relative to the questions listed above. In the second section, the subcommittee suggests alternative methods for deriving ERFs that could eliminate some of the limitations in the current approach. The third section presents the subcommittee's conclusions and recommendations regarding use of the current and alternative approaches to the LATRA-ERF framework.
To assess the appropriateness of the LATRA-ERF model, the subcommittee considered six issues: (1) selection of the independent variable; (2) choice of the exposure-response model; (3) representation of sensitive populations; (4) representation of severity categories; (5) selection of the chemical-specific exposure-response values to which an ERF is fit; and (6) representation and propagation of uncertainties in LATRA. During its evaluation, the subcommittee members identified a few other features of the LATRA model that they felt needed comment.
In the LATRA-ERF model, the independent variable for assessing health effects is the maximum 1-hr TWA concentration or, in some cases, a ceiling concentration. There are alternatives to those independent variables. The subcommittee's review of the sparse toxicity data for the compounds in question (Appendices D, E, and F) indicates that the product of concentration and time (C × T) or (C × T x) appears to be a more appropriate independent variable for exposure times ranging from 5 to 60 min, at least for severe effects and mortality for HCl and NO2 (see Chapter 6). Use of the product of exposure concentration and time as the independent variable allows estimation of the risks associated with exposures of different durations using a single ERF. With C × T x as the independent variable for exposures from 5 to 60 min, it also is not necessary to set a ceiling concentration; the 5-min peak concentration that REEDM can estimate (it does not estimate shorter peak concentrations) is within the range of exposure durations over which the C × T x relationship holds.
For effects for which the relationship between effect and the product of concentration and time is unknown, as might be the case for mild and moderate effects, a TWA concentration for specific exposure dura-