LOGIN 

 

JOIN

 

RENEW

 

CME/MOC

Clinical Cases

Critical Care Cases

American Thoracic Society - Critical Care Medicine

Elderly Man with Abdominal Pain

Burton Lee

Washington Hospital Center

Time 0
An 85-year-old man presents to the emergency room with nausea and severe abdominal pain He denies fever, shortness of breath, chest pain, bleeding, or prior surgical history. The past medical history is significant for atrial fibrillation, congestive heart failure and ischemic stroke. 

The physical exam in the emergency room is notable for temperature of 96.7ºF, heart rate of 73/min, respiratory rate of 24/min, and blood pressure of 117/50 mmHg, and 95% peripheral oxygen saturation on room air.  There were normal bowel sounds with some right lower quadrant tenderness, but no rebound tenderness.  Rectal exam with Guaiac was negative for occult blood.
 
Initial laboratory findings

  • WBC 15,400/mm3
  • Hematocrit 32%
  • Platelets 170/mm3
  • AST 32 units/L
  • ALT 16 units/L
  • Alkaline phosphatase 99 units/L
  • Total bilirubin 0.4 mg/dL
  • Albumin 4.1 g/dL
  • Lipase 137 units/L
  • Na 136 mEq/L
  • K 5.4 mEq/L
  • Cl 101 mEq/L
  • HCO3 20 mEq/L
  • Cr 3.0 mg/dL
  • Serum glucose 367 mg/dL
  • Urine analysis revealed no leukocytes.

Initial radiologic studies:

  • Chest film revealed clear lung fields without free air or cardiomegaly. 
  • Abdominal film revealed few dilated loops of small bowel. 
  • Bedside ultrasound examination reveals no evidence of cholecystitis, cholangitis, hydronephrosis, aortic aneurysm, abscess, gallstones, or ascites.

The patient was administered ciprofloxacin and metronidazole empirically.  A surgical consultation and an abdominal CT scan are requested. 

Time 2 hours
The surgical consultant notes that the diffuse abdominal tenderness to palpation, but does not find rebound tenderness. 

An abdominal CT scan is temporarily deferred because the patient is now noted to have hypotension with a blood pressure of 85/55 mmHg.  The blood pressure normalizes after administration of intravenous fluids and initiation of an intravenous infusion of dopamine. 

Repeat labs are notable for serum HCO3 18 mEq/L, Cr 3.6 mg/dL, and Lactate 4.3 mEq/L.  An arterial blood gas, drawn while the patient is breathing 50% oxygen by face mask, reveals the following:

  • pH=7.28
  • PaCO2=38 mmHg
  • PaO2=71 mmHg

As an experienced intensive care specialist, you are concerned about mesenteric ischemia among other diagnoses and eagerly await the results of the CT scan.  Based on your experience, you estimate the likelihood of mesenteric ischemia to be high, at least 60% if not higher.  While awaiting the CT scan to be completed, you review the literature and find the sensitivity and the specificity of CT scan for mesenteric ischemia to be about 70% and 90%, respectively. 

Abdominal CT scan with PO and IV contrast revealed, “rectal fecal impaction causing ileus and dilated loops of bowel, no appendicitis, abscess, gall stones, perforation, or ascites.”

Question True or False

  1. Had the CT scan been POSITIVE for mesenteric ischemia, the chance of it being FALSELY POSITIVE would have been about 10-20%. 
  2. Had the CT scan been POSITIVE for mesenteric ischemia, the chance of it being TRULY POSITIVE would have been >95%. 
  3. Since the CT scan is NEGATIVE for mesenteric ischemia, patient should be followed closely and no further testing is necessary since the likelihood of mesenteric ischemia is now <10%.    
  4. Further testing (i.e. angiography or surgical exploration) is necessary since the likelihood of mesenteric ischemia is still 30-40%.    

Answer to Question True or False

To answer these questions, one approach is to construct a 2x2 table and assume that 100 patients present with an identical scenario as the patient in this case. 

