Despite a wealth of literature, risk stratifying patients for pulmonary embolism remains subjective
One of the trickiest diagnoses in emergency medicine is pulmonary embolism. The problem is not a lack of information; the amount of literature on this topic is truly staggering. The problem with making the PE diagnosis is that ordering the definitive test, a CT pulmonary angiogram, is a big deal. It is very costly, there is a substantial amount of radiation (probably equivalent to about 500 chest x-rays) and IV contrast is required, which in some patients runs the risk of causing renal issues.
So emergency physicians are often in a quandary when it comes to making the diagnosis. There is a well researched graded pathway to approaching this diagnosis and it is very clear what the usual progression should be. First the physician needs to clinically assess what he/she thinks is the clinical probability that the patient has a PE. To do this the physician needs to consider a significant amount of information – the patients PMH (cancer, previous venous thromboembolic episodes, recent surgery or trauma), current clinical symptoms (shortness of breath, pleuritic pain, hemoptysis, etc), physical examination of the chest and legs, vital signs and oxygen saturations.
Physicians can use their own judgment in regards to risk stratification (although there is some literature that suggests that individual physicians can have grossly different perceptions of the risk of a PE in any one patient) or use any number of validated risk stratification tools – the Wells rules, the Geneva criteria or a number of others. The nice thing about using the validated risk stratification tools is that at least some use very black and white criteria and limit the need for subjective judgments.
The key decision is whether the patient is considered “no risk” or “low risk” – and this is the tough call. Traditionally, if a patient is considered no risk the idea is to not pursue the work-up. If a patient is considered “low risk” traditionally it is advised to do a d-dimer to look for fibrin degradation products. Now here’s the hooker. If the d-dimer is negative (and the hospital is using a good test – the ELISA versions are best), the evaluation is considered over. Nothing further need be done.
The problem occurs if the d-dimer is positive. Unfortunately, there is a high percentage of “false positives” – cases in which patients don’t have any discernible clots of clinical significance. In these cases, further testing is required. The safest test to do in this case is an ultrasound of the lower extremities looking for a clot in the legs. Unfortunately, ultrasonography is not as good at picking up clots in asymptomatic legs compared to symptomatic legs. Fortunately specific is very high. So if the exam is positive, a clot in the legs is present.
Although some feel that treatment for a PE can begin if a clot in the legs is found in a patient with symptoms consistent with a PE, most physicians, at least those in the United States, are going to want definitive proof that there is a clot in the lungs by some visualization process. The concern is that treatment of PE with heparin and then months of warfarin is risky with regarding to the precipitation of bleeding. But the fact that the patient has ultrasound-proven leg clots would still indicate the need for treatment if they are in the thigh (and some say in the calf, and I personally ascribe to this belief).
Therefore, in most cases, a low-risk patient (or moderate risk patient) with a positive d-dimer needs to undergo some sort of lung visualization process – either computer tomographic angiography or radioisotopic lung scanning. This is why it is so important that the d-dimer only be ordered by the treating physician. Ordering d-dimers by triage nurses is not a good idea given how important the risk stratification decision is to determining whether a d-dimer should be ordered or not.
The PERC rule was believed to be a way of excluding the diagnosis of PE without performing a d-dimer test. Many emergency physicians breathed a sigh of relief when this rule was published because it was very unequivocal, easy to apply and, if the patient was PERC-negative, the work-up could stop. Unfortunately, a recent study has cast some doubt on the PERC rule. First of all, in the study abstracted below, only 13% of the patients were PERC-negative. That means in the substantial majority of patients the PERC rule will not be helpful. Secondly, PERC-negative patients were determined to have a 5.4% chance of having a PE – missing one in 20 is just not good enough.
One of the factors to consider in this study that may be responsible for the high false-negative rate was the high incidence of PE – 21%. These numbers are substantially higher than those in U.S. studies where the diagnostic yield for PE is much lower – perhaps in the range of 5%. This being the case, perhaps the PERC rule works better when the incidence of PE is much lower as seen in the U.S. But this is only speculation and would not be very helpful in deciding whether to use the PERC rule or not
THE PULMONARY EMBOLISM RULE-OUT CRITERIA (PERC) RULE DOES NOT SAFELY EXCLUDE PULMONARY EMBOLISM. Hugli, O., et al, J Thromb Haemost 9(2):300, February 2011
BACKGROUND: D-dimer testing has been suggested as a means of excluding pulmonary embolism (PE), but a false-positive rate of 50% or higher has been reported, thus exposing many patients to unnecessary testing for PE. A lowered threshold for diagnostic testing might play a role in the decreasing prevalence of PE in patients undergoing diagnostic work-ups (from 30% to less than 10% according to one study). The “Pulmonary Embolism Rule-Out Criteria” (PERC) rule has been suggested as a means of safely excluding PE without the need for D-dimer testing. According to the PERC rule, the risk of PE is very low in patients below the age of 50 with a pulse rate below 100 and oxygen saturation above 94%, without unilateral leg swelling, hemoptysis, surgery or trauma within the previous four weeks, prior deep vein thrombosis or PE, or oral hormone use.
METHODS: These multinational European authors examined the reliability of the PERC rule as applied to prospectively collected data from 1,675 adult outpatients being evaluated in EDs for suspected PE.
RESULTS: Of the 1,675 patients, 13.2% were “PERC-negative” (i.e., fulfilled all of the PERC criteria) and, of these 221 patients, 85.1% had a low clinical pretest probability of PE while 14.9% had an intermediate probability according to the revised Geneva score. PE was diagnosed at the time of initial evaluation or during three months of follow-up in 21.3% of the patients overall, including 5.4% of the PERC-negative group and 6.4% of the PERC-negative patients with a low pretest probability according to the revised Geneva score.
