Abstract
Background
In-hospital and post-discharge treatment rates for acute coronary syndrome (ACS) remain low in India. However, little is known about the prevalence and predictors of the package of optimal ACS medical care in India. Our objective was to define the prevalence, predictors, and impact of optimal in-hospital and discharge medical therapy in the Kerala ACS Registry of 25,718 admissions.
Methods and Results
We defined optimal in-hospital ACS medical therapy as receiving the following five medications: aspirin, clopidogrel, heparin, beta-blocker, and statin. We defined optimal discharge ACS medical therapy as receiving all of the above therapies except heparin. Comparisons by optimal vs. non-optimal ACS care were made via Student’s t test for continuous variables and chi-square test for categorical variables. We created random effects logistic regression models to evaluate the association between GRACE risk score variables and optimal in-hospital or discharge medical therapy. Optimal in-hospital and discharge medical care was delivered in 40% and 46% of admissions, respectively. Wide variability in both in-hospital and discharge medical care was present with few hospitals reaching consistently high (>90%) levels. Patients receiving optimal in-hospital medical therapy had an adjusted OR (95%CI)=0.93 (0.71, 1.22) for in-hospital death and an adjusted OR (95%CI)=0.79 (0.63, 0.99) for MACE. Patients who received optimal in-hospital medical care were far more likely to receive optimal discharge care (adjusted OR [95%CI]=10.48 [9.37, 11.72]).
Conclusions
Strategies to improve in-hospital and discharge medical therapy are needed to improve local process-of-care measures and improve ACS outcomes in Kerala.
Keywords: acute coronary syndrome, registries, optimal care/health services research
BACKGROUND
In-hospital and discharge medical therapies for acute coronary syndrome (ACS) have been demonstrated to be highly effective and cost effective, including in middle-income countries such as India.1 Despite substantial reductions in ACS-related morbidity and mortality with the combination of antiplatelet agents, beta-blockers, statins, and anti-coagulants for patients with ACS, treatment rates of individual drug classes remain low in both inpatient2, 3 and post-discharge settings4 in India. Quality improvement studies have demonstrated stepwise improvements in outcomes with increases in guideline-based, or optimal ACS in-hospital care.5 However, little is known about the prevalence and predictors of the package of optimal in-hospital or discharge medical therapy for patients with ACS in India.
Our objective was to define the prevalence, predictors, and impact of optimal in-hospital and discharge medical therapy in the Kerala ACS Registry of 25,718 admissions as part of an ongoing ACS quality improvement project in Kerala.
METHODS
Details of the Kerala ACS Registry have been previously described.2 Briefly, representatives from 185 hospitals that treat patient with acute coronary syndrome in Kerala were approached to participate in the Kerala ACS Registry. One hundred and forty hospitals responded, and 125 hospitals agreed to participate. We prospectively collected data on 25,748 consecutive ACS admissions from 2007-2009.
Patients were included if they were 18 years or older, presented with chest pain, and had one of the following: 1) ST segment elevations in two contiguous leads, 2) ST depressions or T wave inversion in two contiguous leads with a history of ischemic heart disease, or 3) troponin or CK-MB elevation. Data were collected in the coronary care units of each hospital. Patients with missing data (n=30) were excluded for a final sample of 25,718 admissions. Informed consent was waived based on the Common Rule.
Trained research staff abstracted demographics, employment, self-reported medical history, prior medications, clinical, laboratory, diagnostic, and treatment data, and in-hospital outcomes. Data were supplemented by patient interviews, when necessary. Data were entered into case report forms in the presence of the treating physicians. Five regional coordinating cardiologists supervised 10 district-level research coordinators who made monthly site visits and resolved outstanding queries through direct contact with the treating physician. Hospital-level characteristics were also evaluated, including urban vs. rural location, cardiologist on staff, onsite cardiac catheterization laboratory, and academic affiliation.
Statistical Analysis
We defined optimal in-hospital medical care as receiving aspirin, clopidogrel, heparin, beta-blocker, and statin since each was considered a class I recommendation at the beginning of data collection (2007). We defined optimal discharge medical care as receiving all of the above except heparin. Continuous data are presented as means (standard deviations) or median (interquartile range) when skewed. Categorical variables are presented as proportions. Comparisons by optimal vs. non-optimal ACS care were made via Student’s t test for continuous variables and chi-square test for categorical variables. A two-sided p value of <0.05 defined statistical significance.
We performed pairwise correlations between in-hospital medications and discharge medications included in each definition of optimal care to explore potential correlations. We then evaluated the proportion of each potential combination of medications for in-hospital and discharge care to explore differences of different drug combinations within and between the in-hospital and discharge setting.
We created univariate and random effects multivariate logistic regression models that adjust standard errors for within-hospital clustering to evaluate the association between optimal in-hospital care and in-hospital death or major adverse cardiovascular events (MACE, defined as death, reinfarction, stroke, heart failure, or cardiogenic shock). We included age, sex, and covariates from the Global Registry of Acute Coronary Events (GRACE) Risk Model.6 We created additional regression models to evaluate associations between patient characteristics and GRACE Risk Model covariates and the outcome of receiving optimal in-hospital or discharge care to better understand potential predictors of optimal care to inform our forthcoming Acute Coronary Syndrome Quality Improvement in Kerala (ACS QUIK) clinical trial. We used STATA v11.2 (College Station, TX, USA) for our analyses.
