentile (P90, 32.85℃) in this study.

Statistical analysis

The effect of apparent temperature on AMI has a lag and is affected by many factors. The quasi-Poisson connected generalised linear regression can process the counting data and allow the non-linear relationship between variance and mean. In this model, multiple factors can be considered at the same time. Therefore, a combined model of quasi-Poisson connected generalised linear regression and distributed lag non-linear regression (DLNM) was employed to examine the association between apparent temperature and the frequency of hospital admissions for AMI in this research.23 The DLNM serves as a modelling framework that enables the simultaneous exploration of non-linear and delayed relationships between predictors and an outcome. The equation was as follows:

Embedded Image

Embedded Image

Embedded Image was the expected number of cases of AMI on day t, and α was the intercept number of the model. Embedded Image was the daily mean apparent temperature matrix obtained from the cross-basis function in DLNM and l was the lag day. Embedded Image was a natural cubic smoothing function for nonlinear variables such as time, rainfall and sunshine duration. Based on previous researches, we defined 10 df per year for time trend and 3 df per year for daily mean rainfall and sunshine duration in the final model, respectively.24 25 Embedded Image was the day of the week, Embedded Image represented the dummy variable (0 indicates a non-holiday and 1 indicates a public holiday).

We used the Akaike information criteria to select the maximum lag days. According to the online supplemental table S1, 30 days was selected as the maximum lag days. Then we selected the expose–response relationship between specific apparent temperature (10th percentile, 90th percentile) and AMI to compare with the risk of AMI admission under the median apparent temperature. The temperature effect was expressed as the relative risk of daily admission to hospital for AMI (risk ratio, RR) and 95% CI. To examine the impact of high and low apparent temperatures on different populations, we stratified analyses by age (<75, ≥75 years old) and sex (female and male).

Supplemental material

To assess the resilience of the findings in our investigation, we have taken the following measures: first, we added six air pollutants (PM2.5, PM10, SO2, NO2, CO and O3) to the model in turn. Then, change the df value of time (9 df–11 df), precipitation (3 df–5 df) and sunshine duration (3 df–5 df).

To investigate how air pollutants may modify the impact of apparent temperature on AMI, a generalised additive model with thin plate splines was constructed by the research team.

All analyses were conducted by using the R V.4.3.1 software, employing packages such as ‘dlnm’, ‘mgcv’ and ‘spline’. All statistical tests were two tailed, with a significance level of p<0.05 considered as statistically significant.


Description of study population, meteorological data and air pollutant

All data were obtained from three hospitals, with a majority of patients being male (582 males vs 290 females). The study population had a median age of 77 years, with an average age of 75.38 years. 46% of patients were over the age of 75. Table 1 depicts daily meteorological and air pollutant data for Chongqing from 1 January 2015 to 31 December 2016. Online supplemental figure S1 shows the trend of meteorological data and air pollutant data. The daily changes of temperature and perceived temperature are in high agreement. The changes of wind speed, relative humidity, NO2 and CO have little significant changes throughout the year. The temporal patterns of O3_8h, rainfall, sunshine duration, ambient temperature and apparent temperature exhibit similar trends. Specifically, there is an increasing trend from July to September, followed by a decreasing trend from October to February in the subsequent year. The changes of PM2.5, PM10 and SO2 are in the same direction, with gradual increases from October to February in second year and gradual decreases from July to September.

Table 1

Characteristics of meteorological data and air pollutants in Chongqing during 2015–2016

Correlation between meteorological factor and air pollutant

Figure 1 elucidates correlation coefficient (r) between the meteorological data and air pollutant. There exists an inverse relationship between temperature and average relative humidity, while a direct relationship can be observed between temperature and average wind speed. Rainfall was negatively correlated with all air pollutants. The average relative humidity is negatively correlated with the average wind speed. PM₂.₅, PM₁₀, SO₂, NO2 and CO are positively correlated with each other, in which PM₂.₅ and PM₁₀ have a very close correlation (r=0.95). The correlation between O₃_8h and temperature as well as apparent temperature was found to be positive, whereas it exhibited a negative correlation with other air pollutants and relative humidity. There is a strong correlation between apparent temperature and atmospheric temperature (r=0.99).

