Highlights
-
•
In this retrospective cross-sectional analysis of STEMI and NSTEMI patients, Median time-to-ECG was 3.0 min shorter for male patients. This disparity in care remained even when assessing a subsample of patients all with a chief complaint of chest pain.
-
•
Male STEMI patients also had a shorter wait time for PCI compared to females. However, sex was not a significant predictor of PCI delay after controlling for time-to-ECG. Indicating time-to-ECG is a critical first step in timely treatment.
-
•
Patients that experienced PCI delay had >2-fold greater likelihood of in-hospital mortality compared to those that achieved time-to-PCI guidelines.
This study sought to assess sex differences in timely diagnosis (time-to-ECG) and treatment (time-to-percutaneous coronary intervention (PCI)) of ST-elevation myocardial infarction (STEMI) and Non-STEMI (NSTEMI) patients utilizing a retrospective cross-sectional analysis of 1,098 STEMI (306 females and 792 males) and 2,179 NSTEMI (747 females, 1,432 males) patients that presented to 2 urban EDs between January 2022 and December 2024 was performed. Sex differences in time-to-ECG were assessed in both STEMI and NSTEMI patients, whereas differences in time-to-PCI were assessed in STEMI patients only. Time-to-ECG and time-to-PCI were compared continuously, as well as categorically (ECG delay = time-to-ECG >10 min and PCI delay = time-to-PCI >90 min or >120 min when a transfer occurred). Median time-to-ECG was 3.0 min shorter for male STEMI and NSTEMI patients. Males also had a reduced likelihood of an ECG delay (OR: 0.64 [95% CI 0.51 to 0.82]). Sex disparities remained when assessing only patients with a chief complaint of chest pain (OR:0.74 [95% CI 0.56 to 0.97]). Male STEMI patients also had a shorter wait time for PCI compared to females (walk-in: 1:26:00 vs 1:41:00, transfer: 2:19:30 vs 2:44:30, respectively). However, sex was not a significant predictor of PCI delay after controlling for time-to-ECG. In conclusion, sex disparities were found in time-to-ECG for STEMI and NSTEMI patents, as well as time-to-PCI for STEMI patients. However, sex was not significantly associated with PCI delay after controlling for time-to-ECG. This highlights the importance of timely diagnosis to ensure timely revascularization in acute myocardial infarction patients.
Early identification and diagnosis are pivotal in the management of acute myocardial infarction (AMI) patients. The American Heart Association (AHA) and American College of Cardiology (ACC) recommend that an electrocardiogram (ECG) be performed within 10 min of arrival to the emergency department (ED) in patients presenting with ischemic symptoms to allow for rapid risk stratification and initiation of appropriate interventions. ,, It is recommended that reperfusion using percutaneous coronary intervention (PCI) is performed in ST-elevation myocardial infarction (STEMI) patients within 90 min from arrival or 120 min if a facility transfer is involved. While crucial for improving patient outcomes, these recommendations are not always met. Further, research suggests sex-based disparities exist in the adherence to these recommendations. ,,,
Past research has shown conflicting findings related to sex-differences in time-to-ECG. ,,,, While several studies have shown male AMI patients experience shorter wait times to ECG, others have shown that sex does not impact time-to-ECG in AMI patients. ,,,,,,,,,, Differences in study sample demographics and clinical characteristics may be the reason behind the conflicting study results, as past research has shown other demographic and clinical factors such as age, race, primary language, chief complaint, etc. can also impact timely care. ,,,, Failure to adequately control for potential confounding variables could obscure the true relationship between sex and time to initial ECG after ED arrival. There is also evidence that sex disparities exist in timely intervention. A systematic review of 34 studies assessing the effect of sex on time-to-PCI in adult patients reported that 91% of the studies observed longer time-to-PCI in female STEMI patients compared to males. However, more information is needed to determine whether these sex-based disparities exist independent of delays in time-to-ECG. Therefore, we sought to assess differences in time-to-ECG and time-to-PCI by sex in AMI patients and determine whether sex differences persist after controlling for additional demographic and clinical characteristics.
