There has been a fair amount of conversation recently on Twitter and on several blogs about how we think in the ED, how we make complex medical decisions, and how we decide what to order.  A recent interaction made me want to add my two cents, and my thoughts on the crucial questions you must ask yourself before you order a test.

A resident and I were seeing a patient who came with a complaint of pleuritic chest pain, shortness of breath, and occasional cough.  She had a history of PE, and 2 months ago had a CT scan showing resolution of her last PE.  She was tachycardic at 105/minute and mildly tachypneic.  The patient said it did not feel exactly like previous PE, but could not say for sure.  She had a non productive cough, but clear lungs, and said this might feel somewhat like previous pneumonia.  The differential was simple: pneumonia and PE topped the list, along with pneumothorax, pericarditis, atypical ACS, all the usual suspects.  The interesting part of the presentation was when the topic of how to evaluate for PE came up:  

"I'm going to order a D dimer and hope it's negative."  This prompted me to ask "What does a negative D dimer mean in her?"  To answer this, you have to know how to properly use and interpret a D dimer.  I have talked about PERC before here on the blog, and you can watch a great debate here  on how we should use PERC and Wells Criteria to help evaluate for PE/DVT.  But let me briefly rehash: Wells Criteria are designed to help you decide if a patient is low risk enough for DVT/PE that a negative D dimer successfully excludes that as the diagnosis.  If a patient is low risk with a negative D dimer, the risk of DVT/PE in the next 90 days is minimal.  If they are moderate or high risk, however, you need to do a study (US for DVT or CT for PE) to truly rule out that diagnosis.  

If you play with the link above to MDCalc, at a minimum with tachycardia and previous PE, this patient is moderate risk, not to mention PE is the number one or equally likely diagnosis.  So a D dimer, even if negative, is not appropriate in this patient.  This brings me to the steps you should go through before you order a test:

1. Am I using this test appropriately? Are you ordering a D dimer on a moderate or high risk patient?  Are you planning on ruling out ACS with a single troponin?  Will a normal WBC tell you your patient does not have appendicitis?  If the test is not appropriate, do not order it. 

2.  What are you going to do with a negative result? In the above case, does a negative D dimer mean anything?  If you are convinced your patient has CHF clinically and that ever-so-helpful BNP comes back at 150, what are you going to do?

3. What are you going to do with a positive result?  Don't get caught with your pants down.  Have a plan if that troponin is positive.  Know who you are going to call if that CT in the patient with atraumatic headache shows a subarachnoid hemorrhage.  This also comes up frequently with the ever popular "I'm going to get a CBC to look at the WBC to see if they have an infection."  What if it's 14.2? 18.9? 10.4?  (This example also harkens back to the "Am I using this test appropriately" point.  As Amal Mattu says, the WBC is the last refuge of the intellectually destitute!)