Assessment outcome

The counter argument to the above rests on the premise that risk assessment outcomes, if not currently testable, might be in the future. Many scientific hypotheses are not testable at the time they are proposed this does not mean they are unscientific, assuming they are based upon and are consistent with all available knowledge. Moreover, risk assessors continue to urge the development of the type of data that will improve the reliability and testability of their predictions. 

The EPA makes decisions about clean-up of abandoned hazardous waste sites under the so-called Superfund law. Risk assessment outcomes are one guide to the decision process. The agency has declared that, for carcinogenic contaminants, clean-up must reach lifetime risks somewhere in the range of one in 10 000 to one-in-one million most decisions seem to aim at risks of one in 100 000 or lower. Hazard index values for non-carcinogens are not expected to exceed one. Costs and technical feasibility figure heavily in these decisions. 

A two- to threefold increased risk of birth defects among children of Vietnam war veterans exposed to Agent Orange has been suggested by several epidemiological studies, but these studies have been criticized on a number of grounds, including exposure assessment, outcome verification, and potential for recall bias. Animal studies have not demonstrated clear-cut adverse effects of phenoxyherbicide exposure on reproductive outcomes. ... 

Similarly, uncertainty can also have substantial implications for assessment outcomes. For example, there is uncertainty about the importance of exposure of birds to pesticides via dermal absorption. This route of exposure is generally ignored in regulatory assessments, but there is evidence that it may be as important as dietary exposure, at least in some circumstances (Driver et al. 1991 Mineau 2002). This uncertainty, therefore, implies a potential 2-fold error in the assessment of exposure. 

The assessment endpoint should be not only measurable (at least potentially) but also modelable. Defining a modelable endpoint is likely to require close discussion between an assessor (who knows what they can model) and a risk manager (who knows what they want to protect). Sometimes the assessment endpoint is only indirectly related to the management goal, for example, if the assessment endpoint is a risk to individuals, but the aim is to protect population sustainability. In such cases, qualitative inference will be required to interpret the assessment result. This inference will need to be done jointly by the risk assessor and risk manager. It is likely to involve substantial uncertainty, which will have to be taken into account qualitatively when producing a narrative description of the assessment outcome. This step should be identified as part of the conceptual model. 

Usually, some lines of evidence will not be suitable for direct incorporation into quantitative analysis. Semiquantitative or qualitative methods will then be needed to weigh the different lines of evidence, including the quantitative assessment, and integrate them for decision making. Methods for assessing weight of evidence were outside the scope of the workshop that developed this book but are discussed by Suter et al. (2000) and were recently the focus of another workshop (Chapman et al. 2002). Whatever method is used for weighing different lines of evidence, it will be important to characterize uncertainties in each line of evidence and show their effect on the overall assessment outcome. 

Furthermore, separating variability and uncertainty can help risk managers and assessors to decide whether to collect additional information and, if so, on which parameters. This is because uncertainty can be reduced by obtaining additional information, but variability cannot. If there is little uncertainty, then the effects are already well characterized and obtaining further data will make little difference to the assessment outcome. If there is much uncertainty, then priority should be given to obtaining better information about those parameters from which it mostly derives. 

The workshop favored the use of graphical representations that combine the key elements of the assessment outcome the magnitude and frequency of effects, together with appropriate confidence bounds. This should always be accompanied... 

Darke S (1998). The effectiveness of methadone maintenance treatment. 3 Moderators of treatment outcome. In Ward J, Mattick RP, Hall W (eds.) Methadone Maintenance Treatment and Other Opioid Replacement Therapies. London Harwood, pp. 75-90 Darke S, Hall W, Wodak A, Heather N Ward J (1992a). Development and validation of a multi-dimensional instrument for assessing outcome of treatment among opiate users the Opiate Treatment Index. British Journal of Addiction, 87, 733-42 Darke S, Hall W, Ross MW Wodak A (1992b). Benzodiazepine use and HIV risk-taking... 

We are using the term inadequate sleep instead of sleep deprivation in our title for a number of reasons. First, few studies have aimed specifically to deprive children or adolescents of sleep. We describe some research on experimental sleep restriction in children but most of these studies fall far short of common deprivation paradigms in animals or even adult humans. Instead, most research in younger humans has assessed outcome measures such as school grades, self-reported sleepiness, and so forth as a function of variations in self-selected or usual sleep patterns with the expectation that children and adolescents who obtain lower than normal amounts of sleep will manifest deficits. Thus, inadequate sleep is defined by sleep characteristics of a sample. We also wanted to note some of the literature on sleep that is disturbed or disrupted due to disease processes such as apnea or periodic leg movements the duration of sleep in sleep disorders may or may not be shortened or restricted although it is likely fragmented and otherwise abnormal. We decided on the term inadequate sleep with the hope that it would encompass these different areas of concern. 

Perhaps the most common model used to assess outcomes uses syngeneic mice given a subcutaneous injection of a transplantable, syngeneic tumor or xenogenic tumor injected into nude or SCID mice. These tumor models allow a broad range of outcome strategies in response to therapeutic intervention. Alternatively, models... 