Methods
We performed a retrospective cross-sectional analysis using data from 1,098 STEMI and 2,179 Non ST-elevation myocardial infarction (NSTEMI) patients that presented to 2 urban hospitals in Portland, OR between January 1, 2022 and December 31, 2024. Patients were included if they had a confirmed diagnosis of a STEMI or NSTEMI during the hospital encounter, including those that presented to the ED or were transferred from another facility for treatment of an STEMI or NSTEMI. Patients were excluded if they were <18 years of age, were transferred to 1 of the 2 participating hospitals >24 hours after arrival at the outside hospital, presented to the hospital with signs, symptoms, or complaints that were not consistent with AMI but were later diagnosed with AMI, and/or showed a suspected prior MI on their ECG without any other positive cardiac marker (Troponin I/ Troponin T results >upper limits of normal or Creatine Kinase-MB peak >4%). This study was approved by the Providence St. Joseph Health institutional review board, with waiver of informed consent.
All data were extracted from patient electronic health records by a trained nurse abstractor; data elements included patient demographics, patient reported chief complaint at ED admission, hospital encounter level data, comorbidities, and clinical outcomes. STEMI and NSTEMI diagnoses, as well as comorbid conditions were defined by International Classification of Diseases, 10th Revision, Clinical Modification codes (ICD-10-CM). ECG and PCI were defined by Current Procedural Terminology codes and Procedure codes (ICD-10-PCS) assigned during hospital encounter or at discharge. Aggregate data from this study are available from the corresponding author upon reasonable request.
Primary outcomes
Time-to-ECG was calculated as the time from arrival at first hospital to the time of first ECG for all AMI patients that had their ECG performed after hospital arrival. Patients were then categorized into having an ECG delay if the time-to-ECG was >10 min. Time-to-PCI was assessed in STEMI patients only and was calculated as time from arrival at first hospital to time of PCI. A PCI delay was defined as time-to-PCI >90 min when no transfer was involved or time-to-PCI >120 min when a facility transfer was involved. When a transfer was involved, distance of transfer was considered in the analysis and patients were grouped into those that transferred within 30 miles and those that transferred over 30 miles. Differences in these time-based outcomes were assessed in patients when stratified by sex, categorized as male or female.
Secondary outcome
In-hospital mortality was defined as a death during index hospitalization and analyzed as a binary term (Y/N).
Analysis
Descriptive statistics were performed to summarize baseline characteristics of the patients. Univariate analysis of variance and chi-square goodness of fit tests were used when appropriate to test for significant differences between males and females. Nonparametric tests were used to compare median values across groups for measures with nonnormal distributions. Univariate and multivariate linear regression analyses were performed to evaluate the significance of sex on study outcomes. Multivariate analyses controlled for the following covariates: age, race (White and all other races, as all other races had small sample sizes), ethnicity (Hispanic and non-Hispanic), body mass index (BMI), language, (English and all other languages, as all other primary languages had small sample sizes), residence (urban and rural defined using patient zip code and based on Rural-Urban Commuting Area Code scores of 1–3 and 4–10, respectively), insurance (public (Medicare and Medicaid) and private (commercial, government, other)), baseline comorbid diagnosis of diabetes, hypertension, hypercholesterolemia, and chronic heart failure (CHF), as well as chief complaint at ED arrival (chest pain and other). When assessing predictors of time-to-PCI, time-to-ECG was also added to the models. Distance transferred was added to the linear regression model assessing time-to-PCI in transfer patients separately. The Shapiro Wilk test was used to evaluate homoscedasticity within the regression analyses. Given the lack of homoscedasticity found, dependent variables (time-to-ECG and time-to-PCI) were transformed using natural log transformation, which eliminated the heteroscedasticity, satisfying the assumptions of linear regression. Following transformation, standard regression diagnostics were used to check for outliers and assess normality of residuals.
