Limitations of point-of-care testing.

Abstract

363 POINT-OF-CARE TESTING (POCT) has become an increasingly popular means of delivering laboratory testing. With the advantage of enhanced turn-around time for obtaining test results, POCT has the potential for faster therapeutic intervention and improved patient outcomes. The portability of POCT devices allows access to laboratory testing in remote healthcare sites, like rural physician’s offices and mobile transport vehicles, helicopters, even cruise ships. Patients at risk for blood loss and transfusions further benefit from the small volume of blood required for POCT, particularly neonatal patients. Recent surveys estimate that POCT currently represents about 25% ($4.9 billion U.S.) of the world-wide diagnostic testing market, and this market share is expected to double in the next decade.1 Yet, for all its convenience, inaccurate POCT results may actually increase overall healthcare costs by triggering further diagnostic testing and procedures, or may fail to detect a significant condition. Complaints about self-monitoring blood glucose meters represent the largest number of reports files with the U.S. Food and Drug Administration for any medical device. Over 3000 reports including 16 deaths have been received.2,3 Without proper training and ongoing supervision, operators of cholesterol analyzers in outpatient clinics have generated results leading to wrong clinical interpretations.4 The quality of POCT is thus directly related to various operator (training, competency), site (light, temperature, storage conditions) and patient-specific factors (hematocrit, drugs), as well as the quality of management oversight of the testing process.5 In one U.S. location, unsupervised bedside glucose testing conducted by nonlaboratory staff differed by over 34% from the clinical laboratory, leading to inappropriate categorization of hypoglycemic and normoglycemic patients. However, after implementation of a quality assurance program involving standardized operator training, quality control and laboratory supervision, the same staff were able to reduce their coefficients of variation from 17 to 8% and improve POCT results to clinically acceptable levels.6 For POCT to be clinically effective, test results must be freely interchangeable with more traditional, centralized laboratory results. Many current clinical management protocols have been based on core laboratory results. Adaptation of these protocols to POCT requires comparable performance within acceptable tolerance for the given clinical situation. This implies an understand of the unique site and patient specific variables that can affect POCT results when testing is conducted with dozens of meters and operators or when devices are used on patients with a variety of acute and chronic diseases. The study by Tang et al.7 in this issue illustrates the need to validate POCT devices on the intended patient population prior to clinical implementation. Previous studies have demonstrated oxygen biases on early generation glucose oxidase based meters.8–10 This has clinical