CONCLUSIONS: Application of the PERC rule, with or without the revised Geneva score, is not a reliable method of identifying those patients with possible PE who do not require further diagnostic testing. 24 references (
- no reprints) Copyright 2011 by Emergency Medical Abstracts – All Rights Reserved 7/11 - #27
Two final points to consider relate to treatment. How important is early administration of heparin in suspected PE and are we giving the right dose? The first paper below seems to indicate that early initiation of heparin treatment is important, yet in this study only 70% of patients got their first dose in the ED. There were gross differences in hospital and 30-day mortality when treatment was started in the ED. The association of attaining a rapid therapeutic PTT with the better outcomes suggests that the speed of treatment would matter.
EARLY ANTICOAGULATION IS ASSOCIATED WITH REDUCED MORTALITY FOR ACUTE PULMONARY EMBOLISM. Smith, S.B., et al, Chest 137(6):1382, June 2010
BACKGROUND: Although guidelines suggest consideration of anticoagulation even before confirmatory testing for some patients with a high clinical suspicion of pulmonary embolism (PE), only limited information is available regarding the timing of anticoagulation and mortality.
METHODS: The authors, from the Mayo Clinic in Rochester, MN, performed an implicit chart review regarding the timing of initiation of heparin therapy, the interval to therapeutic anticoagulation and mortality in 400 patients (median age, 68) with PE diagnosed in the ED with CT angiography.
RESULTS: Heparin was started in the ED for 70% of the patients and prior to establishing the diagnosis of PE in 5%. Most of the patients (85.8%) had a therapeutic aPTT within 24 hours of ED arrival. When compared with initiation of heparin after hospital admission, initiation of treatment in the ED was associated with lower mortality during the inpatient stay (1.4% vs. 6.7%, odds ratio [OR] 0.20) and at 30 days (4.4% vs. 15.3%, OR 0.25). A similar pattern was noted for patients who achieved a therapeutic aPTT within 24 hours (1.5% vs. 5.6% inpatient mortality in those achieving a therapeutic aPTT after 24 hours, and 5.6% vs. 14.8% 30-day mortality, respectively). The median time to achievement of a therapeutic aPTT was 20.9 hours among patients who died in the hospital vs. 10.7 hours among those who did not. Administration of heparin in the ED remained predictive of decreased mortality on multiple logistic regression analyses that controlled for some potential confounders - although the data cannot address the fact that in most if not all cases the delay to heparin was based on a delayed diagnosis of PE, which almost certainly reflects not only a different presentation, but also importantly different comorbidity.
CONCLUSIONS: The authors support initiatives to promote early initiation of anticoagulant therapy in patients with acute PE. 42 references (
- no reprints) Copyright 2010 by Emergency Medical Abstracts - All Rights Reserved 11/10 - #26
Finally, are we giving adequate doses of heparin – particularly in the obese patient? If the speed by which a patient attains a therapeutic PTT matters, as suggested by the previous paper, then how important is it that dosing in obese patients be adequate. The next paper suggests that many patients who are obese (BMI over 30) are being under dosed.
They ascertained that 89% of patients received less than what they deemed to be the appropriate dose (80 units/kg loading and 18units/kg/hr maintenance). The study determined that there only 29% of patients achieved a therapeutic PTT within 24 hours of initiation of therapy and 14% failed to reach the goal within 48 hours. Interestingly, loading doses did not correlate with speed of achievement of a therapeutic PTT, but the infusion dose did.
DOSING OF UNFRACTIONATED HEPARIN IN OBESE PATIENTS WITH VENOUS THROMBOEMBOLISM. Hurewitz, A.N., et al, J Gen Intern Med 26(5):487, May 2011
BACKGROUND: Because the risk of a recurrence of venous thromboembolism (VTE) in overweight or obese patients is nearly double that of normal-weight individuals, aggressive weight-based heparin regimens have been developed for these patients. Although it has been observed that high initial heparin doses reduce the risk of recurrence without significantly increasing the risk of bleeding, physicians often resist prescribing recommended high doses for obese patients.
METHODS: The authors, from Winthrop University Hospital in Mineola, NY, reviewed initial heparin treatment in 84 adult inpatients with VTE and a body mass index (BMI) of 30 or higher. The dosing protocol in this population called for a bolus dose of 80 units/kg followed by infusion of 18 units/kg/hr. RESULTS: The mean initial heparin dose was 58 units/kg, with 89.3% of the patients receiving a dose that was lower than the recommended 80 units/kg. The mean infusion rate was 13 units/kg/hr, with 76.2% of the patients receiving infusion rates below the recommended dose.
Nearly 30% of the patients received a fixed infusion of 1000 units/hr. The gap between the recommended and administered infusion rate increased with increasing body weight (by about 100 units/hr per 10kg increase in weight).
The median interval to a therapeutic PTT (greater than 60 seconds) was 18.5 hours, with 28.6% of the patients failing to achieve this goal within the first 24 hours and 14.3% failing to reach the goal within 48 hours. There was a significant correlation between the interval to a therapeutic PTT and the heparin infusion dose, but not the initial bolus dose. Each 1 unit/kg/hr decrease in the infusion dose translated to a delay of 0.75-1.5 hours in the interval to achieve a target PTT.
Copyright 2011 by Emergency Medical Abstracts - All Rights Reserved 11/11 - #3
Bottom line – it appears that the diagnosis of PE remains elusive (and costly) and it seems important to give the right dose of heparin quickly.
W. Richard Bukata, MD, is the Medical Editor for Emergency Medical Abstracts