Acute coronary syndromes presentations and care outcomes in white, South Asian and Chinese patients: a cohort study.

OBJECTIVES: Successful treatment of acute coronary syndrome (ACS) relies on its rapid recognition. It is unclear whether the accepted presentation of chest pain applies to different ethnic groups. We thus examined potential ethnic variations in ACS symptoms and clinical care outcomes in white, South Asian and Chinese patients.
DESIGN: Cross-sectional survey.
SETTING: Participants were hospitalised at 1 of 12 Canadian centres across four provinces. PARTICIPANTS: 1334 patients with ACS (630 white; 488 South Asian; 216 Chinese). MAIN OUTCOME MEASURES: ACS presentation symptoms (classic/typical midsternal pain/discomfort with or without radiation to the left neck, shoulder or arm) were assessed by self-report. Clinical care outcomes (time to emergency room [ER] presentation, cardiac catheterisation; receipt of cardiac catheterisation, percutaneous coronary intervention [PCI] or coronary artery bypass grafting [CABG]) were obtained by health record audit.
RESULTS: The mean age of the sample was 62 years and 30% had ST-elevation myocardial infarction (STEMI). The most common presenting symptom was midsternal pain/discomfort of any intensity regardless of ethnic status. Yet, a substantial proportion of patients reported atypical symptoms (33% white, 19% South Asian, 20% Chinese; p<0.006). After adjustment for age, sex, education, current smoking, extent of coronary artery disease, presence of diabetes or chronic kidney disease and STEMI vs non-STEMI/unstable angina, South Asians were more likely to present with at least moderate intensity midsternal pain/discomfort (adjusted OR [AOR] 1.44; 95% CI 1.05 to 1.98), whereas Chinese were less likely to present with radiating symptoms (AOR 0.53; 95% CI 0.38 to 0.74) compared with whites. South Asians with atypical pain (relative to those with midsternal pain/discomfort) took significantly longer to present to the ER (p=0.037), and were less likely to receive PCI (p=0.008) or CABG (p=0.041).
CONCLUSIONS: Atypical presentations were associated with greater delays in arrival to the emergency department and reduced invasive cardiovascular care in South Asians. © Author(s) (or their employer(s)) 2019. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

This was a rigorously designed study which included participants who spoke a variety of languages and used highly systematic and validated translation and data collection processes.

A number of mechanisms were used to obtain a data regarding participants’ acute coronary syndrome (ACS) symptoms.

Only participants who survived their ACS event were included in this study. It is not possible to know if patients who succumbed to their ACS event or did not seek care had different symptoms than those who were diagnosed and survived.

There was a noteworthy proportion of South Asian and Chinese patients approached to participate who refused.

Introduction

The burden of acute coronary syndrome (ACS) is substantial and it is now the leading cause of death and disability worldwide.1 2 Life-saving therapies for ACS are dependent on early and timely recognition of symptoms. Atypical presentation of ACS symptoms leads to delays in recognising symptoms by both patients and healthcare providers, and can result in misdiagnosis, delayed treatment, receipt of fewer evidence-based therapies and higher in-hospital morbidity and mortality.3

The ‘classic’ ACS presentation, based on the presentation of men of European descent, includes moderate to severe intensity central chest with or without left arm, neck or jaw radiating pain or discomfort.4 5 This description has been adopted internationally, though some studies suggest that ACS symptoms may differ by ethnicity.6–11 Although these studies yielded discordant findings, some noted that non-white patients may present with additional symptoms (eg, gastrointestinal, dyspnoea, nausea)6–9 11 and South Asians may have more diffuse pain and more frequent back pain10 than whites. It is difficult to draw robust conclusions from the current evidence given variable methods of data collection, inadequate cross-cultural language adaptation, small sample sizes and insufficient adjustment for other factors (eg, sex, age, clinical factors) that may influence ACS presentation.

South Asians have a higher incidence of ACS and earlier onset relative to other ethnic groups.12 13 Though Chinese currently have a lower incidence of acute myocardial infarction (AMI) relative to whites and South Asians,14 15 cardiovascular events in China are expected to rise by more than 50% by 2030.16 A rigorous investigation of ACS symptom presentation, time to arrival to the emergency department and clinical care outcomes between these ethnic groups is warranted to ensure appropriate public health messaging and/or inappropriate stereotyping of presenting symptoms. We thus aimed to examine potential ethnic variations in ACS symptoms using a rigorously cross-culturally adapted symptom questionnaire and visual identifiers in a multicentre cohort of white, South Asian and Chinese patients hospitalised in Canada.

