Abstract
Point-of-care testing (POCT) is now a multi-billion pound international industry which features large in the responsibilities of most clinical biochemistry laboratories. POCT arose from the vision of laboratory scientists, clinicians and the diagnostics industry that the rapid availability of test results might expedite clinical decision-making and improve patient outcomes. Advances in diagnostic testing tend to evolve through a series of overlapping stages. Firstly, there is the development of the required new technology accompanied by excitement and optimism as potential clinical applications are identified. The second stage is marked by rigorous analytical and clinical evaluation to define the benefits and limitations of the new test as compared with existing tests. The third stage is characterized by increasing adoption of the test (often before an evidence base for its impact on patient outcomes has been established).
It is fascinating to look back over the 50 years of the Annals of Clinical Biochemistry and discover the evolution of POCT documented in its pages. The measurement of capillary blood glucose at the bedside was one of the first POCT tests to attain widespread use from the early 1980s onwards. A 1981 Annals paper described how more comprehensive metabolic information could be obtained by patients undertaking self-sampling at home and applying capillary blood to boric acid impregnated filter paper. The strips were sent to the laboratory, the blood eluted and glucose, lactate and 3-hydroxybutyrate measured using continuous flow enzymatic assays. 1 In the small clinical study undertaken in ambulatory subjects with diabetes, the authors noted abnormalities in the concentrations of hydroxybutyrate and lactate and pondered the significance of these. They proposed that this technique would prove a useful research tool to understand better the metabolic disturbance in diabetes. Thirty years on, capillary hydroxybutyrate measurement increasingly forms a routine part of the management of diabetic ketoacidosis and also for patient self-monitoring.
In the late 1980s and early 1990s there followed a series of papers describing analytical assessment of POCT equipment (generally glucose meters). These followed a similar format of comparing analytical performance of the POCT instrument to the laboratory assay and a conclusion as to whether performance was adequate or not. This conclusion was generally based on the authors’ expert opinion as to what constituted an acceptable standard of analytical performance rather than an objective consideration of required performance criteria. The latter issue was addressed in a 1986 review article by Fraser which defined the analytical goals for glucose measurement (both for laboratory assays and POCT analysers) based on biological variation. 2 The goals suggested were a zero bias and a between-batch coefficient of variation (%CV) ≤2.2%. 2 A quarter of a century later the American Association for Clinical Biochemistry/National Academy of Clinical Biochemistry proposed an intermediate goal of limiting total allowable error for 95% of samples to <15% at glucose concentrations of ≥5.6 mmol/L and to <0.8 mmol/L at glucose concentrations <5.6 mmol/L. 3 Both papers, separated by a divide of 25 years, lamented that the proposed targets may not be achievable at present!
As POCT became more established and undertaken in a broader range of clinical settings (including patient self-testing), subsequent contributions to the Annals recognized that analytical requirements varied for different applications. A particular concern has been the use of POCT for glucose measurement in the neonatal intensive care unit (NICU) and the need for reliable detection of hypoglycaemia. This topic featured in two consecutive papers in the September issue of the Annals in 2002 which investigated the performance of different POCT glucose instruments in NICU (in both papers the analysers under investigation were deemed to be satisfactory). 4 , 5 It is interesting to observe that one of these papers used the term ‘near-patient testing’ and the other ‘point-of-care testing’ in their titles. (These terms were used interchangeably for a number of years before the latter emerged as the preferred term for reasons which are unclear to the present author). Both adult and paediatric intensive care unit settings pose analytical challenges for measurement of glucose by POCT as patients frequently exhibit extremes of haematocrit and partial pressure of oxygen alongside acid–base disturbance, all of which may compromise accuracy. These problems have by no means been resolved – in 2012 the Annals carried a paper and accompanying editorial on POCT glucose measurement in critically ill patients.6,7
Contributors to the Annals recognized from an early stage that POCT performed by non-laboratory staff might be more prone to error than central laboratory testing and that this might be a limiting factor in delivering a safe and effective POCT service. 8 In a 2002 ‘Comment’ in the Annals, Freedman, made a call to arms to laboratories to provide leadership in POCT and put in place a robust risk management strategy, supported by a rigorous POCT policy and national guidelines. 9 Ten years on, I believe that this call has been largely heeded in the UK.
The purpose of all laboratory testing is to improve clinical outcomes for individual patients. However, it is often difficult to demonstrate this, particularly so in the case of POCT, given the complexity of clinical care and the fact that clinical outcomes may be far downstream from the testing process. Nevertheless, evidence of the clinical efficacy of POCT as compared with central laboratory testing, in terms of robust clinical outcomes is important in supporting its introduction and demonstrating cost-effectiveness. Despite the widespread use of POCT there have been surprisingly few studies investigating its contribution to better patient outcomes. A notable exception was a paper in the Annals in 2003 presenting the results of a randomized controlled trial showing that the use of POCT for nicotine and its breakdown products in pregnancy was associated with an increase in smoking cessation and an increase in birth weight and body length of the baby. 10 This is the kind of robust evidence that is needed to help define the proper role of POCT in patient care pathways and more such studies are required for all POCT modalities.
The archives of the Annals of Clinical Biochemistry offer an engaging and ongoing narrative on the evolution of POCT from a research concept to widespread clinical adoption. It is the story of enthusiastic development of new technology followed by a period of caution and retrenchment as awareness of problems and limitations emerge, the realization that the vagaries of human behaviour might compromise safe POCT which demand the need for robust governance. Finally there is the search for evidence that it is effective and contributes to better clinical outcomes for patients (often long after it has been brought into routine use). The latter point applies equally to non-POCT tests and it is perhaps time that the evidence base for the clinical benefit of all new tests was established before widespread adoption. I hope this is a theme that will be developed as the Annals moves into its next half century.