Sex differences in time to testing for pulmonary embolism in the emergency department: a survival analysis.

Sex Differences in Time to Testing for Pulmonary Embolism in the Emergency Department: A Survival Analysis - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Clin Ther . Author manuscript; available in PMC: 2026 Jun 4. Published in final edited form as: Clin Ther. 2025 Jun 4;47(8):581–586. doi: 10.1016/j.clinthera.2025.04.014 Sex Differences in Time to Testing for Pulmonary Embolism in the Emergency Department: A Survival Analysis Brandon Maughan 1 , Alexa Redmond 1 , Yoona Shim 1 , Nick Patrick 1 , Mike Hildebrand 1 , Bory Kea 1 , Esther Choo 1 , Angela Jarman 2 Author information Article notes Copyright and License information 1 Department of Emergency Medicine, Oregon Health & Science University, Portland OR, USA 2 Department of Emergency Medicine, University of California – Davis, Sacramento, CA, USA ✉ Address for Correspondence: Brandon C. Maughan, MD, MHS, MSHP, Department of Emergency Medicine, Center for Policy and Research in Emergency Medicine, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, mail code CDW-200, Portland, Oregon 97239-3098, maughabr@ohsu.edu Issue date 2025 Aug. PMC Copyright notice PMCID: PMC12259225  NIHMSID: NIHMS2088010  PMID: 40467347 The publisher's version of this article is available at Clin Ther Abstract Study Objective: Female patients experience delays in diagnostic testing for major cardiovascular diseases such as myocardial infarction and stroke. The objective of this study was to assess sex differences in the time to initial diagnostic testing among emergency department (ED) patients evaluated for pulmonary embolism (PE). Methods: The sample included 15,038 adult patients evaluated for PE in 3 EDs from 2017–2023. The primary outcome was the time from ED arrival to the first order of a diagnostic test. Secondary outcomes included time to CT scan completion and time to admission for patients diagnosed with PE. Data on patient demographics, medical history, diagnostic test orders, and ED-operational factors were extracted from the electronic medical record. We performed survival analysis with a Cox proportional hazards model using an elastic net regression strategy for variable selection. Poisson regression with robust standard errors was used to measure the average marginal association of female sex with time to first PE test order. Results: Female sex was associated with a slower time to first PE test (hazard ratio (HR) 0.92 [95% confidence interval (CI) 0.88, 0.95], p<0.001) with an average delay of 7.2 minutes [95% CI 4.6, 9.8] compared to male patients. Secondary outcomes noted similar delays to CT scan completion (HR 0.85 [95% CI 0.8, 0.89], average 14.1 minutes [95% CI 9.3, 18.8]) and hospital admission (HR 0.83 [95% CI 0.71, 0.98], average 18.7 minutes [95% CI 1, 36.4]) for female patients compared to males. Conclusions: Female patients experience slower times to diagnostic testing for PE in the emergency department. Future research should examine sex-associated delays to anticoagulation and other treatments. INTRODUCTION Rapid recognition and treatment are associated with better outcomes for many emergent medical conditions. Despite this, female patients appear to consistently receive slower and less frequent diagnostic testing and treatment for major cardiovascular (CV) diseases. In patients with chest symptoms, female patients receive fewer electrocardiograms and receive them less quickly after emergency department (ED) arrival compared to male patients, and they are less likely to be referred for invasive cardiac testing. 1 – 4 Similarly, female patients with acute ischemic stroke have longer door-to-CT times and are less likely to have prehospital notification of suspected stroke symptoms compared to men. 5 – 7 These disparities in treatment and subsequent outcomes have variably been attributed to sex-based biological differences, gender-based behavioral and cultural differences, and implicit bias. 8 – 10 Many authors have advocated for research, education, and changes in professional guidelines to address these disparities. 10 – 12 Pulmonary embolism (PE) is the third leading cardiovascular cause of death after myocardial infarction and stroke, 13 and early diagnosis and treatment of PE is associated with improved clot resolution and lower rates of adverse outcomes. 14 – 16 In contrast to other major cardiovascular diseases, female patients evaluated in EDs receive more testing for PE than males. 