Women With Acute Aortic Dissection Have Higher Prehospital Mortality Than Men

Background Acute aortic dissection (AAD) often leads to out-of-hospital cardiac arrest (OHCA) and death before hospital arrival. Objectives The purpose of this study was to investigate differences in AAD incidence by sex. Methods A population-based study in a city with 121,180 residents was conducted using postmortem computed tomography data to identify patients with AAD who died before hospital arrival in 2008-2020. The incidence rate ratio and odds ratio were estimated using Poisson regression and univariable logistic regression, respectively. Results A total of 266 patients with incident AAD were enrolled: 84 patients with OHCA, 137 women [n = 137], and 164 patients with type A AAD. The crude and age-adjusted incidence of AAD was 16.2 and 14.3/100,000 person-years, respectively. The incidence of AAD was comparable by sex (men, 16.7/100,000 person-years; women, 15.7/100,000 person-years; incidence rate ratio: 0.94; 95% CI: 0.74-1.20; P = 0.64). Compared with men with AAD, women with AAD were older (77 ± 11 years vs 70 ± 14 years; P < 0.001), and a higher proportion had type A AAD (76% vs 47%; P < 0.001). Women with AAD had higher prehospital mortality than men with AAD (37% vs 21%; P = 0.004; OR: 2.24; 95% CI: 1.30-3.87; P = 0.004). Among 1,373 patients with OHCA, the proportion of women with AAD was significantly higher than the proportion of men with AAD (11% vs 3.9%; P < 0.001; OR: 2.90; 95% CI: 1.86-4.53; P < 0.001). AAD was most common in women aged 60 to 69 years (16.4%). Conclusions Women had a higher incidence of AAD presenting as prehospital death than men.

A cute aortic dissection (AAD) is a life- threatening condition that can cause sudden death.Therefore, it is challenging to include patients who die before hospital arrival in epidemiological studies of AAD.[3][4] However, the autopsy rate in these studies was low.
In a study from Iceland, for instance, the autopsy rate declined over time, from 24.0% in 1992 to 9.2% in 2013. 2 In a recent study from Sweden, the overall autopsy rate was reported as 11%. 3 Nobeoka City, with a population of 121,180 and 34% of residents aged 65 years or older in 2020, 5 is isolated from other urban areas.It has only one regional highquality resuscitation hospital designated by the municipal government, Miyazaki Prefectural Nobeoka Hospital. 6,7In Japan, emergency medical services (EMS) personnel are not permitted to terminate resuscitation in the field.They are instructed to transport patients with out-of-hospital cardiac arrest (OHCA) to the nearest regional high-quality emergency center if cardiopulmonary resuscitation has been initiated.However, EMS personnel do not start cardiopulmonary resuscitation or transfer patients when they identify postmortem rigidity or lividity. 8S personnel in Nobeoka City transport all patients with OHCA who received cardiopulmonary resuscitation to the Miyazaki Prefectural Nobeoka Hospital (Supplemental Figure 1).At Miyazaki Prefectural Nobeoka Hospital, postmortem computed tomography (PMCT) has been performed since 2008 to identify the cause of death when patients die or are not expected to recover from OHCA as an alternative to autopsy.Generally, PMCT is performed within 2 hours of the pronouncement of death by a physician.
Recently, postmortem imaging was shown to accurately identify the cause of death.[11] Using PMCT data from Nobeoka City, Yamaguchi et al 7 investigated the incidence of AAD between 2016 and 2018.They found that the incidence of AAD was 2fold higher than in previous reports.The aim of this study was to investigate differences in the incidence of AAD by sex among patients with OHCA in a welldefined geographic area.

