Why the Delay? Identification of Factors Which Delay Diagnosis of Acute Aortic Dissection

Article Citation:

Craig E. Strauss, Tesfaye D. Kebede, Brandon R. Porten, Ross F. Garberich, Domenico Calcaterra, Jesse M. Manunga, and Kevin M. Harris (2017) Why the Delay? Identification of Factors Which Delay Diagnosis of Acute Aortic Dissection. Journal of the Minneapolis Heart Institute Foundation: January 2017, Vol. 1, No. 1, pp. 13-18.

Original Article

Craig E. Strauss, MD, MPH, Tesfaye D. Kebede, BBME, Brandon R. Porten, BA, Ross F. Garberich, MS, Domenico Calcaterra, MD, Jesse M. Manunga, MD, and Kevin M. Harris, MD

Minneapolis Heart Institute Foundation at Abbott Northwestern Hospital, Minneapolis MN

Disclosures: None of the authors have a conflict of interest related to this project.

Address for correspondence:
Kevin M. Harris, MDMinneapolis Heart Institute Foundation920 East 28th Street, Suite 40Minneapolis, MN 55407Tel: 612 863 7372Fax: 612 863 2490E-mail: Kevin.Harris@mplsheart.com

ABSTRACT

Patients with acute aortic dissection suffer a high early mortality. However, limited data delineate the factors that contribute to initial delays in establishing the acute aortic dissection diagnosis. We reviewed consecutive patients with acute aortic dissection presenting to a tertiary hospital system over a 10-year period. Patient clinical characteristics, physician assessments, and times from presentation to diagnosis were evaluated. Factors that contributed to “early” (<300 minutes) vs “late” (≥300 minutes) diagnosis were assessed. This dataset included 162 patients with acute aortic dissection: 114 early patients (median age 66, 62% male, 68% type A; median time to diagnosis 102 minutes) and 48 late patients (median age 73, 58% male, 79% type A; median time to diagnosis, 903 minutes). Among the late-diagnosis patients, the initial working diagnosis was acute coronary syndrome, pulmonary etiology, or other etiology in 48%, 23%, and 29%, respectively. For the 48% of patients in the late group, the definitive imaging modality for acute aortic dissection diagnosis was CT scan in 79%, echocardiogram in 13%, aortogram in 4%, magnetic resonance aortography in 4%, and magnetic resonance imaging in 2%. Seventeen (35%) late patients had initial imaging that was inconclusive for aortic pathology. Late-diagnosis patients were less likely to have chest pain (55% vs 89%; P < .0001) or back pain (44% vs 64%; P = .02) and were more likely to present with syncope (30.4% vs 10.1%; P = .002). Mortality was similar for late vs early diagnosis (29% vs 25%; P = NS). Late-diagnosed patients presenting to a community hospital had a median time of diagnosis of 973 minutes compared with 577 minutes in patients presenting to the tertiary center. Delayed acute aortic dissection diagnosis was associated with atypical presenting symptoms, initial suspicion of acute coronary syndrome, or a pulmonary etiology, and presentation to a community referral hospital. Mortality remains high for patients diagnosed late and a heightened suspicion of acute aortic dissection may improve the time to diagnosis, thereby preventing negative outcomes and allowing earlier surgical intervention.

Keywords: aortic dissection, aortic imaging, emergency medicine

INTRODUCTION

Acute aortic dissection (AAD) is the most common acute aortic disease requiring surgical intervention, and up to 35% of patients are misdiagnosed on initial presentation.1 During the initial 48 hours, patients with type A AAD have a 1% to 2% per hour mortality rate.2–4 In contrast, early diagnosis and intervention of AAD limits the risks of aortic rupture, cardiac tamponade, stroke, and death. The 30-day survival rate for patients with efficient diagnosis and optimal care approaches 90%.5 Delays in the time from presentation to diagnosis often begin in the emergency department where atypical symptoms can lead to misdiagnosis and exposure to antithrombotic agents.6 The median time from presentation to diagnosis for all AAD patients in the International Registry of Aortic Dissection (IRAD) is over 4.3 hours.7 According to findings from the same registry (IRAD), factors contributing to the delay of diagnosis include: absence of typical symptoms such as abrupt chest/back pain, pulse deficits on physical examination, hypotension, female sex, and presentation to nontertiary hospitals.7 The purpose of this study was to characterize the initial working diagnosis, the care, and diagnostic algorithm leading to the diagnosis of AAD within a single regional network consisting of referral hospitals and a tertiary care hospital.

