5
DOD's PRIORITYSETTING

INTRODUCTION

The objective of the Department of Defense Priority Model (DPM) is to aid in the ranking of sites according to their relative threat to human health and the environment. According to DOD (1991b),

The DPM relies on site data gathered during the preliminary assessment/site inspection (PA/SI) and remedial investigation/feasibility study (RI/FS) phases described in the National Contingency Plan (40 CFR 300). The product of the DPM is a normalized score from 0 to 100 that is an indication of the relative risk posed by that site to human health and the environment. It is DOD's policy to accomplish site cleanups on a worst case basis, and the DPM has been used to establish this priority order.

The DPM is a mathematical algorithm or model used to compute a numerical score from 0 to 100 that represents the relative potential threat to human health and the environment posed by a contaminated site. Like EPA's Hazard Ranking System, the DPM is a structured-value model. Us-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action 5 DOD's PRIORITYSETTING INTRODUCTION The objective of the Department of Defense Priority Model (DPM) is to aid in the ranking of sites according to their relative threat to human health and the environment. According to DOD (1991b), The DPM relies on site data gathered during the preliminary assessment/site inspection (PA/SI) and remedial investigation/feasibility study (RI/FS) phases described in the National Contingency Plan (40 CFR 300). The product of the DPM is a normalized score from 0 to 100 that is an indication of the relative risk posed by that site to human health and the environment. It is DOD's policy to accomplish site cleanups on a worst case basis, and the DPM has been used to establish this priority order. The DPM is a mathematical algorithm or model used to compute a numerical score from 0 to 100 that represents the relative potential threat to human health and the environment posed by a contaminated site. Like EPA's Hazard Ranking System, the DPM is a structured-value model. Us-

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action ing quantitative data and qualitative estimates, the DPM calculates separate subscores for effects on human and ecological receptors via surface water and groundwater pathways and air and soil pathways for volatiles and dust. The subscores are then combined into an overall site score. Part of the committee's task was to prepare an interim report evaluating the methods, assumptions, and constraints of the DPM. The results of the committee's analysis presented in that report (NRC, 1992) are reflected in this chapter. After completing its analysis, the committee learned that DOD had decided not to use the DPM to aid in ranking sites. BACKGROUND AND HISTORY The DPM is an outgrowth of the Hazard Assessment Rating Methodology (HARM) developed by the Air Force in the early 1980s. The HARM included surface water and groundwater pathways and considered contaminants present at a site and the potential for exposure of receptors to these contaminants. The HARM relied on the aggregation of subscores for the pathway-receptor combination by simple averaging. With the subsequent development of the EPA Hazard Ranking System (FIRS) (discussed in Chapter 4) to establish eligibility for the National Priorities List (NPL), the HARM was revised to improve the scoring for the surface water and groundwater pathways (e.g., by including floodwater transport, depth to groundwater, and infiltration potential); to improve the use of toxicity information to specifically address the relative potency of each significant contaminant; and to obtain better separation of scores by using a root-mean-square algorithm. The resulting modified model, named HARM II, was developed and first tested in 1986 (Barnthouse et al., 1986). In 1987, HARM II was adopted for DOD-wide use and was renamed the

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action Defense Priority Model. In response to comments regarding the DPM, it was further modified (Federal Register, 1989). The first computerized version of the DPM, known as the automated DPM, was developed in 1988 for use on personal computers running under the DOS operating system. This version was released as DPM Version 2.0 in June 1989. A revised version of the DPM was released in June 1991 and is known as the FY 92 version. A rationale document for the DPM was provided to the committee in October 1991 (DOD, 1991c). Throughout this chapter, references are made to the DPM User's Manual for the FY 92 version (DOD, 1991b). THE DPM STRUCTURE Overview The DPM uses a combination of quantitative data and qualitative approximations intended to rank sites according to their potential threats to human health and the environment. Unlike EPA's HRS, which is used for initial screening, the DPM is used after a remedial investigation and feasibility study (RI/FS) has been conducted and the significance of the contamination at a site has been characterized in detail (see Figure 5-1). Contaminant mass, the number of chemicals and their mobility and toxicity, exposure-pathway assumptions, proximity of receptors, and allowable exposure criteria appear to be the important variables that determine the overall ranking of a site. These factors are combined to mimic risk assessment and yield an overall score. The overall scoring method of the DPM is based on a set of product algorithms that account for the exposure pathway, contaminant hazard, and receptors (human, animal, or plant), similar to the HRS (see Chapter 4):

