People of South Asian (SA) descent are particularly susceptible to acute coronary syndromes (ACS). Yet, little information exists regarding their overall prognosis. The purpose of this study was to compare short- and long-term clinical outcomes of SA and European Canadians admitted with an ACS. Using the Alberta Provincial Project for Outcomes Assessment in Coronary Heart Disease registry, 63,393 patients with ACS were reviewed (January 1999 to March 2012). After excluding Chinese patients, 1,825 SAs were compared with 60,791 European Canadians. Both groups were propensity matched, and outcomes were compared. Adjustment was performed using a 3:1 propensity matching technique. Adjusted 30-day and 1-year mortality rates were similar between SA and European patients with ACS (2.6% vs 2.7%, p = 0.93; 5.0% vs 4.8%, respectively, p = 0.75). Repeat angiography did not differ (9.9% vs 9.2%, p = 0.35), yet repeat revascularization within 1 year was greater in SA patients (9.8% vs 7.6%, p <0.01). Improved long-term survival (median 64 months, interquartile range 66 months) was noted with SA patients (hazard ratio [HR] 0.82, 95% confidence interval [CI] 0.71 to 0.95). In particular, long-term survival was observed in SA patients receiving coronary artery bypass grafting (HR 0.75, 95% CI 0.52 to 1.08) and percutaneous coronary intervention (HR 0.75, 95% CI 0.59 to 0.96). In conclusion, SA patients treated with revascularization appear to have improved long-term survival after ACS, compared with European Canadians. As such, clinicians should be cognitive of ethnic-based outcomes when determining therapeutic strategies in patient management.
South Asians (SAs) are individuals whose ethnic roots originate from the Indian subcontinent, which includes India, Pakistan, Sri Lanka, Nepal, and Bangladesh. Although this population is not completely homogeneous, studies continue to show that migrants of SA descent appear to be particularly susceptible to coronary artery disease (CAD). Similar risk has been observed in those from the Indian subcontinent. SA patients present on average 5 years earlier with an acute coronary syndrome (ACS) compared with their European counterparts. As a result, growing awareness about SA ethnicity has become increasingly relevant. Accordingly, using a large, prospective, comprehensive, clinical registry, we compared short- and long-term clinical outcomes of SA and European Canadians admitted with ACS.
Methods
The Alberta Provincial Project for Outcomes Assessment in Coronary Heart Disease (APPROACH) is a prospective clinical data collection initiative encompassing all patients undergoing cardiac catheterization in the province of Alberta, Canada, since 1995. The registry contains detailed clinical information including each patients’ age, gender, ejection fraction, and presence or absence of previous myocardial infarction, congestive heart failure, cerebrovascular disease, peripheral vascular disease, chronic obstructive pulmonary disease, present/previous smoking, renal function, renal dialysis, hyperlipidemia, hypertension, diabetes mellitus, liver disease, gastrointestinal disease, and malignancy as well as indication for revascularization. APPROACH tracks therapeutic interventions including balloon angioplasty and type and location of coronary artery stent. It also includes previous or subsequent revascularization by coronary artery bypass grafting (CABG) surgery or percutaneous coronary intervention (PCI). Extent of CAD is documented and reported through standardized coronary anatomy diagrams. Mortality is tracked through quarterly linkage to data from the Alberta Bureau of Vital Statistics.
According to Statistics Canada (2001 census), most of Alberta’s population is European (85% to 90%) in origin, followed by Chinese (3% to 4%) and SA (3% to 4%). The present study is restricted to patients enrolled in APPROACH of SA or European descent who are admitted to hospital with an ACS. The Nam Pehchan software (version 1.1, Bradford Health Authority) was used to identify patients of SA ethnicity. This software contains a database of SA names that are matched according to complete name or name stem using the first 5 letters of the patient’s surname and has been used in ethnicity-based research across different specialities. Religious origins and language of individuals identified as being SA are reported as well. Validation has shown that Nam Pehchan sensitivity is highest among Muslim (92%) and Sikhs (86%) and somewhat lower in Hindus (62%). However, overall specificity exceeds 95% for all ethnic groups. Patients of Chinese descent were excluded using the surname algorithm validated by Quan et al to define Chinese ethnicity, which reports a sensitivity and specificity of 78% and 99.7%, respectively. The remaining population was identified as being of European lineage.