The Institutional Ethics Committees of Sree Chitra Tirunal Institute of Medical Sciences and Technology in Trivandrum, Kerala and Westfort Hi-Tech Hospital in Thrissur, Kerala approved the study. All participants provided informed consent. The Institutional Review Board of Northwestern University provided an exemption of review for analysis of de-identified data.
RESULTS
The participant flowchart is provided in Supplemental Figure 1. The mean (SD) age of presentation was 60 (12) years, and the majority (77%) of participants were men (Table 1). Unemployed patients were less likely to receive optimal in-hospital (37% vs. 39%, p<0.001) and discharge (38% vs. 39%, p<0.001) medical care. Patients with prior hypertension, diabetes, and smoking were more likely to receive optimal in-hospital and discharge medical care, but patients with prior myocardial infarction were less likely to receive optimal in-hospital (12% vs. 16%, p<0.001) and discharge (13% vs. 16%, p<0.001) medical care. Patients presenting with ST-segment elevation myocardial infarction (STEMI) and higher Killip class were also less likely to receive optimal in-hospital and discharge medical care. Patients receiving optimal in-hospital medical care were also less likely to receive inappropriate thrombolysis in the setting of NSTEMI or unstable angina than patients who did not receive optimal in-hospital medical care (13% vs. 17%, p<0.001).
Table 1.
Patient-level characteristics by presence or absence of optimal in-hospital and discharge medical therapy.
| Optimal in-hospital medical therapy |
Non optimal in-hospital medical therapy |
p value | Optimal discharge medical therapy |
Non optimal discharge medical therapy |
p value | |
|---|---|---|---|---|---|---|
| N (%) | N=10,307 (40) | N=15,411 (60) | N=11,397 (46) | 13,353 (54) | ||
| Demographics, Medical History | ||||||
| Sex (male), N (%) | 8,009 (77.7) | 11,914 (77.2) | 0.31 | 8,790 (77.1) | 10,374 (77.7) | 0.29 |
| Age, years (SD) | 60.3 (12) | 60.6 (12) | 0.12 | 60.4 (12) | 60.5 (12) | 0.24 |
| Unemployed, N (%) | 3,788 (37.1) | 5,974 (39.0) | <0.001 | 4,271 (37.8) | 5,107 (38.6) | <0.001 |
| Prior MI, N (%) | 1,227 (11.9) | 2,428 (15.7) | <0.001 | 1,471 (12.9) | 2,005 (15.0) | <0.001 |
| History of HTN, N (%) | 5,837 (56.6) | 7,443 (48.2) | <0.001 | 6,097 (53.5) | 6,683 (50.1) | <0.001 |
| History of diabetes, N (%) | 4,104 (39.8) | 5,579 (36.1) | <0.001 | 4,495 (39.4) | 4,769 (35.7) | <0.001 |
| History of smoking, N (%) | 4,451 (43.8) | 4,416 (28.6) | <0.001 | 4,741 (41.6) | 3,788 (28.4) | <0.001 |
| Presentation, Clinical Data | ||||||
| STEMI, N (%) | 3,407 (36.0) | 6,162 (42.0) | <0.01 | 3,850 (36.8) | 4,935 (38.9) | 0.001 |
| BMI, kg/m2 (SD) | 23.1 (3.6) | 23.1 (3.6) | 0.44 | 23.0 (3.6) | 23.1 (3.6) | 0.05 |
| Heart rate, bpm (SD) | 80.2 (20) | 80.2 (19) | 0.29 | 79.7 (19) | 80.1 (20) | 0.09 |
| SBP, mmHg (SD) | 140.6 (30) | 141.3 (29) | 0.06 | 141.1 (29) | 140.9 (31) | 0.53 |
| Creatinine, mg/dl (SD) | 1.2 (0.8) | 1.2 (0.6) | 0.003 | 1.2 (0.7) | 1.2 (0.6) | 0.51 |
| Killip class >1 | 1,045 (18.6) | 1,951 (23.8) | <0.001 | 1,062 (17.1) | 1,763 (24.9) | <0.001 |
| Prior/in-hospital ASA, N (%) | 1,650 (16.0) | 2,456 (15.9) | 0.83 | 11,300 (99.2) | 11,735 (87.9) | <0.001 |
| Prior/in-hospital clopidogrel, N (%) | 1,550 (15.0) | 2,384 (15.4) | 0.38 | 11,263 (98.8) | 12,294 (92.1) | <0.001 |
| Prior/in-hospital β-blocker, N (%) | 1,298 (12.6) | 1,792 (11.6) | 0.02 | 10,383 (91.1) | 5,994 (44.9) | <0.001 |
| Prior/in-hospital statin, N (%) | 1,109 (10.8) | 1,832 (11.9) | <0.01 | 10,550 (92.6) | 8,992 (66.8) | <0.001 |
| Procedures, Other Treatment | ||||||
| In-hospital coronary angiography, N (%) | 2,115 (20.5) | 2,896 (18.8) | <0.001 | 2,391 (21.0) | 2,359 (17.7) | <0.001 |
| In-hospital PCI, N (%) | 1,284 (12.5) | 1,776 (11.5) | 0.02 | 14,27 (12.5) | 1,448 (10.8) | <0.001 |
| In-hospital CABG, N (%) | 174 (1.7) | 173 (1.1) | <0.001 | 167 (1.5) | 164 (1.2) | 0.11 |
Table 2 demonstrates the prevalence of optimal in-hospital and discharge medical care by hospital-level characteristics. Rural hospitals were less likely to provide optimal in-hospital (33% vs. 42%, p<0.001) and discharge medical care (35% vs. 50%, p<0.001). On the other hand, academically-affiliated hospitals were more likely to provide optimal in-hospital (64% vs. 39%, p<0.001) and discharge (65% vs. 45%, p<0.001) medical care. There appeared to be no difference in rates of optimal in-hospital (40% vs. 41%, P=0.14) and discharge (46% vs. 46%, p=0.23) medical care among hospitals with or without a cardiologist on staff, respectively. Hospitals with cardiac catheterization laboratories had lower rates of optimal in-hospital medical care (36% vs. 42%, p<0.001) than hospitals without cardiac catheterization laboratories, but the former hospitals had somewhat higher rates of optimal discharge medical care (48% vs. 45%, p<0.001). When we explored the relationship between optimal care and academic status among hospitals with cardiac catheterization labs, we found that optimal in-hospital care was delivered in 31% and 64% (p<0.001) of non-academic (n=120 hospitals) and academic hospitals (n=5 hospitals) with cardiac catheterization labs, respectively. A similar pattern was seen for optimal discharge medical care (45% vs. 65%, p<0.001).
