Abstract
Meeting Summary
The Diabetes Technology Society, Foster City, California, organized a 1-day public meeting on the topic of point-of-care (POC) blood glucose (BG) monitoring in the hospital. The meeting was designed to provide information on the use of POC BG monitors (BGMs) in critically ill patients (CIPs) in the hospital, which is considered an off-label use, and to allow health care professionals (HCPs) and the public to share their respective perspectives and concerns regarding the new FDA Draft Guidance.
Session 1: Introduction to POC BG Monitoring in the Hospital
Mills-Peninsula Health Services, San Mateo, California
David Klonoff, MD, FACP, FRCP(Edin), Fellow AIMBE, from Mills-Peninsula Health Services, San Mateo, California,provided a brief background on POC BG monitoring. Dr Klonoff shared that while POC BGMs have become standard of care for hospitalized patients, they have not been cleared by the Food and Drug Administration (FDA) for CIPs. With regulators promising stricter enforcement of regulations, the goal of the meeting is to find a solution where regulators, manufacturers, and HCPs alike would be confident that they are ensuring optimum delivery of care to CIPs using tools that meet regulatory guidelines.
Mary Korytkowski, MD, from the University of Pittsburgh, Pittsburgh, Pennsylvania, keynote presenter, shared that the availability of current POC BGMs for CIPs has allowed for goal-directed glycemic management, with reductions in length of stay, morbidity, and mortality in patients with diabetes and newly recognized hyperglycemia. Professional organizations such as American Diabetes Association, Society for Critical Care Medicine, American Heart Association, and American Association of Clinical Endocrinologists, all support careful glycemic management in CIPs. Dr Korytkowski added that protocols using POC BGMs and targeting glycemic goals of 110-140 mg/dl and 140-180 mg/dl have been demonstrated to be both safe and effective in patients with and without diabetes, which is why it is the preferred method for guiding ongoing management of individual patients. An unintended consequence of enforcing POC BGMs to not be used off label in CIPs would be a reduction in the amount of glucose monitoring tests performed, with an increase in the risk for hyper- and hypoglycemia and associated adverse patient outcomes.
Session 2: Clinician Perspectives on the Role of POC BG Monitoring in the Hospital
Walter Reed National Military Medical Center, Bethesda, Maryland
Anthony Furnary, MD, from Starr-Wood Cardiac Group of Portland PC, Portland, Oregon, explained that diabetes per se is not a risk factor for increased mortality, length of stay, deep sternal wound infections, and postoperative complication rates in cardiac patients. However, studies have shown that tight glycemic control in a diabetic cardiac surgery patient for 3 full postoperative dates is imperative to minimize morbidity and mortality: it is the gold standard of care and can be safely implemented with POC BGMs. Dr Furnary stressed that greater weight has to be placed on saving lives, preventing infections and complications, and allowing patients to return home while saving on health care costs rather than on POC BGM accuracy alone.
Stanley Nasraway, MD, FCCM, from Tufts University, Boston, Massachusetts, shared that their hospital uses a combination of POC BGMs along with the OptiScanner 5000, which has allowed them to keep 93% of their patients within a target glycemic range of 80-150 mg/dl. Dr Nasraway believes that if POC BGM use is not allowed in CIPs, this will prevent insulin infusion therapy and would require a broadening of the glucose target range. More importantly, Dr Nasraway is concerned that this action will take patient care back to the 20th century and the increased turnaround times for test results would delay prompt delivery of care for CIPs.
Arleta Rewers, MD, from Children’s Hospital Colorado, Aurora, Colorado, stated that POC BGM testing is essential for correct diagnosis and management of a critically child, adding that POC BGM has helped in reducing the frequency of iatrogenic severe hypoglycemia among insulin-treated patients. Dr Rewers shared that POC BGM is also helpful in communicating with her hospital’s patients’ family members. Dr Rewers expressed her concern that discontinuation of POC BGM use in the pediatric emergency department would endanger patients, especially in situations where a quick diagnosis and prompt treatment are needed.
Garry Tobin, MD, from Washington University, St. Louis, Missouri, shared that his hospital provides diabetes care/monitoring to almost 50% of its patients and the use of POC testing in combination with informatics on dosing significantly improves safety for ward patients. Dr Tobin explained that changing the timing and method of glucose measurements could result in errors. POC BGMs facilitate POC testing at appropriate intervals, which is critical for achieving their goal of providing safe and effective diabetes care that meets national standards for glycemic control.
