Article Text

Download PDFPDF

Original research article
Public knowledge of cardiovascular disease and response to acute cardiac events in three cities in China and India
  1. Herbert C Duber1,2,
  2. Claire R McNellan1,
  3. Alexandra Wollum1,
  4. Bryan Phillips1,
  5. Kate Allen1,
  6. Jonathan C Brown1,
  7. Miranda Bryant1,
  8. R B Guptam3,
  9. Yichong Li4,
  10. Piyusha Majumdar3,
  11. Gregory A Roth1,5,
  12. Blake Thomson1,
  13. Shelley Wilson1,
  14. Alexander Woldeab1,
  15. Maigeng Zhou4,
  16. Marie Ng1
  1. 1 Institute for Health Metrics and Evaluation, University of Washington, Seattle, USA
  2. 2 Division of Emergency Medicine, University of Washington, Seattle, USA
  3. 3 Gfk, Delhi, India
  4. 4 Chinese Center for Disease Control and Prevention, Beijing, China
  5. 5 Division of Cardiology, University of Washington, Seattle, USA
  1. Correspondence to Dr Herbert C Duber, Institute for Health Metrics and Evaluation, University of Washington, 2301 Fifth Ave, Suite 600, Seattle, WA 98121, USA; hduber{at}


Objective To inform interventions targeted towards reducing mortality from acute myocardial infarction (AMI) and sudden cardiac arrest in three megacities in China and India, a baseline assessment of public knowledge, attitudes and practices was performed.

Methods A household survey, supplemented by focus group and individual interviews, was used to assess public understanding of cardiovascular disease (CVD) risk factors, AMI symptoms, cardiopulmonary resuscitation (CPR) and automated external defibrillators (AEDs). Additionally, information was collected on emergency service utilisation and associated barriers to care.

Results 5456 household surveys were completed. Hypertension was most commonly recognised among CVD risk factors in Beijing and Shanghai (68% and 67%, respectively), while behavioural risk factors were most commonly identified in Bangalore (smoking 91%; excessive alcohol consumption 64%). Chest pain/discomfort was reported by at least 60% of respondents in all cities as a symptom of AMI, but 21% of individuals in Bangalore could not name a single symptom. In Beijing, Shanghai and Bangalore, 26%, 15% and 3% of respondents were trained in CPR, respectively. Less than one-quarter of participants in all cities recognised an AED. Finally, emergency service utilisation rates were low, and many individuals expressed concern about the quality of prehospital care.

Conclusions Overall, we found low to modest knowledge of CVD risk factors and AMI symptoms, infrequent CPR training and little understanding of AEDs. Interventions will need to focus on basic principles of CVD and its complications in order for patients to receive timely and appropriate care for acute cardiac events.

  • Cardiac arrest
  • acute myocardial infarction
  • global health
  • health services

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.


Cardiovascular disease (CVD) is the leading cause of death and disability worldwide. It is estimated that more than 17 million individuals died from CVD in 2013, accounting for more than 13% of global disability-adjusted life years.1 2 China, with 3.7 million CVD-associated deaths, is estimated to have the largest burden of CVD mortality, followed by India, with more than 2.7 million deaths due to CVD.1

Recognising the increasing burden of CVD, WHO adopted a target of a 25% reduction in CVD-associated premature mortality by the year 2025.3 While many countries have made progress and show declining rates of premature mortality from CVD, there is concern that some, including China and India, will not achieve the WHO targets unless significant and rapid changes are made.4

Although identification and modification of risk factors through behavioural changes and medical intervention will be key to reaching these goals, there also exists a large and increasing number of public health, health system and clinical interventions that have been shown to reduce mortality from acute cardiac events.5 Specifically, a sizeable literature has developed around the multitude of interventions demonstrated to improve patient outcomes for individuals who suffer acute myocardial infarction (AMI) or sudden cardiac arrest (SCA).6 7 Both of these acute cardiac events require rapid recognition and coordination of care beginning at the time of symptoms onset. Increasing the general public’s awareness of how to respond in cases of acute cardiac events, improving prehospital systems and ensuring that hospitals provide timely evidence-based treatment are all critical steps in this process.8 9

The Global HeartRescue project, a multistakeholder public health initiative, is an expansion of a similarly titled USA-focused quality improvement project.10 The underlying goal of the Global HeartRescue Project is to improve patient outcomes for those suffering from acute cardiac events (AMI and SCA) by implementing best practices, particularly in the prehospital setting, while simultaneously increasing public awareness of CVD and providing education regarding actions that the general public can take in instances of acute cardiac events. Demonstration projects will be initiated in select locations in India and China.