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action FIGURE 5-1 DOD's approach to priority setting. Source: Adapted from DOD, 1991b.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action A schematic illustration of the DPM structure, which describes the above components of the DPM, is given in Figures 5-2a and 5-2b and Table 5-1. The final normalized DPM score represents the relative potential threat that a contaminated site poses to human health and the environment. The scores do not constitute the full process of setting priorities for remediation, but they are intended to be one important factor in priority setting (Figure 5-1). The scores can be used by DOD decision-makers with regulatory considerations, program efficiencies, additional risk information, and other factors to determine the relative priority of sites for remedial action (DOD, 1991b). Data Requirements The DPM input data requirements are summarized in Table 5-2; the preference is for site-specific data. The information available for DPM application includes data collected as part of the PA and SI and RI/FS activities. To incorporate a measure of uncertainty in the scoring during the quality assurance (QA) assessment, the DPM uses the concept of confidence factors. These confidence factors, which range from a value of 0 (uncertain) to 1 (certain), are multiplied by the maximum possible score for each factor (DOD, 1991b). Model Documentation and Software The main documentation describing the DPM is the DPM

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action FIGURE 5-2A DPM structure. Source: DOD, 1991b.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action FIGURE 5-2B DPM segments and scoring order. Source: DOD, 1991c.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action TABLE 5-1 Combining scores in the DPM Surface water human health score = Surface water pathway score × Surface human health hazard score × Surface water human receptor score /10,000 Surface water ecological score = Surface water pathway score × Surface water ecological hazard score × Surface water ecological receptor score /10,000 Groundwater human health score = Groundwater pathway score × Groundwater human health hazard score × Groundwater human receptor score /10,000 Groundwater ecological score = Groundwater pathway score × Groundwater ecological hazard score × Groundwater ecological receptor score /10,000 Air/soil volatiles1 human health score = Air/soil volatiles pathway score × Air/soil volatiles human health hazard score × Air/soil volatiles human receptor score /10,000 Air/soil volatiles ecological score = Air/soil volatiles pathway score × Air/soil volatiles ecological hazard score × Air/soil volatiles ecological receptor score /10,000 Air/soil dust1 human health score = Air/soil dust pathway score × Air/soil dust human health hazard score × Air/soil dust human receptor score /10,000 Air/soil dust2eecological score = Air/soil dust pathway score × Air/soil dust ecological hazard score × Air/soil dust ecological receptor score /10,000 1 The higher of these two scores is used in the final Computation. 2 The higher of these two scores is used in the final computation. Source: DOD, 1991c.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action TABLE 5-2. Summary of DPM Data Requirements Data Category Specific Data Needs General Site Site location, name and number Information Distance to nearest installation boundary   Distance to nearest residential, industrial, or commercial land use   Type of waste site, size of waste site (area and depth), type of waste and concentration of contaminant in waste, physical form of waste, waste quantity, and waste containment information   Mean summer temperature, soil porosity, average wind speed, soil bulk density Pollutant Field Data Have non-volatile (vapor pressure less than 10-3 mm Hg) or volatile pollutants detected or releases been observed in the ambient air? (For observed releases to air identify contaminants and measure or calculate maximum concentration levels; for non observed releases to air but observed releases in surface soil, identify contaminants and the maximum observed concentrations   Have volatile or non-volatile pollutants been observed in the surface soil? (For observed releases of non-volatile pollutants determine (via measurements or appropriate models) the maximum concentration levels in air and surface soil for each contaminant. For non observed releases of non-volatile pollutants to air but observed releases in surface soil, identify contaminants and determine (via appropriate models) the maximum concentration levels in air and soil)   Have pollutants been observed in ground water? (For observed releases identify contaminants and determine maximum observed concentrations)

