Evaluation of the Vision Theophylline assays in the emergency department setting.

STUDY OBJECTIVES

To determine whether the use of fingerstick blood samples or venipuncture blood samples affect the accuracy and the precision of the Vision Theophylline I assay and the Vision Theophylline II assay in the emergency department setting.

DESIGN

Prospective clinical trial.

SETTING

ED in a university teaching hospital.

SUBJECTS/INTERVENTIONS: Simultaneous fingerstick and venipuncture blood samples were obtained from adult patients who required baseline theophylline levels in the ED.

MEASUREMENTS

The Vision system was used in the ED to measure theophylline concentrations from fingerstick and venipuncture samples. Vision samples were drawn and analyzed by one of three pharmacists rounding within the ED between March and October 1990. These pharmacists were instructed by laboratory personnel concerning pertinent operating procedures. The vision Theophylline I and Theophylline II assays are enzyme-inhibitor immunoassays contained within a multi-chamber test packet. The high-performance liquid chromatography (HPLC) theophylline assay was used as the reference method.

MAIN RESULTS

Linear regression of theophylline concentrations determined by the Vision Theophylline I assay versus HPLC revealed a significant correlation for fingerstick samples (r = .9440; P less than .05) and for venipuncture samples (r = .9750; P less than .05). The accuracy values for Vision Theophylline I assay venipuncture and fingerstick samples versus HPLC were -0.63 +/- 1.2 mg/l and 0.22 +/- 1.8 mg/L, respectively. The precision for Vision Theophylline I assay venipuncture and fingerstick samples versus HPLC were 0.27 and 0.38, respectively. Linear regression of theophylline concentrations determined by the Vision Theophylline II assay versus HPLC revealed a significant correlation for fingerstick samples (r = .9804; P less than .05) and for venipuncture samples (r = .9875; P less than .05). Accuracy values for Vision Theophylline II fingerstick and venipuncture samples versus HPLC were -0.53 +/- 1.4 mg/L and -0.65 +/- 1.3 mg/L, respectively. Precision values for Vision Theophylline II fingerstick and venipuncture samples versus HPLC were 0.34 and 0.29, respectively.

CONCLUSION

There is a strong correlation between theophylline concentrations measured by Vision Theophylline assay versus HPLC, regardless of whether the Vision Theophylline assay is performed on a fingerstick or venipuncture sample. Vision fingerstick samples were more accurate than Vision venipuncture samples for both Vision Theophylline I and II assays. Theophylline concentrations measured by the Vision system using venipuncture samples were slightly more precise than those using fingerstick samples when compared with HPLC analysis. These differences in accuracy and precision were not clinically relevant.