Understanding Bias in Acute Aortic Syndrome Research

Reference: McLatchie et al and DAShED investigators. Diagnosis of Acute Aortic Syndrome in the Emergency Department (DAShED) study: an observational cohort study of people attending the emergency department with symptoms consistent with acute aortic syndrome. EMJ Nov 2023. Date: February 11, 2024 Guest Skeptic: Nirdosh Ashok Kumar, Emergency Medicine Specialist – Aga Khan University Hospital, Karachi, Pakistan. Case: A 66-year-old female with a history of smoking, hypertension, and type-2 diabetes presents to the emergency department (ED) with syncope while walking her dog. She complains of retrosternal chest pain radiating to her jaw. She is bradycardic, hypotensive, and tachypneic. She is received in the resuscitation room. Monitors have been attached, and intravenous (IV) access has been achieved and IV analgesia has been given. The ECG shows sinus tachycardia with non-specific ST-T changes. The chest x-ray is unremarkable. However, she is still in severe pain. A post-graduate year 2 (PGY-2) resident asks you if it could be a ruptured abdominal aortic aneurysm, aortic dissection, or angina. Background: The diagnosis of acute aortic syndrome (AAS) is commonly delayed or missed in the ED. AAS has been referred to as the “lethal triad” that incorporates aortic dissection (AD), intramural hematoma (IMH), and penetrating aortic ulcer (PAU) [1].  It is a rare condition with a high mortality rate and can present in atypical ways. It affects approximately 4,000 people per year in the United Kingdom [2] and 43,000 to 47,000 people per year in the USA [3]. The annual incidence rate of AD ranges between 2.9 and 7.2 per 100,000. [4-8] The misdiagnosis rate is estimated to be between 16% and 38%6,[9-19] with a diagnostic delay of up to 24 hours for 25% of cases, and mortality follows a linear increase of 0.5% per hour in the first 48 hours. [20] A retrospective observational study from Canadian researcher, Dr. Robert Ohle was published in CJEM in 2023. This study found that between 2003 and 2018, there were 1,299 cases of AAS in Ontario, the largest province in the country. It reported an overall annual incidence rate of 0.61 per 100,000 people which is much lower than previously reported rates. The study also highlighted the significant mortality rate associated with AAS, with a one-year mortality rate decreasing from 47.4% to 29.1%, and ED mortality at 14.9%​​. [21] When looking specifically at atraumatic chest pain presentations to the ED, it is estimated the incidence of AAS is one in 980. [22] It can be like looking for a needle in a haystack of chest pain patients. The gold standard for diagnosing AAS is to perform a CT aorta angiogram (CTA). However, scanning everyone chest pain patient would have a very low diagnostic yield [23,24], expose many patients to unnecessary ionizing radiation and end up being very costly. It would be great if there was a validated clinical decision tool (CDT) to help clinicians be more selective in using CTA to diagnose AAS. Some CDTs have been devised and tested for diagnosing AAS. [25,26] The Aortic Dissection Detection Risk Score (ADD-RS) is one CDT that has been derived and tested. Four studies with methodologic limitations were included in an SRMA of the ADD-RS and published in AEM 2020. [27] The authors concluded that patients with an ADD-RS score of ≤ 1 with d-dimer < 500 ng/mL have high sensitivity for ruling out AASs. However, it is unclear if it is good enough for clinicians to use, better than clinical gestalt [28,29], and an impact analysis has not been done to determine if it would lead to fewer CTAs and d-dimers being performed. Clinical Questions: What are the characteristics of ED attendances with possible AAS, how effective are existing clinical decision tools (ADD-RS, Canadian Guideline, Sheffield, AORTAs) and the use of CTA in an undifferentiated cohort of ED patients? Reference: McLatchie et al and DAShED investigators. Diagnosis of Acute Aortic Syndrome in the Emergency Department (DAShED) study: an observational cohort study of people attending the emergency department with symptoms consistent with acute aortic syndrome. EMJ Nov 2023. Population: Adult patients 16 years of age or older attending one of 27 EDs in England, Wales, or Scotland with onset of symptoms within the past seven days of possible AAS (chest pain, back pain, abdominal pain, syncope, or symptoms related to mal perfusion). Excluded: Absence of any potential AAS symptoms (chest pain, back pain, abdominal pain, syncope, or symptoms related to mal perfusion). Intervention: Clinical judgment, various clinical decision tools characteristics and performance of existing clinical decision tools (ADD-RS, AORTAs, Canadian and Sheffield AAS CDTs), the D-dimer (evaluated separately and in combination with other tools) and CTA. Comparison: None Outcome: Primary