Welcome to the AACC Expert Access Live Online program.

Our topic this month is Evidence Based Medicine and Point-of-Care Testing. POCT is a rapidly evolving tool in laboratory medicine. Today, tests performed with "near-patient"devices include glucose, bilirubin, blood gases,electrolytes, cardiac markers, drugs of abuse, and a myriad of others. Today's expert is James Nichols, PhD, Medical Director,Clinical Chemistry for the Baystate Health System in Springfield, Mass. Dr. Nichols'presentation describes the tenets of evidence-based medicine and shows how you can perform a critical review of the medical literature to make sure your POCT program is contributing to high quality health care in your institution. He is currently online and is awaiting your questions, so please submit them now.

AACC would like to extend its most sincere thanks to the Bayer Corporation for making this program possible.

What is the future of handheld devices and POCT
washington, dc

James H. Nichols, Ph.D., DABCC, FACB: POCT has a very bright future. There is increasing pressure in healthcare for faster turnaround of test results. Inpatients are more acute and sicker than they were 5 years ago, and there is increased utilization of same-day surgery and outpatient/home management of patients. These are all areas where POCT can offer potential for improving care. Recent market surveys of POCT from Emery Stephans group (Enterprise Analysis) indicate that POCT currently represents about 25% of the in vitro diagnostic market and is increasing (depending on the source you use, between 5 and 12% annually). This is already a significant portion of our laboratory responsibility and no matter which statistic one utilizes for growth, all concur that POCT will grow in the future, especially as devices get smaller, have wider menus and there is introduction of new technologies onto the market.

How will you establish that all of the relevant evidence has been reviewed. If this isn't established then cherry picking of evidence is one way that focus groups could bias any conclusions.
Halifax, Nova Scotia

James H. Nichols, Ph.D., DABCC, FACB: The NACB project which is currently underway, proposes to utilize a systematic review process similar to the recommendations that are developed by other professional societies and has been used by the NACB in the past for development of other practice guidelines (diabetes, cardiac, thyroid, etc.) This review process has been defined by the U.S. Preventive Services Task Force (Harris RP et. al. Current Methods of the U.S. Preventive Services Task Force: A Review of the Process, Am J Prev Med 2001;20(3S):21-35.) The systematic review involves defining a clinical question and terms for searching the medical literature databases (primarily Medline OVID). Additional articles will be supplemented based on the collections of the experts who are sitting on each of the focus groups. Articles that are collected will be screened by a minimum of two individuals for relevance to the clinical question and either included or excluded. Included articles will also be reviewed by a minimum of two individuals and graded based on the weight of the evidence (study format and conclusions, concurrence with other studies and benefits/risk of performing the test). Recommendations will be drafted based on the wieght of the evidence and draft recommendations posted on the website for public comment. Any comments will be resolved by the original focus groups and both the comments and resolution incorporated as an appendix to the document. Thus, this NACB project will incorporate and an intial systematic review, followed by a public comment period (similar to the NCCLS consensus process). If any articles have been left out or there is debate about the conclusions, we would expect to resolve those issues during the public comment phase through evidence raised by the public. We are also soliciting for volunteers, thus anyone who has interest and knowledge in a specific area and has the motivation to participate is welcome and we certainly are in need of additional volunteers particularly in the areas of occult blood, infectious disease and pH.

How does the NACB envision the role of the point-of-care diagnostics manufacturer in EBM?
Ottawa, Ontario, Canada

James H. Nichols, Ph.D., DABCC, FACB: The organizing committee of the NACB LMPG 2004 project has divided the POCT field into several more specific focus groups. These groups will have a balance of members representing clinicians, laboratorians, POCT coordinators and manufacturers. The groups are somewhat limited in size by their ability to be able to have a small enough group to get the tasks accomplished, but there are a group of individuals who will be readers to each group and assist the group in development of the appropriate clinical questions, specific terms to be searched, and proofing of the draft recommendations. Essentially helping the groups by commenting as the recommendations develop, before they are available on the web for general public comment. Then, after the recommendations are available on the web, there will be further opportunity for comment Thus, manufacturers can be involved at a variety of levels, from focus group member, actually conducting the systematic reviewes, to reader (expert commenter on focus group development), to commenting after drafts are developed.

