Detecting PEs ============= (much of this is from an Internet emed-l discusssion list conversation between Keith Conover and Craig Feied) þ D-dimer and alveolar dead space in the ED þ Clinical symptoms of PE 1) tachypnea 2) dyspnea (84% of those with PE) 3) pleuritic chest pain (74% of those with PE) and 4) physical evidence of DVT. - According to the PIOPED studie(s) at least one of the previous four findings are present in 97% of patients with PE. Percentages above from Rosen, 4E. þ Risks for PE: - Hypercoagulability: + pregnancy, + oral estrogen, + polycythemia, + protein C deficiency, + protein S deficiency, + antithrombin III deficiency, + malignancy + smoking þ Interpreting VQ scan results þ Spiral CT for PE: þ PE from sites other than proximal thigh þ How good do we have to be at detecting PEs? - Each one of us has to set his/her threshold for missing serious disease. I myself am content to miss 50% of PE's (the national average being to miss 80%) but I would prefer not to send home more than one or two acute MI patients per decade! My position makes no sense from a total mortality standpoint, but I justify it by noting that these two goals are both barely achievable with extreme care, and that the costs of doing better than this are astronomical given current technology. --Craig Feied, MD FACEP cfeied@ncemi.org þ Association of PE and GI bleed - Another interesting PE point is that there is an association between PE and upper GI bleeds for some reason, perhaps related to endogenous thrombolysis. The blood can scare you from wanting to anticoagulate but it's easier to transfuse for bleeding than to raise the dead sometimes. --Don Elton þ Fever and PE - There is one patient with PE discussed on the emed-l discussion list who had a temperature of 102.4 at presentation. (Bob Lang MD, BLang47847@aol.com) þ Hematological risks for DVT and PE - anti-thrombin III, protein S or protein C deficiency. - protein S deficiency is known to occur in AIDS and is a risk factor for PE in this group of patients. þ Coumadin as a Risk Factor for DVT/PE þ Value of the ABG and aA gradient for detecting PEs: þ Value of CXR in suspected PE: - >So, algorithm: >patient with pleuritic chest pain not clearly related to trauma >--> check CXR for pneumo/infiltrate/etc. if negative... Not to be obnoxious, I'd just like to mention that the most commonly given reason for a missed PE found at autopsy is that the cxr showed an 'obvious pneumonia'. The clinical setting still has to carry a lot of weight here... and we'll miss a lot no matter what we do. The real key is AFTER the patient is admitted, when he/she isn't getting better on antibiotics, or when the cultures are negative, or when the infiltrate jumps to another area, or when the sterile effusion appears. At that point, an admission dx that was written as 'probable pneumonia, r/o PE' can sometimes be life-saving! --Craig þ Low pO2 or high aA gradient in healthy person with pleuritic chest paina suggests Mycoplasma - >--> check ABG for Aa gradient; if abnormal, suspect Mycoplasma but go on >to next step anyway Agreed. --Craig - >Only question is why bother with an ABG at all? Why indeed? There is not much value to ABG nor to pulse ox in the _differential_ part of any workup for any disease. It's good for all the obvious things... þ VQ or duplex Doppler of leg first? - >--> if low or high clinical suspicion of PE, get VQ; if intermediate >clinical suspicion, get venous Doppler (see chest pain chapter of latest >Tintinalli for reasoning). (continuing on as per nice flowchart in >Tintinalli) Personally I think the best overall approach is the one outlined in the Oudkerk article (above) which I think Harvey has already cited on emed-l. I think either Color-flow duplex or V/Q is a reasonable starting point for almost any patient, and if the first test is not definitive, I'd add the second. Of course, a positive duplex is diagnostic, while a negative result on duplex is of zero value, in contrast to V/Q, where a negative result is valuable unless there's a _very_ high suspicion. -- Craig þ Risk factor analysis for PE - > Does >risk factor analysis (immobilization, BCPs, obesity, smoking, etc.) >have any legitimate role in the evaluation of pleuritic chest pain? IMHO, no. All of the published studies that purport to show the