A Hemolyzed, Underfilled Sample from a Vomiting Cat
A 12-year-old, 4.1 kg, spayed female domestic shorthair is presented for a 3-week history of weight loss, vomiting, and polyuria. The veterinarian orders a CBC, biochemistry panel, and total T4. You draw about 1.5 mL of blood from the right jugular vein with a 3 mL syringe and a 22-gauge needle, then transfer it into a 0.5 mL EDTA microtainer by pushing the plunger with the needle still attached; the microtainer fills to roughly one-third. You also fill a lithium heparin tube to about half its capacity. The samples sit on the treatment-room counter for 90 minutes before the in-house analyzer is free. Results: hemolysis index 3+ (moderate-severe); K 7.6 mmol/L (reference 3.5-5.8); glucose 168 mg/dL; PCV 41%; TS 8.2 g/dL; platelets 90,000/µL with several clumps noted on the smear; total calcium 7.2 mg/dL (reference 8.0-10.5). The cat is clinically stable, urinating, and an ECG shows a normal sinus rhythm at 180 bpm with no peaked T waves.
1. List the preanalytical errors in this sample and state, for each, the direction in which it biases the reported results.
Read the worked answer
| Error | Effect on results |
|---|---|
| EDTA microtainer filled to about one-third | Excess EDTA (a potassium salt and a calcium chelator) → falsely high K and falsely low total calcium; the dilution also makes the PCV/TS from that tube unreliable |
| Blood pushed through the 22-g needle into the tube | Shear hemolysis → falsely high K, phosphate, AST, LDH; hemoglobin interferes with colorimetric assays and can distort bilirubin and glucose results depending on the analyzer |
| 90-minute delay at room temperature | Glycolysis by red cells → falsely low glucose (roughly 5-10% per hour in a cat); continued K leak from cells → falsely higher K; progressive platelet clumping |
| Heparin tube only half full | Excess heparin → falsely low ionized calcium if that is measured, plus dilutional lowering of analytes |
| Platelet clumping | Analyzer undercounts → falsely low platelet count; 90,000/µL is likely an underestimate |
Note that these biases do not all point the same way, which is why a single 'abnormal' result cannot be interpreted without knowing the sample quality. The hemolysis index of 3+ is the analyzer telling you the potassium in particular cannot be trusted.
2. Which of the reported results would you treat as unreliable, and exactly how would you obtain a valid repeat sample from this cat?
Read the worked answer
Unreliable as reported: potassium 7.6 mmol/L, total calcium 7.2 mg/dL, glucose (probably falsely low), and the platelet count. The strongest evidence is clinical: a cat with a true K of 7.6 mmol/L would be expected to be bradycardic and weak with peaked T waves, and this cat has a normal sinus rhythm at 180 bpm and is stable.
Repeat protocol:
- Calculate the safe draw: blood volume ≈ 65 mL/kg × 4.1 kg ≈ 265 mL, so 10% is about 26 mL; 1-2 mL is more than enough and is safe even in a sick cat. Draw only what the panel requires.
- Use a 23-gauge needle and a short draw with minimal suction; avoid prolonged jugular occlusion. Remove the needle before filling tubes.
- Fill the EDTA tube to its marked line, then invert gently 6-8 times; never shake.
- Submit the chemistry in lithium heparin (plasma), not serum - serum potassium runs higher because clotting releases K from cells and platelets.
- Process within 30 minutes, or centrifuge and separate plasma immediately and refrigerate it.
- Send a fresh smear made at the time of collection so platelets can be estimated (roughly 1 platelet per oil-immersion field ≈ 15,000-20,000/µL).
- Add a note to the record that the first sample was hemolyzed and underfilled so the artifact is not carried forward in the chart.
3. Assume the repeat potassium is 4.6 mmol/L but the total calcium is again 7.3 mg/dL with an albumin of 2.1 g/dL. How do you decide whether the cat is truly hypocalcemic, and how do you resolve the persistent thrombocytopenia?
Read the worked answer
Calcium. Total calcium is a package deal with albumin: roughly 40% of calcium is protein-bound, so a low albumin pulls total calcium down without any change in the biologically active fraction. Correct for it, e.g. corrected Ca (mg/dL) ≈ measured Ca − albumin + 3.5 = 7.3 − 2.1 + 3.5 = 8.7 mg/dL, which is within reference limits. The definitive test is ionized calcium on a minimally heparinized sample analyzed promptly, because heparin itself binds calcium. If ionized calcium is normal, this is pseudohypocalcemia from hypoalbuminemia, not a reason to give calcium.
Platelets. Clumps mean the analyzer's count is a lower bound, not a measurement. Estimate from the feathered edge of a fresh smear (average of 10 oil-immersion fields × 15,000-20,000). If the estimate is adequate and the cat has no bleeding, no petechiae, and no platelet consumption disorder, the '90,000/µL' can be reported as a spurious result of clumping. If the smear genuinely shows few platelets, then pursue true thrombocytopenia (FeLV/FIV status, tick-borne disease, DIC panel) rather than repeating the same clotted collection again.