A COMPARISON BETWEEN ALCOHOL IN SALIVA MEASURED BY AN ENZYMATIC PROCEDURE AND EVIDENTIAL BLOOD AND BREATH ALCOHOL VALUES Ronald C. Denney, University of Greenwich, London. John Matthews, Neil Spencer & John Broad, King's College (KQC), University of London, England. 1. Introduction The use of enzymatic procedures for measuring alcohol in body fluids is well established and has been reported in detail at ICADTS Conferences(1), (2), The application of an enzymatic procedure for the direct measurment of alcohol in saliva samples has also been described (3). Good relationships between saliva alcohol and blood alcohol values have been established, but similar relationships between saliva and breath alcohol results do not appear to have been investigated with this type of system. A recent development for measuring saliva alcohol levels has been the Quantitative Enzyme Diagnostic (Q.E.D.) test. This is based upon the well established alcohol dehydrogenase catalysed oxidation of ethanol, in this case involving the formation of a formazan dye which provides a visible indication of the alcohol level, based upon the following reaction sequence: \ ETHANOL + NAD---- 1------> ACETALDEHYDE + NADH + H* ACETALDEHYDE + TRIS--------- -> COMPLEX NADH + ELECTRON SINK (ES)--------- > NAD + H2 - ES NADH + T S------- 2------- > NAD + FORMAZAN DYE 1 = ALCOHOL DEHYDROGENASE TS = TETRAZOLIUM SALT 2 = DIAPHORASE The Q.E.D. test involves soaking a cotton wool tip with saliva in the mouth and placing the swab in a recess on a plastic test plate on which the enzymatic reaction takes place. As it does 438 Alcohol, Drugs and Traffic Safety - T92 Ed. by Utzelmann / Berghaus / Kroj Verlag TÜV Rheinland GmbH, Köln -1993
so, a blue line develops alongside a scale on the plate and a direct measurement of the saliva alcohol level is obtained from the scale reading. This is illustrated in Figure I, taken from the marketing information provided by the manufacturers. Figure I This investigation was carried out in order to establish three factors: (1) The relationship between Q.E.D. saliva results and blood alcohol values measured by gas chromatography. (2) The relationship between Q.E.D. saliva results and breath alcohol values (converted to blood alcohol figures by multiplying by 2,300). (3) The repeatability of results. 2. Experimental For the first part of this study male and female volunteers were asked to drink various amounts of beer, vodka and whisky in order to produce a range of alcohol (ethanol) values from 0 to 150 mg per 100 ml. In all cases the blood samples were taken at approximately the same time as the subject provided the saliva sample for the Q.E.D. test. By this means 42 comparative results were obtained, and these are listed in Table I. 439
Q.E.D. SALIVA GAS CHROMATOGRAPHY BLOOD (mg/dl) (mg/ dl) 0.0 0.6 0.0 2.1 15.0 5.2 15.0 5.8 20.0 93.7* 26.0 19.4 31.0 13.4 33.0 19.9 35.0 32.7 35.0 34.9 40.0 31.2 42.0 47.7 43.0 24.9 45.0 33.2 45.0 33.3 50.0 35.6 53.0 28.5 53.0 51.2 55.0 51.8 60.0 45.3 60.0 47.4 60.0 53.7 61.0 52.1 75.0 65.4 79.0 62.7 80.0 65.2 80.0 77.1 80.0 83.6 82.0 52.4 82.0 73.7 89.0 62.2 95.0 78.3 98.0 82.8 100.0 77.4 100.0 87.3 100.0 91.0 101.0 86.3 108.0 81.9 110.0 90.7 129.0 110.8 133.0 115.0 140.0 147.5 Table I 440