To inform the demonstration projects, a landscape assessment was completed in three megacities: Beijing and Shanghai, China, and Bangalore, India. In this paper, we present findings from the HeartRescue landscape assessment as they relate to public knowledge of CVD risk factors and AMI symptoms, cardiopulmonary resuscitation (CPR) and automated external defibrillator (AED) awareness and training, and emergency service utilisation.


Setting and population

This was a cross-sectional descriptive study of knowledge, attitudes and practices (KAP) pertaining to CVD risk factors, AMI and SCA symptoms, CPR and AED awareness, and emergency service utilisation. KAP data were collected from the general public, ages 18 and older, through two complementary mechanisms: (1) a structured household survey, and (2) key informant interviews and focus groups. Shanghai, Beijing and Bangalore were selected a priori as potential intervention sites based on numerous factors, including discussions with key health leaders in China and India. Data were collected in Shanghai and Beijing between April and June 2015 and in Bangalore between September and November 2015. Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting of cross-sectional studies were used.11

Study design and instruments

Household survey

Households in Beijing and Shanghai were selected using stratified random sampling to select at least 3000 total households divided evenly among Xicheng and Chaoyang Districts in Beijing, and Changning and Chongming Counties in Shanghai. Within each county/district, streets or townships were stratified into four levels (distant, moderate-distant, nearby and neighboured) according to the distance between the community and the survey hospital. One street or township was selected per stratum. Within each street or township, four residential committees or villages were selected by probability proportional to size, and 48 residents age 18 and older were selected per committee/village using simple random sampling until an age–sex quota was met that matched the age–sex distribution of the county/district.

In Bangalore, a stratified random sampling method was used to select 2400 households from the urban core. First, 80 wards were selected using probability proportional to size. These wards were stratified by Socio Economic Classification (SEC) (high/medium/low), and one polling area representing each SEC strata was then randomly selected from each ward. From the list of households for each polling area, 30 households were selected. One adult per household was then randomly selected from the household roster using the Kish method so as to ensure proper representation of age and sex groups. In case of refusal, the next household was automatically selected using right hand rule. It was ensured that at least 10 respondents were interviewed from each gender.

The instrument used for the household survey asked the selected individual about sociodemographic characteristics, CVD knowledge and emergency health service utilisation. In all locations, the surveys were implemented by trained research assistants who asked survey questions in the local language. Written consent was obtained from all participants and data were collected electronically by trained personnel.

Interview and focus groups

Qualitative data were collected through semistructured individual and focus group discussions with target populations, stakeholders and key informants. In this paper, we focus on findings from community members and patients. In all three cities, focus groups were chosen to maximise variation in sex and age. Individuals invited to participate in the community member focus groups were sampled from the same areas in which household surveys were conducted. Patients were sampled using convenience sampling. In China, interviewers would approach patients outside the hospital. In India, interviewers approached patients in the cardiology department waiting area if the permission of the facility was granted, and outside if permission could not be obtained to enter.

Focus groups and patients were asked to share their experiences and beliefs with respect to health and healthcare in the community, as well as access and behaviour related to acute CVD and general emergency care. After participation, respondents were offered a small monetary or token gift as deemed culturally appropriate by the local research team. Interviews and focus groups were recorded and transcribed. Participant names were not recorded.


Household data were analysed separately for each of the three geographical locations: Beijing, Shanghai and Bangalore. Data on CVD risk factors, AMI symptoms, CPR and AED awareness and training, and emergency service utilisation are described as percentages and fractions of the sampled populations.

We ran several models to assess the sociodemographic determinants of CVD, AMI and CPR knowledge. Given universally low rates of AED recognition and emergency service utilisation, no additional analyses were performed for these outcomes. For CVD and AMI knowledge, we used a Poisson model with a log link function to examine knowledge based on the number of risk factors or symptoms reported by an individual as our outcome of interest. To examine the determinants of CPR awareness, we used logistic regression since this was coded as a dichotomous variable. We ran separate models for each city. Predictors included respondent sex, age and education in all cities. In Bangalore we performed an additional analysis that also included income (this information was not obtained in Shanghai or Beijing based on feedback from local partners). Statistical analyses were conducted using Stata SE V.13.1, and visualised using Microsoft Excel 2013 (Microsoft 2013, Redmond, Washington, USA).