METHODS

The Minneapolis Heart Institute (MHI) at Abbott Northwestern Hospital is a tertiary care center with referral sites throughout Minnesota and Western Wisconsin.8 The Minneapolis Heart Institute is one of 37 sites throughout North America, Europe, and Japan that participate in the IRAD registry.4Study Design

All patients who presented between March 2003 and March 2013 to the tertiary care site (MHI) or an affiliated referral hospital with an AAD and onset of symptoms within 14 days were included in the study. Exclusion criteria include patients who did not consent to participate, patients with iatrogenic aortic dissections, and patients with a history of a chronic aortic dissection. Institutional review board approval was obtained for all data collection and analysis.

Since August 2005, the Minneapolis Heart Institute and Abbott Northwestern Hospital have implemented a standardized protocol for the diagnosis, acute management, and transfer of AAD. The program emphasizes early recognition of AAD via education of emergency department and emergency medical services (EMS) personnel and includes a recommended computed tomography (CT) imaging algorithm for suspected AAD.8Data Extraction

Patient clinical characteristics, medical history, imaging modality performed, dissection type, and definitive management were documented. The time of symptom onset, time of presentation to the initial hospital, transfer times, and time to definitive treatment were obtained through medical record review in accordance with IRAD definitions.4 A subset of this patient group was part of a larger multisite IRAD study.7 Additional data was obtained through a detailed review of the EMS scene description, transport documents, and emergency department staff and physician documentation. Extracted patient data was reviewed independently by 2 physicians and discrepancies were corrected through detailed medical record review.Definition

The time from presentation at the initial hospital to definitive diagnosis was defined as the “presentation to diagnosis” time. Patients with presentation to diagnosis times of <300 minutes were identified as early-diagnosis patients and those with presentation to diagnosis times of ≥300 minutes were termed late-diagnosis patients. The 300-minute breakpoint approximates the mean presentation to diagnosis time for patients in the IRAD registry.7 Definitive diagnosis of an acute aortic dissection was made by CT, magnetic resonance imaging (MRI) or magnetic resonance aortography, echocardiography, aortography, or a combination of these imaging modalities.Statistical Analysis

Fisher’s exact or χ2 test was used to assess the statistical significance of categorical variables and t tests were used to test for differences in continuous variables. Time segments (in minutes) were transformed to natural logarithms (time-segment values) to more closely approximate a normal distribution and are reported as median and interquartile range (25th and 75th percentiles). A value of P < .05 was considered statistically significant and all reported probability values are 2-sided. Statistical calculations were performed using commercial software (SAS 9.3; SAS Institute, Inc., Cary, NC).

Results

Between March 1, 2003 and March 31, 2013, 164 cases of AAD were treated within the referral hospital network. Two patients with iatrogenic dissections were excluded. Among the 162 remaining patients, 114 (70%) had times from presentation to diagnosis of <300 minutes and 48 (30%) had times of ≥300 minutes.Patient Characteristics

Patients with early vs late diagnosis were of similar age and sex with a similar frequency of type A AAD (Table 1). Patients with late diagnosis were more likely to be transferred from referral hospital.

TABLE 1
Baseline characteristics.

Signs and Symptoms at Presentation and Past Medical History

Chest and back pain, especially when abrupt or radiating were characteristics found more frequently in the group with early diagnosis (Table 2). In contrast, syncope was more prevalent in those with delayed diagnosis.

TABLE 2
Presenting symptoms.

Hypertension was the dominant past medical history factor for AAD patients in both groups (Table 3). Presenting hemodynamics did not differ between the 2 groups (Table 4).

TABLE 3
Past medical history.

