Screening for Compensation
As discussed in earlier chapters, amendments in 2000 to the Radiation Exposure Compensation Act (RECA) of 1990 created funds for education, prevention, and early detection of radiogenic cancers and other diseases. Of particular relevance here is the charge to the Health Resources and Services Administration (HRSA) to administer the Radiation Exposure Screening and Education Program (RESEP) to facilitate the RECA compensation program. Chapter 9 clarified that screening activities have two purposes and targets: medical screening to improve health outcomes and compensational screening specifically for RECA. In its interim letter report (NRC, 2003a), the committee addressed medical screening and concluded that no scientific evidence supported the view that medical screening beyond that ordinarily recommended for unexposed populations would improve health outcomes.
In Chapter 9, we examined epidemiologic, statistical, and clinical issues related chiefly to medical screening and explained the numerous drawbacks to programs that use multiple screening tests even for medical purposes. We noted the relative paucity of robust evidence that supports screening for many conditions (including many that RECA covers) and cited the authoritative sources of guidelines for medical screening of populations in general or those known to have specific medical risks. We observed that screening programs can be harmful because of false-positive test results (for example, the physical risks that their workup entails and the psychologic and social consequences of labeling individuals as potentially affected). We explained the problems of false-positive rates with multiple tests in various scenarios. Finally, we discussed the ramifications of these issues specifically for RESEP, giving particular attention to activities
intended to promote health outcomes rather than simply document possible disease for compensation purposes.
In this chapter, we address compensational screening. This is the use of medical tests and administrative qualifications to identify individuals who might be candidates for compensation under RECA. This concept is required because RESEP specifies screening for some diseases not traditionally recommended for medical screening to improve health outcomes. Because a term is required to describe this RESEP concept, we coin the term compensational screening. RECA provides compensation for certain radiogenic and nonradiogenic diseases. Some individuals may wish to be evaluated to establish eligibility for compensation, and RESEP provides a mechanism for such evaluation.
RESEP also provides means of organizing direct or referral care for at least some screened individuals who are found to have abnormalities that need further investigation. It is not, however, specifically mandated to provide care to such persons (who might then be regarded as patients). As noted in Chapter 11, some HRSA grantees with RESEP awards apparently leverage their grant awards to provide at least some medical services to some patients under various arrangements for those with or without health insurance. Nonetheless, even though RESEP is identified as a medical screening program, it is properly seen more as a compensational screening program. Moreover, RESEP was not established to be a provider of medical services (of last resort or otherwise), although one interpretation of the services rendered is that they are meant to enhance prevention.
If populations in the RESEP screening program are medically underserved or reluctant to contact the medical system, diverting resources or providing additional resources to facilitate such contact might be more beneficial than devoting resources to enhance screening (medical or compensational); such screening is unlikely to improve outcomes. The questions of expanded contact, education, and outreach are taken up in Chapter 11.
Regardless of the underlying motivation, screening should be preceded by informed decision making in coordination with a clinician. The reason is that such evaluation has potential benefits and potential harms, and the balance between them may not be clear, especially in RECA-RESEP circumstances. For professional and ethical reasons, an explicit plan for appropriate follow-up services for compensable and noncompensable diseases is needed. Informed decision making is as important, if not more so, for compensational than for medical screening because of the low likelihoods of either monetary payments or health benefits with the former.
COMPENSATIONAL SCREENING ISSUES: THE CORE CONCERN FOR RECA AND RESEP
In compensational screening, individuals being screened may have only a modest likelihood of having a compensable disease. Although the same may be
said of some medical screening tests if the prevalence of the disease in the screened population is low, the motivating effect of potential compensation may well attract individuals with a low likelihood of a compensable disease who, all other things being equal, probably would not elect to be screened for medical reasons alone. Such persons may have only a low likelihood of receiving compensation, but they face the same risks of harm from the screening that individuals with higher likelihoods of disease would face. Hence, they should balance the likelihood of monetary (but not medical) benefit against the health risks of screening.
Reasonable individuals may well disagree about whether being screened solely for the purposes of seeking compensation is a good choice. Sometimes such decisions may reflect a lack of complete understanding of the data that describe the uncertainties engendered by screening; such data may turn on probabilities that themselves are quite low. For example, many individuals do not intuitively understand the implications of uncertainty or small numbers (such as probabilities under 1%), and this is a matter of both literacy and numeracy (discussed in Chapter 11). Many untrained individuals cannot adequately distinguish among, say, 1/100 (1%), 1/1,000 (0.1%), and 1/10,000 (0.01%); all are sometimes seen simply as small numbers. Many adults not only do not understand the meaning of small numbers, but they are also challenged in interpreting probabilities and statistics and applying such data to decisions in their daily lives (Schwartz et al., 1997).