Qualitative data were analysed using the framework approach.12 Initially, a representative sample of interviews and focus group discussions were reviewed. During this process, open coding was used to classify participant responses. After reading and coding the initial group of interviews, a working thematic framework was developed based on emerging themes. The remaining interviews were coded based on this framework. Data were assessed collectively to identify convergent themes, and separately by respondent category to identify additional divergent themes. We present the findings related to patient and community member perceptions of CVD, including acute cardiac events, and emergency healthcare services.

Ethics review

Ethical approval was obtained from local institutional review boards in China (National Center for Chronic and Non-Communicable Disease Control and Prevention, and Chinese Center for Disease Control and Prevention) and India (Centre for Media Studies). All data were deidentified prior to receipt by researchers at the University of Washington.


The household survey was completed by 1524 individuals in Beijing, 1537 in Shanghai and 2400 in Bangalore. Table 1 describes the population of household survey respondents in each city. Consistent with population age patterns, individuals in Bangalore were, on average, younger than respondents in China. Education patterns were similar in Bangalore and Shanghai, while respondents in Beijing were more likely to have attended college/university. In Bangalore, 78% of respondents were uninsured; however, only 3% of respondents in Beijing and <1% in Shanghai reported being uninsured. While qualitative interviews in China suggested that even though insurance coverage approached 100%, many reported that insurance funds ran out quickly if they were using healthcare services, migrant workers were often uncovered in their area of employment, and the perceived quality of care differed by the type of public insurance plan.

Table 1

Household survey respondent sociodemographics

CVD risk factors

Figure 1 shows responses from the household survey in which participants were asked to name as many CVD risk factors as they could. In Beijing, 68% of respondents identified hypertension as a risk factor, and just under half identified high cholesterol, high blood glucose and obesity. In Shanghai these numbers were generally quite similar. Notable differences between these two cities included rates of identifying excessive alcohol consumption (56% vs 39%), physical inactivity (27% vs 15%) and smoking (66% vs 44%).

Figure 1

Household survey respondent knowledge of cardiovascular disease risk factors.

In Bangalore, only 20% of respondents identified high blood glucose as a risk factor, while even fewer named physical inactivity. Hypertension was mentioned by just under half of respondents, while hypercholesterolaemia was noted by 55% of the respondents. Over 90% of individuals surveyed in Bangalore named smoking as a CVD risk factor, the highest percentage of any risk factor identified in all three cities.

In regression analyses, we found education was correlated with CVD knowledge in all three cities (see online supplementary table S1). University/college education was associated with a statistically significant increase in the ability to name a larger number of CVD risk factors, with the most pronounced effect in Shanghai (incidence rate ratio (IRR) 2.95, CI 2.54 to 3.43). Older age was also correlated with the ability to name more risk factors in Beijing, but not in Shanghai or Bangalore.

Supplementary Material

Supplementary material 1

AMI symptoms

Knowledge of AMI symptoms is shown in figure 2. In Beijing, Shanghai and Bangalore, 15%, 18% and 21% were unable to name any AMI symptoms, respectively. The most commonly named symptom in all three cities was chest pain/discomfort, reported by approximately 60% of respondents in each city, although pain/numbness or radiation to back/jaw/neck/arm/hand was listed by 57% of respondents in Bangalore. Other commonly described symptoms such as diaphoresis, dizziness and nausea were mentioned less frequently.

Figure 2

Household survey respondent knowledge of acute myocardial infarction symptoms.

Higher education was again found to be correlated with the ability to name AMI symptoms in Beijing and Shanghai (see online supplementary table S2). However, in Bangalore we found the opposite association, where those who had received a higher education actually named fewer AMI symptoms (IRR 0.88, CI 0.82 to 0.96). This relationship remained when controlling for income as well (IRR 0.91, CI 0.84 to 0.98).

Focus group and patient interviews in Bangalore suggested that relatively little patient–provider interaction focused on AMI symptoms. Interviews with providers and policy experts also suggested this to be true, with many participants stating that this type of education should take place through non-clinic activities. However, members of the lay public with whom we spoke generally felt that there was little public education covering this topic, and most government-sponsored programmes focused primarily on prevention rather than acute CVD events.

