Senior Report 5.15

Case Presentation by Dr. Shereaf Walid



A 33-year-old obese African American female presents with right-sided pleuritic chest pain that started yesterday afternoon suddenly while she was walking at work.  She works a clerical job where she sits down at a desk for the majority of the day; however, she says she does not stay seated for long intervals, as her work requires her to do a lot of walking around the office.  The pleuritic chest pain is maximum at end inspiration.  She feels the pain on the right side of her chest around the anterior axillary line of the chest wall.  The pain radiates to her neck, below her axilla, and in the lateral right rib/chest wall area.  The pain seems worst on her right side below the axilla in the mid-axillary line near the costal margin.  The pain is reproducible by taking a deep breath in, and since it started, she has been taking shallow breaths in to reduce the pain.  The discomfort is also pruritic, but she notes no rash.  She appears to be splinting upon deep inspiration.  She denies any leg swelling or pain and no recent surgery.  No fevers or dyspnea on exertion.  She does not take any oral contraceptive pills.  She has had no trauma to the chest wall.  This pain episode is not associated with any nausea, vomiting or diaphoresis.  She has never experienced pain of this character in the past.  No personal history of hypercoagulable disorders, or family history of blood clots.


CONSTITUTIONAL:  No recent weight loss.

EYES:  No double vision, itching, pain or discharge.

ENMT:  No ear infection, epistaxis or current dental problems.

CARDIOVASCULAR:  No palpitations.

RESPIRATORY:  Right-sided chest pain that is pleuritic since yesterday.

GASTROINTESTINAL:  No recurrent pain, diarrhea or bloody stools.

GENITOURINARY:  No burning or frequency on urination.

MUSCULOSKELETAL:  No muscle pain or weakness.

NEURO:  No loss of consciousness, speech or balance problems.

SKIN:  No rashes.



PAST MEDICAL HISTORY:  Negative for hypertension, diabetes or hypercholesterolemia.



ALLERGIES:  Shellfish.


FAMILY HISTORY:  Hypertension, diabetes.

SOCIAL HISTORY:  Positive for tobacco use and occasional alcohol use, negative for any illicit or IV drug use.


VITAL SIGNS:  On presentation blood pressure 154/79, heart rate is 92, respiratory rate is 20, oral temperature is 36.9, SaO2 100% on room air.

CONSTITUTIONAL:  Obese female in no apparent distress but anxious.  She is taking shallow quick breaths.  Her respiratory rate seems to be only mildly elevated, but for the most part she is sitting comfortably and she speaks in complete sentences.  No tripoding.

PSYCHIATRIC:  A and O x3.  Judgment is sound.

HEAD:  Normocephalic, atraumatic.  No tenderness to palpation.

EYES:  Pupils are equal, round and reactive to light and accommodation.  Extraocular movements are grossly intact.  No conjunctival pallor.  Sclerae anicteric.

ENMT:  Mucous membranes are moist.

NECK:  Supple.  No JVD.  No carotid bruits.  No lymphadenopathy.

CARDIOVASCULAR:  S1, S2.  Regular rate and rhythm.  No murmurs, rubs or gallops.

RESPIRATORY:  Clear to auscultation bilaterally.  No wheezes, rhonchi or rales.  Decreased air entry due to shallow respirations.  No tenderness anywhere over the chest wall.  No palpable crepitus over the chest wall or the neck.

GASTROINTESTINAL:  Abdomen is soft, nontender, nondistended.

MUSCULOSKELETAL:  Strength is 5/5 proximally and distally in both the upper and lower extremities.  No palpable cords in the lower extremities.  Gastrocnemius diameter is equal bilaterally.  No swelling.  No dependent edema.  2+ dorsalis pedis pulses bilaterally.

SKIN:  No acute rashes or lesions.

NEURO:  Normal speech and normal gait.

An ECG and Chest X-Ray were obtained and are shown below:

Question 1:

What is the most specific chest X-ray finding seen in patients with diagnosed pulmonary embolism?

a.   cardiomegaly

b.   elevated hemidiaphragm

c.   hampton’s hump

d.   plate-like atelectasis

e.   small pleural effusion

Question 2:

What is the patient’s Well’s score for pulmonary embolism?

a.  0

b.  1

c.  2

d.  3

e.  4

Question 3:

Which of the following interventions is the most reasonable next step for this patient?

a.  cat scan of the thorax with intravenous contrast

b.  draw blood cultures and start the patient on intravenous antibiotics for community acquired pneumonia

c.  obtain a d-dimer assay

d.  start the patient on PO antibiotics, an incentive spirometer and arrange close outpatient follow-up

e.  ventilation-perfusion (V/Q) scan due to the patient’s shellfish allergy, which puts her at risk for an allergic reaction to iodinated radiocontrast dye




1.  c

2.  d

3.  c


This case is of an actual patient seen in the ED.  The patient has a pulmonary embolism.  This patient presents with no traditional risk factors for PE, although several studies quote obesity as an indirect risk factor due to relative immobility and increased risk for lower extremity DVT.  She has significant pleuritic chest pain and splinting, yet her oxygen saturation is 100%.  An ECG and CXR are indicated in the ED as initial screening tests in evaluating a patient with suspected pulmonary embolism due to being relatively inexpensive and accessible, and to assess for other potential causes.  The American College of Radiology recommends chest radiography as the most appropriate study for ruling out other causes of chest pain in patients with suspected pulmonary embolism.