DIAGNOSIS OF AAD.
The medical records and CT images of all patients in the OHCA and AAD datasets were reviewed to identify patients with AAD.Miyazaki Prefectural Nobeoka Hospital's CT and PMCT protocols have been previously described. 73][14][15] Bloody pericardial effusion was defined as a fluid collection in the pericardial space with 30 to 60 Hounsfield units. 12,16Aortic rupture was diagnosed as periaortic hemorrhage involving the mediastinum, thoracic cavity, or abdominal cavity. 17Aortic rupture was classified as aortic dissection if there were clear CT findings of aortic dissection.To identify patients with aortic dissection, CT images of all patients in the OHCA dataset were reviewed by an experienced cardiologist and an experienced radiologist.A blinded third reader adjudicated cases with disagreement.Among 607 randomly selected PMCT evaluations, 11 required a third reviewer due to disagreement between the 2 reviewers.The k statistic for interobserver agreement on the presence of aortic dissection was 0.87.AAD was distinguished from chronic aortic dissection, defined as occurring <14 days after symptom onset, based on medical history. 18AAD was classified into type A, type non-A non-B, or type B.
Type non-A non-B AAD was defined as AAD involving the aortic arch and the descending aorta but not the ascending aorta. 19ATISTICAL ANALYSIS.Normally distributed continuous variables are expressed as mean AE SD.Categorical variables are expressed as n (%).Groups were compared using Pearson's chi-squared test for categorical variables and the unpaired t-test for continuous variables.To calculate the incidence of AAD by age group and sex, we divided the number of incident AAD cases by the population of Nobeoka per 100,000.
In addition, we performed a direct adjustment to the 2015 Japanese standard population.The incidence rate ratio (IRR) was estimated using Poisson regression.Odds ratios (ORs) were based on univariable logistic regression.Two-tailed P < 0.05 was considered statistically significant.We used SPSS version 24.0 (IBM) and Stata 15 (StataCorp) for statistical analyses.

RESULTS
PATIENT CHARACTERISTICS.A total of 288 AADs were documented.Twenty patients who had recurrent AAD were excluded from this study of incidence.
Table 1 shows the characteristics of the study patients.The mean age was 74 AE 13 years.There were 164 (62%) patients with type A AAD, 78 (29%) patients with type non-A non-B AAD, and 24 (9%) patients with type B AAD. Women comprised 63% of patients with type A AAD.The most prevalent vascular risk factor for AAD was hypertension.Aortic rupture and bloody pleural effusion occurred in 18% and 38% of patients, respectively.Compared with men with AAD, women with AAD were older (77 AE 11 years vs 70 AE 14 years; P < 0.001), and a higher proportion of women had type A AAD (76% vs 47%; P < 0.001).The proportion of current smokers among women with AAD was lower than the proportion among men with AAD (6% vs 31%; P < 0.001).

INCIDENCE OF AAD.
The crude incidence of AAD was 16.2/100,000 person-years.The age-adjusted incidence of AAD for the standard Japanese population in 2015 was 14.3/100,000 person-years.The incidence of type A, type non-A, non-B, and type B AAD was 10.0/ 100,000 person-years, 4.7/100,000 person-years, and 1.5/100,000 person-years, respectively.

DISCUSSION
This study is the first to show that AAD is more common among women than men with OHCA.
Women had a higher incidence of AAD presenting as sudden death than men.One-third of women with AAD died before hospital arrival (Central Illustration).