CPR and AED awareness and training

Household survey respondent awareness and training pertaining to CPR can be found in figure 3. The proportion of respondents who had never heard of CPR was strikingly large in Bangalore (92%), and very few (3%) were CPR-trained. The proportions of CPR-aware and CPR-trained individuals were greatest in Beijing. Even so, less than one-third of respondents reported any prior CPR training in the Chinese capital. Higher education was found to be highly correlated with CPR knowledge (see online supplementary table S3).

Figure 3

Household survey respondent cardiopulmonary resuscitation (CPR) awareness and training.

In interview and focus groups, respondents who had heard about CPR reported having largely heard about it on TV or in the media. When asked about interest in CPR training, interview and focus group respondents in all cities expressed interest in training, although were apprehensive about administering it due to fear of inadvertently injuring someone, incorrectly administering the procedure and/or potential legal ramifications.

Survey participants were also asked if they knew about AEDs, and if so, whether they would be willing to use them. Given a picture of an AED, the proportion of respondents who recognised the device was 22% in Beijing, 15% in Shanghai and only 6% in Bangalore (figure 4). Remarkably, although recognition was lowest in Bangalore, respondents were substantially more likely to report willingness to use an AED (80%) once they were read a definition and informed of its utility. This was in stark contrast to respondents in Beijing and Shanghai, where only 13% in each city reported willingness to use an AED after being given additional information. Of those who reported having received CPR training in the past, the majority in all three cities said that AED use was not explained during trainings (81% in Bangalore, 67% in Beijing and 80% in Shanghai).

Figure 4

Household survey respondent automated external defibrillator (AED) knowledge.

Emergency service utilisation

Looking at emergency department (ED) and emergency medical service (EMS) utilisation, we find overall low rates of use (table 2). In Bangalore, only 1% of household survey respondents reported using the ED within the past 12 months. Higher rates of ED use were noted in Beijing and Shanghai (6% and 7%, respectively). Interview and focus group respondents reported that a primary reason why people sought care in EDs was because they were open during non-work hours. In Bangalore, interviewees commonly reported first calling or visiting their doctor before going to a hospital to seek care. Interestingly, of those who had received care in the ED in the previous 12 months, respondents in all three cities were much more likely to have self-transported rather than use an ambulance: 80% in Bangalore, 93% in Beijing and 94% in Shanghai. In focus groups, respondents in both cities reported slow response times from EMS services due to traffic. Respondents in Beijing and Shanghai also reported a perceived lack of ambulances due to management issues, a lack of supplies available in EMS vehicles and limited capabilities of prehospital providers.

Table 2

Household survey respondent health service utilisation: percent transported via different mechanisms based on the total number of emergency department (ED) visits


To our knowledge, this is the first study to comprehensively examine KAP around CVD risk factors, AMI symptoms and SCA response in China or India. Overall, we found moderate to low rates of CVD risk factor and AMI symptom knowledge in Beijing, Shanghai and Bangalore. In addition, we found low rates of CPR awareness and training, particularly in Bangalore, and nearly non-existent awareness of AEDs in all three cities. Very low rates of EMS utilisation were also noted. The HeartRescue interventions will need to address at least some of these gaps in order to reduce mortality from acute cardiac events in both China and India.

When comparing our findings with other studies examining CVD risk factor and ACS symptoms knowledge, we note a relative lack of awareness in the three cities we surveyed. Knowledge of key risk factors such as hypertension, diabetes, obesity and physical inactivity was far below that documented in other recent studies.13 14 Interestingly, the CVD risk factor knowledge we observe in Beijing, Shanghai and Bangalore is quite similar to findings from US and Canadian populations in the 1980s and 1990s.15 16 In addition, recognition of common ACS symptoms such as chest pain and shortness of breath was far below that found in Singapore and South Korea.17 18 This lack of knowledge is concerning given prior studies in China and elsewhere demonstrating that inappropriate symptom attribution was a common reason for delays in seeking care for ACS.19–21 Education focused on patients at risk for AMI could be a first step towards improved symptom recognition, fewer delays in care and improved patient outcomes.