17 , 18 This observation raises the question of whether the slower speed of testing observed for female patients with other major cardiovascular diseases is replicated in patients with suspected PE or whether there is greater equity in the speed of PE testing after ED arrival. The goal of this investigation is to assess whether patient sex is associated with the time to initial PE testing for patients evaluated in the emergency department (ED). We hypothesize that female patients have a slower time to initial testing than male patients, as seen in other cardiovascular diseases. PARTICIPANTS AND METHODS Study Design: This investigation was a retrospective cohort study of patients evaluated for PE during emergency department (ED) evaluation between January 1, 2017 and December 31, 2023. Study Setting: The study included patients evaluated in EDs at an urban academic medical center (Oregon Health & Science University), an urban hospital without an emergency medicine residency program (Adventist Health Portland [AHP]), and a suburban hospital without an emergency medicine residency program (Hillsboro Medical Center [HMC]). These EDs had approximately 39,000, 42,000, and 28,000 annual patient visits during the study interval, respectively. Eligible dates for HMC and AHP were determined by the dates when they joined the electronic medical record (EMR) system at OHSU; we included data for patients who were evaluated at HMC starting in November 2018 and at AHP starting in March 2023. Patient Population. We included adult patients (age 18–89 years) who underwent D-dimer or CT pulmonary angiography (CTPA) testing their ED evaluation. We included a single visit from each patient; subsequent encounters during the study interval were not included. Exclusion criteria. Based on literature and clinical experience, we excluded several categories of patients whose initial ED triage was likely to bias the speed of initial testing for PE. 19 Patients were excluded if they were triaged with chief complaints of cardiac arrest; altered mental status or agitation; drug overdose; psychosis or suicidality; or if chief complaint was not recorded. Patients with potentially life-threatening complaints (e.g., myocardial infarction, stroke, trauma) are likely to involve immediate consultation and co-management with a specialty service and were therefore excluded. We also excluded patients with a chief complaint of “referral,” “abnormal lab result,” “pulmonary embolism,” “deep vein thrombosis,” or “postoperative complication” as chart review indicated a substantial proportion of these patients were referred from other providers for PE testing. We excluded patients who eloped, left without being seen by a physician, left against medical advice, died in the ED, or were transferred to another facility. Patients with a positive hCG test result or with a chief complaint of “pregnancy” or “pregnancy problem” were excluded, as were other patients whose first PE test was lung perfusion scintigraphy. Both groups were very small and would be expected to have more complex decision-making requirements that could confound the timing of initial test ordering. Since our goal was to assess PE testing patterns among ED physicians, we excluded patients whose initial PE testing occurred within 5 minutes before or after an ED physician order indicating that an admission discussion with an inpatient team had been completed (i.e., during which the admitting team may ask the ED physician to order testing such as for PE). Finally, we excluded patients evaluated when the ED census was above a site-specific 98 th percentile since we anticipated from clinical experience that this context was often associated with testing patterns and delays not representative of routine ED care. Data Collection and Validation The patient cohort was generated by searching an institutional EMR (Epic Systems; Madison, WI) database using SQL (Structured Query Language) to identify eligible patients. We identified test orders based on CPT (Current Procedural Terminology) codes or institutional order codes. Variables . We collected several groups of variables: Patient demographic variables included age, sex (i.e., assigned at birth), race, ethnicity, insurance status, and primary language. Gender was not regularly recorded as a distinct field from sex during the study period and thus was not omitted. Zip code data were used to estimate rates of poverty and unemployment in the patient’s community using Zip code tabulation area (ZCTA) data from the Agency for Healthcare Research and Quality Social Determinants of Health database. 20 ED visit characteristics included a unique identifier for each location, year, hour of ED arrival (daytime arrival between 0700–1859; evening arrival between 1900–0659), ED census at the time of patient arrival (categorized as quartiles based on patient volumes at each ED); Clinical variables included past medical history, chief complaint, and vital signs. Historical diagnoses of interest included chronic lung disease (ICD-10 codes J40-J47), ischemic heart disease (I20-I25), or malignancy (C00-D49). Chief complaints were dichotomized as (1) chest symptoms classically associated with PE including dyspnea, chest pain, or syncope; or (2) any other chief complaint not listed in the exclusion criteria. Triage vital signs (heart rate, blood pressure, respiratory rate, oxygen saturation) were extracted. We expected that vital signs would influence physician testing decisions in a nonlinear fashion, and therefore we included these elements both as continuous variables as well as categorical variables to indicate abnormal vital signs (heart rate ≥ 100 beats/minute, respiratory rate ≥ 22 beaths/minute, or pulse oximetry < 92%, systolic blood pressure < 100 mmHg). Operational variables included indicator and time variables for selected ED diagnostic tests (D-dimer, CTPA, electrocardiogram [EKG], respiratory viral PCR [RVP], and human chorionic gonadotropin [hCG]) and ED disposition (discharge, observation, hospital ward admission, ICU admission, or transfer to another facility). We also collected time data on discharge orders, admission orders, and completion of signout to an inpatient team. Data Validation. Based on an expected final sample of approximately 15,000 charts, we randomly selected a 1% sample (150 charts) for review to validate the accuracy of extracted data elements. Four co-authors were trained on data collection methods based on published standards. 