TABLE 4
Hemodynamics on presentation for recorded.

Initial Imaging and Diagnosis

Computed tomography was the most common initial diagnostic imaging modality, but was employed less frequently in the late-diagnosis group (84% in early vs 63% in late; P = .0024). An abnormal D-dimer test was drawn prior to AAD diagnosis in 77% of the late-diagnosis patients. The results from the initial imaging studies were suggestive of dissection in all patients in the early-diagnosis group and only 64.6% in the late-diagnosis group (P < .001). The intramural hematoma variant of dissection did not differ between the 2 groups (32 [28%] vs 11 [22.9%], P = NS) Patients with late diagnosis presenting to referral hospitals had a mean time to diagnosis of 973 minutes vs 577 minutes at the tertiary hospital.

An evaluation of the initial medical decision making in the late-diagnosis patients revealed that 48% were initially treated for ACS. The diagnosis of ACS was supported by abnormal EKG suggestive of ischemia in 16 (70%) and troponin positivity in 4 (18%). Of this group, 7 (30%) underwent coronary angiography and 7 (30%) received clopidogrel and 11 (48%) received heparin. A total of 23% of late-diagnosis patients were treated for pulmonary etiology prior to the discovery of dissection, generally suspected pulmonary embolism based on elevated D-dimer occurred in 7 (77%). None of these patients received heparin. The remaining 29% were treated for diverse etiologies including gastrointestinal, back and spinal ailments, and syncope. Figure 1shows the initial suspected diagnosis and testing algorithm in late diagnosis patients.

FIGURE 1
The initial clinical diagnosis of the initial medical provider in late diagnosis patients is shown as well as the testing algorithm that ultimately led to the diagnosis of aortic dissection.

Outcomes

The in-hospital mortality rate for early- and late-diagnosis groups was similar (24.6% vs 29.2%; P = NS) as was length of stay (7.5 [5, 13] vs 10 [5, 14]; P = NS).

DISCUSSION

Our findings suggest that a late diagnosis of AAD was associated with atypical presenting symptoms, an initial suspicion of ACS or a pulmonary etiology, and presentation to a nontertiary care hospital. No significant difference in mortality was identified for early- vs late-diagnosis patients. However, the mortality remains high and a heightened suspicion of AAD may improve the time to diagnosis and to definitive treatment.

The earlier diagnosis of AAD is an emphasis of the American College of Cardiology/American Heart Association aortic guidelines with suggestions for clinicians to focus on high-risk conditions placing patients at risk, as well as typical historical and exam features.9–14 This is also an important goal of lay patient advocacy groups including the Thoracic Aortic Coalition and the John Ritter Foundation. Nevertheless, in this cohort, many of the patients in the late-diagnosis group had atypical features that take physicians down alternative diagnostic pathways. Although the reasons for delays in recognition (and treatment) of AAD have been outlined in a prior large IRAD analysis,7 this study allows the specific identification of the initial provider’s working diagnosis and diagnostic algorithm.

Multiple factors, occurring in combination, result in the most significant delays in diagnosis. Patients presenting without typical symptoms or hemodynamic instability are more likely to experience diagnostic delays and to be initially treated for more common etiologies.15 Moreover, dissection in these patients is often discovered incidentally during imaging evaluations to rule out a pulmonary embolism or other etiology.

It is noteworthy that abnormalities of laboratory testing performed in the emergency department, including D-dimer (suggesting acute pulmonary embolism) as well as troponin and EKG (suggesting ACS), led to anticoagulation and in some cases of the latter group, to invasive angiography.15,16 Such treatments are associated with hazard in acute dissection.6 Fortunately, a noteworthy benefit of the widespread use of D-dimer, which was performed frequently in this data set to evaluate PE, may have led to the diagnosis of AAD in this subset of patients.17 Suspicion of ACS was the initial diagnosis in nearly half of the late-diagnosis group and has been felt to be the most common reason for missed or delayed diagnosis of dissection.9 Since the frequency of ACS greatly outnumbers the number of dissections, it is appropriate for emergency physicians to focus on this diagnosis in patients presenting with chest or back pain.10,13

Computed tomography scan of the chest, abdomen, and pelvis remains the most commonly used diagnostic modality for patients presenting with symptoms suggestive of aortic dissection.18 This modality, however, was less frequently used as the initial diagnostic imaging obtained in our late-diagnosis cohort. In fact, our data shows that one-third of patients in the late-diagnosis group did not undergo a CT, transesophageal echocardiogram, or MRI as their initial diagnostic modality, all of which have been proven sensitive in the diagnosis of this pathology.19 Certainly, this contributed to delays in diagnosis this subgroup.