Figure 1

Spearman plot of change trend of meteorological data and air pollutants in Chongqing. ARH, average relative humidity; AT, apparent temperature; AWV, average wind speed; NO₂, nitrogen dioxide; SO₂, sulfur dioxide.

The lagged effect of apparent temperature on the overall study population

Revolving around the median temperature (P50, 20.43°C) as a benchmark, we visualised the temporal relationship between apparent temperature and AMI (online supplemental figure S2). We also used heat maps to demonstrate the lagged effect of apparent temperature on AMI (online supplemental figure S3). The low temperature was established at the 10th percentile of apparent temperature (P10, 7.18°C), while the high temperature limit was set at the 90th percentile (P90, 32.85°C). A graph was generated to illustrate the lag impact of both low and high apparent temperature on the incidence of AMI (figure 2 and online supplemental table S2). The lagged effect of low temperature appeared on the 8th–16th days, with the highest lagged effect on the 11th day (RR =1.094, 95% CI 1.037 to 1.153). At higher temperature, we did not observe a lagged effect.

Figure 2

Lag-effects of specific ATs (10th, 7.19°C, 90th, 32.85°C) on AMI hospital admissions, using the median value of AT (20.43°C) as reference. AMI, acute myocardial infarction; AT, apparent temperature.

The lagged effect of apparent temperature on subgroup population

There was a lagged effect of cold weather on AMI in both individuals over the age of 75 and those under the age of 75, with the lagged effect in the elderly group appearing on 10th–16th day (RR 1.115, 95% CI 1.035 to 1.202) and in the groups under the age of 75 at 7 th–11th day (RR 1.122, 95% CI 1.007 to 1.250). Under high temperature conditions, the young group had a protective effect on AMI on 4th–6th day (RR 0.907, 95% CI 0. to 0.99), while the protective effect disappeared in the elderly group. Both men and women had lagged effects under cold conditions, with the relative risk value and lag time of women (RR 1.122, 95% CI 1.007 to 1.250) being higher than men (RR 1.102, 95% CI 1.013 to 1.198); however, there was no lagged effect in both men and women under high temperature conditions. The results are presented in figure 3 and online supplemental tables S3–S6.

Figure 3

Lagged effect of low AT (10th, 7.19°C and high AT (90th, 32.85°C) on the risk of admission to hospital for acute myocardial infarction in different subgroups (reference AT median value: 20.43℃). AT, apparent temperature.

Effect modification of air pollutants in the association between apparent temperature and AMI

By analysing the interaction between air pollutants and apparent temperature, we found that PM2.5 and NO2 have statistically significant interactions with apparent temperature (figure 4). The analysis of interactions reveals significant associations between PM2.5 and high apparent temperature, as well as NO2 and low apparent temperature. Among the effect modification of PM2.5 on the association between apparent temperature and AMI, there was significant variability in the data, with the highest influence observed when both apparent temperature and PM2.5 levels were at their lowest points. Regarding the effect modification of NO2 in relation to apparent temperature and AMI, a clear upward trend was evident in the data. As both NO2 concentration and apparent temperature increased, the effect modification exhibited a pronounced upward trend as well. The peak effect modification occurred when both apparent temperature and NO2 reached their maximum values.

Figure 4

Effect modification of the association between AT and AMI by PM2.5 and NO₂ in Chongqing,2015–2016. AMI, acute myocardial infarction; AT, apparent temperature.

Sensitivity analysis

Sensitivity analysis showed that after adjusting the time df (df=9–11), precipitation (df=3–5) and sunshine duration (df=3–5), the overall impact of apparent temperature on AMI remained consistent, suggesting that the findings were reliable (online supplemental figures S4–S5). When including six air pollutants successively, although the RR values changed slightly after adjusting the air pollutants one by one, the apparent temperature effect did not change much (online supplemental figure S6).


We chose a sample of 872 individuals from 3 medical facilities in Chongqing and used distributed lag non-linear models and generalised linear additive models to investigate the lagged effect of apparent temperature on AMI, as well as the interaction between air pollutants and apparent temperature. Our research discovered a notable association between reduced apparent temperature and increased hospitalisation due to AMI, and women and individuals older than 75 years were at greater risk. The interaction analysis indicates that there exist significant interactions between PM2.5 and high apparent temperature, as well as NO2 and low apparent temperature. The findings of this study have significant implications for the development of health policies aimed at reducing the prevalence of AMI and alleviating the societal and familial burden associated with it.

