The accuracy of a point of care measurement of activated partial thromboplastin time in intensive care patients.

Point of care (POC) devices are increasingly being used in intensive care units to obtain faster results. Data are limited on the performance of these devices in critically ill patients, especially those on heparin infusion. The objective of this study was to assess the agreement between POC activated partial thromboplastin time (APTT) and laboratory APTT results in patients on heparin infusion and to determine its impact on the clinical decisions regarding heparin dosage. We screened all patients admitted to the intensive care unit (ICU) at St George Hospital, Sydney, over a 7-month period and enrolled those who were receiving intravenous heparin infusion. We measured APTT by two methods: bedside POC test (Hemochron Junior Signature Plus) and central laboratory method (STA analyser). We used the Bland-Altman method to test the statistical agreement between the two measurements and Cohen's kappa statistic to test the clinical agreement regarding heparin dosing decision. A total of 176 paired samples from 44 patients (mean age 63 years, 64% males, mean APACHE 18) were analysed. The mean turnaround time for the point of care APTT result was significantly shorter than the central laboratory result (5.0±0.2 min vs 64.6±2.7 min, p<0.0001). Despite the statistically significant correlation, the overall agreement tested by the Bland-Altman method was poor. The 95% limits of agreement were widest (-27.266 to 64.791) and mean percentage bias was highest (24%) for the comparison between POC APTT using citrated blood and laboratory APTT. When POC APTT results of less than 90 seconds using whole blood were compared to laboratory APTT results, the limits of the agreement became narrower (-23.243 to 28.419), and the mean percentage bias decreased to 5%. The agreement between clinical decisions regarding heparin dosage based on the two methods was poor for plain and citrated blood (kappa 0.35 and 0.11, respectively). The POC APTT results were not sufficiently accurate for use in patients on heparin infusion compared to laboratory APTT assay.