Despite the observed difference in presentation-to-diagnosis time, the morbidity and mortality rates among the 2 groups were not significantly different, though there was a trend toward higher mortality observed in the late-diagnosis group. It is unknown whether patients with late diagnosis represent a subset with a more limited or stable dissection. Furthermore, we are unable to conclude from the data whether or not patients with atypical presentation had “uncomplicated” aortic dissection. However, in the larger IRAD group, the lack of pain was not associated with a benign prognosis.20

The management of AAD requires coordinated, multidisciplinary care across the entire health network to achieve optimal outcomes, an effort which starts with a high index of suspicion of the condition by providers on the front line.8 The majority of patients in this series were evaluated at low-volume referral centers, some of which have diagnosed only 1 or 2 cases of aortic dissection over the last 10 years. This lack of familiarity with this condition by physicians at these low-volume centers certainly contributed to delay in diagnosis. Indeed, our data show that 90% of patients with late diagnosis were initially evaluated at low-volume centers. A program that includes rigorous training of providers and immediate feedback after dissection cases may improve diagnostic times.8 This aspect can be particularly useful for low-volume centers where the delay in diagnosis is compounded by the time needed to transfer patients to tertiary centers where definitive treatment can be offered.Limitations

This study only highlights those cases in which the diagnosis of AAD is established. It is unclear how many cases of aortic dissection were missed within our referral network. Undoubtedly, a large number of patients with aortic dissection go undiagnosed. Indeed, Roberts et al21 found that more than 23% of patients in their cohort had their diagnosis of aortic dissection diagnosed at either autopsy or surgery for an alternative condition. Limitations of this study include the fact that it is a retrospective, single-center study and did not include a specific control group. As a result, assuming a direct cause and effect relationship should be done with caution.

CONCLUSION

Delayed AAD diagnosis was associated with atypical presenting symptoms, initial suspicion of ACS or a pulmonary etiology, and presentation to a community hospital. Mortality remains high for patients diagnosed late and a heightened suspicion of AAD may improve the time to diagnosis and to definitive treatment.