Outcome: Diagnostic accuracy of clinical gestalt, ADD-RS, AORTAs, Canadian and Sheffield AAS clinical decision tools and D-dimer (separately and in combination). Secondary Outcomes: The proportion of patients in whom the ED clinician thought AAS was a possible differential diagnosis, and most likely diagnosis, who had confirmed AAS. The proportion of patients in whom the ED clinician thought AAS was not a possible differential diagnosis but had confirmed AAS Proportion of alternative diagnoses found on CT/CTA and final hospital diagnosis. Median time from hospital presentation to imaging diagnosis Type of Study: This was a multicenter observational cohort study that recruited patients both prospectively and retrospectively with symptoms suspected of having AAS. Authors’ Conclusions: “Only 0.3% of patients presenting with potential AAS symptoms had AAS but 7% underwent CTA. CDTs incorporating clinician gestalt appear to be most promising, but further prospective work is needed, including evaluation of the role of D-dimer.” Quality Checklist for Observational Study: Did the study address a clearly focused issue? Yes Did the authors use an appropriate method to answer their question? Yes Was the cohort recruited in an acceptable way? No Was the exposure accurately measured to minimize bias? Yes Was the outcome accurately measured to minimize bias? No Have the authors identified all-important confounding factors? No Was the follow-up of subjects complete enough? Unsure How precise are the results? Unsure Do you believe the results? Yes Can the results be applied to the local population? Unsure Do the results of this study fit with other available evidence? Yes Funding of the Study - Supported by a Royal College of Emergency Medicine (RCEM) research grant. Dr. Reed is supported by an NHS Research Scotland Career Researcher Clinician award. No conflicts of interest by the authors were reported. Results: They identified 5,548 patients prospectively (37%), retrospectively (48%) and unknown (15%). The mean age was 55 years with 47% being male. The most common complaint was pain (54%) followed by abdominal pain (38%), back pain 23% and syncope (13%). Physician gestalt was available in 74% of the patients. AAS was considered a possibility by the physician in 24% of the patients (n=1,082). A D-dimer was performed 13% of the time with 40% being elevated. A CT scan was performed in 10% of the cases with 78% being CTAs. A total of 14 patients were confirmed to have an AAS (0.3%). The median time from ED arrival to confirmation of AAS was 6 hours. 33 of the patients who had a CT scan were diagnosed with an alternative aortic pathology. This included four ruptured thoracic aortic aneurysms, five ruptured AAA, 21 non-ruptured thoracic or AAA and three previously known stable aortic dissection/intramural hematoma or penetrating ulcers. Key Result: Most patients (99.7%) presenting to the ED with possible AAS do not have the condition and CDT did not improve on clinical gestalt. Primary Outcome: Diagnostic accuracy of clinical gestalt, ADD-RS, AORTAs, Canadian and Sheffield AAS clinical decision tools and D-dimer (separately and in combination). Secondary Outcomes: The proportion of patients in whom the ED clinician thought AAS was a possible differential diagnosis (1%), and most likely diagnosis (3.4%), who had confirmed. The proportion of patients in whom the ED clinician thought AAS was not a possible differential diagnosis but had confirmed AAS (0.06%) The proportion of alternative diagnoses found on CT/CTA (5%) and top 5 final alternative diagnoses (27% pulmonary embolism, 26% lower respiratory tract infection, 21% abdominal aneurysm not ruptured, 15% acute coronary syndrome and 8% cholecystitis) The median time from hospital presentation to imaging diagnosis was 6 hours (IQR 3 - 63; n=13) 1. Recruitment Challenges: Despite extensive advertisement and site engagement, around half of the patients could not be recruited prospectively, which may have introduced bias into the study. Conducting research in this area is challenging, and some patients with AAS may be missed because the diagnosis is not considered. 2. Missing Cases of AAS: It is possible that there were missed cases of AAS as 90 participants did not undergo CTA, and the study focused on patients being primarily investigated for pulmonary embolism, which may have a higher prevalence of AAS. 3. Partial Verification Bias (Referral Bias, Work-up Bias):  This happens when only a certain set of patients who underwent the index test is verified by the reference standard. Only those patients with positive D-dimers would receive a CT/CTA to confirm the diagnosis. This would increase sensitivity but decrease specificity. 4. Hawthorne Effect (Observer Effect): When you conduct a research study with prospectively collected data there is a possibility of introducing the Hawthorne effect.