Clinical Reliability of point-of-care tests to support community based acute ambulatory care

OBJECTIVE: To ensure clinicians can rely on point-of-care testing results, we assessed agreement between point-of-care tests for creatinine, urea, sodium, potassium, calcium, Hb, INR, CRP and subsequent corresponding laboratory tests. PARTICIPANTS: Community-dwelling adults referred to a community-based acute ambulatory care unit. INTERVENTIONS: The Abbott i-STATTM (Hb, clinical chemistry, INR) and the AfinionTM Analyser (CRP) and corresponding laboratory analyses. OUTCOMES: Agreement (Bland-Altman) and bias (Passing-Bablok regression). RESULTS: Among 462 adults we found an absolute mean difference between point-of-care and central laboratory analyses of 6.4g/L (95%LOA -7.9 to +20.6) for haemoglobin, -0.5mmol/L (95%LOA -4.5 to +3.5) for sodium, 0.2mmol/L (95%LOA -0.6 to +0.9) for potassium, 0.0mmol/L (95%LOA -0.3 to +0.3) for calcium, 9.0 μmol/L (95%LOA -18.5 to +36.4) for creatinine, 0.0mmol/L (95%LOA -2.7 to +2.6) for urea, -0.2 (95%LOA -2.4 to +2.0) for INR, -5.0 mg/L (95%LOA -24.4 to +14.4) for CRP. CONCLUSIONS: There was acceptable agreement and bias for these analytes, except for haemoglobin and creatinine.