There are different ways to look at outcomes. One method, the ECHO model, purports three basic types of outcomes economic, clinical, and humanistic (Kozma et al., 1993). Economic outcomes include direct costs and consequences, both medical and nonmedical, and indirect costs and consequences. For example, when assessing outcomes from a patient perspective, a medication copayment would be a direct medication cost, whereas gas money to pick up the medication from the pharmacy would represent a nonmedical direct cost. Lost wages from missed work could be regarded as an indirect cost. 

Clinical outcomes measures can include morbidity and mortality, event rates, and symptom resolution (Ovretveit, 2001). These measures are a direct measure of quality but may be difficult to assess, especially in pharmacy, where their onset could be years following a treatment or intervention (Chassin and Galvin, 1998 Shane and Gouveia, 2000). In these cases, indicators or markers can be used to assess outcomes. These indicators can be condition-specific (e.g., HgAlc) or procedure-specific (e.g., rate of postoperative infection after hip surgery) or address an important issue of patient care. For example, blood pressure may be used as a marker to assess susceptibility to stroke because it is not practical, safe, or ethical to wait and measure the occurrence of stroke. 

On April 1, 2008, the European Court of Justice ruled that the European Commission used improper procedures to exempt deca-BDE from RoHS Directive. The ruling did not question positive EU Risk Assessment outcome of deca-BDE. The outcome of this April 1 ruling was that deca-BDE was banned in the use of electronic and electrical equipment after June 30, 2008. 

Tier 0 analysis may be done just once, but not repeated for every individual assessment. Even though Tier 0 assessments are practical to conduct, they are not suited for addressing problems that require a realistic identification of key factors and exposure conditions contributing to assessment outcomes and uncertainties. Higher-tier assessments are often needed to answer such questions. 

Higher-tier assessments do not require the quantification of every uncertainty. Therefore, a tiered approach is proposed where each individual source of uncertainty in an assessment may be treated at one of three tiers, beginning with qualitative approaches (Tier 1) and progressing to deterministic (Tier 2) or probabilistic approaches (Tier 3) when appropriate. Within a single uncertainty assessment, different sources of uncertainty may be treated at different tiers. Higher-tier methods are targeted on those uncertainties that have most influence on the assessment outcome. It is never practical to treat all uncertainties probabilistically therefore, even in a very refined assessment, some uncertainties will still be treated at the lower tiers. 

It is essential to provide risk managers with an assessment of the overall degree of uncertainty in the assessment outcome. This should integrate all three dimensions of uncertainty, from all parts of the assessment. Integration of qualitative uncertainties is inevitably subjective, so it is important to document the reasoning so that others can evaluate the conclusions that are reached. 

Uncertainties assessed at Tier 1 (qualitative) may be communicated by listing or tabulating them, together with an indication of their direction and magnitude. Possible formats for this are illustrated in chapter 5. In addition, it will generally be desirable to give a more detailed textual discussion of the more important uncertainties in the list and of their combined effect on the assessment outcome. 

The quality of an assessment—and the degree of uncertainty associated with it—should be evaluated in relation to the decision to be made. Is it sufficient to answer the types of questions with which we began this document For example, if an assessment predicts exposure to a contaminant to be below the acceptable or other established exposure limit, is the quality of the assessment sufficiently robust to support that finding Is it possible that the assessment outcome is uncertain enough that the true exposure has a substantial probability of being above the limit ... 

Dennis M, Wellwood I, O Rourke S et al. (1997b). How reliable are simple questions in assessing outcome after stroke Cerebrovascular Diseases 7 19-21... 

Table 10.5 Analysis of impact of measure of central tendency selection on overall occupational risk assessment outcome... 

There are differences as to how safety factors greater than 100 are applied to risk assessment for workers. Some regulatory authorities (e.g. Canada) apply additional uncertainty and safety factors when conducting occupational risk assessments. Other jurisdictions (e.g. the USA) do not. Such divergent approaches can lead to different risk assessment outcomes. Consistent approaches to application of uncertainty and safety factors would greatly facilitate harmonization. 

Level 1 and Level 2 assessment outcomes usually denote pharmaceutical and healthcare companies whose senior management are still not committed to the implementation of validation and rely on their subordinates to enact validation without the practical support they could offer. Computer validation is often characterized by firefighting. 

Another objective of this report is to determine the best method to measure the impact of pharmacotherapeutic decisions made by clinical pharmacists on patient outcomes. There are several approaches that can be used to assess outcomes. These include disease- or treatment-related outcomes (e.g., blood pressure, seizure frequency, medication adherence, target serum concentrations). The Task Force felt that it was not appropriate for this report to delineate specific clinical outcomes such as level of blood pressure control or serum drug concentrations. While these are important outcome measures, the Task... 

TABLE 85—7. Instruments Used for Assessing Outcome Measures in Patients with SLE... 

Compare performance against industry requirements to assess outcome and opportunity Improve measurement precision and accuracy Field test for reliability and lifetime... 

---