Making rational choices about topics as complex as medical screening (wherein longer-term health benefits must be balanced against shorter-term risks) will certainly be near the limits of the ability of some persons to make well-considered decisions. The even more complex tradeoffs in compensational screening (for example, balancing medical risks against the sometimes small likelihood of monetary gain) could well push many people beyond their ability to make informed, rational judgments unless they are carefully supported throughout the process.
Thus, the ethics of screening for rare diseases and conditions among either general populations or populations targeted for screening on grounds other than disease itself (such as compensation) must be taken into account (an issue introduced in Chapter 8). Compensational screening needs to account for both the autonomy and the authenticity of the person; the decision must reflect the person’s desires and must not be coerced.
As now structured, RECA is designed to compensate individuals in three main groups:
Those who were exposed to radiation in the course of their employment.
Those who were exposed during onsite testing of nuclear armaments.
Those who were inadvertently exposed to fallout because they were downwind of nuclear tests.
RECA established criteria for exposure to radiation (based on timing, place of residence, and occupational activities). It also required that claimants demonstrate at least one of more than 2 dozen diseases that were deemed compensable. The list of compensable disorders was based on the best available knowledge in 1990 and was not totally rooted in science; it was modified in 2000 and again in 2002. The most recent regulations from the Department of Justice (DOJ) appeared in mid-2004. Chapter 2 provided details on this legislation.
RESEP was designed to educate and identify individuals who would potentially be compensated under RECA. To receive compensation—assuming that the administrative requirements of documenting the timing of residence and occupational activities were also met—RECA requires an exposed individual to meet specific published criteria for (that is, proof of) the presence of disease. Such proof can rest on a variety of imaging studies, functional studies (such as pulmonary function tests), physician reports, or autopsy reports (see DOJ, 2004). However, these RECA-acceptable proofs would not always establish the presence of disease with certainty.
We consider here several additional epidemiologic, statistical, or clinical issues in screening in the context of the main RECA and RESEP goals. Some of the issues relate to specific diseases and the probability of causation or assigned share (PC/AS—see Chapters 5 and 6); others relate to the information needed to prove the presence of disease.
Base Rates of Disease and Probability of Causation
One problematic issue in requiring exposed individuals to demonstrate the presence of disease is that radiation exposure is only one of several causes of most of these diseases. The RECA standard for compensation for any given individual does not depend on the cause of the disease or even on the probability of such cause (in that individual). Instead, it depends merely on proof that the individual has the disease and is administratively qualified to receive compensation.
Two conditions illustrate the complicated causal pathways that make definitive decisions about causation so difficult for RECA. One is a lifestyle issue: smoking behavior. The other is a highly prevalent, serious condition in some RECA populations: type 2 diabetes.
Lung cancer is the only RECA disease that is compensable in all five RECA populations (miners, millers, ore transporters, onsite participants, and downwinders). As detailed in other chapters, the contribution of radiation exposure to the etiology of lung cancer, now 35 to 60 years after exposure, is likely to be small, even among uranium workers, for whom the case is best made. The contribution is likely to be considerably lower among downwinders and onsite participants.
Tobacco abuse is the most prominent cause of lung cancer. It can be considered a choice that each individual makes. In some ways, then, the less healthy an individual’s lifestyle is, the more likely it is that the individual will have a lung cancer and the more likely that he (or she) will receive compensation. That outcome must be considered a perverse effect of RECA, that is, an effect that would seem contrary to the presumed rationality of RECA.
In a different vein, type 2 diabetes is the most common cause of chronic renal disease in this country. It is highly prevalent among the Native American populations who were frequently uranium workers. Although the renal or other sequelae of diabetes are not a direct consequence of lifestyle choices, they do depend heavily on both self-management and medical care that may or may not be easily accessible for these populations. Chronic renal disease may occur with relatively high frequency in these populations for numerous reasons unrelated to radiation exposure. Hence, decisions about the probability of causation of this compensable disease may be difficult to make.
What Is Disease (RECA Gold Standard)?