Table 2.
Hospital-level characteristics associated with optimal in-hospital and discharge medical care.
|
Rural hospital
N=5,992 |
Non-rural hospital
N=19,756 |
p value | |
| Optimal in-hospital medical care, n (%) | 2,007/5,992 (33) | 8,300/19,756 (42) | <0.001 |
| Optimal discharge medical care, n (%) | 2,014/5,799 (35) | 9,383/18,951 (50) | <0.001 |
|
| |||
|
Academic hospital
N=1,301 |
Non-academic hospital
N=24,447 |
p value | |
| Optimal in-hospital medical care, n (%) | 832/1,301 (64) | 9,475/24,447 (39) | <0.001 |
| Optimal discharge medical care, n (%) | 765/1,181 (65) | 10,632/23,569 (45) | <0.001 |
|
| |||
|
Cardiologist present
N=17,026 |
Cardiologist absent
N=8,722 |
p value | |
| Optimal in-hospital medical care, n (%) | 6,761/17,026 (40) | 3,546/8,722 (41) | 0.14 |
| Optimal discharge medical care, n (%) | 7,504/16,198 (46) | 3,893/8,552 (46) | 0.23 |
|
| |||
|
Catheterization lab present
N=8,061 |
Catheterization lab absent
N=17,687 |
p value | |
| Optimal in-hospital medical care, n (%) | 2,910/8,061 (36) | 7,397/17,687 (42) | <0.001 |
| Optimal discharge medical care, n (%) | 3,643/7,550 (48) | 7,754/17,200 (45) | <0.001 |
We found weak correlations between most in-hospital and discharge medications, and correlation coefficients among discharge medications were higher (Table 3). The correlation coefficients were highest between clopidogrel and beta-blockers for both in-hospital (correlation coefficient=0.44) and discharge (correlation coefficient=0.74) medical care. Figure 1 demonstrates the proportion of each possible in-hospital and discharge medication combination. Single and dual antiplatelet medical therapies were the most common in-hospital and discharge therapies (individually and in combination), and beta-blockers (both individually and in combination) were the least common. Discharge medication rates were lower than in-hospital medication rates for all combinations. Supplemental Figures 2 and 3 demonstrate the variability across hospitals in delivering optimal in-hospital and discharge medical care, respectively.
Table 3.
Pairwise correlation coefficient between in-hospital and discharge medications that comprise the definition of optimal in-hospital and discharge medical care.
| In-hospitals medications | |||||
| Aspirin | Clopidogrel | Beta blocker | Statin | Heparin | |
| Aspirin | -- | 0.1868 | 0.1410 | 0.0941 | 0.1867 |
| Clopidogrel | -- | -- | 0.4352 | 0.1852 | 0.2262 |
| Beta blocker | -- | -- | -- | 0.2488 | 0.2327 |
| Statin | -- | -- | -- | -- | 0.1155 |
| Heparin | -- | -- | -- | -- | -- |
|
| |||||
| Discharge medications | |||||
| Aspirin | Clopidogrel | Beta blocker | Statin | ||
| Aspirin | -- | 0.2783 | 0.2540 | 0.2563 | |
| Clopidogrel | -- | -- | 0.7353 | 0.5643 | |
| Beta blocker | -- | -- | -- | 0.6510 | |
| Statin | -- | -- | -- | -- | |
Figure 1.
Proportion of each drug and drug combination for in-hospital and discharge care in Kerala ACS Registry patients. C=clopidogrel; A=aspirin; B=beta blocker; S=statin; H=heparin.
Patients receiving optimal in-hospital therapy had lower in-hospital death rates (3.6% vs. 4.1%, unadjusted OR [95%CI]=0.86 [0.76, 0.98]) and lower major adverse cardiovascular event rates (MACE, 5.2% vs. 6.1%, unadjusted OR [95%CI]=0.84 [0.75, 0.94]) (Table 4). These estimates were attenuated with adjustment. Patients receiving optimal in-hospital medical therapy had an adjusted OR (95%CI)=0.93 (0.71, 1.22) for in-hospital death and an adjusted OR (95%CI)=0.79 (0.63, 0.99) for MACE.