In our investigation, we observed a lagged effect of reduced apparent temperature on the occurrence of AMI. This finding aligns with prior research conducted in diverse geographical areas characterised by varying climates. In a research conducted in Spain, it was discovered that there is a notable rise in the relative risk of AMI in individuals who perceive lower apparent temperatures.18 A study conducted in Denmark, based on the population, revealed that the winter months were associated with the highest risk of AMI when the lowest apparent temperature was recorded.17 Additionally, studies in Beijing7 and Shanghai26 have found that low temperatures have the greatest impact on admission rates for AMI. When the apparent temperature drops, it can increase the risk of AMI due to two main mechanisms. First, receiving cold stimuli on the skin results in elevated levels of catecholamines in the bloodstream, subsequently inducing vasoconstriction and elevating heart’s blood pressure. These impacts may result in the occurrence of myocardial ischaemia and vulnerability of plaques within the coronary arteries.6 Second, cold air can lead to the onset of pneumonia, and inflammation can in turn induce the rupture of blood vessel plaques, initiating a chain reaction of coagulation.27 In addition, Chongqing, as a non-heating city in China, the risk of cardiovascular disease is also higher in winter compared with heating cities.28 The lack of central heating makes Chongqing’s residents more vulnerable to the external environment. Coupled with the high relative humidity throughout the year in Chongqing, the apparent temperature of the population in Chongqing is relatively low in winter, which increases the risk of AMI.

Multiple studies have consistently demonstrated that the impact of cold weather exhibits a prolonged duration and notable delay in its effects. This study found that the RR continues to increase with a delay time of 30 days and reaches its maximum around the 13th day. Therefore, to prevent AMI, it is necessary not only to take precautions during cold weather, but also to continue monitoring the situation even after the cold weather passes.

In the presence of high apparent temperature, no lagged impact was observed by the research team. Similarly, prior investigations have not established an association between heightened temperatures and AMI.29 30 For example, in a study conducted in major cities in Brazil, researchers discovered that certain cities experienced a delayed impact of cold temperatures on AMI, while high temperatures did not show a lagged effect.31 Nevertheless, several studies have presented contrasting findings. For instance, a research conducted in Brisbane, Australia discovered a notable association between elevated temperatures and hospital admissions for AMI among individuals with pre-existing diabetes (OR 1.19, 95% CI 1.00 to 1.41).32 A research conducted in Tehran has revealed a delayed impact of extreme apparent temperatures, both hot and cold, on the occurrence of myocardial infarction.33 While our study noted a rising pattern in the curve during periods of high apparent temperature, there was no statistically meaningful significance observed. It was speculated that this may be related to the fact that residents in Chongqing tend to stay indoors and use air conditioning more in summer. The reason for the protective effect in younger groups may be that younger people have a stronger metabolism than older people and are better able to reduce the effects of extreme temperatures on themselves.

In subgroup analysis, the elderly are a high-risk group, which aligns with previous studies.34 35 It could potentially be associated with the hardening of blood vessels in older individuals and the dysfunction of their thermoregulatory system, rendering them more vulnerable to temperature fluctuations.36 In terms of gender-specific analysis, women may have a higher risk than men. One possible reason is that the increase in oestrogen-dependent cold-sensitive alpha (2C)-AR expression in women may contribute to the enhanced cold-induced vasoconstrictor activity under conditions of oestrogen sufficiency.37 Simultaneously, the presence of oestrogen can enhance the manifestation of this receptor on the exterior of significant blood vessels. Furthermore, females possess a lower amount of muscular tissue which can result in challenges when it comes to regulating body heat and increase their vulnerability to the impacts of colder temperatures.38