Patient characteristics among those of SA and European descent were compared using chi-square or t tests, where appropriate. A p value <0.05 was regarded as significant. The survival analysis was based on the subject being alive at the end of the study period. The cohort included mortality data on all patients up to April 2013. We have a minimum of 1 year of follow-up data for mortality for all patients, with some patients having data for >10 years.
As in all nonrandomized studies, the direct comparison of 2 distinct groups can be misleading because the groups generally differ systematically. To obtain a comparable distribution of clinical variables among SA patients and Europeans, we used the Rosenbaum and Rubin propensity score-matching technique. This technique allows for a high number of confounding variables and has been defined as being sufficient to produce unbiased estimates of the average group effects. The propensity score was calculated using logistic regression. Multiple variables (age, gender, cerebrovascular disease, congestive heart failure, chronic obstructive pulmonary disease, renal function, renal dialysis, diabetes mellitus, hyperlipidemia, hypertension, malignancy, peripheral vascular disease, present/previous smoker, previous myocardial infarction, previous PCI, previous thrombolytic therapy, indication for catheterization, coronary anatomy, and ejection fraction) were used, with SA ethnicity as an outcome (nonrandomized exposure). Greedy matching techniques were applied (SAS, version 9.2; SAS Institute Inc, Cary, North Carolina) to select 3 European patients (controls) for each SA patient (cases) by choosing the patients with the nearest propensity score, that is, within 3 decimal places of the propensity score for each case. A 3:1 propensity-matched technique was used to reduce variability and improve bias. Overlap of propensity scores between SA and European patients was evaluated using histograms and descriptive statistics. Differences in baseline factors between groups were calculated before and after propensity adjustment using absolute standardized differences in covariate means to assess balance.
Results
After excluding patients of Chinese descent, a total of 62,616 patients presented with an ACS requiring cardiac catheterization (January 1999 to March 2012). Of these, 1,825 patients (2.9%) were of SA descent. Table 1 outlines the unadjusted baseline characteristics of each group. SA patients were younger in age, more likely to be male, and have diabetes and hypertension but less likely to have a previous myocardial infarction, congestive heart failure, and smoke or have other co-morbidities. At cardiac catheterization, SA patients were more likely to have higher risk coronary anatomy.
| Characteristics | Unmatched ACS Patients | 3:1 Propensity Matched ACS Patients | ||||
|---|---|---|---|---|---|---|
| European Canadian (n = 60791) | South Asian (n = 1825) | p-Value | European Canadian (n = 5469) | South Asian (n = 1823) | p-Value | |
| Age (years) | 62.6 | 60.4 | <0.01 | 60.2 | 60.4 | 0.63 |
| Female | 29.4% | 24.9% | <0.01 | 25.5% | 25.0% | 0.65 |
| Chronic obstructive pulmonary disease | 14.2% | 8.7% | <0.01 | 8.6% | 8.7% | 0.89 |
| Cerebrovascular disease | 6.4% | 4.9% | 0.01 | 5.0% | 4.9% | 0.80 |
| Renal disease | 4.3% | 4.0% | 0.60 | 3.9% | 4.0% | 0.83 |
| Congestive heart failure | 12.3% | 10.7% | 0.05 | 11.3% | 10.8% | 0.49 |
| Diabetes mellitus | 21.9% | 35.9% | <0.01 | 35.9% | 35.8% | 0.92 |
| Dialysis | 1.2% | 1.6% | 0.09 | 1.6% | 1.6% | 0.96 |
| Hypertension | 62.5% | 65.5% | 0.01 | 64.3% | 65.5% | 0.37 |
| Dyslipidemia | 69.2% | 70.8% | 0.16 | 70.1% | 70.8% | 0.61 |
| Gastrointestinal/liver disease | 7.2% | 6.2% | 0.09 | 6.5% | 6.2% | 0.68 |
| Malignancy | 3.8% | 2.2% | 0.01 | 2.5% | 2.2% | 0.57 |