Disease Positive

Disease Negative

Test Positive

A

B

CT Positive = A+B

Test Negative

C

D

CT Negative = C+D

Ischemia = A+C

No Ischemia = B+D

TOTAL = A+B+C+D

Since the pretest probability is estimated to be 60%, 60 of the 100 patients are expected to have mesenteric ischemia and 40 patients will not.  Given that the test has a sensitivity of 70%, 42 patients (i.e. 60*0.7) will have a positive CT scan and mesenteric ischemia abdomen and 18 patients (i.e. 60 – 42) will have a negative CT scan but still have mesenteric ischemia.  Given a specificity of 90%, 32 patients (i.e. 40*0.9) will have a negative CT scan and no mesenteric ischemia and 8 patients (i.e. 40 – 32) will have a positive CT scan but no mesenteric ischemia. 

Disease Positive

Disease Negative

Test Positive

A = 42

B = 8

CT Positive = 50

Test Negative

C = 18

D = 32

CT Negative = 50

Ischemia = 60

No Ischemia = 40

TOTAL = 100

After completing the 2X2 table as above, we are now ready to answer the true/false questions. 

  1. Had the CT scan been POSITIVE for mesenteric ischemia, the chance of it being FALSELY POSITIVE would have been about 10-20%. 

True.  As shown in the 2x2 table above, 50 patients will have a positive CT scan, but 8 of the 50 will not have ischemic abdomen.  Thus the chance of false positivity is 8 / 50 or 16%. 

  1. Had the CT scan been POSITIVE for mesenteric ischemia, the chance of it being TRULY POSITIVE would have been >95%. 

False.  As shown in the 2x2 table above, 50 patients will have a positive CT scan, but only 42 of the 50 will truly have ischemic abdomen.  Thus the chance of true positivity is 42 / 50 or 84%. 

  1. Since the CT scan is NEGATIVE for mesenteric ischemia, patient should be followed closely and no further testing is necessary since the likelihood of mesenteric ischemia is now <10%.    

False.  As shown in the 2x2 table above, 50 patients will have a negative CT scan, but only 32 of the 50 will truly not have ischemic abdomen.  Thus the chance of true negativity is 32 / 50 or 64%. 

  1. Further testing (i.e. angiography or surgical exploration) is necessary since the likelihood of mesenteric ischemia is still 30-40%.    

True.  As shown in the 2x2 table above, 50 patients will have a negative CT scan, but 18 of the 50 will still have ischemic abdomen.  Thus the chance of false negativity is 18 / 50 or 36%. 

Further Discussion

This case illustrates one of the most important principles when interpreting any diagnostic test: no test rules in or rules out a diagnosis with 100% certainty.  Rather, a test simply modifies the pre-test probability of a potential diagnosis and offers a new post-test probability of that diagnosis.  In the case above, the negative CT scan changed the probability of mesenteric ischemia from 60% before the test (i.e. pretest probability of 60%) to 36% after the test (i.e. posttest probability of 36%).  Although the likelihood of mesenteric ischemia is lower after the CT scan, there is still sufficient concern to warrant further diagnostic testing such as angiography or surgical exploration. 

Unfortunately, this patient was observed in the ICU without further investigation or surgical intervention.  Patient continued to deteriorate further, requiring intubation and escalation of vasopressor support.  Surgical exploration was reconsidered but it was now thought that the patient was too ill to undergo surgery.  Patient expired approximately 36 hours after admission.  Autopsy

revealed, “gross and microscopic evidence of severe ischemic bowel, extending from the proximal jejunum to distal transverse colon.” 

Formulas


Pretest Probability = (a+c)/(a+b+c+d)
Sensitivity = a/(a+c)
Specificity = d/(b+d)
Positive Predictive Value (PPV) = a/(a+b)
Negative Predictive Value (NPV) = d/(c+d)
False Negative Rate = 1 - NPV
False Positive Rate = 1 - PPV

References

  1. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB.  Evidence-Based Medicine: How to Practice and Teach EBM, 2nd edition, Edinburgh: Churchill Livingtone, 2000. 
  2. Taourel PG, Deneuville M, Pradel JA, Regent D, Bruel JM.  Acute Mesenteric Ischemia: Diagnosis with Contrast-enhanced CT.  Radiology 1996; 199: 632-6.
  3. Park WM, Gloviczki P, Cherry KJ, et. al.  Contemporary Management of Acute Mesenteric Ischemia: Factors Associated with Survival.  Journal of Vascular Surgery 2002; 35: 445-52.