RECA provides specific criteria for establishing the presence of disease, as defined by the legislation and published in the Federal Register (DOJ, 2004). Although the criteria have a basis in good medical practice (such as standards issued by the National Cancer Institute), they and the underlying test results in essence define the presence of disease as far as RECA compensation is concerned. Thus, the results of these tests and processes in fact constitute the gold standards for diagnosis in terms of RECA compensation, although not necessarily in terms of medical certainty. In fact, these standards state that if a disease can be established to a reasonable degree of medical certainty (at or above 50%, as discussed below), then the disease is present as far as RECA is concerned.
If one of these RECA gold standards were used as a (the only) screening test, both the sensitivity and the specificity of the compensational screening test would, by definition, be 100%. The problem of the cascade of false positives, discussed in Chapter 9 as an important issue in medical screening, would disappear, at least with respect to compensational screening. However, the patient would still face the cascade of medical false-positives results and the medical risks they engender, perhaps without medical benefit. Identifying a patient as eligible for compensation does not itself guarantee either that the patient has the disease or that the patient’s health will be improved by such identification.
Tiered Requirements for Compensation
The requirements for proof of the presence of disease have both criteria that can be satisfied by meeting any one of several options and criteria that can be met
only by satisfying all of several requirements. We explored those problems in detail in the discussion of “OR” and the “AND” combinations in Chapter 9. OR criteria in effect increase sensitivity measured against the true medical presence of disease as a gold standard, whereas AND criteria decrease sensitivity. The issue of specificity does not arise here because these criteria are the gold standard for compensation.
For purposes of decreasing the likelihood of (perhaps erroneous) compensation, using AND criteria will be preferred. For purposes of increasing the likelihood of (perhaps warranted) compensation, using OR criteria will be preferred. The latter would provide claimants some element of the benefit of the doubt with respect to the presence of a specified disorder.
Medical Certainty in the Context of RECA
One question that needs clarification is how certain is certain enough to permit compensation, especially when the tests that bear on a diagnosis are imperfect. One interesting phrase appears in the statutory sections about establishing the diagnosis of many compensable diseases: “from which the appropriate authorities … can make a diagnosis to a reasonable degree of medical certainty.”
At least in courts considering such torts as medical malpractice, the concept of “reasonable degree of medical certainty” has often, but not consistently, been held to mean “more likely than not,” or at least 51% (Lewin, 1998; Craig, 1999). Some legal experts question the “appropriateness of the phrase [reasonable medical certainty] as a legal standard” and even regard it as “oxymoronic” (Lewin, 1998; pp. 1-2). Nonetheless, the phrase has been in use since a 1916 opinion issued by the Illinois Supreme Court, and it has gradually become a fairly standard, although not sharply defined, term. It has been applied in worker compensation cases (Lewin, 1998). Textbooks about medical malpractice (Gutheil, 1998, p. 6) also point to the “more likely than not” (greater than 50%) criterion.
At times, the concept is rendered as “as likely as not.” The distinction between “as likely as not” (at least 50%) and “more likely than not” (greater than 50%) may be largely semantic, in that in only very few circumstances would a calculated probability be exactly 50.000%. A probability of 50.001% would satisfy either meaning, whereas a probability of 49.999% would satisfy neither meaning—unless, of course, it was rounded up to 50%.
In the RECA-RESEP context, the problem is compounded because medical probabilities are never precise; indeed, probabilities consistently within a few percentage points of the true (but unknown) value might be regarded as unusually precise. If, as in the policies and procedures for the Energy Employees Occupational Illness Compensation Program Act (EEOICPA), the approach underlying RECA is to provide an exposed individual the benefit of the doubt, then the criterion should be “as likely as not,” with reasonable rounding permitted.
Other Sources of Documentation of Disease for Compensation
Two other mechanisms for screening for compensational purposes might be considered: both appear to reduce risks to individuals, and one provides a way to obtain information that could not otherwise be obtained. We first discuss using autopsy data on individuals who die before any screening can take place; the autopsy yields definitive (diagnostic) information. We then consider whole body magnetic resonance imaging.
The RECA Autopsy
For RECA, autopsy reports are acceptable evidence for most diagnoses of compensable disease. That fact suggests several strategies for maximizing the potential for establishing the presence of compensable disease, assuming a postmortem examination is ethically acceptable to the potentially affected individual and his or her family.