| Current smoking | 34.2% | 19.3% | <0.01 | 19.2% | 19.3% | 0.95 |
| Previous smoking | 34.7% | 15.8% | <0.01 | 16.2% | 15.8% | 0.67 |
| Previous myocardial infarction | 33.0% | 30.4% | 0.02 | 31.4% | 30.4% | 0.44 |
| Previous percutaneous coronary intervention | 5.5% | 5.6% | 0.81 | 5.9% | 5.6% | 0.62 |
| Previous coronary artery bypass surgery | 3.6% | 3.5% | 0.89 | 3.4% | 3.5% | 0.92 |
| Lytic | 8.5% | 7.1% | 0.04 | 7.1% | 7.1% | 0.94 |
| Peripheral vascular disease | 7.6% | 5.1% | <0.01 | 5.0% | 5.1% | 0.90 |
| Coronary anatomy | ||||||
| Normal | 7.9% | 5.3% | 8.6% | 5.3% | ||
| <50% narrowing | 9.2% | 8.1% | 8.1% | 8.1% | ||
| Low risk (Duke 1–6) | 45.4% | 41.9% | <0.01 | 44.0% | 42.0% | <0.01 |
| High risk (Duke 7–11) | 29.8% | 38.5% | 30.8% | 38.6% | ||
| Left main disease | 7.1% | 5.5% | 7.4% | 5.5% | ||
| Missing | 0.5% | 0.6% | 1.0% | 0.5% | ||
| Ejection fraction (%) | ||||||
| >50 | 52.2% | 49.3% | 50.3% | 49.4% | ||
| 35–50 | 19.7% | 18.7% | 19.4% | 18.7% | ||
| 20–34 | 4.5% | 3.8% | <0.01 | 4.2% | 3.8% | 0.34 |
| <20 | 7.9% | 10.5% | 8.9% | 10.5% | ||
| Too unstable | 8.3% | 8.9% | 8.2% | 8.9% | ||
| Missing | 7.5% | 8.8% | 9.0% | 8.8% | ||
Thirty-day and 1-year unadjusted mortality rates did not differ between SA patients and Europeans with ACS (2.6% vs 2.4%, p = 0.32; 5.0% vs 5.0%, respectively, p = 0.92). At 1 year, repeat angiography (a possible surrogate for recurrent angina) did not differ between both groups (9.9% vs 9.0%, p = 0.18); however, repeat revascularization was significantly greater in the SA patient cohort (10.4% vs 7.6%, p <0.01). Kaplan-Meier survival analysis demonstrated SA ethnicity was associated with improved long-term survival (hazard ratio [HR] 0.71, 95% confidence interval [CI] 0.62 to 0.82; Figure 1 ).
Table 1 also shows the adjusted baseline characteristics of the 2 groups. Propensity adjustment yielded excellent balance between the SA and European patient groups as the standardized difference was well below the recommended maximum value of 10% for every risk factor. Adjusted 30-day and 1-year mortality rates were similar between SA and European patients with ACS (2.6% vs 2.7%, p = 0.93; 5.0% vs 4.8%, respectively, p = 0.75). Repeat angiography did not differ (9.9% vs 9.2%, p = 0.35), yet repeat revascularization within 1 year was greater in SA patients (9.8% vs 7.6%, p <0.01). As seen in Figure 2 , SA ethnicity was associated with improved long-term survival for propensity-matched patients with ACS (median 64 months, interquartile range 66 months; HR 0.82, 95% CI 0.71 to 0.95).
Baseline characteristics for propensity-matched patients receiving medical therapy (568 SA patients matched to 1,919 European Canadians), CABG (252 SA patients matched to 737 European Canadians), or PCI (1,003 SA patients matched to 2,813 European Canadians) can be found in the Supplementary index . For all 3 groups of therapy, propensity adjustment yielded excellent balance between the SA and European patient groups. As listed in Table 2 , 30-day and 1-year mortality rates did not differ between SA and European patients with ACS receiving medical therapy, CABG, or PCI. The long-term propensity-matched survival rate according to treatment strategy is outlined in Figure 3 . No difference in long-term outcome was noted in SA patients compared with Europeans with ACS treated medically (HR 0.97, 95% CI 0.77 to 1.22; Figure 3 ). However, there was a trend toward improved survival in SA patients treated with CABG (HR 0.75, 95% CI 0.52 to 1.08; Figure 3 ) and significant improvement in survival with PCI (HR 0.75, 95% CI 0.59 to 0.96; Figure 3 ). However, SA patients treated with PCI had higher rates of repeat revascularization at 1 year compared with European Canadian patients (6.0% vs 3.6%, respectively, p <0.01; Table 2 ).