The ECG provided in this case shows normal sinus rhythm.  At triage the patients HR was 92.  The most common ECG finding in patients with confirmed pulmonary embolism is sinus tachycardia.  It is important to note that this patient’s heart rate is usually in the 70s, this data was not provided in the case, but was ascertained by reviewing the electronic medical record.  Hence, her relative tachycardia is an important sign to recognize and consider in formulating a differential diagnosis. The finding of S1 Q3 T3 is nonspecific and insensitive in the absence of clinical suspicion for pulmonary embolism. Classic findings of right heart strain and acute cor pulmonale include tall, peaked P waves in lead II (P pulmonale); right axis deviation; right bundle-branch block; presence of an S1 Q3 T3 pattern; or atrial fibrillation. Unfortunately, only 20% of patients with proven pulmonary embolism have any of these classic electrocardiographic abnormalities. If electrocardiographic abnormalities are present, they may be suggestive of pulmonary embolism, but the absence of such abnormalities has no significant predictive value.  Some studies have shown that T wave inversions in anterior precordial and inferior leads is the most specific ECG finding in patients with confirmed pulmonary embolism.

The Chest X-ray shown in this case was extremely valuable in this patient’s evaluation and is often under-appreciated and under-utilized.  It was initially read as negative by radiology, however, if you examine closely there are findings that are suggestive of pulmonary embolism in the correct clinical context.  Although CXR cannot establish the diagnosis of PE, it has several useful roles in patients suspected of having a PE.  When examining radiographs, the following findings should be sought.

First, the classic finding is a relatively normal chest radiograph in a patient in “dire straights”.  However, this “classic” teaching is not true, as observations from the PIOPED and PIOPED II studies and several others have shown that a normal CXR is found in only 12-18% of patients with confirmed pulmonary embolism.  So keep in mind that an apparent “normal” CXR is suggestive of PE in that it can lower the suspicion of other disorders such as pneumothorax, pulmonary edema or pneumonia.  Next, there are radiographic abnormalities seen in some cases of PE, although these are non-specific, such as small pleural effusion, plate-like atelectasis, or elevation of a hemidiaphragm.  Occasionally, there are radiographic findings that are characteristic of PE, although they may be subtle and difficult to identify with certainty.  Occlusion of a large pulmonary artery can produce localized oligemia (which looks like diminished lung markings in the region supplied by that vessel) and the occluded pulmonary artery may be dilated proximally due to a large intraluminal thrombus and then taper abruptly (the knuckle sign).  Seen together, localized oligemia and a dilated pulmonary artery with abrupt cut-off are termed the Westermark sign.  Sometimes, you’ll see dilation of the contralateral pulmonary artery because the affected pulmonary artery contains an embolism that is so massive that nearly all the blood flow is directed to the contralateral pulmonary artery.  This is called a “reversed Westermark sign”.

You should note that the chest x-ray in this case is a poor inspiratory film.  Never overlook this as studies that you may be basing your findings on will exclude inadequate films in their data analysis, hence the utility of searching for specific or sensitive findings of pathology may fail before you have even started looking.  With that in mind, if you look closely at the right costophrenic angle on the PA view of the CXR, you’ll notice that the angle is blunted when compared to the left side.  There is clearly a hazy opacity there.  Additionally, there appears to be an elevated right hemidiaphragm, a non-specific finding that is sometimes found in cases of PE, but is neither sensitive nor specific.  It should be noted that the right hemidiaphragm is normally elevated relative to the left due to the liver and this can make the distinction between normal and abnormal difficult.  In this case, the elevated right hemidiaphragm is easier to appreciate on the lateral view, as with most cases that show this finding.  Now after assessing the lateral CXR, you can follow the left and right hemidiaphragmatic lines back to the posterior rib margins near their insertion sites.  You can also clearly see the costophrenic sulcus made by these structures posteriorly.  The “haziness” seen on the PA view is probably anteriorly located, as suggested by the lateral view, or you could argue that its location is obscured (again, this is equivocal possibly due to the inadequate film).  One thing you can definitely conclude is that the haziness seen on the PA view is not a pleural effusion, or else it would have blunted the most dependent area of the lung, the posterior costophrenic sulcus.  This should raise the suspicion of a “Hampton’s Hump.”