Do any currrently available glucometers meet the ADA recommendations for imprecision and accuracy?
North London, UK

James H. Nichols, Ph.D., DABCC, FACB: This is somewhat of a loaded question, I can tell, so I'm not necessarily going to promote one manufacturer's device over another, but I will refer the attendee to a few publications that will assist in answering this question: First, the ADA guidelines. These differ and have changed over the years, the initial guidelines (Consensus Development Panel. Consensus statement on self-monitoring of blood glucose. Diabetes Care 1987;10:95-99) recommend accuracy to within +/- 15% of a core laboratory method and precision <+/-5%, while the later guidelines (American Diabetes Association. Self-monitoring of blood glucose. Diabetes Care 1994;17:81-86) recommend more stringent goals of +/- 5% agreement to a core laboratory method, 100% of the time. The ADA conceeds in this document that no meters were currently able to meet those recommendations. A more recent study conducted by the FDA and quoted in this presentation (Chen ET, et al. A study of the factors that affect the performance of blood glucose monitoring devices. Abstract Clin Chem 2001;47(S6):A189.) demonstrates that some devices still can't meet these recommendations above 100 mg/dL and none of them are acceptable in the hypoglycemic range. These conclusion were further validated by the recommendations in the NACB guideline (Clin Chem 2002;48:436-472) from earlier this year stating that "There are no published data to support a role for portable meters in the diagnosis of diabetes or for population screening. The imprecision of the meters, coupled with the substantial differences among meters, precludes their use in the diagnosis of diabetes and limits their usefulness in screening for diabetes. Level of Evidence: E (peer concensus).

It seems inevitable that there will eventually be a molecular-based diagnostic tool for point-of-care applications. Will NACB efforts address these emerging technologies, and do you see any special considerations in genetic POCT or molecular-based infectious disease testing?
San DIego, CA

James H. Nichols, Ph.D., DABCC, FACB: Currently there is no proposal for addressing molecular diagnostics and POCT, as the literature is just emerging on this topic. The NACB LMPG program hopes to define the extent of our current knowledge of POCT related to patient outcomes and where evidence does not exist, make recommendations on studies that are needed to fill in the gaps in the literature. Molecular diagnostics is too new on the horizon to adequately address at this time

What about the new areas where basically no evidence is available (e.g., noninvasive POCT testing for glucose)? Can we derive conclusions from studies done using current invasive technology?
São Paulo/Brazil

James H. Nichols, Ph.D., DABCC, FACB: I believe that part of the problem the POCT field faces is the extrapolation of conclusions based on one lab method to an alternative method. Thus POCT outcomes are assumed to be at least the same or better than core laboratory outcomes solely because the result is available more rapidly. We know this isn't necesarily true, as each method has it's own inherent sensitivity, specificity and limitations when applied to a specific patient population. Extrapolating patient outcomes from noninvasive glucose testing would present the same problems. We can't assume patients will have the same or better outcomes by looking at the invasive methods, we need to redo the studies in a randomized controlled fashion specifically comparing noninvasive patient outcomes to invasive technologies.

Where is a good place to look for evidence-based reviews of point-of-care technology? Are there certain POC tests that lend themselves more to EBM than others?
Edgewater, MD

James H. Nichols, Ph.D., DABCC, FACB: There is no good place to look for EBM reviews of POCT technologies, this is the reason for starting the NACB. There are sections of other guidelines that are currently available that address POCT in some of the recommendations. The NACB diabetes practice guideline (Clin Chem 2002;48:436-472) specifically addresses the use of glucose meters for diagnosing or screening patients for diabetes, but the guidelines are primarily directed towards core laboratory testing. Additionally, the cardiac guidelines indicate that cardiac markers should be available within 30 - 60 minutes of physician order and presentation of a chest pain patient, yet these guidelines do not specifically address how those markers should be conducted (ie POCT or via pneumatic tube to a core lab). This is the problem and reason for the NACB project.