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action Data Category Specific Data Needs Pollutant Field Data Have pollutants been observed in surface water? (For observed surface water contamination, identify contaminants and determine (maximum observed) concentrations). Climate Net precipitation and representative rainfall intensity east and west of the Mississippi River   Days per year with precipitation greater than 0.25 mm   Flooding potential   Annual average wind velocity Soil Data Soil porosity   Soil permeability   Erosion potential   Neutralization capacity (based on soil chemistry)   Organic content   pH   Average summer soil temperature Groundwater Depth (seasonal high)   "Short circuit potential" (i.e., presence of faults, cracks, etc.)   Distance (downgradient) from waste to supply wells, surface water, habitat or natural areas.   Hydraulic conductivity of aquifer, effective porosity of soil, hydraulic gradient Surface Water Distance from waste site to nearest surface water   Water use of the nearest surface water bodies

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action Population and Ecological Information Sensitive ecosystems located within 4 miles downstream or 1.5 miles radius   Presence of critical environments (e.g., habitat or endangered species, nature preserve, wilderness area, important natural resource, etc.)   Population downstream that obtains drinking water from potentially affected surface water or groundwater   Population within 1/2 mile of the site and population within a 4 mile radius   Land use   Source: Material from DOD, 1991b.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action eight subscores for pathway-receptor combinations (DOD, 1991b). The given aggregation formula uses a root-mean-square combination of the subscores as is also done in EPA's Hazard Ranking System (see Chapter 4). Apparently that choice was made because, compared with the arithmetic mean, it permits "individually higher subscores to have a greater effect on the overall score." The rationale for the above approach is unclear. For example, it would be even more true for roots of higher powers, such as the root-mean-cube. The reasons why root-mean-square is common in statistical and engineering calculations seem irrelevant here; perhaps a larger exponent would yield a spread of site scores more helpful for allocating resources. The given aggregate formula implicitly treats the scores to be combined as independent. The assumption is that there are neither significant interactions (e.g., a health insult through one pathway exacerbating the health impact of some intake via another pathway) nor significant double counting. In the present version of the DPM, volatiles and dust are treated separately for air and soil transport. But for human and ecological receptors, only the larger of the two resulting scores is entered into the aggregation, the other being dropped entirely (DOD, 1991b). This exemplifies the failure of the DPM approach to aggregate the effect of two contaminant release sources that could lead to a higher potential threat than the individual sources alone. Clear justification for various multipliers in scoring algorithms is not provided. Thus, it appears that the individual pathway scores and their aggregation cannot be analyzed along a systematic theoretical basis that might enable one to check the rationale in the scoring methodology and to propose revisions. Of particular concern is the transformation of continuous or cardinal datum (e.g., distance to nearest surface water) to an ordinal score (e.g., 0, 1, 2, 3) for which the rationales for the threshold levels used in the cardinal-to-ordinal transformation are not given. As noted in Chapters 3 and 4, the lack of a rigorous, testable basis for deter-

