ABSTRACT
PURPOSE: The aim of this study was to assess agreement of potassium test results between the point-of-care-testing (POCT) blood gas analyzer and central laboratory (CL) biochemistry analysis associated with cardiopulmonary resuscitation (CPR) cases. METHODS: This was a prospective, comparative study of 108- paired results conducted between January 2009 and December 2010. Blood specimens were collected within 5 minutes of arrival to the emergency department from the femoral artery of cardiac arrest patients, and specimens were used for both arterial blood gas and routine laboratory analysis. Arterial blood gas analysis samples were immediately analyzed using a POCT blood gas machine [GEM Premier 3000 analyzer (Instrumentation Laboratory, Lexington, MA)]. Potassium levels obtained from the POCT arterial blood gas analyzer and subsequent CL biochemistry analysis were then compared. RESULTS: Mean values of potassium measured were 5.17+/-1.72 mmol/l using POCT and 5.37+/-1.79 mmol/l using CL. The mean difference+/-SD between simultaneous POCT and CL measurements was 0.19+/-0.66 mmol/l, with 95% limits of agreement of -1.48 to 1.10 mmol/l. The sensitivity, specificity, positive predictive value, and negative predictive value of POCT in diagnosing life threatening hyperkalaemia (> 6.5 mmol/l) were 85%, 97%, 85%, and 97%, respectively. CONCLUSION: The mean difference between the results obtained from the POCT blood gas analyzer and CL analysis were small. POCT is a thus a useful method for rapidly detecting life-threatening hyperkalaemia during CPR. However, due to wide limits of agreements, caution in interpretation of POCT results is necessary.