Jane Seley, DNP, MPH, MSN, BC-ADM, CDE, CDTC, from New York Presbyterian/Weill Cornell Medical Center, New York, New York, shared that nurses need devices that are quick and user-friendly and provide comfort, quality, and safety. Ms Seley added that insulin dosing is based on a BG range and not on a single BG value. Therefore, a POC BGM is a powerful glycemic control tool to monitor BG and dose insulin in the hospital setting, both in and out of the critical care unit. Moreover, a POC BGM can be used for diabetes self-care management education prior to hospital discharge.
Session 3: Clinical Chemist Perspectives on Performance of Currently Available Hospital POC BGMs
US Food and Drug Administration, Silver Spring, Maryland
Victoria Kark, RN, from Walter Reed National Military Medical Center, Bethesda, Maryland, stated that, in the hospital setting, hundreds of devices are used by thousands of operators to perform an unlimited number of tests. Among these devices, the BGM is one of the most reliable devices used in the hospital to monitor patients and provide instantaneous information to the HCP. In addition, the POC BGM is relatively inexpensive and requires fewer personnel than other test methods to provide reliable results for patient care.
William Clarke, PhD, from Johns Hopkins University, Baltimore, Maryland, listed and explained the 4 components for verifying BGM performance: precision, accuracy, linearity, and clinical decision support. Dr Clarke cited various accuracy studies performed on POC BGMs that show that most differences between reference and BGM values are located in the A zone of both the Clarke and Parkes error grids, which represent clinically accurate readings that pose no clinical consequences.
Deborah Perry, MD, from Children’s Hospital and Medical Center, Omaha, Nebraska, spoke about the history of CAP (College of American Pathologists) and its strong support for quality glucose testing. Dr Perry specified that her institution verifies BGM performance using (1) linearity studies performed on 100% of meters to determine the reportable range of the analyte, (2) accuracy studies performed on 10% of meters (20 samples per meter), and (3) precision studies performed on 10% of meters using 10 replicates of levels 1 and 3 glucose control solutions. In April 2014, CAP wrote a letter to the FDA regarding its new Draft Guidance. In the letter, CAP supported the goal of the FDA in general and expressed concern regarding the vague definition of “critically ill” and how the implementation of the draft document could result in the immediate removal of BGMs from acute care situations. Dr Perry shared that the CAP is working with the FDA to ensure sufficient regulation of laboratory-developed tests (LDTs) and to prevent burdensome regulation that would discourage development of LDTs.
Session 4: Performance Targets for Using POC BGMs
Mills-Peninsula Health Services, San Mateo, California
Dr William Clarke outlined the accuracy standards listed in the new FDA Draft Guidance. Among the standards are (1) 350 samples (minimum) of each sample type in each intended use population; and (2) relative to the reference method (ie, validated central laboratory), 99% of correlation results must be within 10% of reference and 100% within 20% of reference (all outliers must be explained), and the 350 samples must be “native” (up to 50 high BG values and 50 low values can be manipulated to cover the measurable range). Dr Clarke expressed concern that the standard might be so stringent that it cannot be met with current technology and that this degree of stringency may not discernibly increase patient safety. Dr Clarke also cited the possible downstream implications created by the discussions of off-label use by the state of New York Department of Health and the Centers for Medicare & Medicaid Services (CMS).
David Sacks, MB ChB, from the National Institutes of Health, Bethesda, Maryland, provided a general background on accuracy, precision, and bias, and listed previous standards/guidelines for glucose measurement that were released by the Clinical and Laboratory Standards Institute (CLSI), ADA, and Clinical Laboratory Improvement Amendments (CLIA). Dr Sacks mentioned 3 significant modifications to the CLSI standards: (1) tightened accuracy criteria from ±20% to ±12.5%; (2) reduction to 2%, the percentage of results excluded from accuracy criteria; and (3) accuracy studies with samples from areas where device will be used, for example, intensive care unit. Dr Sacks observed that the POCT 12-A3 is substantially better than prior guidelines.