21 Data were extracted from patient charts using standardized electronic data collection forms (REDCap, Vanderbilt University). We compared SQL-collected results with manually-gathered data on patient demographics, time stamps (ED arrival, D-dimer order; CTPA order; ED disposition), ED disposition type (discharge, observation, admission) and PE diagnosis. Since identifying acute PE using retrospective EMR data can be inaccurate, we defined PE diagnosis as patients who had CTPA imaging, an admission or discharge diagnosis of acute PE, and initiation of anticoagulation during their ED encounter. Eighty charts were double-extracted for comparison, with a Cohen’s kappa of 0.98 for interrater reliability. Missingness was reviewed and was less than 1%, which was expected since many demographic and ED visit variables were collected during ED patient registration or were recorded automatically by the EMR (e.g., time stamps for all physician orders). Outcomes The primary outcome of this study was the number of minutes between ED arrival and the patient’s first order for any PE diagnostic test. We also examined each test type separately, including two time-to-event outcomes for D-dimer testing (arrival to D-dimer order; order to D-dimer result) and three outcomes for CT (arrival to CT order; order to CT completed; arrival to CT completed). Time from CT order to CT result was not reported because the latter element is not accurately recorded in the electronic medical record (e.g., preliminary results of PE diagnosis are often provided verbally by radiologists to ED physicians before written reports are finalized in the EMR). Lastly, we examined the time from ED arrival to admission for hospitalized patients with PE. Statistical Analysis We characterized the sample with descriptive statistics including means with standard deviation for normally distributed continuous variables, medians and interquartile ranges for non-normal continuous variables, and proportions for categorical variables. We examined normality of the time outcomes using box plots, tests of skew and kurtosis, and the Shapiro-Francia W’ test, all of which indicated that the primary and secondary outcomes were non-normally distributed. Each time-to-event outcome was analyzed using three approaches. We first assessed Kaplan-Meier survival functions using the log-rank test. Next, we used multivariable Cox proportional hazards models to account for confounders. For the primary outcome, we selected predictor variables for the model using elastic net regression. Elastic net is a hybrid of lasso and ridge regression that offers improved prediction accuracy and better handling of multicollinearity compared to traditional stepwise feature selection. 22 A Bayesian information criterion (BIC) penalty parameter was selected for the elastic net as it had a marginally lower deviance ratio compared to cross-validation. The elastic net approach retained 19 of 30 candidate predictor variables, and these variables were used to generate the Cox model. We tested and confirmed the assumptions of proportional hazards by comparing Kaplan-Meier plots and log-log plots. The average marginal association of the female sex with time-to-event outcomes was examined with a Poisson regression model using robust standard errors with the Stata postestimation margins function. This approach offers more consistent estimators than log-linear regression and avoids the need for back-transformation. 23 The study includes two sensitivity analyses. First, we examined differences in primary results stratified by age (18–49 years vs. 50+ years), based on the concern that two groups could differ due to age-related differences in PE incidence or delays due to pregnancy testing in the younger group. Second, we compared results between the academic and non-academic hospitals (hospital 1 vs. hospitals 2 and 3) to evaluate whether the results were associated with teaching status. All database management and statistical analyses were conducted using Stata/MP v18.0 (StataCorp; College Station, TX). Data presentation is consistent with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. 