In the second part of the study a similar procedure was followed and the subjects were tested on the Lion Intoximeter 3000 evidential breath testing machine immediately after a saliva sample had been taken for the Q.E.D. test. By this means 21 valid results were obtained and the breath alcohol values were converted to the corresponding blood alcohol values in order to make comparison with the previous results easier. The 21 results from this are listed in Table II. LION INTOXIMETER EQUIVALENT BLOOD Q.E.D. 3000 (ug/dl) LEVEL (mg/dl) (mg/dl) 0.0 0.0 0.0 18.0 41.1 52.0 24.0 54.8 70.0 25.0 57.1 63.0 28.0 64.0 70.0 29.0 66.3 70.0 29.0 66.3 83.0 32.0 73.1 89.0 32.0 73.1 94.0 33.0 75.4 76.0 34.0 77.7 99.0 35.0 80.0 80.0 35.0 80.0 103.0 36.0 82.3 105.0 37.0 84.6 89.0 37.0 84.6 112.0 40.0 91.4 115.0 40.0 91.4 115.0 41.0 93.7 120.0 41.0 93.7 130.0 45.0 102.9 130.0 Table II For the third study, a simple repeatability check was carried out by one subject holding six Q.E.D. swabs in his mouth simultaneously for the same length of time before placing each on a separate detector plate for measurement. 441
3. Results Of the 42 results obtained from the saliva/blood comparison there was one result that was totally spurious with an abnormally low saliva alcohol value. For the purpose of the statistical analysis this result was discarded, although all points are presented on Figure II. For the 42 points the correlation coefficient r is 0.903, whilst for the 41 points it is 0.971. So there is a very clear linear relationship between the sets of results A value for the saliva/blood alcohol ratio was also calculated from the results after discarding the spurious value and the zero values. This gave the following figures: N = 39 S/B = 1.34 S.D. = 0.42 Gas Chromatography (mg/dl) Blood F i g u r e I I 442
The 21 results comparing the saliva values with breath alcohol values were treated in exactly the same way, and on this occasion no result was considered to be spurious. In this case the correlation coefficient was 0.959 and a linear relationship between the two was again established. The points are presented graphically in Figure III. In this case a saliva/equivalent blood alcohol ratio (S/EB) was obtained, again found by dividing this equivalent value into the Q.E.D. value with the zero value ignored. The following figures were obtained: N = 20 S/EB = 1.21 S.D. = 0.114 Equivalent Blood Level (mg/dl) From LION Intoximeter F i g u r e I I I 443
The repeat investigation involving the six Q.E.D. swabs produced the following results: Q.E.D. Value (mg/dl) 101 103 103 106 105 100 Range 100-106 mg/dl Range difference = 6' I/O 0/ It was not felt that this set of results was sufficient for the purpose of making any statistical analysis. A more extensive study, probably using simulated samples is probably needed, although the values obtained above do show a good degree of reproducibility. 4. Conclusions The main point which came out very clearly from the results is that the Q.E.D. tends to read high compared with both blood and breath alcohol levels. This is certainly desirable in any system which is likely to be retailed to the general public for self-testing purposes as it serves as a built-in safety factor. However, it is a disadvantage from a law enforcement point of view as it leads to false positives from roadside screening and the unnecessary detaining and evidential testing of some motorists. However, the linearity of the relationships is good and the Q.E.D. is probably of value for roadside screening in those cases in which motorists are unable to provide breath samples due to asthma or bronchitis. 5. References 1. R. C. Denney & P. Daga, Proceedings of the Ninth International Conference on Alcohol Drugs & Traffic Safety, San Juan, Puerto Rico, S. Kaye & G. W. Meier (eds.), (1985), 299-305. 2. R. C. Denney, J. Glover & M. A. Russell, Proceedings of the Eleventh International Conference on Alcohol, Drugs & Traffic Safety, Chicago, 1989,M. W. B. Perrine (ed.), (1990), 809-811. 3. M. W. Timmerman & C. D. Zalesky, Proceedings of the Eleventh International Conference on Alcohol, Drugs & Traffic Safety, Chicago, 1989, M. W. B. Perrine (ed.), (1990), 769-771. 444