If a potentially compensable patient dies, obtaining a routine autopsy might be a reasonable strategy for maximizing the opportunity for the patient’s family to receive compensation. Many families may be reluctant to obtain a postmortem examination; thus, a reasonable step might be for the clinicians caring for the patient to consider discussing this option with the patient and his or her health care proxy or next of kin. This step might be particularly relevant if the patient is chronically or terminally ill and especially if the patient is known to meet the geographic and temporal administrative criteria for compensation. If such an autopsy were approved for the purpose of looking for compensable disease, the pathologist should be so informed and should look specifically for any RECA-compensable diseases. Of course, posthumous establishment of RECA eligibility, whether by autopsy or by a family’s application for posthumous compensation, raises the issue of whether such posthumous eligibility should qualify the individual’s family or estate for retroactive reimbursement for the medical care already received.
In the Radiation-Exposed Veterans Compensation Act (REVCA), “under certain conditions, veterans who are retroactively awarded service-connection may qualify for reimbursement of certain medical expenses back to the date of the original claim filing; a family member could file on behalf of the deceased veteran. These cases would be referred to as “not previously authorized” claims and generally must meet three conditions: (1) treatment was for a service-connected condition or for a condition held to be aggravating an adjudicated service-connected disability; (2) a medical emergency; and (3) Department of Veterans Affairs (VA) or other federal facilities not feasibly available” (information taken from correspondence with Department of Veterans Affairs, March 31, 2005).
Unlike REVCA, under the Energy Employees Occupational Illness Compensation Program Act (EEOICPA), “medical payments are made only to a liv-
ing energy employee if the claim is approved; no reimbursement is made for prior expenses, and no medical payments are made to survivors if the energy employee is deceased” (correspondence with Department of Labor, April 6, 2005)
The committee did not have the opportunity to focus on this conundrum and merely highlights it for careful legislative consideration.
Whole Body Magnetic Resonance Imaging
Although many kinds of imaging studies provide acceptable evidence of specific diseases, many are invasive or involve exposure to radiation. For example, a recent study suggests that the exposure from a routine computed-tomography (CT) body scan carries more risk (from radiation) than the small diagnostic benefit it provides in identifying diseases for which early intervention improves health outcomes (Brenner and Elliston, 2004). Magnetic resonance imaging (MRI), which is not invasive and does not subject patients to ionizing radiation, yields evidence of many diseases that would be acceptable for RECA purposes. That is, “positive” MRI results are acceptable means by which to establish the diagnosis of most RECA-compensable diseases. However, as is the case with most high-sensitivity imaging studies, a whole-body MRI would be associated with an unacceptable number of potentially false-positive test results (for example, incidentalomas), and these would subject the patient to workup and labeling and the risk of physical, social, and emotional consequences.
Nonetheless, if patients were to choose to set aside the issue of costs and the risks of medical, social, and psychologic consequences of false-positive results, they might request to have a “RECA whole-body MRI.” An MRI would not expose the person screened to ionizing radiation. Moreover, it could be considered a gold standard for establishing the presence of RECA-compensable disease insofar as it produces information relevant to the body parts that any RECA-compensable disease might affect, albeit at enormous cost in terms of false-positive results. The feasibility and cost of obtaining an MRI of the entire body must also be questioned.
In short, MRIs might provide both the initial screening information and the diagnostic data relevant to RECA conditions and cause, on balance, less potential harm than the current “screening protocols” impose. If properly explained to the screenee, a negative “RECA whole body MRI” might help reassure patients that they do not have one of the RECA-compensable diseases. Set against this point is that, given the propensity for such an MRI (as an imaging test that identifies multiple diseases) to produce false-positive results, such reassurance for some patients may come at the cost of increased anxiety in many other patients.
Because such a plan would expose the screened individual to the risks of medical false-positive results, workup, and potentially futile treatment, careful, extensive (but culturally sensitive) informed consent is critically important. Regardless of the desirability or feasibility of pursuing such strategies for compen-
sation-oriented screening, the committee reiterates its view of expanding medical screening to compensational screening. Specifically, in considering expansion of compensational screening, the committee recommends that HRSA base decisions about screening primarily for compensation on recommendations drawn from credible scientific evidence that the proposed test provides reliable information about the presence or absence of specified RECA-compensable diseases.