Table 4.
Unadjusted and multivariable adjusted random effects logistic regression model to evaluate the association (OR [95% CI]) between optimal vs. non-optimal in-hospital medical therapy and in-hospital events.
| Optimal in- hospital medical therapy |
Non optimal in- hospital medical therapy (ref) |
Unadjusted OR (95% CI) |
Adjusted OR* (95%CI) |
|
|---|---|---|---|---|
| N (%) | N=10,307 (40) | N=15,411 (60) | ||
| In-hospital death | 366 (3.6) | 632 (4.1) | 0.86 (0.76, 0.98) |
0.93 (0.71, 1.22) |
| In-hospital death, reinfarction, stroke, heart failure, or shock |
532 (5.2) | 938 (6.1) | 0.84 (0.75, 0.94) |
0.79 (0.63, 0.99) |
Adjusted for within-hospital clustering and modified GRACE risk score variables: age, heart rate, systolic blood pressure, serum creatinine, Killip class, cardiac enzyme (positive vs. negative), and ST segment deviation. Cardiac arrest at presentation excluded (N/A).
Patients presenting with Killip class >1 were less likely to receive optimal in-hospital medical therapy (OR [95%CI]=0.56 [0.50, 0.63]) and discharge medical therapy (OR [95%CI]=0.67 [0.59, 0.75]), even after adjustment (Table 5). Patients presenting with STEMI were less likely to receive optimal in-hospital medical care (adjusted OR [95% CI]=0.51 [0.42, 0.62] compared with patients presenting with unstable angina, but patients presenting with STEMI were more likely to receive optimal discharge care (adjusted OR [95%CI]=1.39 [1.15, 1.68]). The unadjusted OR (95% CI) of optimal in-hospital care on the outcome of optimal discharge care is 9.33 (8.80, 9.90), which increases to OR=10.48 (9.37, 11.72, Table 5) after adjustment for within-hospital clustering and patient characteristics, though we note overlap of the 95% CI. We have also explored the correlation coefficient between optimal in-hospital and discharge care across all hospitals, which ranges widely from -0.24 to 1.00 (data not shown).
Table 5.
Multivariable random effects logistic regression model to evaluate the association (OR [95% CI]) between patient and process characteristics with optimal in-hospital and discharge medical care.
| Optimal in-hospital medical therapy Adjusted OR* (95% CI) |
Optimal discharge medical therapy Adjusted OR* (95% CI) |
|
|---|---|---|
| Age | 1.00 (1.00, 1.01 |
1.00 (0.99, 1.00) |
| Women vs. men (ref) | 1.00 (0.90, 1.12) |
1.00 (0.89, 1.13) |
| Heart rate (per bpm) | 1.01 (1.00, 1.01) |
1.00 (1.00, 1.00) |
| Systolic blood pressure (per mmHg) | 1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
| Killip >1 vs. =1 (ref) | 0.56 (0.50, 0.63) |
0.67 (0.59, 0.75) |
| NSTEMI vs. unstable angina (ref) | 1.17 (0.95, 1.45) |
1.71 (1.35, 2.16) |
| STEMI vs. unstable angina (ref) | 0.51 (0.42, 0.62) |
1.39 (1.15, 1.68) |
| Enzyme positive vs. negative (ref) | 2.14 (1.75, 2.62) |
0.74 (0.60, 0.90) |
| Creatinine (per mg/dl) | 1.03 (0.97, 1.10) |
0.97 (0.90, 1.04) |
| Optimal in-hospital medical therapy vs. non optimal in-hospital medical therapy (ref) |
-- | 10.48 (9.37, 11.72) |
Adjusted for within-hospital clustering and modified GRACE risk score variables: age, heart rate, systolic blood pressure, serum creatinine, Killip class, cardiac enzyme (positive vs. negative), and ST segment deviation and within-hospital clustering. Cardiac arrest at presentation excluded (N/A).
DISCUSSION
Summary of Results
In a secondary analysis of the Kerala ACS Registry, we demonstrate that optimal in-hospital and discharge medical care, defined as receiving aspirin, clopidogrel, beta blockade, statin, and heparin (in-hospital only), was delivered in 40% and 46% of admissions, respectively. Wide variability in both in-hospital and discharge medical care was present across the range of participating hospitals with few hospitals participating at consistently high (>90%) levels. When we excluded heparin in our definition of optimal in-hospital medical care, approximately one out of every two patients received the combination of aspirin, clopidogrel, beta-blocker, and statin during ACS hospitalization. The lower rate of inappropriate thrombolysis in patients receiving optimal medical care also suggests that hospital- and provider-level processes of care extend beyond the “optimal care” package. In fact, patients who received optimal in-hospital medical care had a 21% lower rate of in-hospital MACE, after adjustment, and a trend toward lower in-hospital death rates.
In our evaluation of associations with receiving optimal medical care, we found that optimal in-hospital medical care had the strongest association with receiving optimal discharge care (OR [95%CI]=10.48 [9.37, 11.72]). We interpret these results to suggest that patient and hospital-level factors that are associated with in-hospital optimal care are likely associated with optimal discharge care, which highlights the importance of optimal in-hospital care as a key step for improving ACS quality of care. Beta-blocker use had the highest correlation with clopidogrel and statin use, suggesting that targeted efforts to increase the uptake of beta-blockers may lead to greater increases in other medications, particularly since beta-blockers were the least commonly prescribed medication among both in-hospital and discharge groups. On the other hand, beta-blocker prescription may be confounded by the presence of other, unmeasured variables of high-quality care.
