Click here to view next page of this article
Prophylaxis. The standard for prophylaxis for surgical patients is subcutaneous heparin has a substantial effect on the incidence postoperatively of DVT, PE, fatal pulmonary emboli and on total mortality. There are numerous publications now that establish that the risk for surgical patients can be reduced in these things and that it is going to have a significant effect on postoperative mortality.
You should know that in ischemic stroke patients there have been met-analysis now that suggest an 81% reduction in DVT and an 18% reduction in early patient mortality. So it does appear that in these patients, at least in a certain very high risk group of medical patients, stroke patients, you can reduce the incidence of DVT by using prophylactic therapy. It’s not established, even in this very high risk group.
In general applications there are now small series which suggest reductions in DVT’s. This is taken on all comers admitted to a medical service. But again, the evidence for mortality benefit is substantially less. At the same time, the average cost for prophylaxis is less than $100 per patient if you use unfractionated low-dose heparin. If you use low micro weight heparin the cost is substantially greater. And I would say at this point, there is no compelling evidence - although there is an enormous amount of enthusiasm, particularly in European literature - for the use of low micro weight heparin in prophylaxis. There is no compelling evidence that _ heparin is a more effective prophylactic regime than is unfractionated heparin.
Let’s talk about risk factors for pulmonary embolism as we get into diagnostics. The things that feel out as major risk factors for PE were immobilization, surgery within three months - obviously those things may go hand in hand - stroke, a prior history of venous thromboembolic disease and malignancy. In the Nurses Health Study that was done, a huge epidemiologic study which has been a gold mine really for a variety of therapy … of cardiovascular and other risks and therapies. Its traditional risk factors were evident. Obesity, conferred a relative risk of almost 3.
The presenting symptoms for pulmonary embolus were all those things that you typically think of. So in the _ study, of patients who had no prior cardiovascular, cardiopulmonary disease, the presenting symptoms were shortness of breath, dizziness, in about 75%, pleuritic chest pains in about 65%, 66%, so just about 2/3rds had pleuritic chest pain. Cough in about one-third and hemoptysis in a relatively small percent, a little less than 15%. So those are the presenting symptoms of patients with pulmonary embolus in the populace. Presenting signs, were tachypnea in about three-quarters, crackles on exam, multiple crackles on exam in about half, tachycardia in about one-third.
Although clinical judgment is helpful - and I think extremely useful as you work up patients for pulmonary embolism, the majority of patients are assigned an intermediate probability. So that we, as we think about this, we are able to say it’s not low, it’s not very low, but it’s not very high either and I’m just not sure. And as you’ll see in a second, one of the problems you have is trying to develop a diagnostic strategy for PE. Our clinical judgment really doesn’t discriminate that well among the patients that we are evaluating.
The other thing is that the clinical judgment is fallible. We all know this from a variety of experiences. But this is probably indeed where my fallibility has been proven repeatedly during the course of the year. One-third of the patients were thought to have classic signs and symptoms for pulmonary embolism, were assigned a high P-test probability, turned out not to have the disease. So in those cases in whom the clinicians were absolutely certain, "This is it, there is nothing else it could be" - and you are often in that situation where you are debating how to proceed and someone says, "What else could it be besides a pulmonary embolus?" and everybody scratches their head and says, "I can’t think of anything else" - one-third of those patients turned out not to have PE when they went through diagnostic evaluations. So while they couldn’t think of anything else, it turned out to be something else. So this is not a disease that we are very good at diagnosing based on our clinical judgment.
Remember also that fewer than one-third of the patients with pulmonary embolism in the study had symptoms of DVT, so that didn’t help much. The majority of cases who turned out to have pulmonary embolism had no symptoms or signs of deep venous thrombosis. Remember also that 50% of the patients with DVT without symptoms of PE will have abnormalities on a perfusion scan or lung scan suggestive of pulmonary embolus, when they are scanned. So the symptoms that we look for to tell us that someone may be having a coronary are not terribly sensitive. Patients can have disease. They can have significant pulmonary emboli without having symptoms and that is the other danger of this disease.
Remember also, non-invasive tests for DVT: impedance plethysmography and IVG are the two that I am going to talk about. And Doppler ultrasound. Repetition here to increase the sensitivity so if did the test, you have a high clinical suspicion and you did the test and it’s negative, don’t be dissuaded by your radiologist but do the test again. Because in fact we will pick up some patients the second or even the third time around.
What about ultrasonography? A point to remember about it is it’s sensitive in symptomatic patients with their first episode of DVT. I’ll explain that in a second. And then its accuracy is less in asymptomatic patients. But the particular point about Doppler ultrasonography that you need to keep in mind is that it has difficulty … it’s not a useful test particularly in patients who have had prior deep venous thrombosis. Doppler studies remain positive really indefinitely in patients who have prior clotting. So if you are evaluating patients having second or third episode, if you’ve got prior IPG’s and there’s a change, that’s helpful. But the Doppler ultrasonography is not going to distinguish old clots from new clots and therefore really can’t help you in the patient who is coming back for a second or third episode of symptomatic venous distention.