As suggested in Chapter 9, even if screening is performed primarily for purposes for identifying RECA-compensable disease, if such tests reveal any information about the potential presence (or absence) of other diseases, that information should, in general, always be shared with the screenee. In conveying such information, however, clinicians must be sensitive to the cultural traditions and other preferences for individuals to be told bad news or have negative information discussed in their presence (Gostin, 1995).
Considering the Medical Risks Posed by Screening in the Context of Compensation
Another complexity arises in the context of compensational screening. Because each of the screening tests except a postmortem examination places the patient at risk for a false-positive finding or even at risk for a true-positive result that identifies a disease that does not benefit from early intervention, such compensational screening creates some degree of medical risk. The risk may be physical or psychologic (or both). Labeling the patient or even the patient’s family may produce adverse social effects.
Thus, clinicians should obtain truly informed consent about the consequences (medical, psychologic, and social) of compensational screening before undertaking such a program (see Beauchamp and Childress, 2001). Such consent is enhanced through techniques of shared, informed decision-making, perhaps especially for cancer diagnoses (Barry, 2002; Whitney et al., 2003; van Roosmalen et al., 2004; Sheridan et al., 2004).
During such informed consent and shared decision-making activities, the individual to be screened will be weighing the likely small chance of substantial monetary gain without medical gain against a more substantial chance of physical, psychologic, or social harm. Society allows people to gamble with their own health, but such decisions should be well counseled and fully informed. One might consider that undertaking such compensational screening is playing a lottery of sorts, in which one buys his or her ticket of entry not with money but by undertaking risk. In some ways, because such a gamble involves a risk of physical harm (from unnecessary tests and treatments), it could be seen as akin to playing a sort of Russian roulette with a monetary payoff attached to some outcomes.
Untrained people are not likely to understand the meaning of small numbers, in terms of both risk and payoff. Indeed, as elaborated in Chapter 11, many
people have low numeracy skills in general. In this context, ensuring that individuals are informed of the likelihood that they will receive RECA compensation, the likelihood and consequences of false-positive results, and the likely medical benefit of early detection, is essential.
One part of appropriate counseling about compensational screening might be to use a decision-support tool that helps patients to understand their willingness to risk health consequences in exchange for the opportunity for monetary gain. Willingness to risk is, in effect, the obverse of one’s willingness to pay for healthy life years—a concept and measurement method with decades of theory and application (Gold et al., 1996, p28). Developing and standardizing such a decision-support tool might be the type of activity that HRSA could usefully pursue for use by its grantees (or contractors), as outlined in Chapter 11.
Although such numeracy issues are beyond the scope of this report, the committee expresses its support for and concern about thorough discussion with potential screenees of the advantages and drawbacks of compensational screening. Specifically, the committee recommends that any screening carried out under RESEP auspices be preceded by detailed counseling and informed consent that reflects an understanding of and sensitivity to the culture of the potential screenee. When screening is primarily or exclusively for compensation (and when the likelihood of medical benefit is low or absent), a special obligation exists to be as certain as possible that the potential screenee is aware of all the risks and benefits of screening. Although compensational screening is unlikely to provide medical benefit, a successful applicant for compensation might receive such compensation sooner, perhaps while he or she still had the health to enjoy its benefits. Such early compensation might in many ways be analogous to the early treatment that motivates medical screening. Those issues should be made explicit when counseling a potential screenee.
Therefore, the committee also recommends that counselors, when dealing with screening for compensation, ascertain that individuals proposed to be screened fully understand the associated risks, benefits, and likelihood of potential outcomes of screening. Although ascertaining such comprehension is complex, a minimal standard might be to have the patient verbalize those risks and benefits in his or her own words. When the level of understanding seems inadequate, counselors would be expected to engage in a more detailed explanation couched in terms and using examples tailored to the individual’s literacy and numeracy capabilities and taking into account cultural issues. Moreover, if a screening test produces a positive result, even one that suggests the possibility of a disease that is not compensable under RECA, the clinician or counselor should communicate that result to the screenee, in a culturally sensitive manner, and make an appropriate referral for followup.
The drawbacks of compensational screening, especially in the event of low prevalence rates and multiple tests, led the committee to consider the desirability of what is, in effect, a two-stage process. Specifically, the committee recom-
mends that RESEP screening be undertaken only if individuals satisfy administrative criteria for compensation before screening. By satisfying administrative criteria, we mean that patients need to be certain that they are already administratively eligible before any screening takes place. Meeting these criteria can include documenting employment or residence history (as appropriate), and, for cancers, obtaining an individually calculated PC/AS for the cancer being proposed for screening.