I’d like to combine a couple of previous questions. Do you think IVD manufacturers have any incentive to follow EBM-based recommendations? I think the evidence is pretty strong that handheld glucose meters can’t match high-performance, high-throughput analyzers when it comes to accuracy and precision. Nevertheless, health care professionals (as well as patients) have been using the handheld devices for years and devising systems to work around this limitation, yet manufacturers have done little to improve performance. Why would this scenario be any different for any other POCT device?
St. Louis, MO

James H. Nichols, Ph.D., DABCC, FACB: This is an excellent question, because it gets at the heart of Evidence Based Medicine. There is no direct means of forcing a clinician to change their practice or to adopt a specific practice if they choose to do something different. This is part of the clinical interpretation and judgement required in patient care. Most physicians are data driven, but the amount of information that is being generated each day can be overwhelming for a clinician to keep current. Guidelines are peer reviewed and consensus driven recommendations for "best" use of a specific device and help summarize for clinicians the current state of the art in knowledge on the application of a test. Additionally, when something goes wrong with a patient, or there is a threat of lawsuit, guidelines and peer practice tend to weigh heavily with juries. Thus, to answer this question, there is no means of forcing any one clinician to adope a specific recommendation, the force comes from peer-pressure and definition of "standard" practice.

What is required to confirm the value of a test by EBM?
Waltham, MA

James H. Nichols, Ph.D., DABCC, FACB: Confirming the value of a test requires review of the research and clinical literature, as well as weighing the benefits/risk of a specific procedure or test. In this presentation, I summarized the current methods defined by the U.S. Preventive Services Task Force (Am J Prev Med 2001;20(3S):21-35) that the NACB has adopted for evaluating the evidence of individual studies. The group looks at the weight of the study (randomized controlled trials get more weight that case control or peer consensus), the concurrence of one study with other studies in terms of conclusions, and the degree to which the study linkg the test directly to health outcomes (rather than indirectly with other confounding factors). This weight of the evidence is then balances with net benefit/risk to the patient to make final recommendations supporting a specific application of a test or not. I would refer the attendee to the original manuscript for more details on the process.

Will an EBM approach be used to evaluate whether current consensus recommendations are also valid. For example, the NACB has recommended that assays for cardiac markers should have an imprecision (CV) <10% at the AMI decision limits. If the LMPG is prepared to set recommendations for POC testing based on whether systems meet this criteria, it must assure that the criteria being measured are evidence-based as well.
San Francisco, CA

James H. Nichols, Ph.D., DABCC, FACB: Absolutely, I agree totally and it is intended that the focus groups not take any current guidelines at face value, but to update and revalidate if those guidelines are still applicable or not. Your example is a case in point.

What are the first test(s) or disease condition for which you think best practice guidelines should be developed from a laboratorian's viewpoint and why this choice??
Chattanooga, TN

James H. Nichols, Ph.D., DABCC, FACB: The first test or disease condition that should be addressed is the one that is most applicable in any institution. Thus, if occult blood utilization nis causing an institution problems, that would be the test of highest priority for that institution to address. The NACB intends not to address a single specific test or disease condition, but has divided the POCT field into several specific areas, hoping to address comprehensively the existing evidence for best application of POCT.

What are the references of the over utilization study (it leads to problems)
St. George, UT

James H. Nichols, Ph.D., DABCC, FACB: In the presentation, I utilized the term overutilization in a more general sense, rather than referring to data from a specific study. However, consider the problem of POCT glucose on hospitalized inpatients. If a physician orders q6hr glucoses on a patient and does not DC the glucose in the morning and evening chemistry panels, there is the chance that one of these are not going to agree with the core laboratory. We all have experienced this in day-to-day practice because of the imprecision and differences in results between POCT methods and lab methods. The clinician is then faced with the dilemma, they treated based on the POCT result, but the core lab result comes back different. Now what are they to do? Recollect and hope for the best 2 out of 3? In essence, the mismatch triggers additional testing and diagnostic management which increases healthcare cost and adds risk from further phlebotomy, when it would have been better to select the optimal test in the first place through realization that switching between methods increases the risk of obtaining results that do not clinically agree.

