A systematic evaluation of immunoassay point-of-care testing to define impact on patients' outcomes.

Background Point-of-care testing has been developed to provide rapid test results. Most published studies focus on analytical performance, neglecting its impact on patient outcomes. Objective To review the analytical performance and accuracy of point-of-care testing specifically planned for immunoassay and to evaluate the impact of faster results on patient management. Methods A search of electronic databases for studies reporting immunoassay results obtained in both point-of-care testing and central laboratory scenarios was performed. Data were extracted concerning the study details, and the methodological quality was assessed. The analytical characteristics and diagnostic accuracy of six points-of-care testing: troponin, procalcitonin, parathyroid hormone, brain natriuretic peptide, C-reactive protein and neutrophil gelatinase-associated lipocalin were evaluated. Results A total of 116 scientific papers were analysed. Studies measuring procalcitonin, parathyroid hormone and neutrophil gelatinase-associated lipocalin reported a limited impact on diagnostic decisions. Seven studies measuring C-reactive protein claimed a significant reduction of antibiotic prescription. Several authors evaluated brain natriuretic peptide or troponin reporting faster decision-making without any improvement in clinical outcome. Forty-four per cent of studies reported analytical data, showing satisfactory correlations between results obtained through point-of-care testing and central laboratory setting. Half of studies defined the diagnostic accuracy of point-of-care testing as acceptable for troponin (median sensitivity and specificity: 74% and 94%, respectively), brain natriuretic peptide (median sensitivity and specificity: 82% and 88%, respectively) and C-reactive protein (median sensitivity and specificity 85%). Conclusions Point-of-care testing immunoassay results seem to be reliable and accurate for troponin, brain natriuretic peptide and C-reactive protein. The satisfactory analytical performance, together with an excellent practicability, suggests that it could be a consistent tool in clinical practice, but data are lacking regarding the patient outcomes.