Methods

Study design

We conducted a cross-sectional survey of a cohort of white (ie, of European descent), South Asian (ie, descended from India, Pakistan, Bangladesh, Sri Lanka) and Chinese (ie, descended from China, Hong Kong, Taiwan, Macau) patients hospitalised with physician-confirmed ACS (unstable angina, ST-elevation myocardial infarction [STEMI] or non-STEMI5).

Setting and sample

Following institutional ethics review at each participating hospital, participants were recruited consecutively from 12 urban hospitals in four Canadian provinces (>90% of Canada’s immigrants reside in urban areas and the majority reside in these provinces17) between September 2010 and December 2015. Inclusion criteria were age >19 years; admission to hospital with a confirmed diagnosis of ACS (as identified on health record); self-reported ethnicity for either white (European), South Asian or Chinese; and speaking English, Punjabi, Tamil, Urdu, Hindi or Gujarati (among the most common languages of Canadian residents from South Asia),17 Cantonese or Mandarin. Nearly 95% of South Asians and all Chinese residing in Canada would have been able to speak one or more of the languages used in this study.17 Exclusion criteria were multiple ethnic origins (ie, mixed race) or known cognitive deficits (ie, Alzheimer’s disease, dementia, uncontrolled psychiatric disorder), as identified on the health record. All participants provided informed consent to participate in this study.

Patient and public involvement

Patients were not involved in the conception, design or interpretation of this study. However, we have a detailed plan to disseminate this information widely to healthcare providers as well as potential patients, through traditional and social media.

Data collection

Data were collected while participants were in hospital (within the first 5 days), in the patient’s preferred language (as listed above) by highly trained like-speaking research assistants.

As reported elsewhere,18 rigorous translation processes were undertaken for study materials.

Questionnaires were developed based on the current literature and our earlier work.10 19–21 We, like Teoh et al,10 asked participants to mark, on a gender-neutral torso silhouette (pictograph), all the locations of the pain/discomfort that brought them to hospital. Each participant used the same type of pen to ‘colour in’ the areas. Then, the participant was prompted to identify if he/she experienced other symptoms such as nausea, diaphoresis, dyspnoea, dizziness or weakness. Next, they were asked to identify their ‘chief’ or ‘main’ symptom, describe its nature by pointing to pictorial identifiers (ie, stabbing, heavy, shooting, burning, squeezing)9 21 and finally identify its intensity (Likert-type scale, 0–10 where 0 was no pain). Pain/discomfort severity rated as <5 was considered mild and >5 or higher was considered moderate to severe pain.

A health record audit was undertaken to collect additional demographic and as well as clinical data, once the participant had been discharged from hospital.

The presence of classic (typical) ACS presentation was defined as having midsternal or having midsternal with radiating left neck, shoulder or arm pain/discomfort of any or at least moderate intensity. The clinical care outcomes of interest were time to emergency room (ER) presentation and cardiac catheterisation, as well as receipt of cardiac catheterisation, percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG).

There was potential for bias when using this design and data collection method. This was a self-selected group of patients (ie, selection bias) who agreed to participate in the study. Also, the retrospective nature of the design could lead to recall bias given that patients were asked about their symptoms up to 5 days following hospital admission.

Statistical analysis

Descriptive statistics and χ2 tests were used to identify differences in participant characteristics, as well as symptoms and location(s) of pain/discomfort between ethnic groups. Each variable had a ‘not entered’ category. In binary variables of present/not present, ‘not entered’ was assumed to be not present. To analyse presenting symptoms data, a grid was placed over the original torso silhouette to identify the location(s) of pain/discomfort in a standardised manner.10 These data were also presented as proportions of each participant group who identified any pain/discomfort in the various areas as well as moderate to severe pain/discomfort rated as >5/10 on a Likert-type scale. Crude and adjusted OR (AOR) with 95% CI were examined using logistic regression for South Asians and Chinese relative to whites having any midsternal pain/discomfort or any midsternal with radiating left neck, shoulder or arm pain/discomfort, as well as moderate to severe (>5/10) midsternal pain/discomfort or midsternal with radiating left neck, shoulder or arm pain/discomfort. Finally, crude and AOR (as above) with 95% CI were examined to determine the likelihood of receiving cardiac catheterisation, PCI and CABG for each ethnic group, based on having any midsternal pain/discomfort or any midsternal with radiating left neck, shoulder or arm pain/discomfort.