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action mining such factor scores is a problem common to structured-value models such as the HRS and DPM. VALIDATION Model validation encompasses the soundness and accuracy of the model as a means of establishing priorities for remedial action, as well as the mathematical and numerical aspects of the model computer code. The DPM has not been validated, even though validation is recognized as critical in the development and application of models for use in policy and regulatory decisions (Naylor and Finger, 1967; Chapra and Reckhow, 1983; Reckhow et al., 1990; Arula, 1987; Shaeffer, 1980; ASTM, 1984; EPA, 1989b; NRC, 1990b). A broad validation can be performed fully only in the context of the intended model use; that is, does the model give good advice in establishing priorities for remedial action? The direct output of the model, the DPM score, is intended to provide a measure of relative site risk or threat, which is intended to be used in the setting of priorities for resource allocation. Validation efforts need to address not only the relative measure of risk provided by the output scores, but also the quality of the rankings that result. Possibilities for the latter might involve classifying the sites simply into groups (e.g., high, medium, or low priority for remediation) as a result of the scoring or making a finer detailed ranking of the site scores. An appropriate validation study-comparing model results with what they should be-would involve perhaps 10 to 30 sites and the comparison of scores and rankings from the DPM with those from another approach, assumed a priori more likely to yield the right answer. Ideally, the sites used would be authentic ones, perhaps including some already selected for rapid cleanup outside the DPM framework, but they could also include hypothetically de-

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action signed sites. The purported right answers could be based on the results of independent risk analyses performed according to well-established procedures or on the judgment of experts who can give those sites intense consideration. Similar comparisons have been made for the EPA Hazard Ranking System models (Dory and Travis, 1990). For authentic sites that have advanced to later stages of investigation, the right answers could be based on experience involving the manifestation of a hazard or the benefits realized through remediation. A record of the validation study's procedures and results should become part of the DPM's documentation. SENSITIVITY AND UNCERTAINTY ANALYSES An important step in evaluating the performance and reliability of priority-setting models is to determine through sensitivity and uncertainty analyses the magnitude of uncertainties in the model site scores and the implications of the uncertainties for site ranking. Detailed sensitivity and uncertainty analyses are yet to be performed on the direct model output (the DPM score) and on the resulting site rankings or priorities. The latter can be examined by determining how uncertainties in DPM scores affect an overall ranking and inclusion on the short list of sites identified for highest-priority consideration. Uncertainties in model output can be derived from assumed uncertainties in model inputs (or structure) or evaluated directly by analyzing how site scores vary among different analysts. To illustrate the potential impacts of uncertainties in the DPM model results on ranking, a preliminary analysis was recently conducted (NRC, 1992) based on the available set of FY 1991 DPM

