Optimization of an implantable coated wire glucose sensor.

An effective glucose sensor should have a short response time and show a strong linear relationship with concentration changes. This paper describes the methods used in an optimization experiment designed to minimize the response time and maximize the coated wire glucose sensor output (voltage or current for glucose concentrations ranging from 40 to 200 mg%). The sensor used for the experiments was a coated wire type. The object was to find the optimum membrane composition performance in relation to different glucose solution concentrations. Four experimental designs were used to analyse the effects of Aliquat: decanol (vol%), glucose salt concentration, and polymer: solvent ratio on the response time and voltage or current differences. Regression models for response time and output were used to establish the optimum experimental conditions. These models were calculated using the statistical analysis system (SAS) general linear model (GLM) procedure. Both polarographic and potentiometric models were used to determine the sensitivity of the sensor. The coated wire sensor developed during the present study had a response time of about one minute, a marked improvement over earlier designs. In addition, it yielded reproducible voltage or current differences with changes in glucose concentration.