Univariate and multivariate logistic regression models were also run to calculate the odds ratios for ECG delay and PCI delay, again controlling for the same demographic and clinical covariates as in the linear regression models. Additionally, logistic regression models were performed to assess the association of PCI delay with in-hospital mortality in STEMI patients. These models were controlled for sex, age and clinical covariates that have been shown to influence patient outcomes (BMI, diabetes, hypertension, hypercholesterolemia, and CHF diagnoses). All analyses were performed using SPSS V. 29 (Armonk, NY: IBM Corp) and R program (R Core Team (2024). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. < https://www.R-project.org/ >). An alpha of.05 was used to assess significance.
Results
Patient characteristics by AMI diagnosis and sex are presented in Table 1 . Patients were predominately male (68%) and more likely to have been diagnosed with an NSTEMI (67%). The mean age of the overall sample was 68.2 ± 13.3 years (range 21–109 years); 76% of patients were White, 94% were non-Hispanic, and 92% spoke English as their primary language. Further, patients in the sample were more likely to have public health insurance (70%) and reside in an urban area (88%). The median time-to-ECG for the overall cohort was 9.0 min, with STEMI patients reporting slightly shorter wait times than NSTEMI patients (8.0 min and 9.0 min, respectively). Median time-to-PCI (STEMI patients only) when no transfer was involved was 1.2 hr, whereas when a facility transfer occurred the median time was 2.4 hr.
Table 1
Characteristics of study sample by AMI diagnosis and sex
| Overall (n = 3,277) | NSTEMI (n = 2,179) | STEMI (n = 1,098) | Male (n = 2,224) | Female (n = 1,053) | |
|---|---|---|---|---|---|
| Sex (% female) | 32.1% | 34.3% | 27.9% | – | – |
| Diagnosis | |||||
| NSTEMI | 66.5% | – | – | 64.4% | 70.9% |
| STEMI | 33.5% | – | – | 35.6% | 29.1% |
| Age (mean ± SD) | 68.4 ± 13.2 | 69.9 ± 13.0 | 65.3 ± 12.9 | 67.0 ± 12.7 | 71.3 ± 13.7 |
| Race | |||||
| Alaska native/American Indian/Pacific Islander | 1.1% | 1.0% | 1.9% | 0.9% | 1.7% |
| Asian | 5.2% | 4.6% | 8.7% | 5.3% | 4.1% |
| black/African American | 3.0% | 3.6% | 1.8% | 2.6% | 4.0% |
| white | 76.3% | 76.4% | 76.0% | 76.8% | 75.2% |
| other | 1.9% | 2.2% | 1.4% | 1.8% | 2.5% |
| unknown | 11.9% | 11.7% | 16.0% | 11.9% | 11.8% |
| Ethnicity | |||||
| Hispanic | 3.6% | 4.0% | 2.8% | 3.6% | 3.6% |
| non-Hispanic | 93.7% | 93.8% | 93.5% | 93.6% | 93.9% |
| unknown | 2.7% | 2.2% | 3.6% | 2.7% | 2.5% |