presence of risk factors in most patients with PE are a posteriori studies, and knowledge of these factors in a 'typical' population at risk fails to discriminate between those with and those without the disease. The early reports were from incredibly bad science, as patients without the risk factors were preferentially assigned to a 'presumptive' negative diagnosis and were refused the diagnostic opportunity for detection of the disease. The ultimate in self-fulfilling hypothesis-testing. - Finally, the question of the utility of risk-factor analysis has been raised. Although estimates are variable, it is clear that an absence of risk factors does not obviate the need for work-up of patients with presenting with symptoms or signs consitant with PE. Absence of risk factors have been documented in from 12% to 25% of cases of documented PE. - The issue of Mycoplasma pneumonia entering into the differential diagnosis of PE has been rasied. Although it it true that hypoxemia in these cases is often out of proportion to the radiographic findings, it is also true that pleuritic chest pain is a very uncommon presentation for mycoplasma. --H. Louzon MD references: þ When do you get Dopplers for DVT or V/Q for PE? Any time you think of the possibility of DVT or PE. You can't rule them out by H+P. - >My feeling at this point about venous Dopplers (as limited as they >are) is that if you have _any_suspicion of DVT based on history or >physical, get a Doppler. We are lucky in having good U/S techs on >call 24 hours a day, and we use them all the time. It turns out that >history and physical are terrible at ruling out DVTs, and Dopplers are at >least moderately good at ruling them out (comparatively that is). I totally agree. Clinical diagnosis of DVT is almost exactly half-reliable in both directions. - >I'm starting to get the same feeling about V/Q scans: if you have >any suspicion of a PE, get a VQ. It's not quite as easy to get as >Dopplers (may be unique at our institution), but the radiation >exposure is low, and it's almost noninvasive. Too bad it's such a >poor test, and costs so much. (And we have to get radiologists to >agree to it rather than just calling in the vascular ultrasound >technicians as with Dopplers.) Again, I agree. The problem is that a negative V/Q is very valuable to me as a clinician, when I'm trying to make sure that it's reasonable for me to send home a patient with unexplained chest pain that's probably nothing -- and the risk-benefit calculation is overwhelmingly in favor of working these people up as much as possible. Unfortunately, the radiologist thinks of every negative V/Q the way the surgeon thinks of a negative appy -- they are forced to admint that a few are inevitable, but they have the sneaking feeling that a really good doctor would never have any negatives... --Craig Feied, MD FACEP cfeied@ncemi.org þ Use of D-dimer to detect PE - summary: ELISA may be valuable, but not the cheap latex agglutination that is generally available. AN MKYA-711817 AU Heit, John A. Nichols, William L. Goldhaber, Samuel Z. Simons, Grant R. TI Letters: Plasma D-Dimer Levels and Diagnosis of Pulmonary Embolism. IN Mayo Clinic, Rochester, Minn. Brigham and Women's Hospital, Boston, Mass. SO JAMA. 1994 May 11. 271(18). p 1404. PU Copyright 1994 by the American Medical Association, 515 N State St, Chicago, IL 60610 To the Editor.--Dr Goldhaber and colleagues *RF 1 * report a 93.3% sens and 91.4% negative predictive value for a fibrin D-dimer value less tha ng/mL (Asserachrom D-Di enzyme immunoassay kit, Diagnostica Stago, Asnieres-sur-Seine, France) for acute pulmonary embolism using pulmonar angiography as the diagnostic reference standard. The authors report "n significant relationship between heparin use and D-dimer levels (P=.78) do not report the duration of heparin therapy or duration of symptoms p D-dimer measurement. Using the same assay, Bounameaux et al *RF 2 * fou D-dimer levels to be significantly lower on days 3 and 7 compared with of presentation (day 1). Speiser et al *RF 3 * reported that elevated l of D-dimer decrease within 24 hours in patients with acute deep vein thrombosis who are treated with heparin. Therefore, the D-dimer sensiti for acute pulmonary embolism may be appreciably reduced in patients who been symptomatic for