Calling 911 during acute coronary syndromes (ACS) decreases time to treatment and may improve prognosis. Women may have more atypical ACS symptoms compared to men, but few data are available on differences in gender and ACS symptoms in calling 911. In this study, patient interviews and structured chart reviews were conducted to determine gender differences in calling 911. Calls to 911 were assessed by self-report and validated by medical chart review. Of the 476 patients studied, 292 (61%) were diagnosed with unstable angina and 184 (39%) with myocardial infarctions (MIs). Overall, only 23% of patients called 911. Similar percentages of women and men with unstable angina called 911 (15% and 13%, respectively, p = 0.59). In contrast, women with MIs were significantly more likely to call 911 than men (57% vs 28%, p <0.001). After adjustment for sociodemographic factors, health insurance status, history of MI, the left ventricular ejection fraction, Global Registry of Acute Coronary Events (GRACE) score, and ACS symptoms, women were 1.79 times more likely to call 911 during an MI than men (prevalence ratio 1.79, 95% confidence interval 1.22 to 2.64, p <0.01). In conclusion, the findings of the present study suggest that initiatives to increase calls to 911 are needed for women and men.
Previous studies on differences in the proportion of men and women having acute coronary syndromes (ACS) who call 911 have produced inconsistent results. Although some studies have found no difference in calls to 911 between women and men with ACS, other studies restricted to patients with myocardial infarctions (MIs) found that women call 911 more often than men. However, few data have described gender differences in calls to 911 by ACS type or by specific ACS symptom experienced. Differences in the presenting characteristics and pathobiology of patients with MIs versus unstable angina pectoris (UAP) may influence patterns of 911 calling. Therefore, we analyzed data from an ongoing, observational cohort of patients with ACS to determine gender differences in calls to 911 and investigated whether clinical characteristics, ACS type, or ACS symptoms modified these differences.
Methods
Patients from the Prescription Use, Lifestyle, Stress Evaluation (PULSE) study at Columbia University Medical Center constituted the study population. PULSE is an ongoing, observational, single-site, prospective cohort study of the prognostic risk conferred by depressive symptoms and clinical depressive disorders at the time of ACS. Patients with UAP, ST-segment elevation MIs, or non–ST-segment elevation MIs by published American College of Cardiology and American Heart Association definitions were included and were recruited from Columbia University Medical Center within 1 week of hospitalization for their ACS. From February 1, 2009, and June 30, 2010, 500 patients were recruited; 24 (5%) were excluded from these analyses because of missing data on calls to 911. The present analysis included 476 English- or Spanish-speaking patients aged ≥18 years who either presented to the emergency department at Columbia University Medical Center or were transferred from nearby hospitals. The institutional review board of Columbia University Medical Center approved this study, and all participants provided informed consent.
Calling 911 was self-reported during an in-hospital interview within 7 days of admission and was verified by review of 100 randomly selected medical records. During the interview, patients were asked whether they (1) called 911, (2) went to the emergency department, or (3) called or went to a physician’s office at ACS onset. No other information on calling 911 was collected. Demographic, psychosocial, and clinical factors were assessed within 7 days of enrollment. Age, gender, ethnicity (Hispanic or Latino vs other), English fluency, marital status, high school education, health insurance over the previous 2 years and insurance with Medicaid or Medicare were assessed during a study interview. ACS symptom assessment was restricted to 2 typical (chest pain, arm and/or jaw pain) and 3 atypical (dyspnea, nausea and/or vomiting, syncope) symptoms, because these 5 symptom clusters have been identified as independent predictors of hospital mortality ; time course of ACS symptoms was assessed and dichotomized as constant or intermittent for analysis. ACS severity was determined using the Global Registry of Acute Coronary Events (GRACE) risk score. The GRACE score includes age and clinical parameters at presentation (heart rate, systolic blood pressure, serum creatinine, congestive heart failure, and the presence of cardiac arrest, ST-segment elevation, and cardiac enzymes or markers) and provides an estimate of mortality within 6 months of ACS. The Beck Depression Inventory was used to assess depressive symptoms; a score ≥10 was used to identify clinically significant depression, as this score has been independently associated with poor cardiovascular prognosis. History of MI and the left ventricular ejection fraction (LVEF) during admission were abstracted from the medical chart.
Previous studies have documented potential differences in prognosis and presentation for patients with MIs versus UAP. Therefore, gender differences in calls to 911 were first examined by ACS type. Because rates of 911 calling in this study were markedly different for patients with MIs versus UAP, all analyses were stratified by ACS type. Patient characteristics were calculated separately for women and men. The percentage of study patients calling 911 was calculated by characteristics including age, race, ethnicity, native English status, marital status, completion of high school, the LVEF, GRACE risk score, history of MI, insurance over the past 2 years, Medicare or Medicaid insurance, depressive symptoms, nausea and/or vomiting, syncope, constant symptoms, chest pain, arm and/or jaw pain, and dyspnea, with the statistical significance of differences determined using Student’s t tests for continuous variables and chi-square tests for categorical variables. Binomial regression was used to calculate prevalence ratios of calling 911 for women compared to men. Prevalence ratios are recommended instead of odds ratios for cross-sectional studies with common outcomes. An initial model was unadjusted (model 1). Subsequent models included progressive adjustment for age, race and ethnicity, public health insurance status, health insurance status within the past 2 years, high school education, and marital status (model 2); model 2 variables and history of MI model 3); model 3 variables and ACS symptoms (nausea and/or vomiting, syncope, and constant vs inconstant; model 4); and model 4 variables and the LVEF and GRACE risk score (model 5). Variables were selected for adjustment if significant at the p <0.10 level on calls to 911 in the unadjusted model or if identified in previous studies as potential determinants of calls to 911 (GRACE score, marital status, insurance status). Among patients with MIs, subgroup analyses were performed to examine the consistency of the relation between gender and calls to 911. Multiplicative interaction was assessed using the full population and including main effects and interaction terms (e.g., race × gender). Data on covariates were missing for 100 participants (21%). We used multiple imputation with chained equations and 5 data sets to account for the missing data. The association between gender and 911 calling was similar using a complete case analysis (data not presented). All analyses were conducted using Stata version 11 (StataCorp LP, College Station, Texas).