In some cases of PE, there is an area of focal airspace filling.  This represents focal intraparenchymal hemorrhage due to ischemia or infarction of lung tissue that occurs distal to a large embolus.  The area of hemorrhage is typically located at the lung periphery and appears as a wedge-shaped pleural-based opacity with its apex pointing toward the lung hilum.  This has been termed “reversible infarction” because it clears rapidly over several days, reducing in size progressively without residual scarring.  This focal area is called a Hampton’s Hump and should be suspected in this patient.  Also note that the elevated right hemidiaphragm is much easier to appreciate on the lateral CXR.  Finally, non-specific findings radiographic findings in PE are:  “normal”, small pleural effusion, plate-like atelectasis and elevated hemidiaphragm.  Specific abnormalities are the Westermark sign and Hampton’s hump, hence the answer to question 1 is “c”.


Now after obtaining the ECG and CXR the most reasonable next step in the patient work-up should be to evaluate the patient’s pre-test probability for a pulmonary embolism. Evidence-based literature supports the practice of determining the clinical probability of pulmonary embolism before proceeding with testing. One study assessed the performance of 4 clinical decision rules in addition to d-dimer testing to exclude acute PE. All 4 rules – Wells rule, simplified Wells rule, revised Geneva score, and simplified revised Geneva score, showed similar performance for excluding acute PE when combined with a normal d-dimer result.  Several studies have validated the use of the Wells criteria for pulmonary embolism in the ED setting.  The Wells scoring system is as follows:

Hence, based on this scoring system what would your score be?  There are no clinical signs of DVT.  On exam, the patient’s gastrocnemius diameter is equal without palpable cords, something that should be clearly documented in all cases of suspected pulmonary embolism.  The heart rate is less than 100 bpm.  There is no history of immobilization.  Please note that to qualify for “history of immobilization” according to Wells, and Jeff Kline’s PERC score, one must have uninterrupted immobility in a knees-flexed position for 6 consecutive hours.  So if the patient gets up to stretch their legs half way through a trans-Atlantic flight there is no history of immobilization as per these scoring systems.  However, these scoring systems are not, under any circumstance, designed to replace clinical “gestault”, but to supplement it.  In other words, never abandon your clinical suspicion in lieu of a clinical scoring system.  The patient has no objectively diagnosed PE or DVT.  There is no history of hemoptysis or active malignancy.  Also, note that prior malignancy is not a risk factor (another point of confusion).  The Wells score includes malignancy with treatment within 6 months because only active malignancy is a risk factor, not remission.  Finally, is PE the #1 diagnosis or equally likely?  If you answer no for this then your Wells score is zero.  If you answer yes then it is 3.  My Wells score on this patient was 3.  If you are not suspecting PE in this patient, then what is the more likely diagnosis?  Hence, the answer to question 2 is “d”.  More on this later if you answered zero.

Upon calculating a Wells score of 3, we now have to decide what to do with this data.  In a seminal study by Jeff Kline he used a test threshold formula to derive the test threshold for PE to be 1.8% (this is essentially the derivation of the PERC rule).  What this means is that when the pre-test probability for PE is less than 1.8% then pursuing testing is more likely to do more harm than good (ie. the false positives and potential harm inflicted by treatment outweighs treatment of true positives).  During grand rounds I will discuss the concept of test threshold and treatment threshold more thoroughly.  For now, recognize that for every test there is a test threshold, below which, testing does more harm than good.  And, on the other end of the spectrum, there is a treatment threshold, above which testing does more harm than good (ie. your pre-test probability is so high that a negative d-dimer has poor negative predictive value and you should go straight to CT/ treatment).  The patients that fall between the test threshold and the treatment threshold are the ones where d-dimer should be used.

There are 3 studies that I will comment on at this point, two of which are cited on MD-calc (the website from which the above figure was obtained).  The first is “Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the ED by using a simple clinical model and d-dimer.”  Wells PS et al.  Ann Intern Med. 2001.  In this study, low risk patients, defined as Wells score of 0-2 had a 1.3% risk of pulmonary embolism, hence, they were below the testing threshold which means that testing is more likely to do harm than good.  The second study is “Prospective validation of Wells Criteria in the evaluation of patients with suspected pulmonary embolism.”  Wolf SJ et al.  Ann Emerg Med. 2004.  In this study, low risk patients, also defined as Wells score of 0-2 had a 1.7% risk of pulmonary embolism, also below the testing threshold indicating that testing is more likely to do harm than good.  Finally, in one of the most impressive studies on PE decision rules, “Effectiveness of managing suspected PE using an algorithm combining clinical probability, d-dimer testing, and CT.” Christopher Study Investigators. JAMA, 2006, in a large ED patient population this impact study tested the validity of using d-dimer on patients with Wells scores of 3 or 4 and found it to be safe.  Remarkably, 6 months later, there were less adverse outcomes in using d-dimer in Wells ≤ 4 than CT scan.  Additionally, 15% of the patients studied were inpatient, a population known to be at higher risk than those seen in the ED, and the data, in the context of PE/DVT studies was compelling.  This study used a simplified decision tree: if Wells ≤ 4 use d-dimer; if > 4 use CT.  For more details, read the article.