Models were adjusted for demographic and clinical characteristics associated with atypical symptoms (age, sex, education, current smoker, extent of coronary artery disease [based on cardiac catheterisation], presence of diabetes or chronic kidney disease [CKD; as defined by the Charlson Comorbidity Index22], and STEMI vs non-STEMI/unstable angina). All data were analysed using SAS V.9.4.

Pictographic analysis

We used a heat-mapping technology to portray where participants identified that they pain/discomfort on the torso silhouettes.10 Given that the numbers of participants varied by ethnic group, the mapping was based on proportion of participants who coloured areas on the torsos versus the absolute number.

Results

Of 3243 persons screened for eligibility, 1042 did not meet eligibility criteria and 903 patients declined enrolment in the study. Of the final 1334 patients (58.8% men), 47.2% were white, 36.6% were South Asian and 16.2% were Chinese (see table 1).

Table 1

Patient characteristics by ethnicity

	White (n=630)	South Asian (n=488)	Chinese (n=216)	P value
Age, mean (SD), years	65.9 (13.0)	62.2 (12.5)	65.1 (12.6)	<0.0001
Male	276 (43.8%)	355 (72.8%)	154 (71.3%)	<0.0001
English Language Survey	630 (100%)	440 (90.2%)	161 (74.5%)	<0.0001
Immigration status				<0.0001
Canadian-born	455 (72.1%)	11 (2.3%)	10 (4.6%)
Immigrant <20 years	15 (2.4%)	189 (38.8%)	81 (36.6%)
Immigrant >20 years	160 (25.4%)	288 (59.0%)	123 (56.9%)
None documented	1 (0.25)	0	2 (0.9%)
Marital status				<0.0001
Married/Common law	362 (57.4%)	383 (78.5%)	174 (80.6%)
Other	269 (42.7%)	105 (21.6%)	42 (19.5%)
Highest level education				0.0001
None	8 (1.3%)	31 (6.4%)	7 (3.2%)
<High school	94 (14.9%)	83 (17.0%)	41 (19.0%)
>High school	529 (83.9%)	374 (76.7%)	168 (77.8%)
None documented	0	0	0
Current smoker	123 (19.5%)	53 (10.9%)	27 (12.5%)	0.0002
Extent of coronary artery disease				<0.0001
Normal/<50% disease	92 (14.6%)	50 (10.3%)	36 (16.7%)
1–2-vessel disease	148 (23.5%)	91 (18.7%)	33 (15.3%)
3-vessel disease/LAD	272 (43.2%)	293 (60.0%)	109 (50.5%)
Left main	11 (1.8%)	4 (0.8%)	10 (4.6%)
Not entered/missing	107 (17.0%)	50 (10.3%)	28 (13.0%)
Diabetes	193 (30.6%)	224 (45.9%)	64 (29.6%)	<0.0001
Chronic renal disease	74 (11.7%)	45 (9.2%)	18 (8.3%)	0.229
Acute coronary syndrome				0.187
Unstable angina	205 (32.5%)	132 (27.1%)	61 (28.2%)
Non-STEMI	228 (36.2%)	208 (42.6%)	83 (38.4%)
STEMI	188 (29.8%)	145 (29.7%)	68 (31.5%)
Unspecified	9 (1.5%)	3 (0.6%)	4 (1.8%)

LAD, left anterior descending artery; STEMI, ST-elevation myocardial infarction.

Baseline characteristics

Over 72% of whites were Canadian-born whereas the majority of South Asians and Chinese had immigrated to Canada >20 years earlier. Among these three ethic groups, whites were most likely to be a current smoker and have 1–2-vessel disease, South Asians were most likely to have 3-vessel disease and diabetes and the greatest proportion of each patient group (36.2% whites, 42.6% South Asian, 38.4% Chinese) were admitted with non-STEMI.

ACS symptom presentation

The most common presenting symptom across ethnic groups was midsternal pain/discomfort of any intensity (89.3%) followed by left shoulder pain/discomfort (46.6%) (table 2). Most of the pain or discomfort was rated as moderate to severe (>5/10) across ethnicities. More than 78% of all patients reported the classic presentation of at least moderate intensity midsternal pain/discomfort, and more than 58% of all patients reported moderate to severe intensity midsternal chest with any of radiating left neck, shoulder or arm pain/discomfort. In these unadjusted analyses, whites were less likely to report having the classic presentation of at least moderate intensity midsternal pain/discomfort while the Chinese were less likely to report having at least moderate intensity midsternal pain/discomfort with radiation to left neck, shoulder or arm. Extent of coronary artery disease was not associated with having classic symptoms in whites (p=0.973), South Asians (p=0.562) or Chinese (p=0.304) patients.