Tying Screening to Probability of Causation and Assigned Share
Criteria for Probability of Causation (Assigned Share)
In addition to considering the medical certainty in establishing the presence of a RECA-compensable disease, one must take into account the administrative requirements for eligibility for compensation. In Chapters 5 and 6, the committee proposed that Congress amend RECA by requiring the use of a calculated PC/AS for each exposed individual; in determining compensation, that value and its uncertainty (credibility interval) would be compared to the criteria that Congress establishes for a cutoff. As discussed by several experts (Lagakos and Mosteller, 1986; Robins and Greenland, 1991), such calculations may introduce a bias against individuals harmed by their exposure to ionizing radiation because such exposure can be a contributing cause of disease without being the predominant cause of disease. As a case in point: if an individual so exposed to fallout were to develop a thyroid cancer but, based on her residence history and age at exposure, her calculated PC/AS were only 35%, then exposure to ionizing radiation from fallout could have been an important factor in her cancer. Nonetheless, if the rules for compensation were to be set at a PC/AS of at least 50% (as likely as not), then the contribution of the fallout would go unrecognized and uncompensated. Of course, this shortcoming might be managed by setting the cutoff criterion for PC/AS at some value below 50%, by employing credibility limits around PC/AS estimates, or both.
As discussed in Chapter 5, we relied on contemporary scientific evidence and reasoning in recommending that RECA be amended to use a PC/AS approach. We did not, however, find a scientific basis for establishing a particular threshold value of PC/AS or a specific allowable uncertainty in an individually calculated PC/AS that should be used for determining administrative eligibility for compensation. Calculating an individual PC/AS and comparing its distribution with such threshold criteria are administrative activities that are predicated on the presumed presence of a RECA-compensable disease (see also Chapter 11).
Making such a calculation, however, does not require that screening or diagnostic tests already have been performed in that individual. The individual will need to identify the disease or diseases for which he or she wishes to be screened for (or obtain) the PC/AS calculation to be performed. Thus, calculating a PC/AS
and making a judgment about administrative eligibility do not require that any patient first be screened and (thereby) subjected to the costs and risks of such screening tests.
Clearly, a RECA compensational screening program could look for the presence of any or all of the RECA-compensable diseases. Which tests or combinations of tests to perform should involve the patient’s informed consent about risks and the often-small likelihood of benefits (primarily from compensation) of screening. The patient’s choice of the diseases for which to seek compensation might reflect several factors: the preassessment of the maximum reasonably expected PC/AS in a population or geographic region (as called for in the explication of the committee’s first recommendation); the likelihood that such a disease is present in a particular patient; the calculated PC/AS value for that patient; and the patient’s desire to receive information about ailments about which he or she may be anxious.
Using the same threshold criterion for PC/AS as for establishing the diagnosis of a RECA-compensable disease may be tempting. Doing so requires interpreting the criterion; for example, it might be taken to mean “with reasonable medical certainty”—either “more likely than not” or “as likely as not.” Because calculating PC/AS and making a diagnosis are distinct issues, however, the criteria (thresholds or cutpoints) may well be different. Currently, RECA does not use the PC/AS approach (although it had been proposed before RECA for other compensation programs), so it is silent on what the criterion or cutpoint should be.
Apart from a point estimate for a threshold is the question of uncertainty around the point estimate. For example, although EEOICPA uses a “more likely than not” criterion, it gives the exposed individual an enormous benefit of the doubt by requiring that the person have only at least a 1% chance that the true PC/AS exceeds 50%. That is, EEOICPA, like some other compensation programs, compensates on the basis of an upper credibility limit, analogous to the upper bound of a confidence interval, set in the EEOICPA case at 99%. Currently, when applicable in REVCA, radiation doses and information in the 1985 NIH radioepidemiological tables from the National Institutes of Health are used to calculate a PC using the version of the Interactive Radioepidemiological Program (IREP) software from the National Institute of Occupational Safety and Health. The PC is used in parallel with consideration of other factors, to make the determination of eligibility for compensation (correspondence with Department of Veterans Affairs, March 31, 2005).
