By Katheryn Inglis, MS(MT) 

This article is a sidebar to CLP‘s June 2014 feature, “Glucose, Under Analysis.”

Katheryn Inglis, MS(MT)

Katheryn Inglis, MS(MT)

On January 13, 2014, the New York State Department of Health sent to its licensed New York laboratories a “Dear Laboratory Director” letter regarding off-label use of glucose meters. The department’s letter followed by just a few days the issuance of two FDA guidance documents in which the agency established new premarket submission requirements for blood glucose monitors.

FDA has long approved glucose meters for the measurement of whole blood glucose, and for use by healthcare professionals as an aid in monitoring the effectiveness of a diabetes control program. In the past, however, manufacturers’ validation data has not been sufficient for FDA to extend this approved intended use to other patient populations or settings. The intended use statements found in the package inserts or manuals of many such products do not include the monitoring of glycemic control for non-diabetic patients in hospitals, or use on critically ill patients.

As clarified in a frequently asked questions document issued on February 18, the purpose of the Department of Health’s letter was to “inform facilities on the policies and requirements surrounding off-label use of these devices.” Specifically, laboratories that use glucose meters for purposes or populations beyond the approved intended use are considered to be engaging in off-label use.

In such a situation, noted the department, the glucose meter defaults to a high-complexity categorization under the terms of the Clinical Laboratory Improvement Amendments of 1988 (CLIA). In turn, the laboratory is required to fully establish (ie, validate) the analytic and clinical performance specifications of all such devices for any change of intended use in their patient population. Until such requirements have been met, the department’s letter advised, “the use of glucose meters in health fairs, other community screening events, and/or critical care settings must be discontinued.”

In its communications, the Department of Health provided labs with guidance for investigating the safety of their blood glucose meters. To show that such meters are sufficiently accurate to be used safely by healthcare professionals, the lab should demonstrate that 99% of all values are within ±10% of the reference method for glucose concentrations >70 mg/dL, and within ±7 mg/dL for glucose concentrations <70 mg/dL. Further, to avoid critical patient management errors, no individual result should exceed ±20% of the reference method for samples >70 mg/dL, or ±15 mg/dL for samples <70 mg/dL. Labs must investigate and justify any test result that exceeds these criteria, and describe why the potential for such results does not render the device unsafe and ineffective in the intended use setting (eg, when extrapolated to the actual volume of testing performed in the intended use setting).

To make it possible for units within the St Elizabeth Medical Center to continue using the Accu-Chek Inform II glucose meter for all patient populations, laboratory staff designed and conducted a risk-based validation study. The study was designed to establish the analytical and clinical performance of the glucose meters when used in any circumstances or populations other than those defined by the FDA-cleared intended use.

Materials and Methods 

Utilizing the Roche Accu-Chek Inform II, fingerstick glucose tests were performed on 353 patients in the intensive care unit, cardiothoracic intensive care unit, emergency department, and other inpatient units (see Figure 1). The study used test strips from three different lot numbers. For the patient samples collected for the study, the breakdown by source was as follows:

  • 57 patient samples from the emergency department.
  • 65 patient samples from the intensive care unit (see Figure 2).
  • 92 patient samples from the cardiothoracic intensive care unit (see Figure 3).
  • 139 patient samples from inpatient units dispersed throughout the medical center.

Plasma glucose samples were also collected and tested in the laboratory on the Siemens Dimension Vista 1500 chemistry analyzer (see Table 1 and Table 3).

Testing was performed on all shifts and all days of the week, and represent intended use conditions from September 1, 2013, through January 29, 2014. All told, 118 different nurses performed the fingerstick glucose tests at the patient bedside in the nursing units (see Table 2 and Table 4).

Among the study participants, 74% of the patients had a previously documented diagnosis of diabetes; 26% did not. The nondiabetic patients were following tight glycemic control protocols, a practice that has been shown to decrease postoperative infections and shorten length of stay.

To verify the effectiveness of meter disinfection, researchers used an ATP monitor, which works by detecting the amount of adenosine triphosphate (ATP) that is present on any given surface. ATP offers a quick way to detect unseen organisms such as organic debris, bacteria, parasites, and spores. With proper cleaning and disinfection, all shared equipment should have an ATP reading below 200.


Comparing the test results from the Inform II meter with those from the laboratory’s Vista analyzer, all Inform II values <70 mg/dL were within ±7 mg/dL of the Vista values. Among Inform II test values >70 mg/dL, 92% were within ±10% of the Vista glucose result. No individual result exceeded a difference of ±20%. Among the 353 patient test results produced by the two methods:

  • 3 patients had glucose test levels >600 mg/dL on the Inform II, confirmed on the Vista analyzer.
  • 22 patients had glucose levels <80 mg/dL.
  • 23 patients had glucose levels >300 mg/dL.

Data from all subjects in the study were submitted. The three glucose levels >600 mg/dL were excluded from the graph.

Sodium and hematocrit levels were also recorded for the patients. The sodium levels ranged from 124 to 152 mEQ/L, and the hematocrit ranged from 21% to 57%.

Ongoing monitoring with the ATP monitor demonstrated that the disinfection process is effective.


This study demonstrated that the Roche Accu-Chek Inform II device is sufficiently accurate to be used safely by healthcare professionals in all nursing units within the St Elizabeth Medical Center, including the intensive care unit and the cardiothoracic intensive care unit.

The study also demonstrated that the Roche Accu-Chek Inform II device is sufficiently accurate to be used safely for maintaining tight glycemic control.

Katheryn Inglis, MS(MT), is laboratory manager at St Elizabeth Medical Center, Utica, NY. This article originated in a poster presentation to the Clinical Laboratory Management Association KnowledgeLab 2014 Conference (Las Vegas, May 3-7, 2014).