RESUMO
BACKGROUND: Thrombocytosis is one of several pre-analytical factors which contribute to spuriously high serum potassium concentrations, yet there is little published data to guide analysts in the selection of a specific platelet count threshold above which serum potassium results become unreliable. We have studied the sensitivity and specificity of blood platelet count as a predictor of false elevations in potassium. METHODS: Paired serum and plasma potassium measurements together with full blood count were performed for 300 patients. All samples were stored at room temperature and analysed within 4 h of collection. The difference between serum and plasma potassium was plotted against blood platelet count. RESULTS: When the difference (serum-plasma) in potassium concentration was plotted against platelet count, there was a direct linear relationship. Blood platelet counts of >500 x 10(9)/L will detect elevations in serum relative to plasma potassium of >0.5 mmol/L, with a sensitivity of 71% and a specificity of 89%. CONCLUSION: It is recommended that where blood platelet count is above 500 x 10(9)/L, potassium measurements should be repeated using lithium heparin plasma. When serum potassium results of >5.4 mmol/L are obtained, it is our policy to check the platelet count if a sample is available before reporting results. If available and above 500 x 10(9)/L, potassium results are withheld and plasma requested.