Perhaps most notable was the lack of CPR and AED awareness and training. Studies from the USA, Japan and various other countries have generally found much higher rates of CPR and AED awareness and training.22–25 Given the preponderance of evidence showing increased survival among patients who suffer SCA when early CPR and defibrillation are performed, this is clearly an area for future intervention.26 27 In all three locations we saw a strong correlation between CPR knowledge and education, suggesting that interventions focused on training community members with less education, or focusing interventions within primary and secondary schools, may be of benefit.

However, interventions will also need to recognise and address important cultural, social, financial and even legal barriers to performing CPR, using AEDs and accessing emergency care. The lack of willingness to perform CPR and use AEDs even after receiving explanations about them was striking in both Beijing and Shanghai. Furthermore, concerns about quality of care, particularly in the prehospital setting, were rampant, likely contributing to low EMS utilisation. Interventions akin to those focused on quality of care in the hospital setting in China and India could have a significant impact on both quality and confidence.28–30 Concerns about payment, insurance coverage and other financial burdens associated with receipt of emergency care will also need to be addressed if utilisation of time-sensitive services is to be meaningfully increased.

This study should be considered in light of certain limitations. First, all three cities have metro populations greater than 10 million, and our sample sizes were small by comparison. Although we attempted to use a methodologically sound sampling strategy, we were limited by financial and logistical realities, which could have influenced the representativeness of our sample. Second, interview and focus group participants were few and were selected using a convenience sample. As a result, we chose to present these findings qualitatively. Lastly, cultural and language barriers could have influenced the way in which participants understood survey questions. Despite using local partners for data collection and translating instruments into the appropriate language, it is possible that some answers may reflect lack of technical understanding rather than a true knowledge gap.


Knowledge of CVD risk factors and AMI symptoms was modest at best in three of the largest cities in the world with some of the highest burden of CVD. CPR awareness and training was poor, and AEDs were largely unknown. Significant demand-side barriers to seeking emergency care for acute cardiac events will require a dedicated longitudinal commitment by key stakeholders to reduce AMI and SCA mortality. Interventions aimed at improving outcomes for patients with acute cardiac events in these settings will need to improve general public knowledge, bystander responses, access to care and EMS capabilities.

Key messages

What is already known on this subject?

  • Data describing public knowledge of cardiovascular disease (CVD), symptoms of acute myocardial infarction (AMI) and how to respond in cases of acute cardiac events are very limited outside of high-income countries.

What might this study add?

  • In three of the largest cities in the world—Bangalore, India, and Beijing and Shanghai, China—we found low to modest knowledge of CVD risk factors and AMI symptoms. Furthermore, we found very low rates of CPR training and nearly non-existent awareness of automated electronic defibrillators.

How might this impact on clinical practice?

  • In addition to creating systems of care designed to improve outcomes for patients with acute cardiac events, a robust and sustained educational effort to increase public knowledge of CVD and acute cardiac events will be necessary in order to address the growing disease burden of CVD in China and India.


The authors would like to acknowledge the support and insight of our local data collection teams from GFK India and the China CDC. We would also like to thank the HeartRescue implementation partners at the University of Illinois-Chicago, RTI International, MS Ramaiah Medical Center, Arizona Department of Public Health and Shanghai Jiao Tong University for their comments and suggestions. Special thanks to Dr Bentley Bobrow for reviewing the manuscript, Ms Joan Mellor and Ms Lyrae Myxter for their guidance, and the Medtronic Foundation for their continued support.



  • Contributors HCD oversaw all parts of the analysis, interpreted results and wrote the first draft of the manuscript. CRM contributed to development of methods, produced all figures and assisted with interpretation of findings and manuscript writing. AWoll, BP, KA, GAR, BT and AWold contributed to project development, data analysis and interpretation. JCB, MB and SW contributed to project development, management and manuscript writing. RBG, PM, YL and MZ led country-level facility data collection and interpretation of results. MN conceptualised the project and provided guidance on data analysis and manuscript writing. All authors read and approved the final manuscript.

  • Funding This study was supported by a grant from the Medtronic Foundation as part of the Global HeartRescue Project. The funder had no role in study design, data collection, analysis, interpretation of data, writing of the manuscript or the decision to submit for publication.

  • Competing interests None declared.

  • Ethics approval Ethical approval was obtained from local institutional review boards in China (National Center for Chronic and Non-Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention) and India (Centre for Media Studies).

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