Table 2

Reported acute coronary syndrome symptoms by ethnicity

	White (n=630)	South Asian (n=488)	Chinese (n=216)	P value
Pain/Discomfort location
Midsternal	559 (88.7%)	438 (89.9%)	194 (89.8%)	0.828
Left shoulder	310 (49.1%)	239 (49.0%)	72 (33.3%)	0.0006
Left arm	206 (32.7%)	140 (28.7%)	22 (10.2%)	<0.0001
Left jaw	122 (19.3%)	67 (13.7%)	21 (9.7%)	0.003
Left neck	114 (18.1%)	69 (14.1%)	19 (8.8%)	0.0006
Midsternal with radiating left neck, shoulder or arm pain/discomfort	435 (69.0%)	316 (63.8%)	123 (57.0%)	<0.0001
Midsternal pain/discomfort with intensity ☹ >5	482 (76.5%)	394 (80.7%)	172 (79.6%)	0.0435
Midsternal with radiating left neck, shoulder or arm pain/discomfort with intensity ☹ >5	385 (61.6%)	285 (58.4%)	111 (51.4%)	<0.0001
Nature of pain/discomfort
Pressure	312 (49.5%)	174 (35.7%)	93 (43.1%)	<0.0001
Squeezing	118 (18.7%)	120 (24.6%)	57 (26.4%)	0.016
Burning	82 (13.0%)	71 (14.6%)	25 (11.6%)	0.533
Stabbing	48 (7.6%)	68 (14.0%)	17 (7.9%)	0.001
Shooting/Moving	30 (4.8%)	39 (8.0%)	11 (5.1%)	0.065
None	40 (6.4%)	16 (3.3%)	13 (6.0%)	0.059
Not documented	0	0	0	–
Other symptoms
Shortness of breath	386 (61.2%)	303 (62.1%)	132 (61.1%)	0.952
Diaphoresis	313 (49.7%)	239 (49.0%)	103 (47.7%)	0.877
Dizziness	329 (52.2%)	221 (45.3%)	99 (45.8%)	0.047
Nausea or vomiting	235 (37.3%)	145 (29.7%)	56 (25.9%)	0.002

The majority of patients described their pain or discomfort as pressure, squeezing or burning (81.3% whites, 74.8% South Asians, 81.0% Chinese), although south Asians reported significantly more stabbing pain than their counterparts. Other symptoms were also prevalent among all ethnic groups: shortness of breath (62.4%), diaphoresis (49.8%), dizziness (49.3%) and nausea/vomiting (33.1%). There were significant differences between groups in reporting these symptoms with whites reporting more dizziness and nausea or vomiting.

The pictographic analysis of the torsos (figure 1) shows that the profiles were very similar for whites and South Asians, whereas the Chinese tended to report more central as well as less back or left arm pain/discomfort.

Figure 1

Pictographic for each ethnic group. The colour gradient (white/purple to red) indicates no/few participants to most (>65%) of participants identified pain/discomfort in that location.

In the analysis stratified by ethnic group (table 2), ethnicity was not significantly associated with reporting midsternal pain/discomfort of any intensity. However, in adjusted (age and sex; age sex, education, current smoker, extent of coronary artery disease, diabetes, CKD and STEMI vs non-STEMI/unstable angina) models (table 3), South Asians were significantly more likely to report having midsternal pain/discomfort of moderate to severe intensity. The Chinese participants were less likely to report having radiating ACS symptoms (ie, midsternal, with left neck, shoulder or arm pain or discomfort) of any as well as moderate to severe intensity, relative to whites in crude and adjusted models.