As explained in Chapter 5, Congress might consider two different criteria in determining who might be due compensation. The first is the value of a specific threshold or cutoff of PC/AS above which compensation can be awarded (such as 50% for EEOICPA), but perhaps 30% or 10% or any other value (below 50%). The second is the use of an upper credibility limit such that the value of some specified upper percentile of the PC/AS uncertainty distribution is compared to
the designated threshold. Compensation would be awarded when the value of the upper credibility limit exceeded the threshold.
With some knowledge or expectation about the distribution of PC/AS values and their associated uncertainties, Congress should be able to determine how the number of individuals compensated and the cost of the program (both RECA and RESEP) might vary with different criteria. The committee did not have the time or resources to examine the cost issues in detail.
From the perspective of populations now receiving compensation or possibly eligible for compensation, it is instructive to note that many individuals would have calculated PC/AS values well below a 50% threshold. Whether the upper percentiles of their uncertainty distributions might fall above the 50% threshold is less clear.
Should the Calculated PC/AS Affect How Screening Tests Are Interpreted?
As discussed in Chapter 9 and Appendix D, many medical tests (both screening and diagnostic) can have sensitivities and specificities that vary depending on the criterion of positivity chosen in a particular circumstance. By analyzing a test’s receiver operating characteristic (ROC) curve with knowledge of the prevalence of disease in the population to be tested and the burdens of false-positive and false-negative classifications that the test results may impose, one can determine the optimal operating point on the ROC curve and its corresponding cutoff criterion.
That principle is often applied to screening. If the population to be screened has a low disease prevalence or if the burden of missing a disease is low, one uses a strict criterion of positivity. If the population to be screened has a high disease prevalence or if the burden of missing disease is high, one uses a less strict (more lax) criterion of positivity. Similarly, in the context of RECA and RESEP (after administrative eligibility has been established as suggested earlier), the criterion of positivity for the screening tests performed to identify compensable disease could be adjusted according to the prevalence of disease in the population. The burden of a false-negative result will be largely the denial of compensation (and potential emotional effects of such denial) because, as stated previously, most of these screening tests will not produce any improvement in health outcomes. The burden of a false-positive test, however, will be the physical, psychologic, and social risks described in Chapter 9.
A more difficult question—which may not even be practical to address—is whether the criterion of positivity of a medical screening test should be adjusted to reflect the individually calculated value of PC/AS or its uncertainty. If one is compensating for radiogenic disease, the calculated value of PC/AS might represent the pretest likelihood (for the screening test) of a specific disease caused by radiation in a given individual (such as thyroid cancer). If that is the case, an
individual with a higher PC/AS might logically be screened with a more sensitive test—for example, thyroid ultrasonography with a cutoff for nodule size set at 2 mm rather than 5 mm.
One might take the perspective that a disease should be worthy of compensation only if it has clinical effects. In that case, a tiny or inconsequential cancer would not be compensable. However, currently RECA apparently does not adopt this perspective. RECA compensates for the presence of pathologically or clinically established disease, even if that disease has no bearing on the patient’s prognosis. For example, the family of a patient with an asymptomatic pulmonary nodule, shown to be cancer, will be compensated even if the patient is struck and killed by a car, so that the cancer had no effect on his survival and produced no morbidity.
In addition to the challenging mathematics of how one should adjust the pretest probability of “disease” on the basis of PC/AS (although adjustments in the odds domain seem attractive at first blush), at least two practical challenges arise. First, in a screening program, how does one manage the logistics of adjusting test interpretation on an individual basis? Second, would using a more sensitive criterion change the apparent base rate of disease in the population, and would that affect the excess relative risk and the calculated PC/AS? Notwithstanding those challenges, whether and how to think about adjustments to a PC/AS approach warrant further examination.
Many patients dealing with the tradeoffs among medical risks, health benefits, and potential monetary compensation will require decision-support tools. Broadly speaking, such tools embody ways for patients and clinicians to share their decision-making logic. Other intellectual and emotional supports are also important.
We are not aware of proven models of shared decision-making protocols that are adequately designed to help people to manage those difficult tradeoffs in the RECA-RESEP context. Some tools have been developed for behavioral and decision-making research (for example, for prostate cancer screening decisions). The entire field of shared decision making and the tools to support it are important topics for future research, both in the domain of collaborative communication and in the domain of explicit mathematical modeling.
Thus, the committee recommends that the Department of Health and Human Services support development of explicit decision models and approaches to shared decision making and related tools that enhance the ability of patients to participate in decisions that affect their care and prognosis. We recommend, in particular, that HRSA take responsibility for similar activities in the domain of compensational screening.
