Evaluation of a noninstrumented disposable method for quantifying serum theophylline concentrations

To compare the performance of a new point-of-care theophylline assay (AccuMeter) with that of a standard laboratory assay (TDx), and another point-of-care method (AccuLevel).

Prospective evaluation of consecutive patients receiving theophylline.

University-based, ambulatory, allergy-pulmonary clinic.

Forty subjects receiving maintenance theophylline therapy for asthma.

Theophylline concentrations obtained from AccuMeter, TDx, and AccuLevel were compared.

The error, or difference, between TDx and AccuMeter results in 40 subjects on maintenance theophylline described accuracy. Mean error, an estimate of bias, was 1.1 (95% CI 0.72-1.5), 0.67 (0.34-1.0), and 0.98 (0.79-1.2) microg/ml for AccuMeter capillary, serum, and heparinized blood samples. Square root of the mean squared error, an estimate of precision, was 1.6 (1.2-2.0), 1.22 (0.90-1.5), and 1.14 (0.96-1.3) microg/ml for AccuMeter capillary, serum, and heparinized samples. Difference between AccuMeter and AccuLevel ME, an estimate of relative bias, was 0.59 (0.04-1.1) microg/ml. The difference in mean squared errors, an estimate of relative precision, was 0.86 (-0.54-2.3) microg/ml.

AccuMeter demonstrated good precision and minimal bias compared with TDx and AccuLevel . Method of sample collection had no effect on its accuracy.