Rural, non-academic hospitals were less likely to provide both in-hospital and discharge optimal medical care. In subsequent interviews and focus group discussions with participating cardiologists after these data were collected, we have learned of the near impossibility of administering unfractionated heparin in the crowded wards of public hospitals where measuring partial thromboplastin times (PTT) is infeasible, and the cost of low molecular weight heparin is unaffordable. Such local factors likely play crucial roles in determining the definition of optimal care. We also note similarities in receiving optimal in-hospital and discharge medical care from hospitals with and without cardiologists, which suggests that task shifting in the care of low-risk ACS patients to non-cardiologists may be one strategy to provide high-quality, high-throughout care in regions with limited resources. However, we remain cautious in our interpretation of the association between hospital-level characteristics and optimal care given the potential for reverse causality, such that sicker patients may be more likely to seek care at facilities with cardiac catheterizations labs, they may be more likely to undergo invasive procedures, and they may have more contraindications to components of optimal medical care.
Comparison with Prior Studies
One small study of 137 patients at a single center in Mangalore, India reported similar rates of discharge optimal medical care (41%) for ACS patients, though the definition included a single antiplatelet agent, beta-blocker, statin, and ACE-I/ARB use.7 The CREATE investigators have not reported analyses evaluating combination therapy for ACS patients, but beta-blocker and statins rates were lower than in our analyses; on the other hand, anti-coagulant (and ACE-I/ARB) rates were higher in CREATE compared with the Kerala ACS Registry.3 Optimal combination secondary prevention of CVD (aspirin, beta-blocker, statin, and ACE-I/ARB) in the outpatient setting appear low as well, ranging from 7% to 44% in a recent report of 2,993 prescriptions from the Indian state of Rajasthan, compared with rates of 54% at the time of CVD discharge from tertiary medical centers.8
Investigators from other low- and middle-income countries have also evaluated the prevalence of optimal medical care in their registries. For example, the prevalence of optimal medical care at discharge, defined by the concurrent use of aspirin, beta-blockers, statin, and ACE-I/ARB was evaluated in the Clinical Pathways for Acute Coronary Syndromes in China (CPACS) registry of 2,901 ACS admissions from 49 hospitals. In this registry, optimal discharge medical care was present in approximately one out of every two patients (48%).9 Similar to our data, patients with higher GRACE scores had lower rates of optimal discharge medical care.
Investigators from the Gulf Registry of Acute Coronary Events (Gulf RACE) also evaluated the prevalence of participants receiving optimal discharge medical therapy as defined by concurrent use of aspirin, beta-blocker, statin, and ACE-I/ARB.10 In this registry of 8,176 ACS admissions, a similar proportion of approximately one out of every two patients (49%) received this combination, but there was substantial heterogeneity among the six countries that participated with rates ranging from 38% (Kuwait) to 68% (United Arab Emirates). While higher GRACE scores were minimally associated with lower rates of optimal care at discharge (OR per 1-point increase in GRACE score [95%CI]=0.99 [0.99, 0.99]), patients with cardiogenic shock in the Gulf RACE were far less likely to receive optimal discharge medications (OR [95%CI]=0.27[0.21, 0.36]).
Building upon these previous publications, investigators from the Brazilian Intervention to Increase Evidence Usage in Acute Coronary Syndromes (BRIDGE ACS) cluster-randomized clinical trial of 1,150 ACS patients in 34 clusters demonstrated increased rates of optimal acute (first 24 hours) and discharge medical therapy through a multifaceted quality improvement intervention.11 The intervention led to an increase in acute medical care, defined by the concurrent use of aspirin, clopidogrel, anti-coagulant, and statin from 50% to 68% (OR [95%CI]=2.64 [1.28, 5.45]). Optimal discharge medical therapy, defined as the use of aspirin, beta-blocker, statin, and ACE-I/ARB, also increased from 57% in control arm to 66% in the intervention arm (OR [95%CI]=1.55 [0.75, 3.18]).
Reasons for the differences between our data and other investigators are likely multi-factorial and include: 1) Differences in definitions of optimal care. We included clopidogrel given its class I recommendation for all patients in 2007, which is not the case with ACE-I/ARB, which requires concomitant heart failure, left ventricular systolic dysfunction, diabetes, or hypertension for a class I indication;12 2) Wider range of hospitals. We included 125 hospitals compared with 49, 65, and 34 hospitals from CPACS, Gulf RACE, and BRIDGE ACS, respectively. These registries included a higher proportion of academically-affiliated hospitals (all >70%, compared with 8% of hospitals in the Kerala ACS Registry), which may bias their results to centers more likely to provide optimal medical care and limit regional generalizability; and 3) Longer duration of data collection. Our study included data collection for 25 months, compared with 19 months, 14 months, and 8 months in CPACS, Gulf RACE, and BRIDGE ACS, respectively. We hypothesize that initial enthusiasm for collecting quality of care data may lead to changes in process of care measures through a Hawthorne effect, which may wane over time and lead to lower rates of optimal process of care measures through regression to the mean, particularly in lower performing hospitals.