Both ultrasound and IPG are more sensitive when used to evaluate symptomatic patients than when used to evaluate asymptomatic patients. Constantly, in our institution, when people talk about Doppler ultrasonography they cite a greater than 90% sensitivity and specificity. That’s true but it’s only true for symptomatic patients and as we said earlier, many of the patients with pulmonary embolic disease have asymptomatic deep venous thrombosis. In that situation the sensitivity in these studies is about 50-65%. So there are a lot of false negative studies in asymptomatic patients. If you are using Doppler ultrasonography or IPG to screen somebody who has symptoms of DVT it’s extremely sensitive and very specific. If you are screening somebody, as we often do because someone has had symptoms consistent with a PE and trying to tighten your pretest probability if you will, it’s a lot less useful. A negative test doesn’t tell you as much. A positive test is still useful. It’s a relatively specific test but it’s not a sensitive test.
As I said, they are sensitive and specific if the patient is symptomatic and the evidence for that are several studies. One study in Canada following patients out … patients with symptoms suggesting deep venous thrombosis who had negative Doppler studies, they followed them out for 12 months and there were no fatal PE’s and only 2% of the patients developed proven or established DVT in the next 12 months. Suggesting that in negative studies actually pretty helpful. Similarly in another study in the Netherlands there were only less than 1% of DVT and no fatal coronary emboli in six months in patients who were untreated after a Doppler study suggested that their symptoms were not DVT. So it’s a useful test to exclude the diagnosis of DVT in patients who are symptomatic. It’s not a useful test to exclude the diagnosis of DVT in patients who are asymptomatic.
One of the problems, as I said, is local skills make a big difference here. One of the problems with the approach now to patients is if you get a negative study in a patient and you are still suspicious in an asymptomatic patient, you are still suspicious of the disease in the venous system, venography isn’t all that helpful anymore because the local expertise has disappeared in most institutions. Venography was always a difficult test to interpret. It was always a test that required a lot of … relied a lot on the skills of the radiologist, mainly to interpret the study and most of the skills have been lost as people have gone more and more to non-invasive testing. So venography is much less available to us in most institutions and much less helpful than it once was.
Okay, what about diagnosis of pulmonary embolus as this emerges with diagnosis of PE? I’m going to start by talking about lung scanning. Recognize that lung scanning is often diagnostic alone. Most serious, more than half of the lung scans that are done are considered non-diagnostic scans. They are most helpful if they are either normal or high probability but that’s a minority. The thing that I think is also important to recognize is that within institutions there is substantial both intra-observer and inter-observer variability. If you show the same scan to the radiologists repeatedly you get … there is some variation in terms of their reporting and there is also a significant inter-individual variability. So you need to know the abilities of your nuclear medicine people as you interpret lung scan information as a clinician.
A normal lung scan firstly excludes the diagnosis of PE. This is one thing we can say. Now theoretically you can have a normal lung scan and have a PE. But that’s a relatively uncommon occurrence to have those two things simultaneously. Most people who develop PE will have prior warning and a lot of their PE will show up. So a truly normal lung scan virtually excludes the diagnosis. In one series over 500 patients with negative scans, normal scans, followed out six months. The rate of pulmonary embolus was less than 1%. So a very small rate of PE considering the fact that the patients were considered to have some risk factors for PE at the time they were first studied.
There is a set of high probability scans reasonably specific. If you join a high probability scan with a high pre-test probability, that’s the clinicians judgment, you get about a 95% specificity. But they are not that sensitive. There are a lot of people who have pulmonary emboli who don’t have high probability scans. It’s only the sensitivity is around 42% in high probability studies. So that’s not tremendous. So if you get a high probability scan in a patient in whom you have a high clinical suspicion you are in good shape. Most of us would feel comfortable acting upon that judgment and treating the patient. But if you have a high clinical suspicion and the scan is less than high probability, and anything other than normal.
But one reasonable diagnostic strategy that many of us have adapted is that if we have had an intermediate probability scan and an intermediate or high probability clinical suspicion, you might go ahead and do lower extremity studies to try and increase your probability. If they are positive, that is helpful. If they are negative it’s somewhat helpful, recognizing though that as I said, the lower extremity studies are much more sensitive in patients who have symptomatic disease. So if the patient is asymptomatic for DVT but has clinical signs and symptoms of PE, you got an intermediate probability scan, do you really feel comfortable ruling out the diagnosis with the lower extremity studies?