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action scores (M. Read, DOD, pers. comm., July 1991). In this analysis, the uncertainty in the composition of the list of FY 1991 sites that had the 50 highest scores was evaluated. The changes in scores that resulted from the FY 1991 DOD quality-assurance (QA) review were used to scale the uncertainty in site scores. The uncertainty assumed for the scores was based on the magnitude of the changes that occurred for the 50 sites for FY 1991 that underwent the QA correction. This uncertainty was superimposed on the full set of 284 sites scored in the DPM by DOD in FY 1991. The DPM scores for FY 1991 ranged from 1 to 64, with the distribution shown in Figure 5-4. Because all the scores that will ultimately be used for DOD priority setting will already have undergone QA, an additional analysis was performed on the assumption that the uncertainty in site scores is only one-fourth FIGURE 5-4. Grouped frequency distribution of DPM scores for 284 sites in Fiscal Year 1991. Of 284 sites scored, 15 have DPM scores in an interval from 8 to 12. Highest-ranked site is Rocky Mountain Arsenal, which has a score of 64. Two sites, Riverbank Army Ammunition Plant E/P Ponds and Richards Gebaur Hazardous Waste Drum Storage Site 923, have a low DPM score of 1.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action that reflected in the QA corrections. Although this change in assumption did significantly reduce the variation in rankings, in both cases the composition of the group of sites with the highest rankings proved subject to considerable variation. To implement the analysis, fifty simulations were performed by adding random-error values with the properly scaled variance to the FY 1991 DPM scores. The modified scores and overall ratings were determined for the 284 sites. The results of the analysis are summarized in Figure 5-5, which presents the uncertainty in the composition of the top-50 list for the two cases. It shows the probability, or fraction, of simulations in which each of the 284 FY 1991 sites is included among the top-50 ranked sites. As shown in Figure 5-5a, the variation in the full-QA-uncertainty case is quite large. Sites with baseline scores as low as 12 have a nonzero chance of being included among the top-50 sites (corresponding to a least one simulation in which the site was among the top-50 scores). Fully 219 of the 284 sites were included among the top 50 sites in at least one simulation. Furthermore, the top baseline site with a score of 64 was included among the top-50 sites only 84% of the time (corresponding to 42 of the 50 simulations). If the magnitude of site-score uncertainty is in fact comparable with that reflected in the FY 1991 QA modifications, then decisions based on the DPM ranking (e.g., to begin remediation with the top 50 sites) are subject to considerable uncertainty. The variation in the top-50 list for the one-fourth-QA-uncertainty case is shown in Figure 5-5b. The impact of site-score uncertainty is shown to be greatly reduced. Sites with baseline scores above 50 are virtually assured of being on the list, whereas sites with baseline scores below 30 are virtually assured of being left off the list. The number of sites included in the top-50 list at least once is reduced to 107, compared with 219 in the full-QA-variability case. Site rankings and remediation decisions are thus more robust with this lower level of site-score uncertainty.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action FIGURE 5-5 (a: top; b: bottom) Probability of inclusion among top-50 ranked sites as function of final Fiscal Year 1991 DPM score. Results based on 50 Monte Carlo simulations.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action The analysis presented in Figure 5-5 can be extended to focus attention on sites that warrant additional study and effort. For example, in the case of one-fourth QA uncertainty, sites with scores above 50 are so likely to belong to the top-50 list, and sites with scores below 30 so unlikely to belong, that they require little additional study to reduce the uncertainty in their scores. The composition of the top-50 list is sensitive only to uncertainties in sites with scores of 30 to 50. It is those transition sites that should be targeted for further study to reduce their site-score uncertainty. The uncertainty analysis thus provides a mechanism for focusing further data collection and study efforts. This analysis illustrates the kinds of sensitivity and uncertainty evaluation that could be performed. The uncertainty in site scores is shown to have a considerable impact on the composition of the top-50 list, although this impact is sensitive to the magnitude of the assumed site-score uncertainty. The analysis demonstrates that the uncertainty in DPM scores could potentially limit the use of the model for setting priorities among sites for remediation. It is essential that uncertainty evaluation of this type be performed for priority-setting models in the context of their intended use. SUMMARY The DPM is structured as a user-friendly model, and QA/QC approaches are used in its application. The DPM approach of using the product of pathway potential, hazard, and receptor is one reasonable approach to defining an overall site score. A detailed review of the DPM, however, reveals that the some of the transport and fate algorithms, toxicologic and exposure assumptions, and methods embedded in the DPM have weak theoretical foundations. For example, the fate algorithms used for the surface water, groundwater, and air and soil characteristics do not have an acceptable theoretical basis.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action The pathway algorithms in the DPM use a summation formula to reflect the combined effect of the various pathway parameters, whereas theory suggests that a product of characteristics or a sum on a logrithmic scale would be the preferred approach to score the pathway potential. The DPM does not appear to explicitly address social and economic impacts on site characterization, and it is not clear whether DOD addresses these issues through an evaluation process external to the DPM. The DPM scoring scale is linear, and results of DPM site scores reveal that the score intervals for the FY 1990 and FT 1991 sites are small. Thus, based on the above results, it appears that the DPM may have a limited capability to discriminate between sites (NRC, 1992). Score compactness on the 0-100 score interval and factors in the pathway summation algorithms that mitigate against discrimination on pathway potential suggest that model algorithms should be restructured to produce a logarithmic scoring scale. Spreading out the numerical scores with alternate algorithms to combine scores, such as product algorithms (see NRC, 1992), might allow better discrimination between site scores. A simple sensitivity analysis for 50 highest DPM scoring sites (NRC, 1992) demonstrated that uncertainties can have large effects on the composition of the top-50 list. To provide a summary evaluation of the tools used in DOD's priority-setting process, the criteria identified in Chapter 2 for effective model development and application and for the specific technical desired features are examined with primary focus on the DPM model. Given that the DPM was undergoing development when the committee performed its analysis, the comments given below should be viewed as committee recommendations of needs for such future efforts, rather than as evaluative of a completed product.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action General Issues in DPM Model Development and Application Defined Purpose: The DPM has a well-defined purpose within DOD's overall priority setting process. It is responsive to the stated DOD policy of giving site-cleanup priority to sites which present the greatest potential threat to human health and the environment. Credibility and Acceptability: The extensive scientific peer review, public participation, and public comments that are needed for establishing credibility and acceptability of a model to be used in priority-setting have not yet been conducted with the DPM. Appropriate Logic and Implementation of Mathematics: The calculation methods used in DPM are fairly straightforward. However, the logic of the choices made for particular operations and for combining quantities appears somewhat arbitrary. The basis for combining and weighting is not clear from the information which was provided to the committee. Model Documentation: The documentation for the model was limited at the time the committee performed its analysis. That which was available to the committee does not adequately explain why the model is designed as it is. More extensive documentation is needed to describe the whole modeling process, as well as its product. For example, many of DPM's features, instead of being self-evidently correct, appear to be choices among many possible options. Such choices call for explanation and support. Documentation is also needed for evaluation of whether some particular kind of risk is being quantified consistently and whether the model's default values were chosen to be consistent with some explicitly stated policy. Model Validation: The DPM has not yet been validated. Complex models such as this need to be checked carefully to determine whether in fact they perform sufficiently well for their intended