Do you foresee better assays for POC testing that will increase the precision and accuracy of the POC testing compared to laboratory testing for such tests as prothrombin times?
Roseburg, Oregon

James H. Nichols, Ph.D., DABCC, FACB: Yes, testing does improve over time. Just look back at the history of lab testing or POCT as an example. Clearly, our coefficients of variation and accuracy have improved over time as reagent manufacturing and instrumentation has improved. As new technology is developed, there would not be an advantage to adopting the new technology unless there were significant improvements in accuracy, precision, cost, or other feature of the device.

Are there currently lists of diagnostic tests for the ER, OR, and ICU that have been confirmed by the principles of EBM?
Waltham, MA

James H. Nichols, Ph.D., DABCC, FACB: No, I know of no lists. The problem with POCT literature is that the test is found to be comparable to a core lab method in one patient population and there is pressure to readily adopt the device in other populations without reexamining if the test is really achieving the desired outcomes. This is why the NACB project could be so beneficial to clinicians, laboratorians, POC coordinators and the industry in general, because it will define the tests and areas where there is proven benefit to patient care, as well as those where patient benefit is not so clear and requires more research.

In a very broad sense, what should I hope to gain from evidence-based evaluation of the components of my POCT program? Should I expect to adjust or change the testing procedures? Our testing process? The hardware? Personnel?
Lisbon, Portugal

James H. Nichols, Ph.D., DABCC, FACB: A document that outlines the existing evidence for POCT and patient outcomes would be useful when clinicians request new testing as well as to the continued reexamination of existing testing. Consider a physician requesting Test X that we may have some knowledge of in the core lab but now wants to implement the test on inpatients and widely throughout the outpatient clinic environment. The first thing I help our POCT team collect is the literature that indicates others experience with the test and allows us to predict what we can expect after implementation. I tend to look critically at the studies to determine if I will obtain similar outcomes, but its often difficult. Wouldn't it be great if there were a document that represented best practice of POCT that you could also refer to, indicating whether there is or isn't evidence for the new application. This would help committees both direct implementation of new devices, but also raise questions whether current testing needs to be reevaluated in specific applications

What level of interest is the quality of sample processing, aka "preanalytics", relative to POCT? Would phlebotomized collection of a 3 or 4 ml blood sample into a separator tube that could provide virtual quantitative separation into harvestable fractions of plasma, nucleated cells, and RBCs within 60 seconds after collection be a desirable product goal?
Pittsburgh, PA

James H. Nichols, Ph.D., DABCC, FACB: Preanalytics is everything to POC and could have the same impact that analytical or postanalytical problems could have on the quality of a test result. In POCT, specifically, preanalytical sample collection, could have an even greater impact on POCT results than preanalytical variation on the core lab results because of the use of capillary blood and the potential for poor collection technique with fingersticks. From the standpoint of current POCT, a device that separates plasma in the collection tube without operator interaction would probably be of little value, since all current devices are based on whole unprocessed blood or body fluid. For future technologies, requiring plasma, such a device may have better applications, but you will have to remember that capillary or noninvasive collections (urine, saliva, etc.) tend to be preferred samples for POCT over venous phlebotomy that cannot be performed on by patients on themselves.

Coming back to the non-invasive technology for glucose testing, I completely agree with your answer. Not only the methods are different from the analytical point-of-view (precision, accuracy, spec, sens., etc.), but the way they will be used may also lead to different patient outcome (sporadic vs. almost-continuous monitoring). However, current studies are only focusing in the analytical part of the story, which may delay the progress in terms of other clinical-oriented outcomes. For sure new studies are needed, addressing the classical and also these new questions. Will the NACB recommendations address these new challenges?
São Paulo, Brazil

James H. Nichols, Ph.D., DABCC, FACB: The analytic focus of the majority of the POCT literature is one of the limitations in defining patient outcomes. The NACB project intends to define the current extent of the literature and/or lack of evidence and indicate potential areas that need to be investigated. Since continuous monitoring is such a hot topic, I'm sure that it will be one of the clinical questions addressed by the diabetes group.