24 The study protocol was approved by the OHSU Institutional Review Board and research oversight committees at HMC and AHP. RESULTS Our sample included 15,038 eligible patients with PE testing, of which 8,518 (56.6%) were female (sc 1). D-dimer testing was ordered in 9,029 patients (60%) and CTPA was ordered in 9,347 patients (62.2%); 3,338 patients (22.2%) had both tests. A total of 960 patients were diagnosed with PE, resulting in a CT test yield of 10.3%. Seventy percent of patients (n=10,514) had a classical symptom of PE as their chief complaint. Primary Outcome. Kaplan-Meier curves are shown in Figure 2 . The log-rank test indicated a statistically significant difference in the time-to-test functions by sex (p < 0.0001). In the Cox proportional hazards model, female sex was independently associated with longer times to initial PE testing (HR 0.92 [95% CI 0.88, 0.95]) ( Table 2 ). Other factors associated with a longer time to testing included D-dimer as the initial PE diagnostic test, higher ED census, ED arrival during afternoon or evening hours, Medicaid or uninsured status, and history of chronic lung disease. Orders for respiratory viral panels were associated with a longer time to testing whereas pregnancy testing was not. Higher hazard ratios (e.g., faster arrival-to-testing outcomes) were observed in patients who presented with classical PE symptoms as a chief complaint, spoke English as a primary language, were evaluated at a non-teaching hospital, or had abnormal vital signs. Figure 2. Open in a new tab Kaplan Meier plots Table 2. Cox Proportional Hazards Model. Variable HR (95% CI) p Patient Demographics Female sex 0.92 (0.88, 0.95) <0.001 Age (per decade) 1.02 (1.01, 1.04) 0.003 Medicaid or Uninsured 0.97 (0.94, 1.02) 0.34 Primary language English 1.12 (1.05, 1.20) 0.028 Hospital & ED Visit Characteristics Hospital 1 (academic, urban) (reference) 2 (nonacademic, suburban) 1.60 (1.55, 1.66) <0.001 3 (nonacademic, urban) 1.45 (1.35, 1.56) <0.001 ED Census at Time of Arrival Bottom quarile (reference) Second quartile 0.87 (0.82, 0.91) <0.001 Third quartile 0.72 (0.68, 0.76) <0.001 Top quartile 0.64 (0.60, 0.68) <0.001 Year 2017–19 (reference) 2020–21 1.25 (1.19, 1.32) <0.001 2022–23 1.33 (1.25, 1.41) <0.001 Time of Arrival 0700–1859 (reference) 1900–0559 0.92 (0.89, 0.96) <0.001 Clinical Details Chest symptoms as chief complaint 1.30 (1.25, 1.36) <0.001 Medical history Chronic lung disease 0.93 (0.89, 0.97) <0.001 Coronary disease 1.03 (0.99, 1.07) 0.208 Cancer 1.06 (1.02, 1.11) 0.006 Vital Signs in Triage RR > 22 1.20 (1.23, 1.35) <0.001 SpO2 < 92% 1.21 (1.14, 1.28) <0.001 HR > 100 1.07 (1.03. 1.11) <0.001 SBP < 100 1.29 (1.11, 1.34) <0.001 Diagnostic Tests Ordered hCG 0.97 (0.93, 1.03) 0.28 Respiratory Viral PCR 0.86 (0.83, 0.89) <0.001 CXR 0.69 (0.60, 0.72) <0.001 D-dimer 1.32 (1.27, 1.38) <0.001 ICU admission 1.12 (1.05, 1.20) 0.001 Open in a new tab The unadjusted median time from ED arrival to first PE test order was 82 minutes for female patients and 75 minutes for male patients. Using Poisson regression, the average marginal association of female sex with time to initial testing was 7.2 minutes (95% CI 4.6, 9,8). Secondary Outcomes. In patients with D-dimer testing, female patients experienced longer times from ED arrival to initial test orders (HR = 0.93 [95% CI 0.9, 0.97], p < 0.001) compared to male patients, with an average difference of 6.1 minutes (95% CI 2.6, 9.7, p = 0.001). The time between D-dimer order and D-dimer result did not differ by sex ( Table 3 ). In patients with initial CT testing (i.e., no antecedent D-dimer), we observed longer times after ED arrival for female patients to have initial CT orders placed (HR 0.89 [95% CI 0.84, 0.94], average 8.4 minutes [95% CI 7.9, 9]) and to have the CT scan completed (HR = 0.85 [95% CI 0.8, 0.89], average 14.1 minutes [9.3, 18.8]). Female patients who were diagnosed with PE and admitted to the hospital experienced longer times between ED arrival and admission to an inpatient service (HR = 0.83 [0.71, 0.98]; average 18.7 minutes (95% CI 1, 36.4) compared to male patients, after adjustment for age, vital signs, ED census, and other diagnostic tests including D-dimer. Table 3. Primary and Secondary Time-to-Event Results. Patient Group Time Outcome Test Female vs. Male Sex All Arrival to first test order Log-rank test p < 0.0001 Cox multivarible model HR = 0.92 [0.88, 0.95], p < 0.001 Poisson average marginal effect 7.2 min [4.6, 9.8], p < 0.001 D-dimer first (n = 9029) Log-rank test p < 0.0001 Cox multivarible model HR = 0.93 [0.9, 0.97], p = 0.03 Arrival to D-dimer order Poisson average marginal effect 6.1 min [2.6, 9.7], p = 0.001 D-dimer order to result Log-rank test p = 0.90 Cox multivarible model HR = 1 [0.96, 1.04], p = 0.94 Poisson average marginal effect -0.5 min [−2, 1], p = 0.5 CT first (n = 6009) Arrival to CT order Log-rank test p < 0.0001 Cox multivarible model HR = 0.89 [0.84, 0.94], p < 0.001 Poisson average marginal effect 8.4 min (7.9, 9), p < 0.001 CT order to CT complete Log-rank test p = 0.003 Cox multivarible model HR = 0.92 [0.89, 0.97], p = 0.001 Poisson average marginal effect 5.6 min (2.6, 8.3), p < 0.001 Arrival to CT complete Log-rank test p < 0.0001 Cox multivarible model HR = 0.85 [0.8, 0.89], p < 0.001 Poisson average marginal effect 14.1 min (9.3, 18.8), p <0.001 Admitted with PE (n = 799) Arrrival to Admission Log-rank test p = 0.02 Cox multivarible model HR = 0.83 [0.71, 0.98], p < 0.001 Poisson average marginal effect 18.7 min (1, 36.4), p = 0.04 Open in a new tab Sensitivity analysis. Sensitivity analysis of the primary outcome identified that time-to-testing outcomes for males and females aged 18–49 years were not