| Primary language | |||||
| English | 91.9% | 91.1% | 93.3% | 92.8% | 89.8% |
| All other languages | 8.1% | 8.9% | 6.7% | 7.2% | 10.2% |
| Insurance | |||||
| Public | 69.7% | 73.7% | 61.8% | 64.2% | 81.4% |
| private | 30.3% | 26.3% | 38.2% | 35.8% | 18.6% |
| BMI (mean ± SD) | 29.6 ± 6.3 | 29.7 ± 6.5 | 29.4 ± 6.0 | 29.5 ± 5.7 | 29.8 ± 7.5 |
| Residence | |||||
| Urban | 87.8% | 89.5% | 84.4% | 87.4% | 88.4% |
| Micropolitan | 9.9% | 8.8% | 12.0% | 10.1% | 9.2% |
| rural | 2.4% | 1.7% | 3.6% | 2.5% | 2.4% |
| Comorbidities | |||||
| Hypertension | 59.4% | 63.6% | 51.1% | 58.4% | 61.4% |
| Hypercholesterolemia | 51.3% | 48.9% | 55.9% | 53.6% | 46.3% |
| Diabetes | 11.4% | 12.9% | 8.3% | 11.5% | 11.2% |
| CHF | 34.2% | 38.0% | 26.7% | 31.7% | 39.5% |
| Chief Complaint | |||||
| Chest Pain | 67.6% | 74.4% | 64.5% | 69.9% | 62.8% |
| Other | 32.4% | 25.6% | 35.5% | 30.1% | 37.2% |
| Time to ECG (Median; mM:ss [IQR]) | 09:00 (06:00–15:00) | 08:00 (05:00–12:00) | 09:00 (06:00–16:00) | 08:00 (05:00–13:00) |
10:00
(07:00–19:00) |
| Time to PCI (Median; h:mm:ss [IQR]) | |||||
| No transfer | 1:10:00 (0:50:00–1:36:00) | 1:10:00 (0:50:00–1:36:00) | – | 1:26:00 (1:12:00–1:57:00) | 1:41:00 (1:15:00–2:46:00) |
| Transfer | 2:24:00 (1:53:45–3:28:00) | 2:24:00 (1:53:45–3:28:00) | – | 2:19:30 (1:50:00–3:15:45) | 2:44:30 (2:00:00–3:44:45) |
Compared to males, females were on average older (71.3 ± 13.7 vs 67.0 ± 12.7) and more likely to have public health insurance (81% vs 64%). Females were also more likely to have a history of hypertension (61% vs 58%) and CHF (40% vs 32%), but less likely to have a history of hypercholesterolemia (46% vs 54%). Chief complaint reported at the ED was also significantly different between males and females, with females less likely to report a chief complaint of chest pain (63% vs 70%) and more likely to report other symptoms such as dyspnea, jaw pain, nausea, fatigue, etc. (38% vs 30%) compared to males.
Time-to-ECG
Comparisons in the median time-to-ECG by sex for both STEMI and NSTEMI patients are provided in Table 2 . Time-to-ECG after ED arrival was significantly longer in both female STEMI and NSTEMI patients compared to male patients. Specifically, median time-to-ECG was 4.0 min longer for female compared to male STEMI patients (11.0 min vs 7.0 min, respectively), and 3.0 min longer for female versus male NSTEMI patients (12.0 min vs 9.0 min, respectively). Female STEMI and NSTEMI patients were also more likely to experience an ECG delay (time to ECG >10 min) compared to males (STEMI: 42% vs 30%; NSTEMI: 52% vs 37%, respectively).