Table 3

Crude and adjusted OR for ACS symptom presentation

Ethnic group	Any intensity			Moderate to severe intensity
	Crude OR (95% CI)	Adjusted* OR (95% CI)	Adjusted† OR (95% CI)	Crude OR (95% CI)	Adjusted* OR (95% CI)	Adjusted† OR (95% CI)
At least moderate intensity midsternal pain/discomfort
White	1.0	1.0	1.0	1.0	1.0	1.0
South Asian	1.11 (0.76 to 1.63)	1.04 (0.70 to 1.56)	1.04 (0.68 to 1.57)	1.29 (0.96 to 1.72)	1.44 (1.06 to 1.96)	1.43 (1.04 to 1.97)
Chinese	1.12 (0.68 to 1.86)	1.12 (0.67 to 1.89)	1.18 (0.70 to 2.00)	1.20 (0.82 to 1.75)	1.40 (0.95 to 2.07)	1.41 (0.95 to 2.09)
At least moderate midsternal, with left neck, shoulder or arm pain/discomfort
White	1.0	1.0	1.0	1.0	1.0	1.0
South Asian	0.86 (0.68 to 1.10)	0.83 (0.64 to 1.06)	0.81 (0.62 to 1.05)	0.94 (0.74 to 1.19)	0.95 (0.74 to 1.21)	0.92 (0.71 to 1.75)
Chinese	0.44 (0.32 to 0.61)	0.45 (0.32 to 0.62)	0.46 (0.33 to 0.63)	0.50 (0.36 to 0.69)	0.53 (0.38 to 0.74)	0.53 (0.38 to 0.75)

*Adjusted for age and sex.

†Adjusted for age, sex, education, current smoker, extent of coronary artery disease (no disease vs low risk or high risk), diabetes, chronic kidney disease and ST-elevation myocardial infarction (STEMI) vs non-STEMI/unstable angina.

ACS, acute coronary syndrome.

Clinical care outcomes

The mean time to ER presentation, for those who had a distinct time of symptom onset, was 5.53 to 7.41 hours (table 4). South Asians with atypical symptoms had significantly longer delays in arrival to the emergency department than those who had any typical symptoms. The mean time to receipt of cardiac catheterisation following ER presentation was 3–3.93 hours. Whites with atypical symptoms had significantly longer delays in time to receipt of cardiac catheterisation than those with typical symptoms. South Asian participants who had any typical midsternal pain/discomfort were more likely to receive PCI but less likely to receive CABG than those with atypical symptoms. Table 4 also shows that South Asian participants who had any typical midsternal with left neck, shoulder or arm pain/discomfort were more likely to receive PCI than those who had atypical symptoms.

Table 4

Clinical care outcomes by symptoms and ethnicity

	White			South Asian			Chinese
	Typical (n=482)	Atypical (n=148)	P value	Typical (n=394)	Atypical (n=94)	P value	Typical (n=172)	Atypical (n=44)	P value
Any midsternal pain/discomfort	n=451	n=116		n=359	n=71		n=144	n=32
Time to ER presentation (hours; if distinct time of onset) Mean (SD)	6.43 (5.77)	6.11 (5.53)	0.591	5.69 (5.35)	7.17 (5.94)	0.037	5.92 (5.49)	7.41 (6.10)	0.176
Median (IQR)	3.0 (6.5)	3.0 (6.0)	0.872	3.0 (6.0)	5.0 (6.5)	0.127	3.0 (6.0)	6.0 (6.5)	0.318
Time to CATH (hours) Mean (SD)	3.35 (1.76)	3.93 (1.57)	0.001	3.10 (1.80)	3.10 (1.72)	0.975	3.09 (1.82)	3.58 (1.72)	0.134
Median (IQR)	4.0 (2.0)	5.0 (1.0)	0.002	3.0 (2.0)	2.0 (2.0)	0.862	3.0 (2.0)	5.0 (1.5)	0.109
	n (%)	n (%)		n (%)	n (%)		N (%)	n (%)
CATH	402 (83.4%)	124 (83.8%)	0.913	359 (91.1%)	81 (86.2%)	0.148	151 (87.8%)	38 (86.4%)	0.826
PCI	228 (47.3%)	60 (40.5%)	0.149	211 (53.6%)	36 (38.3%)	0.008	81 (47.1%)	19 (43.2%)	0.777
CABG	44 (9.1%)	17 (11.5%)	0.396	38 (9.6%)	16 (17.0%)	0.041	13 (7.6%)	5 (11.4%)	0.636
Any midsternal with left neck, shoulder, or arm pain/discomfort	n=330	n=237		n=234	n=196		n=68	n=108
Time to ER presentation (hours; if distinct time of onset) Mean (SD)	6.41 (5.82)	6.31 (5.57)	0.836	5.53 (5.35)	6.42 (5.60)	0.092	6.0 (5.65)	6.31 (5.62)	0.726
Median (IQR)	3.0 (6.5)	3.0 (3.0)	0.926	2.0 (6.0)	3.0 (6.5)	0.037	3.0 (6.0)	3.0 (6.0)	0.649
Time to CATH (hours) Mean (SD)	3.40 (1.76)	3.61 (1.68)	0.154	3.0 (1.77)	3.2 (1.79)	0.241	3.28 (1.81)	3.14 (1.81)	0.615
Median (IQR)	4.0 (2.0)	5.0 (1.5)	0.192	3.0 (2.0)	3.0 (2.0)	0.202	4.0 (2.0)	3.0 (2.0)	0.635
	n (%)	n (%)		n (%)	n (%)		n (%)	n (%)
CATH	306 (86.0%)	220 (80.3%)	0.058	235 (91.1%)	205 (89.1%)	0.469	65 (82.3%)	124 (90.5%)	0.126
PCI	172 (48.3%)	116 (42.3%)	0.135	143 (55.4%)	104 (45.2%)	0.024	41 (51.9%)	59 (43.1%)	0.361
CABG	33 (9.3%)	28 (10.2%)	0.690	30 (11.6%)	24 (10.4%)	0.675	4 (5.1%)	14 (10.2%)	0.307

