Can D-Dimer Risk Stratify Potential Aortic Dissection Patients?
Search Strategy: Using the search term “aortic dissection” you conduct a broad/sensitive search using PUBMED Clinical Queries for diagnosis and obtain 3541 citations on July 8, 2009. Next, you conduct an unfiltered search using “D-dimer” (4637 citations) and then combine these two search results to obtain 36 citations. These results include several evidence-based reviews (Best Bets) and systematic reviews, providing all of the research material below.
You are working at a small Level II hospital evaluating a 37 year old man with chest pain. He states that he’s been having constant 5/10 right lower sternal pain with radiation into the upper chest and around the lower rib margins for 2-days. Today he presents with worsening pain and new light-headedness. Pain is described as heavy, non-pleuritic, and not associated with exertional dyspnea, movement, or palpation. He denies any fever or respiratory symptoms nor can he link the symptoms to any dietary indiscretions. He rarely drinks and has not had any alcohol in several weeks. His past medical history is unremarkable except for an episode of diverticulitis 10-years ago at which time IV dye allergy was identified. Periodically, he uses over-the-counter non-steroidal anti-inflammatory medications. He has a family history of HTN, CAD, and DM all of which were present in his parents only after age 60. He does not smoke nor use drugs. He works as a car mechanic and believes that the pain may have manifest after he finished working on a car.
His physical exam is notable only for a presenting blood pressure of 180/110. His ED workup includes a CBC, BMP, PT/PTT, troponin, lipase, and LFT’s — all normal. His CXR is unremarkable and his EKG shows NSR at a rate of 90. Subsequently, a RUQ ultrasound demonstrates a normal liver and gallbladder. Simultaneous administration of a GI cocktail, metoprolol, ketorolac and baclofen significantly improve his discomfort to 2/10 and his blood pressure to 170/95.
Working diagnosis: atypical chest pain, hypertension, and GERD. His PMD is willing to set up both an EGD and a stress test for the next morning. The patient is more than eager to go home and get his follow-up as scheduled. As you are about to discharge the patient, you walk by the physician lounge where a lone physician is grabbing a quick bite while watching a “Three’s Company” rerun. As you thoughtfully contemplate Suzanne Somers’ rollercoaster career, you recall the mediocre comedy of John Ritter and his untimely demise in 2003 of an occult aortic dissection. Suddenly, you are really questioning whether sending your mechanic home with timely follow up is such a good decision. Have you really risk-stratified him as low-risk for aortic dissection?
You inform the patient that you want to make sure he does not have a rare vascular condition called an aortic dissection. The patient asks how you will proceed so you explain that his dye allergy will necessitate either a transfer to another MRI-capable hospital or consult the on call Cardiologist to perform a trans-esophageal echocardiogram. Both options will take several hours. Since he would miss a playoff basketball game, the patient is not really interested in being a patient much longer. He asks you if there’s a blood test that can be done to make sure he doesn’t have aortic dissection. You decide to do a quick literature search on blood test markers in acute aortic dissection.
Population: Emergency Medicine patients with suspected aortic dissection
Intervention: D-dimer screen to risk-stratify aortic dissection
Comparison: Unaided clinical gestalt (no D-dimer)
Outcome: Diagnostic accuracy for aortic dissection
Fourth years: Diagnosis of acute aortic dissection by D-dimer: The International Registry of Acute Aortic Dissection Substudy on Biomarkers (IRAD-BIO) Experience, Circulation 2009; 119: 2702-2707. (http://pmid.us/19433758)
Article 1: Serum D-dimer is a Sensitive Test for the Detection of Acute Aortic Dissection: A Pooled Meta-analysis, J EmergMed 2008; 34: 367-376
Article 2: A Rapid Bedside D-Dimer Assay (Cardiac D-Dimer) for Screening of Clinically Suspected Acute Aortic Dissection, Circ J 2005; 69: 397 –403
Article 3: D-dimer in ruling out acute aortic dissection: a systematic review and prospective cohort study, Eur Heart J 2007; 28: 3067–3075
Article 4: Diagnosis of Acute Aortic Dissection by D-Dimer: The International Registry of Acute Aortic Dissection Substudy on Biomarkers (IRAD-Bio) Experience Circulation 2009;119: 2702-2707
The Succinct Answer
Because of its relative rarity (best pre-test probability ~0.2% for general ED patient population) but high lethality, acute aortic dissection remains a highly litigated disease with accusations of malpractice against treating physicians and hospitals. Von Kodolitsch (http://pmid.us/14715319) previously described that the absence of acute tearing or ripping pain, pulse or blood pressure differentials, or mediastinal or aortic widening on CXR reduced the aortic dissection risk from 50% to 7%. Future research should derive and validate clinical decision rules to reliably define and identify low-risk populations. Without a validated decision-aid to identify low-risk subsets appropriate for D-dimer screening, premature acceptance of this test as a valid screening test could paradoxically increase (http://pmid.us/19307380) expensive, time-consuming, risky diagnostic testing for the elusive aortic dissection diagnosis without improving diagnostic accuracy. Therefore, randomized controlled studies (http://pmid.us/16754927) will then need to prospectively assess the additional benefit D-dimer offers in sensitivity, specificity, reliability and impact on test-ordering.
Multiple studies using different D-dimer assays and cut-offs consistently demonstrate high, clinically relevant sensitivity (pooled sensitivity estimate 97%, LR- 0.06 with significant heterogeneity I2 = 47%) and negative predictive value for D-dimer to exclude acute aortic dissection among ED patients with chest pain. Whether higher D-dimer levels with lower sensitivity or NPV meet standard of care (http://pmid.us/15520714) (i.e. optimal test-treat threshold, http://pmid.us/7366635) remains to be determined. For now, D-dimer for aortic dissection is where D-dimer for pulmonary embolism was 10-years ago without validated CDR’s to differentiate no-risk from low-risk patient subsets.