In addition, the committee’s proposed strategy for using individually calculated PC/AS values for RECA eligibility determinations will require tools to make such calculations more convenient. The committee also calls for further exploration of the benefits, limitations, potential biases, and necessary extensions of the PC/AS approach.
HEALTH-CARE ISSUES BEYOND SCREENING
Followup Diagnostic and Therapeutic Services
One question that arises for of any medical screening situation and presumably for RECA and RESEP even for compensational screening concerns the obligations to patients inherent in the programs themselves and in the agency roles of health care providers. When applicants or patients are found, through RESEP screening, to be at special risk for a particular RECA-compensable condition, the expectation (and apparently generally the practice) is to refer them for appropriate diagnostic followup and therapy. When applicants or patients are found, through similar screening, to be at special risk for a noncompensable disease, those patients should be advised to get appropriate medical management, independent of its possible RECA connection.
For that process to be effective, health care practitioners have to be informed and stay abreast of knowledge in several domains. These include the special risks known to be associated with mining and related activities and health care issues that may be related to radiation doses received by covered individuals at special risk (for example, children in selected areas who drank large amounts or milk or consumed foodstuffs contaminated with iodine-131). In addition, if Congress amends RECA to adopt a PC/AS strategy for populations, geographic areas, and diseases not now covered, clinicians need to be knowledgeable about the method and the implications for their patients. The committee urges HRSA and appropriate professional societies and other groups to assist in educational activities for health care professionals related to these topics. We return to these matters in Chapter 11.
In Chapter 8, we noted a distinction between the duty to restore a loss to health and a duty to compensate for the effects of the loss. These are distinct duties; compensation for the effect of the loss does not fulfill a duty to ameliorate the loss of health by providing health care.
The remedy for the loss sustained is amelioration or restoration. Medical services are probably the only, if not the best, manner in which to provide restoration. Because of the nature of the health loss, in many cases discovering the
loss may be possible but restoring health impossible. In other cases, some measure of restoration may be possible through medical services. Services begin with discovery through first evaluation and diagnosis and proceed to referral, workup, and care that are medically indicated (Sheridan et al., 2004).
The committee received many written and oral statements, both at its information-gathering meetings and while reviewing the programs of HRSA grantees, that many RECA applicants or patients are uninsured or underinsured for the costs of followup diagnostic or therapeutic interventions. Other populations covered by radiation-compensation programs (REVCA and EEIOCPA) provide medical services; uranium workers covered under RECA are also covered under EEIOCPA for medical expenses.
Hence, for consistency with other similar compensation programs and to overcome any inequities, the committee believes that similar relief is owed to RECA population, and, therefore, the committee recommends that, if an individual has established eligibility for compensation, RECA cover the costs of screening, complications of screening, referrals (followup), diagnosis (workup), and treatment for the RECA-compensable diseases for which such eligibility has been established. As stated earlier in this chapter, establishing such eligibility for people exposed to fallout from the US nuclear-weapons program may include obtaining an individually calculated PC/AS for the cancer being proposed for screening.
Screening in an Era of Scarce Resources
Apart from the issues discussed to this point about compensational screening, the cost of expanding RECA is a matter raised by some committee members, by some people who testified at the committee’s information-gathering meetings, and by some reviewers of this report. The fundamental question is appropriate allocation of social resources, especially when those funds are limited. Some commentators posed the question in terms of expanding RESEP, in a (present) configuration that focuses so heavily on compensational screening (as well as outreach and education), to a national program. The basic problem is whether resources so directed, if now expanded to cover the country, would be a cost-effective use of HRSA appropriations. More importantly, one must ask whether expanding RECA to additional populations, geographies, or diseases is a potential threat to the “medical commons” of societally shared limited resources (Hiatt, 1975). The committee did not pursue these matters in depth; we view them as beyond both our charge and our time constraints. We note, however, that these matters are subjects that will warrant attention in the larger social debates about the changes in RECA suggested in our report.
This chapter has examined the conceptual and practical elements of screening for compensational purposes. We have linked these issues to the committee’s adoption of a PC/AS model and we have noted some clinical and epidemiologic issues related to compensable diseases. Some of our recommendations are directed at agents that would need to amend RECA (or RESEP) legislation; others are directed at steps that HRSA can consider without statutory changes. The next chapter addresses on the remaining core elements of the RESEP program: education and outreach.