Monte Minute‎ > ‎Monte Minute‎ > ‎

Pulmonary Embolism - Hypercoagulability Workup/Overdiagnosis and Autoimmune Hepatitis - 8/22/2012

posted Aug 16, 2012, 5:48 PM by Rohit Das   [ updated Dec 27, 2012, 7:10 AM by Purnema Madahar ]

So I did my first intern report today – craziness. We had a very nice discussion about a frequently encountered inpatient issue – pulmonary embolism. We went over Well’s simplified clinical prediction rule, the related pre-test probabilities, and how those probabilities are impacted by the different diagnostic options we have. A lot to talk about in one hour…hopefully I covered it well. Attached is a good review article from the Rational Clinical Exam series, as well as the original study from which the beloved Well’s criteria was created.

One of our smart interns brought up a very interesting question – what is the role for screening for hypercoagulability in patients with unprovoked VTE events? Here’s the data:

  • In a post-hoc analysis of the ELATE trial (attached), patients who had recurrent VTE were looked at to see if hypercoagulability posed any increased risk towards getting their recurrent episode.
  • In summary, the hazard ratios for all the mutations you can possibly think of were 0 to 0.7 (i.e., not an increased risk). Antiphospholipid trended towards an increased risk (HR – 2.9), but the confidence interval was very wide and crossed 1…
  • In addition, as compared to patients without any defects, patients with more than one thrombophilic defect were also NOT at increased risk of recurrent DVT (HR – 0.7)

What does this all mean? Hypercoagulability does not significantly alter your risk for recurrent DVT, and therefore its presence does not alter our management, so probably not worth ordering...

I’d like to divert a bit, though, towards a very important, and I feel underappreciated, phenomenon in medicine – overdiagnosis. This concept has been mainly part of the cancer screening literature, particularly prostate cancer, but is also important for some inpatient issues, including pulmonary embolism.

We all love technology. In medicine, it gives us diagnostic options that have increased accuracy, and therefore better predictive performance. But, it is important to realize that this sometimes comes with a price… Figure 1 in the attached Annals article depicts it very nicely. Super awesome tests are leading to increased incidence of disease, but NOT a matching significant change in the mortality related to that disease. This is the essence of overdiagnosis – we are diagnosing more frequently, but not significantly changing clinical outcomes, which is ultimately what we should all care about. Imagine being a patient, and being told by a doc that you have a disease/illness (“labeling”) that, in the long run, won’t have any impact on your life whatsoever. Wouldn’t that suck!? Additionally, we should consider the harms of treating such cases (especially anticoagulation, in regard to PE…). Well that’s why overdiagnosis is important, and why we should all try to understand it. Some key points from the annals article, specifically related to PE:

  • In a 5 year period after the advent of CT angio (1998-2006), the incidence of PE rose by 81%. In a 5 year period before the advent of CT angio (1993-1998), incidence did not change significantly.
  • Mortality has decreased by only 3% after the advent of CT angio, whereas the mortality decrease was MORE pronounced before its introduction – 8%
  • Case fatality rate (i.e., deaths/people diagnosed) has decreased by 36% in the post-period, as compared to an 8% decrease in the pre-period. Since there hasn’t been much improvement in mortality, this can only be explained by the denominator going up – i.e., diagnosing more clinically insignificant cases.
  • Finally, treatment related complications have gone up significantly since the advent of CT angio (3.1 to 5.3 per 100,000 patients), whereas they did not change significantly during the pre-CT period.

Trials looking at observation vs. anticoagulation in stable patients with small PEs are necessary to really tease this out further. Nevertheless, always keep the above data in mind when you’re faced with a patient with a PE and asking yourself…REALLY???!!!...

At resident report, acute liver injury in the setting of autoimmune hepatitis was the topic at hand. Cruff presented a case of an elderly woman presenting with 2 weeks of painless jaundice, with a history of previous similar episodes, who was found to have marked abnormalities in her LFTs. She had a positive anti-SMA, and a biopsy showing plasma cell infiltration and evidence of chronic liver disease. Few words on this interesting illness:

  • Autoimmune hepatitis is a relatively rare illness, effects women more than men, distributed fairly equally in all age groups/ethnicities, and has a wide variety of presentations – ranging from asymptomatic to fulminant hepatic failure.
  • There are two classification schema for autoimmune hepatitis. Type 1 autoimmune hepatitis is characterized by positive ANA and/or anti-smooth muscle antibodies, whereas Type 2 is characterized by antibodies to liver/kidney microsomes (ALKM-1).
    • Type 1 is distributed worldwide across all age groups and has a broad range of presentations.
    • Type 2 is a bit more selective – it involves almost exclusively young females (95%), and generally has a more consistently severe presentation.
  • Diagnosis is made in the setting of a consistent biopsy (plasma cell infiltrates), with associated circulating antibodies and elevated serum globulin levels.
  • Treatment (with steroids and azathioprine) is clearly indicated in patients with marked LFT abnormalities, advanced histologic features, or cirrhosis. It is also effective, with nearly 60-80% achieving histologic remission by 18 months.
    • 50% of patients with severe disease, if untreated, will die within 5 years. Patients with asymptomatic disease are a different story – their natural history is not completely clear, and the decision to treat is on a case-by-case basis.

Cool stuff…hope you guys learned something over the course of the week. Now it’s time to turn my brain off for a few days…Until Monday!!




Derivation of a Simple Clinical Model to Categorize Patients Probability of Pulmonary Embolism
Wells et. al., Thromb Haemost 2000; Volume 83: 416-420

Chunilal et. al., JAMA 2003; Volume 209 (21): 2849-58

Kearon et. al., BLOOD 2008, Volume 112 (12): 4432-4436

Wiener et. al., Arch Intern Med 2011; Volume 171 (9): 831-837

Krawitt, NEJM 2006; Volume 354: 54-66