several days or who have received several days of therapy. Do the authors have additional data regarding the impact of ei duration of symptoms or duration of heparin therapy on the operating characteristics of the D-dimer assay for acute pulmonary embolism? We agree with the authors' conclusion that "plasma D-dimer levels measured using an ELISA (enzyme-linked immunosorbent assay) hold promise as a useful tool in the evaluation of medical patients with suspected acute PE." However, it is important to emphasize that this study used a very sensitive laboratory method (ELISA) for measuring the D-dimer level. The ELISA is relatively time consuming and generally not amenable to rapid turnaround laboratory reporting on an around-the-clock basis. Most laboratories measure D-dimer levels using latex agglutination, which may have a lower sensitivity for acute pulmonary embolism *RF 4 *. Furthermore, the correlation among commercial ELISA or latex D-dimer kits is poor, possibly attributable to recognition of different epitopes or different fibrinolytic fragment analytes by D-dimer antibodies *RF 4,5 *. Additional studies of the influence of D-dimer assay methods on test operating characteristics for acute pulmonary embolism are needed. We caution physicians to be aware of the D-dimer laboratory method being used by their respective laboratories in order to use the test appropriately in the diagnostic evaluation of patients with clinically suspected acute pulmonary embolism. John A. Heit, MD William L. Nichols, MD Mayo Clinic Rochester, Minn In Reply.--We emphasize, as do Drs Heit and Nichols, that in the diagnostic work-up of pulmonary embolism a quantitative plasma D-dimer ELISA should be used, rather than the more commonly employed but semi-quantitative latex agglutination assay. Though rapid and simple, the latex agglutination D-dimer is just not sufficiently sensitive to be useful clinically as a screening test for pulmonary embolism. (It is adequate, however, to test for disseminated intravascular coagulation.) Fortunately, skilled technologists can set up and perform the quantitative ELISA for D-dimer in about 4 hours. The resources and training required to use the D-dimer ELISA on short notice are substantial, but not nearly as demanding as the maintenance of around-the-clock availability of pulmonary angiography. Even at hospitals capable of performing angiography, it often takes 4 hours to orchestrate the logistics and to perform this invasive diagnostic test. We observed no difference (P=.78) in the average (+- SD) D-dimer levels of those patients who received heparin (2560 (+- 2421) ng/mL) compared with those who did not (2382 (+- 2287) ng/mL). Although Bounameaux et al *RF 1 * reported a diminution in the sensitivity of plasma D-dimer ELISA over time, this trend was actually quite modest. They found a sensitivity of 98% at baseline, 96% on day 3, and 93% at 1 week after admission among patients who presented to their emergency department with suspected pulmonary embolism. We have no additional data on duration of symptoms or of heparin therapy. Samuel Z. Goldhaber, MD Grant R. Simons, MD Brigham and Women's Hospital Boston, Mass RF 1. Goldhaber SZ, Simons GR, Elliott CG, et al. Quantitative plasma D-dimer levels among patients undergoing pulmonary angiography for suspected pulmonary embolism. JAMA. 1993;270:2819-2822. 2. Bounameaux H, Cirafici P, de Moerloose P, et al. Measurement of D-dimer in plasma as diagnostic aid in suspected pulmonary embolism. Lancet. 1991;337:196-200. 3. Speiser W, Mallek R, Koppensteiner R. D-dimer and TAT measurement in patients with deep venous thrombosis: utility in diagnosis and judgement of anticoagulant treatment effectiveness. Thromb Haemost. 1990;64:196-201. 4. van Beek EJR, Schenk DE, van den Ende A. Comparative analysis of D-dimer assays in patients with clinically suspected pulmonary embolism. Thromb Haemost. 1993;69:624. 5. Ellis DR, Eaton AS, Plank MC, Butman BT, Ebert RF. A comparative evaluation of ELISAs for D-dimer and related fibrin (ogen) degradation products. Blood Coagul Fibrinolysis. 1993;4:537-549. 1. Bounameaux H, Cirafici P, de Moerloose P, et al. Measurement of D-dimer in plasma as diagnostic aid in suspected pulmonary embolism. Lancet. 1991;337:196-200. IS 0098-7484 PT Letter (LET). UP 94053