The mistake often made with d-dimer is that it’s used on the wrong population (ie. those below the testing threshold = Wells 0-2).  If you’re doing this, you are practicing dangerous medicine.  What the Christopher study showed was that our collective “gestault” is greatly skewed and we teach the application of d-dimer grossly incorrectly.  If you look at the table for Wells score for PE, if a patient presents with flu-like symptoms but they have a history of DVT, hemoptysis and are currently being treated for cancer, then their Wells score is 4 and you should order a d-dimer.  Let me repeat this validated point: a patient with history of DVT, hemoptysis and active cancer should still get a D-dimer.  And here is the take home point:  this is what d-dimer was designed for.  There are large, validated studies that show that doing less is safer than doing more.  Additionally, from a medico-legal standpoint, it is my belief and practice that the Christopher study should be mentioned in your dictation or macro for PE and is so easily defensible as to be a powerful enough deterrent to keep litigation at bay.  In other words, a prosecuting attorney will be hard pressed to take a case with this referenced in the patient’s chart.  Let’s never forget that we don’t practice medicine for attorneys, we practice for patients.  However, as we all know, we don’t practice in a vacuum.

Finally, the answer to question 3 is “c”.  Initially, my Wells score was zero (was I right?).  After reviewing the CXR, initially read by radiology as negative, I noted a Hampton’s hump (specific, relatively), not a pleural effusion, and an elevated right hemidiaphragm (non-specific).  Since the patient’s clinical presentation did not fit pneumonia, I concluded that this was more likely a Hampton’s hump than a pneumonic infiltrate.  I re-calculated her Wells score to be 3, ordered a d-dimer, electrolytes and gave her 1L of normal saline for renal protection in case the d-dimer came back positive.  Should I have just ordered a serum creatinine?  As an aside, regarding the allergic cross-reaction of shellfish and radio-iodinated contrast dye, the literature has shown that this is a fallacy.  To date, I have not found any literature to support this and there is a robust amount of literature to refute this correlation.  Please feel free to read the following review,  “The relationship of radiocontrast, iodine, and seafood allergies: a medical myth exposed.”  Schabelman E, Witting M. J Emerg Med. 2010 Nov; 39(5):701-7.

The d-dimer came back elevated at 1.33. CT scan – PE protocol was ordered and was reported by radiology as, “positive for pulmonary emboli in the superior segment of the right lower lobe pulmonary artery with associated pulmonary infarction and in a posterior basilar branch of the right lower lobe pulmonary artery.  There is also a triangle shaped airspace opacity in the anterior aspect of the right upper lobe without definite associated pulmonary embolus.  This finding could represent a focal area of atelectasis.”  The lung window of the CT scan is found below, clearly showing the associated pulmonary infarct, or Hampton’s hump.



“Therefore, at this point we sent a set of coagulation studies on the patient and started the patient on fondaparinux and called to have the patient admitted.  Results of the CT PE protocol were explained to the patient and the reason for admission has also been explained to her.  She is obviously agreeable to the necessity of this admission.”  In the near future, the validity of the last sentence of the quoted dictation will be challenged, but for now all confirmed pulmonary embolisms are to be admitted.  Warfarin was also appropriately started in the ED and the patient was discharged within 48 hours.  Outpatient follow-up was arranged and a repeat CXR showed resolution of the pulmonary infarct with anticoagulation rather than antibiotics further confirming the suspicion of a Hampton’s hump.  Six months out the patient had no recurrences and has made lifestyle modifications.  As of one month ago the patient continues to do well and, as per her routine, frequently flexes her gastrocnemius muscles while working at her desk.


2 Responses

  1. There, I did it this week.

  2. Shereaf- Outstanding discussion. One area of disagreement. You state that there are patients that are high pre-test probability and should not have a d-dimer, but should get automatic CT. This is true for the older d-dimer tests, but with the high-sensitivity d-dimer that we, and many EDs, are now using we can use d-dimer on high pre-test prob patients and if negative you are done with your work-up. You may not save a huge number of CTs in this population, but will definitely spare some unecessary radiation. This is supported by Jeff Kline, etc. AG

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: