Evaluation of point-of-care glucose testing accuracy using locally-smoothed median absolute difference curves

Abstract

Background: We introduce locally-smoothed (LS) median absolute difference (MAD) curves for the evaluation of hospital point-of-care (POC) glucose testing accuracy.

Methods: Arterial blood samples (613) were obtained from a university hospital blood gas laboratory. Four hospital glucose meter systems (GMS) were tested against the YSI 2300 glucose analyzer for paired reference observations. We made statistical comparisons using conventional methods (e.g., linear regression, mean absolute differences).

Results: Difference plots with superimposed ISO 15197 tolerance bands showed bias, scatter, heteroscedasticity, and erroneous results well. LS MAD curves readily revealed GMS accuracy patterns. Performance in hypoglycemic and hyperglycemic ranges erratically exceeded the recommended LS MAD error tolerance limit (5 mg/dl). Some systems showed acceptable (within LS MAD tolerance) or nearly acceptable performance in and around a tight glycemic control (TGC) interval of 80-110 mg/dl. Performance patterns varied in this interval, creating potential for discrepant therapeutic decisions.

Conclusions: Erroneous results demonstrated by ISO 15197-difference plots must be carefully considered. LS MAD curves draw on the unique human ability to recognize patterns quickly and discriminate accuracy visually. Performance standards should incorporate LS MAD curves and the recommended error tolerance limit of 5 mg/dl for hospital bedside glucose testing. Each GMS must be considered individually when assessing overall performance for therapeutic decision making in TGC.

Figure 1

Modified Bland-Altman Plots with Superimposed ISO 15197 Tolerance Bands

Figure 2

Locally-Smoothed Median Absolute Difference Curve: GMS 1. The graphs show LS MAD curves for GMS 1A and 1B. The 95% confidence intervals appear above and below the curve. The dashed horizontal line at 5 mg/dl represents the recommended error tolerance. Values of reference glucose where the LS MAD curve crosses the error tolerance, so-called “breakout points,” are identified by the black arrows.

Figure 3

Locally-Smoothed Median Absolute Difference Curves: GMS 2. LS MAD curves for GMS 2A and 2B. Same format as Figure 2.

Figure 4

Locally-Smoothed Median Absolute Difference Curves: GMS 3 and 4. LS MAD curves for GMS 3 and 4. The LS MAD curves float above the error tolerance throughout the glucose range from 35 to 220 mg/dl, which reflects paired observations with reference glucose values from 20 to 235 mg/dl. Each system produced one dangerous Class I discrepant value (see Table 2) in the TGC range from 80 to 110 mg/dl.

Figure 5

Tight Glucose Control (TGC) Intervals—Evidence-Based Clinical Investigations. To maintain TGC, nurses adjust therapy quickly after obtaining bedside glucose results each hour or more frequently. This short therapeutic turnaround time generally requires bedside testing. Hence, POC results must be accurate, because the clinical team uses them for immediate decision making. The TGC ranges at the top were obtained from published papers (8-18). The mark in the top interval represents mean glucose (138 mg/dl) in the study by Lazar et al. [8]. Hospital TGC ranges shown below were obtained by calling.