Attaining specificity in the measurement of ethanol in breath.

In most cases, new breath alcohol measuring devices satisfy the requirements for accuracy and precision. All instruments however, have responded to some substances other than ethanol. In this area of instruments specificity for ethanol there is scope for further development. Substances other than ethanol may be distinguished as interfering substances by the use of two methods of ethanol detection. One example of this is dual-wavelength infra-red detection. Alternatively, infra-red detection could be coupled to a fuel-cell or Taguchi cell detector. Comparison of the two readings obtained should allow most substances other than ethanol to be distinguished as interfering substances. Two instruments using two modes of ethanol detection are the Intoximeter 3000 (3.4 um infra-red plus Taguchi cell) and the BAC DataMaster (dual wavelength infra-red). These have proved unsuitable because of incomplete specificity for ethanol, largely because acetone appears to be considered the only potential interfering substance. In the selection of two modes of ethanol detection, each system must be chosen such that a comparison of the response from each mode would indicate a wide range of possible interfering substances, not just acetone. A performance specification has been developed, against which future commercial instruments will be evaluated.