CABG, coronary artery bypass graft surgery; CATH, catheterisation; ER, emergency room; PCI, percutaneous coronary intervention.

However, after adjustment, South Asians with atypical symptoms were least likely to receive PCI than those with typical symptoms (table 5).

Table 5

Crude and adjusted OR for clinical care outcomes

Atypical versus typical	Crude OR (95% CI)	Adjusted* OR (95% CI)
Any midsternal pain/discomfort
CATH
White	1.03 (0.62 to 1.69)	1.15 (0.53 to 2.49)
South Asian	0.61 (0.31 to 1.20)	0.66 (0.22 to 1.98)
Chinese	0.89 (0.33 to 2.35)	2.77 (0.64 to 11.90)
PCI
White	0.76 (0.52 to 1.10)	0.73 (0.45 to 1.19)
South Asian	0.54 (0.34 to 0.85)	0.57 (0.32 to 1.00)
Chinese	0.87 (0.45 to 1.69)	1.10 (0.45 to 2.68)
CABG
White	1.29 (0.71 to 2.34)	1.59 (0.81 to 3.09)
South Asian	1.92 (1.02 to 3.62)	1.90 (0.96 to 3.77)
Chinese	1.63 (0.56 to 4.73)	1.76 (0.53 to 5.80)
Any midsternal, with left neck, shoulder or arm pain/discomfort
CATH
White	0.67 (0.44 to 1.02)	0.90 (0.47 to 1.73)
South Asian	0.80 (0.44 to 1.46)	1.37 (0.55 to 3.41)
Chinese	2.08 (0.92 to 4.68)	4.94 (1.41 to 17.31)
PCI
White	0.79 (0.57 to 1.08)	0.84 (0.56 to 1.26)
South Asian	0.66 (0.46 to 0.95)	0.74 (0.47 to 1.18)
Chinese	0.69 (0.40 to 1.21)	0.49 (0.23 to 1.05)
CABG
White	1.11 (0.66 to 1.89)	1.24 (0.69 to 2.21)
South Asian	0.89 (0.50 to 1.56)	0.92 (0.50 to 1.70)
Chinese	1.82 (0.63 to 5.31)	2.10 (0.65 to 6.85)

*Adjusted for age, sex, education, current smoker, extent of coronary artery disease (no disease vs low risk or high risk), diabetes, CKD and STEMI vs non-STEMI/unstable angina.

CABG, coronary artery bypass graft surgery; CATH, catheterisation; PCI, percutaneous  coronary intervention.

Discussion

The classic (typical) presentation of moderate to severe intensity midsternal chest pain or discomfort was the most common presenting ACS symptom across ethnic groups. Yet, a substantial proportion of patients, up to one-third, in each group studied had non-classic (atypical) ACS presentations. Differences in ACS symptom presentation among whites, South Asian, and Chinese patients were statistically and clinically significant, as South Asians were more likely to present with classic ACS symptoms and Chinese were less likely to present with radiating symptoms in all models relative to whites. Our study extends the findings from previous work by employing rigorous cross-cultural and language adaptation of study questionnaires, and including visual diagrams across multiple centres.