Post-discharge rates of optimal medical care in the outpatient setting are unfortunately even lower. Data from the PURE registry suggest that less than half of patients with prevalent cardiovascular disease take at least one medication for secondary prevention.4 Even large contemporary, US-based clinical trials (FREEDOM, BARI 2D, and COURAGE) suggest that research participants simultaneously reach clinical cholesterol, blood pressure, smoking, and hemoglobin A1c targets only ~20% of the time, due at least in part, to low adherence.13 While others have demonstrated nearly 40% lower mortality rates (HR [95%CI] = 0.61 (0.43, 0.88)] at 12 months at a single US center in patients with at least 80% adherence, or concordance, to guideline-based care,14 dismal rates of optimal medical care in our data, in other observational studies from other low- and middle-income countries, and in carefully monitored clinical trials in high-income countries argue for both comprehensive, continuous quality improvement, as well as different paradigms (inpatient and post-discharge use of three drug [aspirin, clopidogrel, statin] fixed dose combination, or polypill, therapy, which is already available in India, or alternative [non-oral] drug delivery systems, e.g.) for secondary prevention of CVD. These data may also have implications for the ongoing debate regarding universal health insurance in India and other low- and middle-income countries since patients frequently bear high out-of-pocket expenses for treatment, including for in-hospital medications. More than one out of every two patients hospitalized for ACS or stroke in Kerala experience financial catastrophe as a result of their medical expenditures (or “catastrophic health spending”, defined as spending > 40% of total, non-food expenditures on health),15 which likely affects not only in-hospital medical treatment but also discharge medical treatment. Lack of private or social health insurance has been associated with a four-fold higher risk of catastrophic health spending (OR [95%CI]=3.93 [2.23, 6.90]) within 15 months of CVD hospitalization in India as well as Argentina, China, and Tanzania.15 Scalable strategies to increase the proportion of ACS patients who receive the optimal combination of in-hospital and discharge medications in low- and middle-income country settings appear to have substantial potential impact, particularly since ischemic heart disease is the leading cause of death worldwide and the majority (80%) of global CVD deaths occur in lowand middle-income countries.
Strengths/Limitations
These data represent the largest ACS registry in India to date and provide the first large-scale evaluation of optimal in-hospital and discharge medical care in India. We will leverage these data in the development of our regional ACS quality improvement program in Kerala (ACS QUIK) for improved process-of-care measures. We also aim to explore reasons for successes, including high rates of dual antiplatelet use, which are comparable to contemporary registries from high-income countries.16 However, our data also have several limitations. First, patients receiving optimal medical care may have had fewer contraindications or been less ill than patients who did not receive optimal care, as demonstrated by patients with Killip class >1 being 44% and 33% less likely to receive optimal in-hospital and discharge care, respectively, compared with patients who were Killip class=1. This example of reverse causality may also partially explain the lower mortality and MACE rates in patients receiving optimal medical care. Second, these data are susceptible to residual confounding, which might change the strength or direction of the associations between process-of-care measures and outcomes, which may be reflected by the imprecision of our results. However, we restricted our evaluation to contemporary medications that have been shown to improve in-hospital and post-discharge outcomes and are class I recommendations in major cardiovascular society guidelines for all ACS patients.12, 17, 18 While we appreciate the potential for residual confounding, even after controlling for GRACE risk score variables, we argue that these process-of-care measures are generally considered the standard of care. Third, we did not include ACE-I/ARBs in our definition of optimal medical care, which some may argue for inclusion. Rates of left ventricular systolic dysfunction (EF<30%) were very low (<2%), and we were missing data on ejection fraction on one-third of participants. Given these low rates of severe left ventricular systolic dysfunction and the lack of a class I recommendation of administering ACE-I/ARB in patients with preserved ejection fraction, we excluded ACE-I/ARB in our definition. Fourth, our outcome data are limited to in-hospital events, and so potential outcome benefits from optimal discharge medical care could not be explored. Fifth, our data are geographically limited to the state of Kerala and thus cannot represent the whole of India. However, the Kerala ACS Registry has broad coverage throughout the state.
Implications/Conclusions
Rates of optimal in-hospital (40%) and discharge (46%) medical care for ACS are sub-optimal in Kerala, and these process-of-care gaps represent opportunities for improvement. Even after adjustment, patients at higher risk (Killip >1, STEMI) were less likely to receive optimal in-hospital medical therapy, which was associated with increased MACE rates. In-hospital ACS medical therapy is a powerful predictor of discharge medical therapy and provides an important target for quality improvement of discharge medical therapy. Strategies to improve in-hospital medical therapy (checklists, audit/feedback systems for continuous quality improvement) as well as discharge/post-discharge adherence are needed to improve local process-of-care measures and to improve ACS outcomes. Novel paradigms might also be useful to increase optimal medical care for patients with ACS, including inpatient and post-discharge use of fixed dose combination or polypill therapies, alternative drug delivery mechanisms, and universal health insurance to improve access to essential medicines.
Supplementary Material
Acknowledgements
PPM, MDH, and DP had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
We would like to thank Professors Sankar Sharma and K.R. Thankappan (Sree Chitra Tirunal Institute for Medical Sciences and Technology) for contributions as advisors in the conception and design of the study and review of the final manuscript for the Kerala ACS Registry. Neither received compensation for their work.
Funding Sources The Cardiological Society of India-Kerala Chapter (CSI-K) funded the Kerala ACS Registry. The study investigators include leaders within the CSI-K and, as such, participated in the study design, data collection, analysis, writing, review, and approval of the manuscript. MDH is supported by a National Heart, Lung, and Blood Institute Pathway to Independence grant (1 K99 HL107749-01A1). DP receives partial salary support from a contract award (HHS N268200900026C) from National Heart, Lung, and Blood Institute and a grant award (1D43HD065249) from National Institute for Child Health and Development.
