INTRODUCTION No single non invasive test is both sensitive and specific enough Test: “ruling in” PE, e.g. helical CT or “ruling out” e.g. D-dimer, others do both, often non-diagnostic e.g. VQ scans Choice of initial diagnostic test guided by clinical assessment of probability & patient characteristics that may influence test accuracy Clinical assessment alone unreliable, objective testing crucial Failure to Dx PE high mortality, incorrect Dx of PE exposes patient risks of anticoagulation. OBJECTIVES OF THIS REVIEW Outline approach to Dx of PE that minimises the use of Pulmonary Angiography (PA) Based on 2 guiding principles Accurate test/combination of tests should have a positive predictive value 85% & a negative predictive value of 95% OR Be associated with no more than 2% VTE during F/U if it is the basis of withholding treatment CLINICAL ASSESSMENT 2 Categories of clinical assessment ◦ Empirical clinical assessment: Hx, Examination, CXR, ECG, ABG – low, intermediate & high probability categories PIOPED & McMaster studies, prevalence of PE established by PA was 15%, 38%, 79% ◦ Standardized clinical model (or prediction rules): Wells & colleagues used S&S, alternative Dx possible & presence of risk factors for VTE- low, intermediate & high probability categories CLINICAL ASSESSMENT Wells Score ◦ Variable Points S&S of DVT 3.0 Alternative Dx less likely than PE 3.0 HR>100 b/min 1.5 Immobilization/Sx past 4/52 1.5 Previous DVT/PE 1.5 Haemoptysis 1.0 Malignancy(on Rx/in past 6/12) 1.0 ◦ High >6, Moderate 2-6, Low <2 Clinical assessment SUMMARY Wells scoring, prevalence of PE 2% low probability group, 19% intermediate & 50% in high probability group Evidence shows that clinical assessment
( empirical/standardised) can stratify patients’
probability of having PE Prevalence expected: <10% low probability, 20% intermediate group & >60% high probability category NATURAL HX OF VTE Most cases DVT (~90%) start in the calf Isolated calf DVT rarely causes leg symptoms or PE 25% unRxd calf DVT will extend to proximal veins, do so within a week of presentation 75% of pts with PE have DVT Most pts with symptomatic PE have incr D-dimer ~50% symptomatic PE involve lobar or main pulmonary arteries Without Rx,1/2 pts symptomatic DVT/PE have recurrence within 3/12 With Rx of PE, ~50% resolution of perfusion defects in 2-4/52. Complete resolution occur in 2/3 of pts With Rx of proximal DVT, residual thrombus is seen on USS in ½ of pts at 1 yr D-DIMER TESTING D-dimer – cross-linked fibrin lysed by plasmin Elevations are non-specific: infection, inflammation, ageing, cancer, cardiac ischemia Wide variety of assays Valid assays for PE Dx: 2 Categories ◦ Very Highly sensitive D-dimer test Sensitivity >98%; Usu low specificity ~40%- high false positives Used to rule out PE ◦ Moderate-Highly sensitive D-dimer test Sensitivity 85-98%, not high enough to rule out PE, needs to be combined with another assessment ◦ HBH: Simple D-dimer assay: 100% sensitive, ~50% specific OTHER DIAGNOSTIC TESTS VQ SCAN ◦ Normal scan excludes PE ◦ Perfusion defects are non-specific-1/3 pts with defects have PE ◦ Probability increases with increase in number and size of perfusion defects & presence of normal ventilation scan CT ◦ Spiral/Helical scans with contrast (CT-PA) MRI Tests for DVT: USS, Venography: indirect way of Dx PE OTHER DIAGNOSTIC TESTS Pulmonary Angiography: Requires more expertise & support staff; invasive, time consuming, more expensive, and less available Echocardiography: Transthoracic/TEE ◦ directly visualise thrombi in right heart chamber or central pulmonary arteries ◦ show right heart hemodynamic changes that indirectly suggest PE ◦ TEE visualise thrombi in central pulmonary arteries with a specificity of >90% DX OF PE IN PREGNANCY MODIFICATIONS IN MXM ◦ 1ST: USS of proximal veins – initial test ◦ 2nd : Amount of radio-isotope used for VQS reduced & duration of scanning extended ◦ 3rd: If PA performed, brachial approach