Results
Statistically significant differences in baseline characteristics were present between women and men with UAP and between women and men with MIs ( Table 1 ). There were also significant differences in calls to 911 by baseline characteristics in patients with UAP versus MI ( Table 2 ).
| Variable | UAP | MI | ||||
|---|---|---|---|---|---|---|
| Women (n = 101) | Men (n = 191) | p Value ∗ | Women (n = 60) | Men (n = 124) | p Value ∗ | |
| Age (yrs) | 65.9 ± 11.8 | 63.1 ± 10.9 | 0.184 | 64.3 ± 11.3 | 61.3 ± 11.6 | 0.101 |
| Black | 24% | 12% | 0.010 | 28% | 18% | 0.099 |
| Hispanic | 30% | 25% | 0.347 | 50% | 32% | 0.020 |
| Native English speaker | 73% | 72% | 0.888 | 53% | 65% | 0.127 |
| Married | 51% | 73% | <0.001 | 37% | 64% | <0.001 |
| At least high school education | 72% | 84% | 0.014 | 60% | 81% | 0.003 |
| LVEF (%) | 55 ± 9 | 51 ± 10 | <0.001 | 47 ± 13 | 47 ± 12 | 0.789 |
| GRACE score | 87.6 ± 25.1 | 85.3 ± 24.4 | 0.463 | 100.3 ± 33.5 | 92.6 ± 30.3 | 0.117 |
| STEMI | NA | NA | NA | 27% | 32% | 0.440 |
| History of myocardial infarction | 29% | 31% | 0.669 | 32% | 23% | 0.163 |
| Insured over past 2 yrs | 91% | 90% | 0.888 | 92% | 89% | 0.634 |
| Medicaid/Medicare insurance | 62% | 56% | 0.378 | 65% | 52% | 0.099 |
| Current depression † | 28% | 11% | <0.001 | 26% | 12% | 0.016 |
| Nausea/vomiting | 23% | 9% | 0.001 | 31% | 16% | 0.025 |
| Syncope | 11% | 9% | 0.575 | 12% | 11% | 0.909 |
| Constant symptoms | 26% | 25% | 0.851 | 65% | 51% | 0.069 |
| Chest pain | 90% | 85% | 0.237 | 78% | 81% | 0.714 |
| Arm/jaw pain | 41% | 27% | 0.020 | 42% | 42% | 0.955 |
| Dyspnea | 60% | 50% | 0.099 | 43% | 39% | 0.549 |
∗ Comparing differences across gender for UAP and MI using chi-square analysis for categorical data and analysis of variance for continuous data.
| Variable | UAP | MI | ||
|---|---|---|---|---|
| (n = 292) | p Value ∗ | (n = 184) | p Value ∗ | |
| Age (yrs) | 0.79 | 0.61 | ||
| <65 | 13% | 36% | ||
| ≥65 | 14% | 40% | ||
| Race | 0.74 | 0.045 | ||
| Not black | 13% | 34% | ||
| Black | 15% | 51% | ||
| Ethnicity | 0.026 | 0.034 | ||
| Not Hispanic | 11% | 32% | ||
| Hispanic | 21% | 47% | ||
| Native English speaker | 0.014 | 0.147 | ||
| No | 21% | 44% | ||
| Yes | 10% | 33% | ||
| Marital status | 0.111 | 0.870 | ||
| Not married | 18% | 38% | ||
| Married | 11% | 37% | ||
| Completed high school | <0.001 | 0.052 | ||
| No | 28% | 50% | ||
| Yes | 10% | 34% | ||
| LVEF | 0.506 | 0.643 | ||
| <50% | 13% | 40% | ||
| ≥50% | 16% | 36% | ||
| GRACE risk score | 0.955 | 0.643 | ||
| <87 | 14% | 40% | ||
| ≥87 | 13% | 36% | ||
| History of myocardial infarction | <0.001 | 0.281 | ||
| No | 9% | 35% | ||
| Yes | 24% | 45% | ||
| Insured over past 2 yrs | 0.889 | 0.504 | ||
| No | 15% | 44% | ||
| Yes | 14% | 36% | ||
| Medicare/Medicaid insurance | 0.003 | 0.134 | ||
| No | 7% | 30% | ||
| Yes | 19% | 41% | ||
| Depressive symptoms † | 0.586 | 0.352 | ||
| No | 13% | 37% | ||
| Yes | 16% | 46% | ||
| Syncope | 0.009 | 0.125 | ||
| No | 11% | 35% | ||
| Yes | 29% | 52% | ||
| Constant symptoms | <0.001 | 0.001 | ||
| No | 7% | 24% | ||
| Yes | 32% | 48% | ||
| Chest pain | 0.282 | 0.419 | ||
| No | 8% | 43% | ||
| Yes | 14% | 36% | ||
| Arm/jaw pain | 0.560 | 0.617 | ||
| No | 13% | 39% | ||
| Yes | 15% | 35% | ||
| Dyspnea | 0.722 | 0.313 | ||
| No | 13% | 35% | ||
| Yes | 14% | 42% | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