Abbreviations
- ACS
acute coronary syndrome
- CSI-K
Cardiological Society of India-Kerala Chapter
- GRACE
Global Registry of Acute Coronary Events
- MACE
major adverse cardiovascular event rates
- OR
odds ratio
- SD
standard deviation
- STEMI
ST-segment elevation myocardial infarction
Footnotes
Disclosures None.
Author Responsibilities: Conception and design: DP, AKA, MNK, PPM. Acquisition of data: AKA, MNK, PPM. Analysis and interpretation of data: MDH, DP, PPM. Drafting of manuscript: MDH. Critical revision of manuscript for important intellectual content: All. Statistical analysis: MDH, DP, PPM. Obtaining funding: MNK, PPM. Administrative, technical, or material support: All. Supervision: DP, MNK, PPM.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Gaziano TA. Cardiovascular disease in the developing world and its cost-effective management. Circulation. 2005;112:3547–3553. doi: 10.1161/CIRCULATIONAHA.105.591792. [DOI] [PubMed] [Google Scholar]
- 2.Mohanan PP, Mathew R, Harikrishnan S, Krishnan MN, Zachariah G, Joseph J, Eapen K, Abraham M, Menon J, Thomas M, Jacob S, Huffman MD, Prabhakaran D, on behalf of the Kerala ACS Registry Investigators Presentation, management, and outcomes of 25 748 acute coronary syndrome admissions in Kerala, India: Results from the Kerala ACS Registry. Eur Heart J. 2013;34:121–129. doi: 10.1093/eurheartj/ehs219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Xavier D, Pais P, Devereaux PJ, Xie C, Prabhakaran D, Reddy KS, Gupta R, Joshi P, Kerkar P, Thanikachalam S, Haridas KK, Jaison TM, Naik S, Maity AK, Yusuf S. Treatment and outcomes of acute coronary syndromes in India (CREATE): A prospective analysis of registry data. Lancet. 2008;371:1435–1442. doi: 10.1016/S0140-6736(08)60623-6. [DOI] [PubMed] [Google Scholar]
- 4.Yusuf S, Islam S, Chow CK, Rangarajan S, Dagenais G, Diaz R, Gupta R, Kelishadi R, Iqbal R, Avezum A, Kruger A, Kutty R, Lanas F, Lisheng L, Wei L, Lopez-Jaramillo P, Oguz A, Rahman O, Swidan H, Yusoff K, Zatonski W, Rosengren A, Teo KK, Prospective Urban Rural Epidemiology Study Investigators Use of secondary prevention drugs for cardiovascular disease in the community in high-income, middle-income, and low-income countries (the PURE study): A prospective epidemiological survey. Lancet. 2011;378:1231–1243. doi: 10.1016/S0140-6736(11)61215-4. [DOI] [PubMed] [Google Scholar]
- 5.Peterson ED, Roe MT, Mulgund J, DeLong ER, Lytle BL, Brindis RG, Smith SC, Jr., Pollack CV, Jr., Newby LK, Harrington RA, Gibler WB, Ohman EM. Association between hospital process performance and outcomes among patients with acute coronary syndromes. JAMA. 2006;295:1912–1920. doi: 10.1001/jama.295.16.1912. [DOI] [PubMed] [Google Scholar]
- 6.Granger CB, Goldberg RJ, Dabbous O, Pieper KS, Eagle KA, Cannon CP, Van De Werf F, Avezum A, Goodman SG, Flather MD, Fox KA. Predictors of hospital mortality in the Global Registry Of Acute Coronary Events. Arch Int Med. 2003;163:2345–2353. doi: 10.1001/archinte.163.19.2345. [DOI] [PubMed] [Google Scholar]
- 7.Kademane KC, Manjurath SE, Rai M, Nayak RP. Use of secondary prevention drug therapy in patients with coronary artery disease in a tertiary healthcare centre in Mangalore, South India. International Journal of AJ Institute of Medical Sciences. 2012;1:36–40. [Google Scholar]
- 8.Sharma KK, Gupta R, Agrawal A, Roy S, Kaslwal A, Bana A, Tongia RK, Deedwania PC. Low use of statins and other coronary secondary prevention therapies in primary and secondary care in India. Vascular Health and Risk Management. 2009;5:1007–14. doi: 10.2147/vhrm.s8017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bi Y, Gao R, Patel A, Su S, Gao W, Hu D, Huang D, Kong L, Qi W, Wu Y, Yang Y, Turnbull F. Evidence-based medication use among Chinese patients with acute coronary syndromes at the time of hospital discharge and 1 year after hospitalization: Results from the clinical pathways for acute coronary syndromes in China (CPACS) study. Am Heart J. 2009;157:509–516. e501. doi: 10.1016/j.ahj.2008.09.026. [DOI] [PubMed] [Google Scholar]
- 10.Al-Zakwani I, Zubaid M, Panduranga P, Rashed W, Sulaiman K, Almahmeed W, Al-Motarreb A, Al Suwaidi J, Amin H. Medication use pattern and predictors of optimal therapy at discharge in 8,176 patients with acute coronary syndrome from 6 Middle Eastern countries: data from the Gulf Registry of Acute Coronary Events. Angiology. 2011;62:447–454. doi: 10.1177/0003319710394163. [DOI] [PubMed] [Google Scholar]