Findings from other studies have some similar conclusions. Like the south Asian and Chinese participants in our study, Blacks have been reported to have a similar prevalence of chest pain relative to whites,6–9 while other studies suggest that non-white patients with ACS are more likely to have chest pain than whites. A study from a single hospital in the UK reported South Asians are more frequently to have ‘classic chest pain with radiation’ than whites (90 vs 82%).10 A study of 390 patients with ACS residing in the Asia-Pacific rim revealed patients from India, China and Korea were significantly more likely to report typical pain than their white counterparts.11 A small study of whites, blacks (US residents) and Koreans (Korea residents) revealed Koreans reported significantly more frequent and greater radiation of chest pain.8 Yet, our earlier study from a single health region revealed that South Asian and Chinese patients with AMI were less likely to report classic symptom presentation.19 These inconsistencies may be explained by the controlled analyses we used, and the lack of cross-cultural (and language) adaptation in other studies.

We found significant differences among ethnic groups in the intensity and nature of the pain/discomfort as well as the presence of additional symptoms. Ethnicity and culture may influence patients identifying, recognising and acknowledging urgent symptoms as such, as well as making the decision to seek care. Though not consistent, study findings suggest that people who represent ethnic minorities (who are not residing in their country of origin) have greater pain sensitivity in both clinical and experimental conditions relative to the ethnic majority (eg, whites). Though most of this research has been undertaken in African Americans and Hispanics relative to whites,23 what little research that has focused on south Asians or Chinese suggests that both groups may report less pain tolerance relative to whites.24 25 Single-centre studies have revealed that south Asians and Chinese with AMI tended to report atypical symptoms and did not recognise or accept their symptoms as ‘urgent’.19 26 Research findings suggest that expression of symptoms, and language and semantic differences of symptom descriptors may also account for discrepancies in reporting symptoms. Given that pain/discomfort are subjective experiences, how they are portrayed are culturally bound.23–30 For example, the Worchester Heart Attack Study revealed that south Asian participants were significantly less accepting of overtly expressing pain,21 and both south Asians and Chinese are viewed as more stoic in their presentations.29

We identified little overall ethnic variation in time to emergency room presentation, though south Asians who had atypical symptoms were more likely to have a longer time to presentation to the emergency room (mean of 1.48 hours longer) relative to those who had typical symptoms. In a study of 440 398 registry patients, Ting et al31 demonstrated that delays from symptom onset to ER presentation were associated with reduced likelihoods of receiving any reperfusion therapy in the general population. Authors of other American studies have demonstrated ethnic (eg, African Americans, Hispanics, Pacific Islanders) differences relative to whites in longer time to ER presentation and reduced access to treatment (reperfusion therapies) for ACS events.30–33

Overall, and after adjustment, South Asians with atypical symptoms were least likely of the ethnic groups to receive PCI than those with typical symptoms. The reasons for this difference are not clear. It is possible that South Asians with atypical presentations arrive too late for emergent coronary revascularisation, there are biases from healthcare providers or this population may be more likely to refuse invasive cardiovascular procedures.

Study limitations

There were some limitations to this study. First, we studied participants who survived their ACS event. It is not possible to know if patients who succumbed to their ACS event or did not seek care had different symptoms than those who were diagnosed and survived. Second, we studied only those who agreed to participate. The proportion of south Asian and Chinese patients approached to participate and who refused was noteworthy. Third, we studied only patients in Canada who were admitted to selected hospitals. Finally, the cross-sectional design renders only observations and no basis for attributing causation. We undertook the following to balance these limitations: we employed a rigorous approach to comparing symptom presentation between cultural groups in this large cross-sectional study; we included participants who spoke a variety of languages, using highly systematic and validated translation and data collection processes; and had a large sample size.

Conclusions

A substantial proportion of white, south Asian and Chinese patients with ACS do not report classic ACS symptoms. Differences in ACS symptoms presentation among whites, South Asians and Chinese patients are significant with south Asians being more likely to report midsternal pain/discomfort and Chinese being less likely to report radiating pain/discomfort, relative to whites. On the whole, the mean time to ER presentation was unacceptably long for each ethnic group and especially so for south Asians. These delays can reduce the opportunity to receive evidence-based reperfusion therapies. From a public health perspective, we need to include non-chest pain symptoms for screening and public awareness campaigns for ACS among South Asian, Chinese and white persons.