SEX DIFFERENCES IN THE INCIDENCE AND
CHARACTERISTICS OF AAD.The International Registry of Acute Aortic Dissection enrolled 1,078 patients with AAD.In that registry, 32% of women had AAD, 20 which was cited in recent guidelines as a sex difference in AAD incidence. 21,22Similarly, the Spanish Registry of Acute Aortic Syndrome showed a lower proportion of patients with AAD were women (27%). 23However, these studies did not enroll patients who died before hospital admission.
Population-based studies of AAD using autopsy data to enroll patients who died before hospital admission have shown a higher proportion of women (36%-40%) with AAD 1-4 than the registries mentioned above, but the incidence of AAD was lower in women than in men, [2][3][4] In the present study using PMCT data, 52% of patients with AAD were women, and the incidence of AAD was comparable between men and women.
Women with AAD had higher prehospital mortality in this study.A previous nationwide population-based study showed higher 30-day mortality in female patients admitted with aortic dissection than male patients admitted with aortic dissection and that the proportions of women were higher among patients deceased without hospital admission than in hospitalized patients.The authors of that study suggested the higher mean age of women might have contributed to differences by sex. 3 A higher short-term mortality rate in women compared to men has also AAD ¼ acute aortic dissection; OHCA ¼ out-of-hospital cardiac arrest.
been reported in patients with acute myocardial infarction.Women with acute myocardial infarction are typically older and have more comorbidities. 24ese findings might be helpful in assessing the reason for sex-specific mortality differences in AAD.
In our study, women with AAD were older and less likely to be current smokers than men with AAD.The prevalence of hypertension was comparable.Notably, the higher proportion of women with type A AAD might have strongly contributed to higher mortality in women.Further studies are needed to assess the reasons for sex-specific mortality differences in AAD.
The incidence of AAD in women could have been previously underestimated because more than one-third of women with AAD died before hospital arrival.
The incidence of AAD in the present study might have been affected by the higher age of this study population relative to other study populations because age is a risk factor for AAD.By including a substantial number of patients with AAD and OHCA identified by PMCT, we found that the incidence of AAD was lower in women in younger age groups, but the incidence in women caught up with the incidence in men as age increased (Figure 2).This suggests that the incidence of AAD in women and men will be comparable in other areas with aging populations.
Our results might be generalizable for Japan because our study population distribution is similar to that of Japan as a whole.Our results can help predict the trajectory of other countries with aging populations.These results help inform future studies aimed at better understanding the biology of AAD and AAD prevention and treatment strategies throughout the world.
Prehospital mortality in this study (29%) was similar to mortality before hospital arrival or admission from previous reports based on autopsy data (17.6%-38%). 1,3,4Considering the low autopsy rates in other studies, mortality due to AAD could have been underestimated.Population-based studies in other countries with a higher autopsy rate or studies with PMCT data are warranted for comparisons of AAD incidence and mortality with those found in the present study.

SEX DIFFERENCES IN THE PROPORTION OF PA-
TIENTS WITH OHCA AND AAD.This study is the first to show that AAD is more common among women with OHCA than men with OHCA.The proportion of patients with OHCA who have AAD was recently investigated in other Japanese studies.Two recent studies evaluated approximately 90% of all consecutive patients with OHCA who were transported to hospitals.They showed that 7.6% of patients with OHCA who underwent CT examination had AAD; 7.0% to 7.1% had type A AAD. 12,25 We showed a similar proportion of patients with OHCA had AAD (6.7%) and type A AAD (6.1%)There is a possibility that patients with no clear findings of AAD on CT were not included.For instance, in patients with aortic rupture, diagnosis of AAD based on PMCT was particularly difficult.Second, not all individuals who die in Nobeoka City undergo PMCT.For example, family members of patients with terminal diseases who did not want to be resuscitated did not call EMS or transport them to hospitals.In addition, if postmortem rigidity or lividity was evident, EMS personnel did not initiate resuscitation or transport the patient to a hospital.
During the study period, 1,096 people with nontraumatic OHCA in Nobeoka City were not transported to hospitals because of apparent postmortem rigidity or lividity.Among 1,373 patients with OHCA who were transported to hospitals, 138 (10%) did not undergo PMCT evaluation.Of these, 52 (38%) patients had an obvious cause of OHCA.The remaining 86 (62%) patients could have had AAD.In the field of epidemiological studies of AAD, there is a need to continue evaluating more prehospital deaths in order to increase the accuracy of epidemiological information.Third, the incidence of AAD observed in our study might be a reflection of unidentified genetic factors, environmental factors, or both.Japanese people have the longest life expectancy in the world. 26Moreover, climate can affect the incidence of AAD. 27However, comprehensive information from Japan, which has a rapidly aging population, can be crucial for developing perspectives in other countries.Fourth, for patients with OHCA, it was

CONCLUSIONS
Women have a comparable incidence of AAD as men, but they have higher prehospital mortality.A higher proportion of women with OHCA had AAD than men with OHCA, with the highest proportion in women aged 60 to 69 years, at 16.4%.A population-based study in a city with a population of 121,180 used postmortem computed tomography data from out-of-hospital arrests over 13 years (2008-2020) to investigate differences in the incidence of AAD by sex.The overall incidence of AAD in women and men was comparable (men, 16.7/100,000 person-years; women, 15.7/100).The incidence of AAD without OHCA was significantly lower in women than in men (men, 13.1/100,000 person-years; women, 9.1/100,000 person-years).In contrast, the incidence of AAD with OHCA was significantly higher in women than in men (men, 3.6/100,000 person-years; women, 6.4/100,000 person-years).