- 11.Berwanger O, Guimaraes HP, Laranjeira LN, Cavalcanti AB, Kodama AA, Zazula AD, Santucci EV, Victor E, Tenuta M, Carvalho V, Mira VL, Pieper KS, Weber B, Mota LH, Peterson ED, Lopes RD. Effect of a multifaceted intervention on use of evidence-based therapies in patients with acute coronary syndromes in Brazil: The BRIDGE-ACS randomized trial. JAMA. 2012;307:2041–2049. doi: 10.1001/jama.2012.413. [DOI] [PubMed] [Google Scholar]
- 12.Hamm CW, Bassand JP, Agewall S, Bax J, Boersma E, Bueno H, Caso P, Dudek D, Gielen S, Huber K, Ohman M, Petrie MC, Sonntag F, Uva MS, Storey RF, Wijns W, Zahger D, Guidelines ESCCfP, Bax JJ, Auricchio A, Baumgartner H, Ceconi C, Dean V, Deaton C, Fagard R, Funck-Brentano C, Hasdai D, Hoes A, Knuuti J, Kolh P, McDonagh T, Moulin C, Poldermans D, Popescu BA, Reiner Z, Sechtem U, Sirnes PA, Torbicki A, Vahanian A, Windecker S, Document R, Windecker S, Achenbach S, Badimon L, Bertrand M, Botker HE, Collet JP, Crea F, Danchin N, Falk E, Goudevenos J, Gulba D, Hambrecht R, Herrmann J, Kastrati A, Kjeldsen K, Kristensen SD, Lancellotti P, Mehilli J, Merkely B, Montalescot G, Neumann FJ, Neyses L, Perk J, Roffi M, Romeo F, Ruda M, Swahn E, Valgimigli M, Vrints CJ, Widimsky P. ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: The task force for the management of acute coronary syndromes (ACS) in patients presenting without persistent ST-segment elevation of the European Society of Cardiology (ESC) Eur Heart J. 2011;32:2999–3054. doi: 10.1093/eurheartj/ehr236. [DOI] [PubMed] [Google Scholar]
- 13.Farkouh ME, Boden WE, Bittner V, Muratov V, Hartigan P, Ogdie M, Bertolet M, Mathewkutty S, Teo K, Maron DJ, Sethi SS, Domanski M, Frye RL, Fuster V. Risk factor control for coronary artery disease secondary prevention in large randomized trials. J Am Coll Cardiol. 2013;61:1607–1615. doi: 10.1016/j.jacc.2013.01.044. [DOI] [PubMed] [Google Scholar]
- 14.Allen LA, O’Donnell CJ, Giugliano RP, Camargo CA, Jr., Lloyd-Jones DM. Care concordant with guidelines predicts decreased long-term mortality in patients with unstable angina pectoris and non-ST-elevation myocardial infarction. Am J Cardiol. 2004;93:1218–1222. doi: 10.1016/j.amjcard.2004.01.063. [DOI] [PubMed] [Google Scholar]
- 15.Huffman MD, Rao KD, Pichon-Riviere A, Zhao D, Harikrishnan S, Ramaiya K, Ajay VS, Goenka S, Calcagno JI, Caporale JE, Niu S, Li Y, Liu J, Thankappan KR, Daivadanam M, van Esch J, Murphy A, Moran AE, Gaziano TA, Suhrcke M, Reddy KS, Leeder S, Prabhakaran D. A cross-sectional study of the microeconomic impact of cardiovascular disease hospitalization in four low- and middle-income countries. PLoS One. 2011;6:e20821. doi: 10.1371/journal.pone.0020821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Roe MT, Messenger JC, Weintraub WS, Cannon CP, Fonarow GC, Dai D, Chen AY, Klein LW, Masoudi FA, McKay C, Hewitt K, Brindis RG, Peterson ED, Rumsfeld JS. Treatments, trends, and outcomes of acute myocardial infarction and percutaneous coronary intervention. J Am Coll Cardiol. 2010;56:254–263. doi: 10.1016/j.jacc.2010.05.008. [DOI] [PubMed] [Google Scholar]
- 17.Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey DE, Jr., Chavey WE, 2nd, Fesmire FM, Hochman JS, Levin TN, Lincoff AM, Peterson ED, Theroux P, Wenger NK, Wright RS, Smith SC, Jr., Writing Group M. ACCF/AHA Task Force M 2011 ACCF/AHA focused update incorporated into the ACC/AHA 2007 guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2011;123:e426–579. doi: 10.1161/CIR.0b013e318212bb8b. [DOI] [PubMed] [Google Scholar]
- 18.Task Force on the Management of ST Elevation Acute Myocardial Infarction of the European Society of Cardiology. Steg PG, James SK, Atar D, Badano LP, Blomstrom-Lundqvist C, Borger MA, Di Mario C, Dickstein K, Ducrocq G, Fernandez-Aviles F, Gershlick AH, Giannuzzi P, Halvorsen S, Huber K, Juni P, Kastrati A, Knuuti J, Lenzen MJ, Mahaffey KW, Valgimigli M, van ’t Hof A, Widimsky P, Zahger D. ESC guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J. 2012;33:2569–2619. doi: 10.1093/eurheartj/ehs215. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.