Interaction refers to the phenomenon that the difference in the amount of response between the various levels of a factor changes with the different levels of other factors. Previous research has shown the role of air pollutants in AMI. Claeys et al found that PM2.5 and NO₂ have an interaction with perceived temperature.6 The interaction analysis reveals significant interactions between PM2.5 and high apparent temperature, as well as NO2 and low apparent temperature. In a study conducted in South Korea, researchers have discovered that the admission rate for AMI during the summer is primarily influenced by elevated temperatures and increased levels of NO₂.39 Another study has shown that PM2.5 has the greatest impact on AMI during the winter months.40 Previous studies have indicated that the association between PM2.5 and low perceived temperature could be attributed to the extensive utilisation of coal, oil, diesel or wood as heating sources during colder weather conditions and spending more time indoors in inadequately ventilated spaces. These factors can potentially amplify the impact of PM2.5 particles.41 42 The potential association between the robust interaction of NO₂ and elevated perceived apparent temperature could be attributed to variations in temperature (either high or low) that are associated with inflammatory reactions, increased levels of cholesterol, heightened blood viscosity and enhanced coagulation propensity. Exposure to NO₂ results in oxidative stress and inflammation in the lungs, as well as throughout the body. This leads to an increase in homocysteine levels and inflammatory markers such as interleukin-1, interleukin-6, reactive protein and fibrinogen. Consequently, it contributes to the development of atherosclerosis and the formation of blood clots.43

Our research examined the delayed impact of apparent temperature on AMI and its association with air pollutants. Our study has several advantages. First, the utilisation of DLNM enables a more accurate representation of the non-linear and delayed impacts of the exposure–response association, resulting in enhanced dependability of findings. Second, apparent temperature is selected as the primary independent variable, which better reflects human perception of ambient temperature. More importantly, this study divided the population into subgroups based on age and gender for investigation, allowing for the identification of at-risk populations and the development of targeted prevention measures. Finally, this study also investigates the interaction between air pollutants and perceived temperature, which helps to explore the role of exposure factors on AMI under real-world conditions.

However, there are still some limitations. In the initial phase of data collection, the team focused on examining the overall population, inadvertently overlooking other individual variables like diabetes that exhibit a significant association with AMI. Second, our study did not consider the influence of temperature difference between day and night and seasonal factors on AMI. Nevertheless, we considered the impact of rainfall and duration of sunlight on AMI. Finally, our study was limited to the area with subtropical monsoon humid climate, and therefore, cannot be generalised to regions or societies with similar characteristics. Nevertheless, considering the absence of widespread adoption of central heating in southern China, including Chongqing where of low apparent temperature on AMI as observed in our investigation, it is evident that these research findings hold substantial value for future reference. Furthermore, despite the timeliness issues with the data from 2015 to 2016, analysis of meteorological and air pollutant data for Chongqing from 2017 to 2023, as measured by the China Meteorological Administration, reveals no significant changes. Therefore, this research retains its practical significance.


Low apparent temperature is an important reason for the increased admission rate of AMI. And it has obvious lagged effect, which could last more than 10 days. Females and people older than 75 years old were identified as susceptible subgroups. In our investigation, no significant association was observed between elevated apparent temperature and hospitalisations due to AMI. The relationship between air pollutants and apparent temperature indicates that high levels of PM2.5 exhibit the most pronounced association with low apparent temperature, while elevated NO2 concentrations demonstrate the strongest association with high apparent temperature. In general, our research offers numerical proof regarding the impact of apparent temperature and air pollutants on the occurrence of AMI. We anticipate that the findings of this investigation can serve as a dependable foundation for policymakers to devise efficacious strategies.

Data availability statement

Data are available on reasonable request. Data are available on reasonable request. Researchers can apply for deidentified data and biomaterial by submitting a proposal to the author LZ.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by the Ethics Committee of Chongqing Medical University (No. 2021043).


We thank other researchers in the Department of Epidemiology for providing suggestions and comments for this study. We thank the Medical Data Science Academy Platform at Chongqing Medical University for providing the data of AMI cases.


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.


  • XB and XM contributed equally.

  • Contributors XB: data analysing, writing original manuscripts—review and editing. XM: data analysing, writing original manuscripts—review and editing. MZ: investigation, data curation, resources, validation. LZ: funding acquisition, conceptualisation, supervision, project administration, writing—review and editing. LZ is acting as the guarantor.

  • Funding This study was supported by the funds of the Chongqing Yuzhong District Natural Science Foundation (20210139) and the National Natural Science Foundation of China (Grant No.81773519 to LZ).

  • Competing interests None declared.

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

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.