significantly different (log-rank test p = 0.11), whereas the difference was significant among older adults (p < 0.001). We examined Cox proportional hazards models for each age group, stratified by type of initial PE test ( Table 4 ). Among younger patients (18–49 years), there was no sex difference in testing time among those whose initial test was D-dimer (HR = 1 [95% CI 0.93, 1.07]), but there was a difference among patients who had CT without antecedent D-dimer testing (HR = 0.89 [95% CI 0.80, 0.99]). In older patients (50–89 years), female patients experienced slower testing than males regardless of initial test type. Table 4. Sensitivity Analysis by Age and PE Test Type Age Group Initial PE Test Cox model HR for Female sex (95% CI]) p 18–49 years D-dimer 1.0 (0.93, 1.07) 0.99 CT 0.89 (0.80, 0.99) 0.04 50–89 years D-dimer 0.92 (0.87, 0.98) 0.01 CT 0.88 (0.83, 0.94) <0.001 Open in a new tab Our second sensitivity analysis identified statistically significant differences in the primary outcome by sex at both teaching and nonteaching hospitals (p = 0.002 for both). DISCUSSION In this multicenter study, we examined records from over 15,000 ED patients to identify sex differences in the timeliness of diagnostic testing for PE. Our study has three principal findings. First, we find that female sex is associated with slower time to initial PE testing in the ED, which holds true even after accounting for potential confounders related to demographics, medical history, and ED-level characteristics. This difference persists regardless of whether the patient’s initial testing is D-dimer or CTPA. To our knowledge, this is the first report of sex differences in timeliness of time to testing for PE, but these results are consistent with patterns observed in other cardiovascular diseases. In patients with ST-elevation myocardial infarction (STEMI), female sex is associated with delays to first EKG, 2 – 4 , 25 as well as slower EMS transport times, 1 , 26 and slower door-to-balloon times. 2 , 26 , 27 Similarly, female patients presenting to an ED with acute ischemic stroke undergo CT imaging in a delayed fashion compared to men. 7 , 28 Our second finding is that sex differences in PE time-to-testing outcomes are generally small in magnitude but appear to accumulate during sequential steps in the patient’s ED encounter. Compared to male patients, females had CT orders placed 8.4 minutes later, CT scans completed 14.1 minutes later, and admission orders placed 18.7 minutes later. Similar patterns of cumulative delay for female patients have been observed in the components of door-to-balloon time for STEMI. 2 A notable exception in our secondary analysis is that we observed no sex differences in the time between a D-dimer order and its final result. One possible explanation for this difference is automation: obtaining a CT scan involves more direct care of a patient (e.g., by a physician, nurse, or hospital transport personnel) in interactions that could be influenced by subtle implicit bias (e.g., lower suspicion of disease; less urgency to complete testing), whereas the act of processing a blood specimen in a laboratory may have a lower risk of implicit bias as it involves more automation and less patient contact. Our third notable result is that time-to-testing patterns differ somewhat by age. Female patients experience delays to initial CT testing regardless of age, whereas delays to D-dimer testing are only seen in female patients who are older than age 50. The routine use of pregnancy testing in reproductive-age female patients before CT imaging is common and could potentially be expected to cause a testing delay (and thus a greater difference) in this age group; however, hCG testing was included in the multivariable Cox model and was not found to be a significant contributor to time outcomes. Similarly, physicians may initially be more reluctant to pursue CT testing due to concerns about radiation exposure in young females. Another possibility is that physicians may be reluctant to perform CTPA testing in older patients due to the theoretical risk of contrast nephropathy in the setting of chronic kidney disease, which is more common in older adults. Our data set did not include patient-level data on renal function, but epidemiologic data suggest severe chronic kidney disease is more common among older males rather than females, which would likely lead to results opposite from those we found. 29 Biological sex and sociocultural gender are not synonymous, and it remains unclear how each element may contribute to observed differences in PE testing. For instance, female patients are more likely to present with complaints of dyspnea, whereas males are more likely to report chest pain. 30 – 32 This difference could slightly alter physicians’ perceptions of disease and lead to differences in the timeliness or chronological order of diagnostic tests. Physicians may also perceive a higher risk of potential harm from ionizing radiation in female patients (e.g., possibility of pregnancy; risk of breast cancer). 