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action purpose. The question is whether or not the DPM gives good advice in establishing priorities for remedial action. This has not yet been adequately determined. Model Sensitivity and Uncertainty Analysis: There are many uncertainties in the data collected from a site for use in priority setting models. The effect of such uncertainties in model outcome and the priorities that sites receive should be known. An adequate sensitivity and uncertainty analysis for the DPM model is needed. Specific DPM Technical Features Applicability to All Waste Sites: The DPM is broadly applicable to essentially all DOD sites for which the model might be used. Allowance for Dynamic Tracking: The DPM has not been developed as a tool for dynamic tracking. Discrimination between Immediate and Long-Term Risk: The purpose of the model is to address primarily the longer-term risks. Immediate risks will be addressed as a first priority of the DOD and will not be subjected to priority-setting through the DPM. Inclusion of Cost Estimates of Remediation: The DPM does not consider cost issues or timing with respect to remediation. Transparency: The DPM is a highly transparent model. The mathematical formulation used are well described, and the procedures used for weighing of health and environmental risks are readily apparent. Model transparency and simplicity in use are major advantages of this model. User-Friendliness: The DPM scoring procedures are straightforward and are described in an easy-to-follow procedure. Appropriate Security: It is not clear at the moment how security issues will be addressed.

OCR for page 135
Ranking Hazardous-Waste Sites for Remedial Action DOD Priority-Setting Process Unlike the EPA's HRS, the DPM is used later in the priority-setting process after a more detailed site characterization, representative of a remedial investigation and feasibility study (RI/FS), has been completed. The DPM is not intended for use in ranking of all contaminated sites at DOD facilities. Apparently, sites posing imminent threats from hazardous or toxic substances will receive top priority for cleanup, and will not become part of the DPM evaluation. In addition, DOD plans to place higher priority on cleanup of sites on DOD installations that are subject to closure. Up until recently, site cleanup has not been restricted by a lack of funds. However, with increasing number of sites with detailed characterization completed, competition for funds is expected to become evident soon. The DPM does not explicitly evaluate the social and economic effects often associated with hazardous waste sites. It is intended only to provide a relative ranking of sites based upon their relative threat to human health and the environment. The ranking provided by DPM is to be used "along with additional risk information and other factors such as regulatory requirements and program efficiencies" to establish cleanup priorities among the DOD sites. The process by which other factors will be considered in setting priorities is not known to the committee, thus the committee is unable to comment on how well the overall priority-setting process will address the several features and technical issues noted above.