AISC 2004 Homepage for team 002 (Matthew Woller, Levi Blackstone)

We conducted a review of current literature on ethanol pharmokinetics, and learned from Kapur (1991) that most existing BAC prediction models are based on Widmark’s (1932) mathematical equation (Appendix A). We also discovered that Watson et al. (1980) had developed regression equations to calculate total body water (TBW) to make an accurate estimation of the actual distribution volume of alcohol for each individual; this was the most notable update to Widmark’s work (Appendix A). Pieters et al. (1990) developed a three-compartment model (stomach, small intestine, lean body mass) to simulate ethanol metabolism (Appendix A). Their model assumed that first-pass metabolism was insignificant and that ethanol is released from the stomach at approximately a first-order rate (rate of transfer to the small intestine is dependent on the concentration of alcohol in the stomach). This model also assumed that ethanol absorption from the small intestine followed a first-order rate and that ethanol metabolism from the lean body mass follows Michaelis-Menton kinetics (Appendix A).

We based our new model on the information from our research. Our model uses three compartments (stomach, intestine, lean body mass) to implement our equations. In the stomach compartment, we account for changes in liquid volume based on gastric secretions and the volume of alcohol being consumed. Ethanol leaves the stomach compartment either through first-pass-metabolism, simulated by a Michaelis-Menton kinetics equation, or through the pyloric sphincter to the small intestine at a first-order rate. We assume that the volume of liquid in the stomach leaves at a zero-order rate (constant rate) because of feedback inhibition in the small intestine controlling the rate of release. In the small intestine, we assume that the ethanol is absorbed into the lean body mass at a first-order-rate. Liquid is also assumed to leave the small intestine at a first-order-rate as it is absorbed by the intestines. Finally, we assume that ethanol is eliminated from the lean body mass following Michaelis-Menton kinetics, being dissolved in a liquid volume equal to the TBW. We use Watson et al. TBW equations to calculate TBW in our model. We considered modeling the transport of ethanol through the bloodstream (allowing ethanol to return to compartments it had left previously), but ultimately decided not to include this factor in our model because of the difficulty of obtaining specific blood flow rate values. We felt that the inclusion of this factor without reliable data would adversely affect the accuracy of our predictions, rather than improving them.

Our model last year did not account for several factors important in the process of ethanol metabolism. Previously, we ignored first-pass metabolism in the stomach, and assumed that it was merely a zero-order transport compartment (serving only to delay the absorption of ethanol into the body). We also ignored the presence of food in the stomach and the drinking history of the subject, which both affect the rate of ethanol absorption into the body. Additionally, we did not calculate separate liquid volumes for each compartment, instead dividing the amount of ethanol in each compartment by the TBW to calculate concentration (consequently affecting our transfer rates).

We developed a computer program to test our model in an efficient fashion and allow a user to input values for each subject to be tested. To estimate errors associated with the prediction model, we used a root-mean-square calculation, a statistical method of finding deviation from a mean value (see Results).


				Materials and Methods
			We conducted a review of current literature on ethanol pharmokinetics, and learned from Kapur (1991) that most existing BAC prediction models are based on Widmark’s (1932) mathematical equation (Appendix A). We also discovered that Watson et al. (1980) had developed regression equations to calculate total body water (TBW) to make an accurate estimation of the actual distribution volume of alcohol for each individual; this was the most notable update to Widmark’s work (Appendix A). Pieters et al. (1990) developed a three-compartment model (stomach, small intestine, lean body mass) to simulate ethanol metabolism (Appendix A). Their model assumed that first-pass metabolism was insignificant and that ethanol is released from the stomach at approximately a first-order rate (rate of transfer to the small intestine is dependent on the concentration of alcohol in the stomach). This model also assumed that ethanol absorption from the small intestine followed a first-order rate and that ethanol metabolism from the lean body mass follows Michaelis-Menton kinetics (Appendix A). We based our new model on the information from our research. Our model uses three compartments (stomach, intestine, lean body mass) to implement our equations. In the stomach compartment, we account for changes in liquid volume based on gastric secretions and the volume of alcohol being consumed. Ethanol leaves the stomach compartment either through first-pass-metabolism, simulated by a Michaelis-Menton kinetics equation, or through the pyloric sphincter to the small intestine at a first-order rate. We assume that the volume of liquid in the stomach leaves at a zero-order rate (constant rate) because of feedback inhibition in the small intestine controlling the rate of release. In the small intestine, we assume that the ethanol is absorbed into the lean body mass at a first-order-rate. Liquid is also assumed to leave the small intestine at a first-order-rate as it is absorbed by the intestines. Finally, we assume that ethanol is eliminated from the lean body mass following Michaelis-Menton kinetics, being dissolved in a liquid volume equal to the TBW. We use Watson et al. TBW equations to calculate TBW in our model. We considered modeling the transport of ethanol through the bloodstream (allowing ethanol to return to compartments it had left previously), but ultimately decided not to include this factor in our model because of the difficulty of obtaining specific blood flow rate values. We felt that the inclusion of this factor without reliable data would adversely affect the accuracy of our predictions, rather than improving them. Our model last year did not account for several factors important in the process of ethanol metabolism. Previously, we ignored first-pass metabolism in the stomach, and assumed that it was merely a zero-order transport compartment (serving only to delay the absorption of ethanol into the body). We also ignored the presence of food in the stomach and the drinking history of the subject, which both affect the rate of ethanol absorption into the body. Additionally, we did not calculate separate liquid volumes for each compartment, instead dividing the amount of ethanol in each compartment by the TBW to calculate concentration (consequently affecting our transfer rates). We developed a computer program to test our model in an efficient fashion and allow a user to input values for each subject to be tested. To estimate errors associated with the prediction model, we used a root-mean-square calculation, a statistical method of finding deviation from a mean value (see Results).

Materials and Methods