This is where I think we may be safe, and the problem, as I said, with pulmonary angiography has been that many institutions didn’t have the ability to do it easily or well. We were all reluctant to subject a patient in whom we thought was clinical suspicion, intermediate scan. Three-quarters of those patients are going to turn out not to have PE. Should you really subject them to pulmonary angiography and its attendant risks? Plus, even in our institution which is a pretty large referral institution, we … pulmonary angiography is actually incredibly difficult and the number of pulmonary angiograms that get done at the institution over the course of a year is relatively small. So the local expertise in smaller institutions really doesn’t exist. And I will tell you, having done this over 15 years, reading pulmonary angiography is not easy. It is not uncommon for us to get a group of angiographers gathered around a scan arguing about the interpretation. And while they don’t talk about it much, places that do a lot of angiography and are honest about their interpretations will tell you that there are a number of pulmonary angiograms that are truly indeterminate. They can’t really diagnose PE but they can’t rule it out either. I think that what’s going to happen over the next several years though is that CT-angiography.
One of the criticisms of CT-angiography initially was that it only defined clot out to the third order vessels. That was thought not to be too much of a problem because clot in smaller vessels was thought to be insignificant, not hemodynamically significant. And yet negative or extremity studies that would really rule out high risk of further clot. But I’ll tell you, as the scanners have gotten faster this is no longer true. You can define vessels well out, easily out to the pleural surface with the newer scanners. So you can see clots in vessels or even out to the pleura.
As I said earlier, this kind of throws a monkey-wrench into what used to be our sequence of studies. No one has done a study that tells you what the role of lung scanning is when you are also going to do CT-angio. I have essentially stopped ordering conventional pulmonary angiography in the last few months. But I haven’t stopped ordering lung scans because my feeling is that for CT-angio, just as for conventional angiography, the lung scan is helpful if for no other reason than it localizes the attention of the CT-radiologist to an area on the scan that they are going to pay more attention to as they interpret the study. So the lung scan is still helpful, giving us some direction as to where we are interested in going. As I said, I think it may ultimately prove to be better than conventional pulmonary angiography and it’s clearly going to be more available than pulmonary angiography.
Presenting signs - pioped (no preexisting cardio-pulmonary disease)
• Tachypnea 70%
• Crackles 51%
• Tachycardia 30%
• S4 24%
• Accentuated P2 23%
• Circulatory collapse 8%
Symptoms and signs did not distinguish patients with and without PE - pioped
Remember:
• Fewer than 30% of patients with PE have SX of DVT (pioped)
• 50% of patients with DVT without SX of PE have defects on lung scan
Likelihood of Pulmonary Embolism According to Scan Category and Clinical Probability in PIOPED Study | |||
Scan Category | Clinical Probability of Pulmonary Emboli | ||
High | Intermediate | Low | |
High | 95 | 86 | 56 |
Intermediate | 66 | 28 | 15 |
Low | 40 | 15 | 4 |
High | |||
Near normal through normal | 0 | 6 | 2 |
Although clinical judgement is helpful, the majority of patients are assigned an "intermediate" probability
Clinical judgement is fallible
• 1/3 of patients thought to have classic SX and assigned high pretest probability did not have PE diagnosed
A normal lung scan virtually excludes the diagnosis of PE
• 515 patients with negative scan had 0.6% rate of PE in 6 months
• A high probability scan is specific (particularly with high pretest probability - 95%).
• High probability scans are not sensitive, however (42% sensitive)
A minority of patients with clinically suspected DVT actually have the disease (17-32%)
Impedance plethysmography and doppler ultrasonography are sensitive (90+%) and specific (95%) if patient is symptomatic
• No fatal PE, 2% DVT in 311 pts followed 12 months -Canada
• No fatal emboli, 0.3% DVT in 6 months - Netherlands
Lower extremity studies may be helpful in patients with non-diagnostic scans (without C-P Disease)
• Nonfatal PE in only 3% of patients untreated 6 mos
Routine laboratory studies
• CBC
• Peripheral smear
• Liver & renal function tests
• PSA- MEN >50 yrs
• Urinalysis
Routine laboratory studies:
• Increased HCT or PLT - Myeloproliferative disorder
• Decreased HCT, WBC, PLT - PNH, marrow replacement
With tumor
• Increased PTT - Antiphospholipid syndrome
Screening for thrombophilia
• 24-37% of pts with DVT (10% of controls)
• Large spanish study - -2132 pts WITH DVT - 7% protein S EF, 3% protein C DEF, 4% antiphospholipid AB
• Factor V leiden 12-21% of PTS with DVT, 5% of controls
• Prothrombin gene mutation - 7% PF DVT PTS, 2% controls
Screening for thrombophilia - conservative approach
• Early age (<40 years)
• Recurrent disease
• Familial setting