33 Despite this, pregnancy testing was not statistically associated with our primary outcome, suggesting that the results of these may have routinely been available before physicians decided to pursue testing. Several of our results met expectations. Females represented a majority of patients tested for PE and yet males had higher yields from CT testing, as seen in prior literature. 31 , 34 – 37 Abnormal vital signs were associated with faster times to testing, as were classical chest-related chief complaints for PE. Higher ED volumes were associated with slower testing. Patients with chronic lung disease and respiratory viral testing also had a longer time to testing, which could support this hypothesis that physicians may first pursue an alternate and potentially more common diagnosis, such as pneumonia or Covid, prior to testing for PE. Chest x-ray testing was also associated with slower testing. We suspect this variable may indirectly reflect an estimate of patient illness severity insofar as patients who appeared stable were allowed to wait for routine chest x-rays, whereas those who appeared ill or clinically unstable might “bypass” the routine chest x-ray and instead have expedited CT imaging. It is also possible that clinicians who suspected an alternate diagnosis started with a chest x-ray and only progressed to PE testing if they did not identify an alternate explanation for patients’ symptoms. The study has several limitations related to its retrospective nature. The EMR source for our data had sparse data on gender as a separate characteristic from biological sex, and we were unable to screen for or separately analyze patients who identified as transgender. Although we extracted chief complaint data recorded during ED triage, we were unable to identify other symptoms noted by physicians during their assessment of the patient. We were unable to identify time intervals for all key diagnostic steps, especially the time at which CT results were made available. We were also unable to identify patients with a history of venous thromboembolism prior to arrival at their index ED visit. Finally, we lacked data on the mode of ED arrival (e.g., private vehicle vs. ambulance), which has been associated with differences in wait times and could influence the timeliness of testing. 38 CONCLUSION In multivariable Cox time-to-event models, female patients evaluated for PE in an ED experienced delays to initial testing compared to males, especially in adults aged 50 years or older. Future research should evaluate sex-associated delays to anticoagulation and other treatments, as well as the effect on patient-centered outcomes such as mortality. Figure 1. Open in a new tab Sample Selection Diagram Table 1. Characteristics of Patient Sample. Female Male Patients 8518 (56.7%) 6520 (43.3%) Patient Demographics Age (SD) 53.1 (15.6) 56.1 (16.9) Caucasian 7010 (82.3%) 5333 (81.8%) Hispanic ethnicity 997 (11.7%) 711 (10.9%) English primary language 7760 (91.1%) 5972 (91.6%) Medicaid or Uninsured 2700 (31.7%) 2041 (31.3%) Hospital & ED Visit Characteristics Hospital 1 (academic, urban) 4429 (52%) 3397 (52.1%) 2 (nonacademic, suburban) 3475 (40.8%) 2601 (39.9%) 3 (nonacademic, urban) 613 (7.2%) 522 (8%) Year 2017–19 1755 (20.6%) 1291 (19.8%) 2020–21 2871 (33.7%) 2204 (33.8%) 2022–23 3893 (45.7%) 3025 (46.4%) Time of Arrival 0000–1159 3041 (35.7%) 2575 (39.5%) 1200–2359 5477 (64.3%) 3945 (60.5%) Disposition Discharge from ED 4992 (58.6%) 3038 (46.6%) ED Observation Unit 537 (6.3%) 424 (6.5%) Admission to hospital ward 2308 (27.1%) 2295 (35.2%) Admission to ICU 562 (6.6%) 613 (9.4%) Transfer to another facility 119 (1.4%) 150 (2.3%) Clinical Details Medical history Chronic lung disease 2274 (26.7%) 1330 (20.4%) Coronary disease 1525 (17.9%) 1773 (27.2%) Cancer 1797 (21.1%) 1408 (21.6%) Chest symptom as chief complaint 5758 (67.6%) 4460 (68.4%) Vital signs Heart rate ≥ 100 3382 (39.7%) 2758 (42.3%) Respiratory rate ≥ 22 1985 (23.3%) 1845 (28.3%) Pulse Oximetry ≤ 92% 920 (10.8%) 880 (13.5%) Systolic blood pressure <100 mmHg 341 (4%) 333 (5.1%) ED Testing Ordered Chest x-ray 3663 (43%) 2673 (41%) Flu/Covid viral PCR 3084 (36.2%) 2719 (41.7%) D-dimer 5402 (63.4%) 3627 (55.6%) CTPA 5062 (59.4%) 4285 (65.7%) Diagnostic Test Results PE Diagnosis 454 506 CTPA yield 8.9% 11.8% Open in a new tab Highlights. We retrospectively examined 15,000 ED patients tested for pulmonary embolism (PE). Female patients receive testing for PE more slowly than males. Disparities related to sex/gender were observed in each step of PE diagnosis. Future research should examine delays in anticoagulation for patients with PE. FUNDING Brandon Maughan was supported by a Career Development Award from the American Heart Association award #941303, a Pilot Research Grant from the Collins Medical Trust, and the Scholars in Diagnostic Excellence Program from the National Academy of Medicine and Council of Medical Specialty Societies. Angela Jarman is supported by Building Interdisciplinary Research Careers in Women’s Health at UC Davis. Bory Kea is supported by the National Heart Lung and Blood Institute R01HL157598. Footnotes Publisher's Disclaimer: 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 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. An abstract of these results was presented at the Society for Academic Emergency Medicine meeting on May 17, 2024 (Phoenix, AZ). Declaration of interests The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Brandon Maughan reports financial support was provided by American Heart Association. Brandon Maughan reports financial support was provided by Collins Medical Trust. Brandon Maughan reports financial support was provided by National Academy of Medicine. Angela Jarman reports financial support was provided by National Institutes of Health. Bory Kea reports financial support was provided by National Institutes of Health. Angela Jarman serves as a Consulting Editor for Clinical Therapeutics. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. REFERENCES 1. Aguilar SA, Patel M, Castillo E, et al. Gender differences in scene time, transport time, and total scene to hospital arrival time determined by the use of a prehospital electrocardiogram in patients with complaint of chest pain. J Emerg Med 2012;43(2):291–297. [ DOI ] [ PubMed ] [ Google Scholar ] 2. Dreyer RP, Beltrame JF, Tavella R, et al. Evaluation of gender differences in Door-to-Balloon time in ST-elevation myocardial infarction. Heart Lung Circ 2013;22(10):861–869. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Muhrbeck J, Maliniak E, Eurenius L, Hofman-Bang C, Persson J. Few with ST-segment elevation myocardial infarction are diagnosed within 10 minutes from first medical contact, and women have longer delay times than men. Int J Cardiol Heart Vasc 2020;26:100458. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Zou Y, Zhu W, Zeng J, Lin J, Dai S. Sex-differences in the management and clinical outcome among patients with acute coronary syndrome. BMC Cardiovascular Disorders. 2021;21(1). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Hsieh M-J, Tang S-C, Chiang W-C, Tsai L-K, Jeng J-S, Ma MH- M. Effect of prehospital notification on acute stroke care: a multicenter study. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine. 2016;24(1). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Leung WC, Teo KC, Kwok W, et al. Pre-hospital stroke screening and notification of patients with reperfusion-eligible acute ischaemic stroke using modified Face Arm Speech Time test. Hong Kong Medical Journal. 2020. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Polineni SP, Perez EJ, Wang K, et al. Sex and Race‐Ethnic Disparities in Door‐to‐CT Time in Acute Ischemic Stroke: The Florida Stroke Registry. Journal of the American Heart Association. 2021;10(7). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Daugherty SL, Blair IV, Havranek EP, et al. Implicit Gender Bias and the Use of Cardiovascular Tests Among Cardiologists. Journal of the American Heart Association. 2017;6(12). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Hussain A, McDonald N, Little N, Weldon E. Paramedic perspectives on sex and gender equity in prehospital electrocardiogram acquisition. Paramedicine. 2023;20(1):23–34. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Gauci S, Cartledge S, Redfern J, et al. Biology, Bias, or Both? The Contribution of Sex and Gender to the Disparity in Cardiovascular Outcomes Between Women and Men. Current Atherosclerosis Reports. 2022;24(9):701–708. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Bastian K, Rohmann J, Piccininni M, Kelly Irving M, Bajos N. Gender bias in cardiovascular practice guidelines: the example of ischemic heart diseases. European Journal of Public Health. 2022;32(Supplement_3). [ Google Scholar ] 12. Mackey C, Diercks DB. Gender Bias in the Management of Patients Still Exists. Academic Emergency Medicine. 2018;25(4):467–469. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Office of the Surgeon General (US), National Heart Lung and Blood Institute (US). The Surgeon General’s Call to Action to Prevent Deep Vein Thrombosis and Pulmonary Embolism. Rockville, MD. 2008. Available at: https://www.ncbi.nlm.nih.gov/pubmed/20669525 . Accessed November 1, 2020. [ PubMed ] [ Google Scholar ] 14. Fernandes T, Auger W, Fedullo P. Epidemiology and risk factors for chronic thromboembolic pulmonary hypertension. Thromb Res 2018;164:145–149. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Klok FA, Dzikowska‐Diduch O, Kostrubiec M, et al. Derivation of a clinical prediction score for chronic thromboembolic pulmonary hypertension after acute pulmonary embolism. Journal of Thrombosis and Haemostasis. 2016;14(1):121–128. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Sanchez O, Helley D, Couchon S, et al. Perfusion defects after pulmonary embolism: risk factors and clinical significance. Journal of Thrombosis and Haemostasis. 2010;8(6):1248–1255. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Mongan J, Kline J, Smith-Bindman R. Age and sex-dependent trends in pulmonary embolism testing and derivation of a clinical decision rule for young patients. Emerg Med J 2015;32(11):840–845. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ 3rd. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med 1998;158(6):585–593. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Kline JA, Hernandez-Nino J, Jones AE, Rose GA, Norton HJ, Camargo CA. Prospective Study of the Clinical Features and Outcomes of Emergency Department Patients with Delayed Diagnosis of Pulmonary Embolism. Academic Emergency Medicine. 