Robert Rushakoff, MD, from the University of California, San Francisco, San Francisco, California, shared that given all the variables of a patient’s condition and treatment, it is impossible to predict the insulin requirements for any specific patient. Dr Rushakoff also stressed that, after millions of glucose tests performed in his institution, a technical issue has yet to occur where the use of the meter caused patient harm. Therefore, Dr Rushakoff believes that POC BGM use should not be an issue. Dr Rushakoff explained that there are no different targets for accuracy for different populations, with the exception of insulin-sensitive type 1 diabetes patients who are trying to stay in “tight” control. However, Dr Rushakoff added that the glucose levels of these patients would most likely not be controlled as tightly in a hospital setting.
Session 5: Regulation of POC BG Monitors in the Hospital
Mills-Peninsula Health Services, San Mateo, California
Karen Dyer, MT(ASCP) DLM, from CMS, Baltimore, Maryland, provided the CMS perspective on the new FDA BGM limitation. Ms Dyer stated that the CMS was made aware of the new FDA required limitation in BGM package inserts that prohibit use of the meter for CIPs. Ms Dyer noted that this is not a new CLIA regulation and that it is incumbent on each laboratory/facility to define “critically ill” for its specific patient populations. Otherwise, the option would be to use POC test systems that do not have the “critically ill” limitation or to send glucose tests to the central laboratory. As part of standard survey practices, CLIA surveyors may visit laboratories and if POC BGM practice is identified during the survey, then (1) laboratories will be advised of other testing options, (2) citations will be written per the annual Certificate of Waiver package instructions and the State Operations Manual, and (3) laboratories will be given an opportunity to correct the deficiencies.
Courtney Lias, PhD, from the US Food and Drug Administration, Silver Spring, Maryland, provided the FDA perspective regarding regulation of POC BGMs in the hospital. Dr Lias stated that the issue of off-label use of POC BGMs has always existed, but there has been an increase in awareness soon after Roche Inform II was cleared and laboratories noticed the limitation language in its package insert. Dr Lias shared that when manufacturers seek over-the-counter clearance for BGMs commonly used in hospitals, they effectively assign responsibility for their use and validation to the laboratories. Dr Lias added that if the guidances were implemented, then manufacturers would have to validate use in CIPs so that laboratories would need only to verify performance. The lack of appropriate controls and the risk of patient harm will remain even if the FDA decides against issuing the final guidance documents. The intention of the FDA in writing the guidance was that all meters be CLIA-waived. The studies and performance criteria described in the draft guidance were designed to enable companies to meet the criteria for waiver at the same time that they met the criteria for clearance. Dr Lias acknowledged that all stakeholders (FDA, CMS, industry, HCPs) want a solution that enables BGMs to be used safely where they are needed and encouraged all parties to work together in finding applicable solutions.
Session 6: Comments from the Public on POC BG Monitoring in the Hospital
Walter Reed National Military Medical Center, Bethesda, Maryland
Maxwell Booker, from the United Medical Center, Washington, DC, explained repeat verification testing using the calculated 10% range of the critical glucose. Kelly Close, BA, from Close Concerns, San Francisco, California, provided a patient’s perspective on the critical need for acute care BG monitoring. Jeffrey DuBois, PhD, MBA, from Nova Biomedical, Waltham, Massachusetts, gave a brief presentation on method traceability and concordance. Bennet Dunlap from StripSafely, Bryn Athyn, Pennsylvania, also provided his perspective as a diabetes patient. Sharon Geaghan, MD, from Stanford University, Palo Alto, California, gave a detailed presentation on alternative technologies to POC testing for glucose monitoring and its impact on diagnostic testing sample volume and health outcomes. Roma Gianchandani, MD, from the University of Michigan, Ann Arbor, Michigan, provided details on a remote digital entry meter. Eli Ipp, MD, from the University of California, Los Angeles, Los Angeles, California, provided a perspective on POC BGM use in a public hospital system.
Session 7: Alternatives to POC BG Monitoring for Measuring Blood Glucose in the Hospital
US Food and Drug Administration, Silver Spring, Maryland
James Nichols, PhD, DABCC, FACB, a clinical chemist at Vanderbilt University, Nashville, Tennessee, shared his perspective on the use of blood gas analyzers (BGAs) and nonstrip testing as an alternative to POC BGMs. Dr Nichols explained that POC BGMs have the distinct advantage of being accurate and lower in cost relative to BGAs and nonstrip devices. A hospital that uses only BGAs and nonstrip devices for measuring BG will see its annual costs for tests increase tenfold from hundreds of thousands to millions of dollars. In addition, a high-complexity designation of POC BGMs by CLIA will, in effect, preclude all nurses and other health care providers, except physicians and PhD-level personnel, from administering BG tests and greatly increase the burden of hospitals to provide documentation. Dr Nichols indicated that reducing access to POC BGM use in hospitals will have a significant impact on patient care because it will create countless challenges not only in the flow and costs of hospital operations, but also in clinics and other settings where patient care is administered.
