Dr. Rabih Geha’s excellent post on Closler – Every Piece of Data Matters – has stimulated much thought. He makes one important point that I would like to expand.

My mind finds it much harder to attach diagnostic significance to pertinent negatives. Even if their impact on a diagnostic hypothesis is equal, I find that a positive test result sways my reasoning more so than an equally impactful negative test.

My students and residents know that I want to hear the numbers. Sometimes normal is not normal. Sometimes seemingly normal gives clues.

The problem comes from our laboratory definition of normal.The lab uses large data sets to estimate a “normal” range. Sometimes the clinical condition says otherwise.

Several years ago I had a patient with an unknown “pneumonia” who was not improving. On examining his labs I noted that his creatinine has risen from 0.8 to 1.2. Both numbers were technically normal, but a 50% increase in creatinine should grab out attention. When we finally obtained a urinalysis, he had evidence of acute glomerulonephritis. The “pneumonia” was really GPA.

What is a normal platelet count? What is a normal WBC? How do we interpret a serum calcium if we do not know the serum albumin? What does it mean when the BUN is 2? What is the BUN/creatinine ratio – and does that help us understand the patient’s story?

The patient has a history of severe vomiting, but has a normal bicarbonate level. What is the anion gap? Does the computer flag an abnormal anion gap?

When students present, I expect to hear the numbers. When reading MKSAP questions or reading (or listening to) Human Diagnosis Project cases, too often labs are reported as normal. Yet the numbers may still provide some value.

So here is my call for presenting the numbers and let me decide if they provide information. Perhaps I can teach the team something from these labs. But let me decide if the numbers are actually normal. Please!