Table 2
Sex comparison of time to ECG and time to PCI
| Time to ECG | Male STEMI | Female STEMI |
|---|---|---|
| Time to ECG (mm:ss [IQR]) | 07:00 [03:00–13:00] (n = 792) | 10:00 [07:00–19:00] (n = 306) * |
| Chief complaint: Chest pain—Time to ECG (mm:ss [IQR]) | 7:00 [5:00–10:00] (n = 505) | 8:00 [5:00–11:00] (n = 175) |
| Chief complaint: Other—Time to ECG (mm:ss [IQR]) | 10:00 [6:30–17:00] (n = 287) | 22:30 [12:00–31:15] (n = 131) * |
| ECG Delay (%) | 29.8% | 42.4% * |
| Male NSTEMI | Female NSTEMI | |
| Time to ECG (mm:ss [IQR]) | 08:00 [05:00–14:00] (n = 1,432) | 11:00 [07:00–20:00] (n = 747) * |
| Chest pain- Time to ECG (mm:ss [IQR]) | 7:00 [5:00–11:00] (n = 895) | 9:00 [6:00–14:00] (n = 436) * |
| Other symptoms- Time to ECG (mm:ss [IQR]) | 11:00 [7:00–23:00] (n = 449) | 15:00 [9:00–35:00] (n = 285) * |
| ECG Delay (%) | 37.1% | 51.5%* |
| PCI | Male STEMI | Female STEMI |
| Time to PCI (no transfer) (h:mm:ss [IQR]) | 1:26:00 [1:12:00–1:57:00] (n = 421) | 1:41:00 [1:15:00–2:46:00] (n = 129) * |
| Time to PCI (transfer) (h:mm:ss [IQR]) | 2:19:30 [1:50:00–3:15:45] (n = 265) | 2:44:30 [2:00:00–3:44:45] (n = 109) * |
| Time to PCI (≤30 miles) | 2:00:00 [1:41:45–2:39:15] (n = 177) | 2:10:00 [1:48:00–3:31:00] (n = 69) * |
| Time to PCI (>30 miles) | 3:03:00 [2:28:15–4:02:15] (n = 88) | 3:06:00 (2:37:00–4:08:30] (n = 40) |
| PCI delay—Overall (%) | 45% (n = 311) | 54% * (n = 129) |
| PCI delay—no transfer (%) | 38% (n = 162) | 47% * (n = 60) |
| PCI delay—transfer (%) | 56% (n = 149) | 63% * (n = 69) |
Tables 3–5 display the results from the univariate and multivariate linear regression analyses assessing significant associations between sex, demographic and clinical covariates and log time-to-ECG. Sex had a significant effect on log time-to-ECG in the univariate adjusted model and this effect remained significant after controlling for patient demographic and clinical covariates. Specifically, male sex was associated with significantly shorter wait times to ECG. The multivariate linear regression model also found chief complaint of chest pain and previous diagnosis of hypercholesterolemia to be significantly associated with shorter wait times to ECG in the overall sample. When assessing STEMI and NSTEMI patients separately, female sex was still significantly associated with longer wait time to ECG in the univariate and multivariate models. A chief complaint of chest pain was significantly associated with faster time-to-ECG for both STEMI and NSTEMI patients. Additionally, a previous diagnosis of CHF was also found to be significantly associated with time to ECG in NSTEMI patients.
Table 3
Linear regression models assessing the association between sex and log time to ECG in the overall study sample
| Regression coefficient [95% CI] | p-value | |
|---|---|---|
| Univariate analysis | ||
| Sex [male] | −0.320 [−0.410 to −0.230] | <0.001 |
| R 2 = 0.024, n = 1,969 | ||
| Multivariate analysis | ||
| Sex [male] | −0.203 [−0.299 to −0.107] | <0.001 |
| Age [yr] | 0.003 [−0.001 to 0.007] | 0.16 |
| Race [White] | −0.014 [−0.109 to 0.082] | 0.77 |
| Ethnicity [non-Hispanic] | 0.133 [−0.097 to 0.362] | 0.26 |
| Residence [urban] | 0.100 [−0.035 to 0.234] | 0.15 |
| Insurance [public] | −0.025 [−0.135 to 0.085] | 0.66 |
| Language [not English] | 0.155 [−0.031 to 0.340] | 0.10 |
| BMI [Kg/m 2] | −0.006 [−0.013 to 0.001] | 0.12 |
| Diabetes [yes] | −0.106 [−0.240 to 0.028] | 0.12 |
| Hypertension [yes] | −0.008 [−0.103 to 0.087] | 0.87 |
| CHF [yes] | 0.073 [−0.026 to 0.172] | 0.15 |
| Hypercholesterolemia [yes] | −0.106 [−0.200 to −0.012] | 0.03 |
| Chief complaint [chest pain] | −0.743 [−0.842 to −0.644] | <0.001 |
| R 2 = 0.193, n = 1,522 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