2007;14(7):592–598. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Social Determinants of Health Database. Agency for Healthcare Research and Quality,; 2023. https://www.ahrq.gov/sdoh/data-analytics/sdoh-data.html . Accessed May 1, 2024. [ PubMed ] [ Google Scholar ] 21. Kaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. Ann Emerg Med 2014;64(3):292–298. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Zou H, Hastie T. Regularization and Variable Selection Via the Elastic Net. Journal of the Royal Statistical Society Series B: Statistical Methodology. 2005;67(2):301–320. [ Google Scholar ] 23. Talbot D, Mésidor M, Chiu Y, Simard M, Sirois C. An Alternative Perspective on the Robust Poisson Method for Estimating Risk or Prevalence Ratios. Epidemiology. 2023;34(1):1–7. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. PLoS Medicine. 2007;4(10):e296. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Pacheco C, Boivin-Proulx L-A, Bastiany A, et al. Impact of STEMI Diagnosis and Catheterization Laboratory Activation Systems on Sex- and Age-Based Differences in Treatment Delay. CJC Open. 2021;3(6):723–732. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Stehli J, Martin C, Brennan A, Dinh DT, Lefkovits J, Zaman S. Sex Differences Persist in Time to Presentation, Revascularization, and Mortality in Myocardial Infarction Treated With Percutaneous Coronary Intervention. Journal of the American Heart Association. 2019;8(10). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Brush JE Jr., Chaudhry SI, Dreyer RP, et al. Sex Differences in Symptom Complexity and Door-to-Balloon Time in Patients With ST-Elevation Myocardial Infarction. Am J Cardiol 2023;197:101–107. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Bushnell C, Howard VJ, Lisabeth L, et al. Sex differences in the evaluation and treatment of acute ischaemic stroke. Lancet Neurol 2018;17(7):641–650. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Kattah AG, Garovic VD. Understanding sex differences in progression and prognosis of chronic kidney disease. Ann Transl Med 2020;8(14):897. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Barrios D, Morillo R, Guerassimova I, et al. Sex differences in the characteristics and short-term prognosis of patients presenting with acute symptomatic pulmonary embolism. PLOS ONE. 2017;12(11):e0187648. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Robert-Ebadi H, Le Gal G, Carrier M, et al. Differences in clinical presentation of pulmonary embolism in women and men. Journal of Thrombosis and Haemostasis. 2010;8(4):693–698. [ DOI ] [ PubMed ] [ Google Scholar ] 32. Tanabe Y, Yamamoto T, Murata T, et al. Gender Differences Among Patients With Acute Pulmonary Embolism. The American Journal of Cardiology. 2018;122(6):1079–1084. [ DOI ] [ PubMed ] [ Google Scholar ] 33. Shyu JY, Sodickson AD. Communicating radiation risk to patients and referring physicians in the emergency department setting. Br J Radiol 2016;89(1061):20150868. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Jarman AF, Mumma BE, Singh KS, Nowadly CD, Maughan BC. Crucial considerations: Sex differences in the epidemiology, diagnosis, treatment, and outcomes of acute pulmonary embolism in non‐pregnant adult patients. Journal of the American College of Emergency Physicians Open. 2021;2(1). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Geersing GJ, Erkens PMG, Lucassen WAM, et al. Safe exclusion of pulmonary embolism using the Wells rule and qualitative D-dimer testing in primary care: prospective cohort study. BMJ 2012;345(oct04 2):e6564–e6564. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Sharp AL, Vinson DR, Alamshaw F, Handler J, Gould MK. An Age-Adjusted D-dimer Threshold for Emergency Department Patients With Suspected Pulmonary Embolus: Accuracy and Clinical Implications. Ann Emerg Med 2016;67(2):249–257. [ DOI ] [ PubMed ] [ Google Scholar ] 37. Van Der Hulle T, Cheung WY, Kooij S, et al. Simplified diagnostic management of suspected pulmonary embolism (the YEARS study): a prospective, multicentre, cohort study. The Lancet. 2017;390(10091):289–297. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Ataman MG, Sariyer G, Saglam C, Karagoz A, Unluer EE. Factors Relating to Decision Delay in the Emergency Department: Effects of Diagnostic Tests and Consultations. Open Access Emergency Medicine. 2023;Volume 15:119–131. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] ACTIONS View on publisher site PDF (695.1 KB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Patient and public involvement in pulmonary embolism research: insights from the development of a pulmonary embolism follow-up care model.#157206
• Frequency of Diagnostic Symptomatic Pulmonary Embolism's in Patients Hospitalized for Clinical Exacerbation of Chronic Obstructive Pulmonary Disease (COPD)#804395
• Clinical Pulmonary Embolism#870372
• The value of electrocardiography in predicting inpatient mortality in patients with acute pulmonary embolism : a cross sectional analysis#281932
• Examination of the Standardized Use of the Pulmonary Embolism SOP in the Clinical Routine of the Emergency Department.#918187