REFERENCES

1.Spittell PC, Spittell JA Jr, Joyce JW, et al. Clinical features and differential diagnosis of aortic dissection: experience with 236 cases (1980 through 1990). Mayo Clin Proc. 1993;68:642–651. [Crossref] [Google Scholar]
2.Hirst AE, Johns VJ, Kime SW. Dissecting aneurysm of the aorta: A review of 505 cases. Medicine (Baltimore). 1958;37:217–279. [Crossref] [Google Scholar]
3.Strauss CE, Harris KM, Hutchison S, Ehrlich MP, et al. “Time is Life”: Early mortality in type A acute aortic dissection: insights from the IRAD registry. J Am Coll Cardiol. 2013;61:E516. [Crossref] [Google Scholar]
4.Hagan PG, Nienaber CA, Isselbacher EM, et al. The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease. JAMA. 2000;283:897–903. [Crossref] [Google Scholar]
5.Bavaria JE, Pochettino A, Brinster D, et al. New paradigms and improved results for the surgical treatment of acute type A dissection. Ann Surg. 2001234:336–343. [Crossref] [Google Scholar]
6.Hansen MS, Nogareda GJ, Hutchison SJ. Frequency of and inappropriate treatment of misdiagnosis of acute aortic dissection. Am J Cardiol. 2007;99:852–856. [Crossref] [Google Scholar]
7.Harris KM, Strauss CE, Eagle KA, et al. International Registry of Acute Aortic Dissection (IRAD) Investigators. Correlates of delayed recognition and treatment of acute type A aortic dissection: the International Registry of Acute Aortic Dissection (IRAD). Circulation. 2011;124:1911–1918. [Crossref] [Google Scholar]
8.Harris KM, Strauss CE, Duval S, et al. Multidisciplinary standardized care for acute aortic dissection: Design and initial outcomes of a regional care model. Circ Cardiovasc Qual Outcomes. 2010;3:424–430. [Crossref] [Google Scholar]
9.Hiratzka LF, Bakris GL, Beckman JA, et al. 2010 ACCF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM guidelines for the diagnosis and management of patients with thoracic aortic disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, American Association for Thoracic Surgery, American College of Radiology, American Stroke Association, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society of Interventional Radiology, Society of Thoracic Surgeons, and Society for Vascular Medicine. Catheter Cardiovasc Interv. 2010;121:e266–e369. [Google Scholar]
10.von Kodolitsch Y, Schwartz AG, Nienaber CA. Clinical prediction of acute aortic dissection. Arch Intern Med. 2000;160:2977–2982. [Crossref] [Google Scholar]
11.Klompas M. Does this patient have an acute thoracic aortic dissection? JAMA. 2002;287:2262–2272. [Crossref] [Google Scholar]
12.Rosman HS, Patel S, Borzak S, Paone G, Retter K. Quality of history taking in patients with aortic dissection. Chest. 1998;114:793–795. [Crossref] [Google Scholar]
13.Harris KM, Strauss CEDiagnostic delay in acute aortic syndromes: how sensitive and specific are clinical features in disease recognition. In: Bonser RS, Pagano D, Haverich A, Mascaro J, eds. Controversies in Aortic Dissection and Aneurysmal Disease. London England: Springer Verlag;, 2014:107–114. [Google Scholar]
14.Rogers AM, Hermann LK, Booher AM, et al. Sensitivity of the national thoracic aortic disease guideline diagnostic algorithm for identification of acute aortic dissection at initial presentation: results from the International Registry of Acute Aortic Dissection (IRAD). Circulation. 2011;122:2213–2218. [Crossref] [Google Scholar]
15.Biagini E, Lofiego C, Ferlito M, et al. Frequency, determinants, and clinical relevance of acute coronary syndrome-like electrocardiographic findings in patients with acute aortic syndrome. Am J Cardiol. 2007;100:1013–1020. [Crossref] [Google Scholar]
16.Rapezzi C, Longhi S, Graziosi M, et al. Risk factors for diagnostic delay in acute aortic dissection. Am J Cardiol. 2008;102:1399–1406. [Crossref] [Google Scholar]
17.Ranasinghe AM, Bonser RS. Biomarkers in acute aortic dissection and other aortic syndromes. J Am Coll Cardiol. 2010;56:1535–1541. [Crossref] [Google Scholar]
18.Pape LA, Awais M, Woznicki EM, et al. Presentation, diagnosis, and outcomes of acute aortic dissection: 17-year trends from the International Registry of Acute Aortic Dissection. J Am Coll Cardiol. 2015;66:350–358. [Crossref] [Google Scholar]
19.Shiga T, Wajima Z, Apfel CC, et al. Diagnostic accuracy of transesophageal echocardiography, helical computed tomography, and magnetic resonance imaging for suspected thoracic aortic dissection: systematic review and meta-analysis. Arch Intern Med. 2006;166:1350–1356. [Crossref] [Google Scholar]
20.Park SW, Hutchison S, Mehta RH, et al. Association of painless acute aortic dissection with increased mortality. Mayo Clin Proc. 2004;79:1252–1257. [Crossref] [Google Scholar]
21.Roberts WC, Vowels TJ, Mi Ko J, Guileyardo JM. Acute aortic dissection with tear in ascending aorta not diagnosed until necropsy or operation (for another condition) and comparison to similar cases receiving proper operative therapy. Am J Cardiol. 2012;110:728–735. [Crossref] [Google Scholar]
Print ISSN: 
2475-0190
eISSN:
2475-0204