Mark J. Rice, MD, from the University of Florida, Gainesville, Florida, discussed the merits of the common technologies used for BG monitoring in the hospital. Dr Rice echoed the observations and concerns raised by Dr Nichols that central laboratories are an impossible alternative because they are very cost-prohibitive and have a long turnaround time (TAT). Likewise, BGAs have a moderately slow TAT and are relatively expensive. Possible alternatives such as nonstrip devices have faster TATs but still cost significantly more than POC BGMs per test, and the i-Stat requires a larger sample volume and a capillary tube. Dr Rice shared that the major issue for all HCPs is ease of use, which translates to significant time and cost savings. Dr Rice anticipated that elimination of POC BGMs in the hospital would lead to higher costs despite fewer tests being performed, and to more adverse events in CIPs.
Jeffrey Joseph, DO, from Thomas Jefferson University, Philadelphia, Pennsylvania, shared that measuring BG frequently is key to avoiding hypoglycemia, hyperglycemia, and glycemic variability. Dr Joseph presented data for arterial, venous, and capillary blood samples along with a comparison of common technologies used in the hospital to measure BG. Dr Joseph provided assessments by simulation models that showed BG measurements made every 5 minutes were associated with a lower frequency of true hypoglycemia and a reduction in the frequency of hypoglycemia.
Session 8: Professional Organization Perspectives
Walter Reed National Military Medical Center, Bethesda, Maryland
William Clarke, PhD, from the American Association for Clinical Chemistry, Washington, DC; Helena Duncan from the College of American Pathologists, Northfield, Illinois; Donna Ryan, RN, RD, MPH, CDE, from the American Association of Diabetes Educators, Chicago, Illinois; and Jason Wexler, MD, from the Endocrine Society, Washington, DC, all provided their comments on behalf of the respective professional organization they represented.
Overall, similar concerns were expressed on the potential enforcement of off-label guidelines, but all representatives remained hopeful that regulatory agencies would be able to work with all interested parties to find solutions that would help HCPs provide care for CIPs in a safe and effective manner.
Session 9: Reaching Consensus on Safe and Effective POC BG Testing in the Hospital
Mills-Peninsula Health Services, San Mateo, California
Dr Klonoff initiated a panel discussion to reach a consensus on safe and effective POC BG testing in CIPs. Members of the discussion panel were William Clarke, PhD; Karen Dyer, MT(ASCP) DLM; Anthony Furnary, MD; Jeffrey Joseph, DO; Victoria Kark, RN; Mary Korytkowski, MD; Courtney Lias, PhD; Stanley Nasraway, MD, FCCM; Deborah Perry, MD; Arleta Rewers, MD; Mark J. Rice, MD; Robert Rushakoff, MD; David Sacks, MB ChB; Jane Seley, DNP, MPH, MSN, BC-ADM, CDE, CDTC; and Robert Vigersky, MD.
The panelists discussed the benefits of using POC BGMs for CIPs. They discussed the scarcity of realistic alternate methods for rapid determination of BG in CIPs who require frequent adjustments in insulin dosage. The panelists felt that this is not a good time to suddenly restrict the use of POC BGMs in CIPs.
Dr Klonoff summarized the meeting by calling on CMS to extend a 5-year moratorium on enforcing sanctions against the use of POC BGMs in hospitalized CIPs. HCPs could use this time to see which patients from the ICU, operating room, recovery room, and ER should not be managed with POC BGMs. HCPs can also use the moratorium periodto develop new workflow systems in case the current generation of POC BG monitors cannot attain FDA clearance or CLIA-waived status for CIPs. FDA could use the time to work with Industry to develop a pathway for clearnce of existing products for CIPs. Industry could use the time to develop next-generation products that will meet new performance standards. The moritorium will providetime for the unintended consequences of removing current POC BGMs from hospitalized CIPs to be mitigated, and it will allow a smooth transition toward new and better methods for monitoring blood glucose in hospitalized CIPs.