CENTRAL ILLUSTRATION Incidence of AADs in Women and Men
Women with AAD had higher prehospital mortality than men with AAD (37% vs 21%).AAD ¼ acute aortic dissection.KEY WORDS acute aortic dissection, incidence, prehospital mortality, sex difference APPENDIX For a supplemental table and figure, please see the online version of this paper.

Marume et al
STUDY DESIGN.All patients with AAD and patients with OHCA in Nobeoka were included using a registry of patients with AAD based on 2 datasets: the AAD dataset for Miyazaki Prefectural Nobeoka Hospital and the OHCA dataset from Nobeoka.The study period was January 2008 to December 2020.This study was conducted in accordance with the Declaration of Helsinki and its amendments.The research ethics committees of Miyazaki Prefectural Nobeoka Hospital (No. 20191004-1) and Kumamoto University (No. 2491) approved this study.As individual patients were not identified, the requirement to obtain informed consent from each study participant or their surviving family members was waived.However, we publicized the study by posting an easy-to-understand summary on the hospital's website.Participants or their surviving family members were allowed to refuse participation at any time.ACUTE AORTIC DISSECTION DATASET.A database of patients with AAD but not OHCA who visited or were admitted to Miyazaki Prefectural Nobeoka Hospital between January 2008 and December 2020 was generated from the hospital's medical records based on International Classification of Diseases-10th Revision codes.Patients were extracted using the following key words: aortic dissection, dissecting aortic aneurysm, and ruptured aortic aneurysm.Information on demographics, comorbidities, risk factors, and computed tomography (CT) findings were collected.OHCA DATASET.Miyazaki Prefectural Nobeoka Hospital has collected the following data on patients with OHCA since 2008: demographic characteristics, initial cardiac rhythm, comorbidities, risk factors, CT findings, and prehospital death.Prehospital death was defined as death in an emergency department after OHCA.Data were reviewed monthly for the cause of death and missing values by a medical committee comprised of physicians, surgeons, emergency physicians, and emergency medical technicians.This committee reviews all OHCAs that involved transport to hospitals in the city.During the study A B B R E V I A T I O N S A N D A C R O N Y M S AAD = acute aortic dissection IRR = incidence rate ratio OHCA = out-of-hospital cardiac arrest PMCT = postmortem computed tomography The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors' institutions and Food and Drug Administration guidelines, including patient consent where appropriate.For more information, visit the Author Center.Manuscript received March 3, 2023; revised manuscript received June 29, 2023, accepted July 26, 2023.
SEX DIFFERENCES IN PRE-HOSPITAL MORTALITY FROM AAD.Among 266 patients with incident AAD, 84 experienced OHCA.Of 84 patients with AAD and OHCA, 78 died in an emergency department and were diagnosed based on PMCT findings.The prehospital mortality rate was 29% (78 of 266).The prehospital mortality rates of type A, type non-A non-B, and type B AAD were 45% (73 of 164), 4% (3 of 78), and 8% (2 of 24),

FIGURE 1
FIGURE 1 Study Flow Chart in this study, which evaluated 90% (1,235/1,373) of transported patients with OHCA in a city with a population of 120,000.It should be noted that a substantial proportion of patients with OHCA aged 50 to 59 years or 60 to 69 years had AAD; the highest proportion was 16.4% in women aged 60 to 69 years.To develop a practical treatment strategy, it is important to strongly consider AAD in women with OHCA.STUDY LIMITATIONS.First, autopsies were not performed in all cases.Although the diagnostic value of PMCT for AAD is high, there are limitations in diagnosing aortic dissection based on CT, especially noncontrast PMCT.In the present study, AAD was diagnosed based on the presence of clear CT findings.

J
A C C : A D V A N C E S , V O L . 2 , N O .8 , 2 0 2 3 Sex Differences in Acute Aortic Dissection O C T O B E R 2 0 2 3 : 1 0 0 6 2 3

TABLE 1
Characteristics of Patients With AAD by Sex (N ¼ 266) Values are mean AE SD or n (%).

TABLE 2
Characteristics of Patients With OHCA Transported to a