used with abdominal screening ◦ 4th: In the absence of safety data, helical CT is discouraged CLINICAL SITUATIONS ALTER DIAGNOSTIC APPROACH/ TEST INTERPRETATION In-hospital patients Inpatients, especially after surgery, often have increased D-dimer levels that markedly reduce the value of D-dimer testing (e.g., specificity of 7% in inpatients versus 47% in outpatients). Treatment of presumptive pulmonary embolism D-dimer levels are estimated to decrease about 25% after 24 hours of heparin therapy, and this is expected to reduce the sensitivity of D-dimer testing (e.g., from 96% to 89%) High clinical probability ◦ D-dimer testing has little clinical utility in patients with a high clinical probability of pulmonary embolism, because specificity is lower in this group (e.g., 28% compared with 54% with low clinical probability) ◦ The combination of a lower specificity and high prevalence of embolism results in a low frequency of negative D-dimer results (e.g., 17% compared with 51% with low probability), which have a lower negative predictive value(e.g., 77% compared with 100% with low probability). Previous venous thrombo-embolism ◦ Imaging abnormalities associated with previous DVT or PE may persist and be misdiagnosed as recurrent VTE(e.g., decrease in positive predictive value of a high probability lung scan from 91% to 74% with a history of PE). ◦ In about half of patients with recently diagnosed DVT who present with suspected PE and have a high-probability lung scan, the abnormalities predate the onset of chest symptoms Influence of age on accuracy of diagnostic tests ◦ The specificity of D-dimer testing and lung scanning decreases with age (e.g., D-dimer specificity: 67% at ≤ 50 years versus 10% at ≥ 80 years ◦ Proportion of lung scans that are non- diagnostic: 32% at ≤ 40 years versus 58% at ≥ 80 years) Cardiopulmonary disease Cardiopulmonary disease (particularly lung disease) is associated with a high proportion of non-diagnostic lung scans (e.g., 78% [91% with COPD] versus 64%) and a lower positive predictive value with a high-probability defect (e.g., 83% versus 93%) Malignant disease Malignancy reduces the specificity of many tests for PE(e.g., D-dimer: 48% versus 82%) & may also result in false- positive results (e.g., high-probability lung scans or abnormal helical CT with intra-thoracic malignancy). Central venous catheters The arms and central veins should be considered as a source for emboli & as a target for diagnostic testing in patients with central venous catheters who are suspected of having PE Pregnancy As compared with non-pregnant patients, the prevalence of PE among pregnant patients who are investigated for PE is low (about 5% versus about 20%) and the prevalence of normal perfusion scans is high (about 70% versus about 25%) SUMMARY PE IS CONFIRMED BY: ◦ Pulmonary angiography: intra-luminal filling defect ◦ Helical CT: intra-luminal filling defect in a lobar or main pulmonary artery ◦ Ventilation–perfusion scan: high-probability scan and moderate/high clinical probability ◦ Diagnostic tests for DVT: evidence of acute DVT with non-diagnostic ventilation– perfusion scan or helical CT SUMMARY PE IS EXCLUDED BY: ◦ Pulmonary angiogram: normal ◦ Perfusion scan: normal ◦ D-dimer test: normal test with a very high sensitivity(≥ 98%) & at least moderate specificity (≥ 40%) ◦ Normal D-dimer that has at least moderately high sensitivity (≥ 85%) and specificity (≥ 70%) AND (a) low clinical suspicion for PE OR (b) normal alveolar dead space fraction ◦ Non-diagnostic VQS or normal helical CT, and normal proximal venous ultrasound scans AND (a) low clinical suspicion for PE OR (b) normal D-dimer test that has at least moderately high sensitivity (≥ 85%) and specificity (≥ 70%) QUESTIONS TAKE HOME MESSAGE When individual tests are non- diagnostic, it is possible to combine their results to confirm or exclude pulmonary embolism Assessment of clinical probability is of vital importance