|
||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||
|
Public Attitudes About Genetic Testing For Alzheimers Disease
In a general population survey (N = 314), 79 percent of respondents stated that they would take a hypothetical genetic test to predict whether they will eventually develop Alzheimers disease. The proportion fell to 45 percent for a "partially predictive" test (which had a one in ten chance of being incorrect). Inclination to obtain testing was similar across age groups. Respondents were willing to pay $324 for the completely predictive test. Respondents stated that if they tested positive, they would sign advance directives (84 percent), get their finances in order (74 percent), and purchase long-term care insurance (69 percent). Only a third of respondents expressed concern about confidentiality. The results suggest that people value genetic testing for personal and financial reasons, but they also underscore the need to counsel potential recipients carefully about the accuracy and implications of test information.
Media reports about the genetic underpinnings of Alzheimers disease, as well as publicity surrounding the effective completion of sequencing the human genome, have heightened hopes for genetic testing and therapy. In reality, the genetic foundations of Alzheimers disease are complex. Advances in genomics may in time permit accurate tests for genetic predisposition and treatment. Preventive programs tailored to persons based on genomic analyses may emerge but are still many years away. Some patients are already demanding testing to predict if they carry the "Alzheimers gene."1 The anticipation of genetic testing for Alzheimers and other diseases raises a host of policy questions, with large social, economic, and ethical ramifications. Early diagnosis could offer patients and families more time for planning for future care and financing. On the other hand, it could lead to increased anxiety or false reassurance if tests are inaccurate or if results are misinterpreted. Widespread testing also could greatly increase health costs and could lead to job or insurance discrimination. This paper reports on a general population survey of attitudes toward genetic testing for Alzheimers disease. We examine five questions: (1) What share of people would take a genetic test for Alzheimers disease? (2) How much does the choice depend on the accuracy of the test? (3) How much would people pay for the information? (4) What would people do with information from a positive test? (5) How worried are people about confidentiality? Genetic aspects of Alzheimers disease. Knowledge about genetic aspects of Alzheimers disease has advanced considerably in the past decade. Family members with early-onset dementia (under age sixty) have led to the discovery of three genes in which mutations cause Alzheimers disease.2 Other than its early onset, this genetic form of Alzheimers disease is nearly indistinguishable from sporadic disease. While all known mutations in the three genes can cause early-onset Alzheimers disease, not all early-onset disease has been linked to mutations. Also, mutations are rare, likely representing fewer than 1 percent of Alzheimers disease cases.3 A second means by which genetic factors influence disease expression is by enhancing the risk for late-onset disease. The apolipoprotein E4 allele (APOE*E4), which has received attention as a genetic risk factor, increases the risk of developing Alzheimers disease. Apolipoprotein E is a monomeric glycoprotein originally identified as an important intermediary in lipid metabolism. The APOE gene, on chromosome 19, has three well-studied alleles: APOE*E2, APOE*E3, and APOE*E4. Several studies have shown an increased frequency of the APOE*E4 allele in persons with Alzheimers disease compared with controls.4 But not all persons with APOE*E4 alleles will develop Alzheimers disease, nor do all persons with Alzheimers disease carry an APOE*E4 allele. Therefore, APOE is neither necessary nor sufficient for the development of Alzheimers disease. There is also an increased risk of developing Alzheimers disease associated with increasing copies of the APOE*E4 allele. However, the relatively low frequency of the allele in the general and Alzheimers disease populations limits the utility of APOE genotype testing. Based on published data, the presence or absence of the APOE*E4 allele is not very useful to distinguish Alzheimers disease from other causes of dementia, given its low sensitivity and specificity.5 Recommendations from leaders in the field state that APOE genotype testing not be ordered to diagnose Alzheimers disease in persons with dementia or possible dementia.6 They also state that it should not be ordered as a screening test to predict Alzheimers disease in asymptomatic persons, although they state that research on APOE genotype testing is appropriate. Other genetic risk factors are beginning to emerge; they require further evaluation.7 Public attitudes about and understanding of genetic testing. Relatively little is known about attitudes regarding genetic testing for Alzheimers disease. One study of a convenience sample of cognitively normal persons found that 69 percent would obtain predictive testing; the desire for testing did not vary with age, sex, race, or income but was higher among those with a family history of Alzheimers disease.8 Studies of family members of Alzheimers patients have revealed a high interest in predictive testing, although other studies have found that many persons, both at risk and not, would not favor testing, because of concerns about the effect of unfavorable results on children, spouses, or themselves.9 Researchers have found that most older persons would want to be told of a diagnosis of Alzheimers disease.10 Those wanting to know were more likely to know someone who had dementia. Reasons for wanting to know the diagnosis included financial planning (84.7 percent) and settling family matters (76.6 percent). Some research on genetic testing for cancer has found strong support for presymptomatic genetic testing among at-risk persons.11 Others have reported that many persons at low risk of hereditary cancer would seek testing as a means of obtaining reassurance.12 Perceived risk of getting cancer has been found to be predictive of who obtains or states an intention to get genetic testing.13 However, a body of studies underscore the confusion and anxiety surrounding genetic testing.14
Survey design. A telephone survey was administered to a random sample of U.S. adults in FebruaryMarch 2001. The survey instrument was refined based on 200 respondents at a public festival in Boston. The instrument briefly described Alzheimers disease, noting that most people who get Alzheimers are older than age sixty-five and that the prevalence rate is one in ten Americans. The instrument also stated that there is no prevention or cure. Respondents were then asked about their willingness to take a blood test that would tell them now whether they would one day develop Alzheimers disease. Respondents maximum willingness to pay for the test was elicited for those willing to take the test. Willingness to pay is a standard measure of how people value health interventions, consistent with individual preferences and welfare economics.15 The willingness-to-pay questions followed a binary (yes/no) bidding game format, with respondents randomized to one of four initial amounts: $100, $500, $1,000, and $1,500. If respondents answered yes to the initial bid, they were asked whether they would pay double that amount. Those answering no to the initial bid were asked whether they would pay half. The double-bounded, dichotomous-choice format permits an analysis of the willingness-to-pay data as a survival analysis with censoring.16 Respondents also were asked if they would take an "imperfect" or partially predictive blood test to predict Alzheimers disease. They were told to assume that there was a one in ten chance that they would not get the disease, despite a positive test, and that there was a one in ten chance that they would get it when the test said they would not. Willingness to pay for this test was also assessed with respondents randomized to $50, $200, $400, or $800.17 Finally, respondents were asked what they would do with information from a positive test, whom they would tell, how worried they would be about others gaining access to test information, and a series of background questions about whether they had a family history of Alzheimers disease and whether they had ever personally cared for someone with the disease. Study sample. Data were collected by CODA Inc., a survey research firm with expertise in health-related data collection. Potential respondents were identified using random-digit-dialing techniques. Contact with an adult in the household was obtained during the first call, and the potential survey respondent was identified as the household resident age eighteen or older, with the most recent birthday. The survey was administered during the initial call if possible, or at a later time when the identified adult respondent was available. Respondents were offered $5 as an incentive for their participation. Information was obtained from 314 interviews47 percent of eligible households. The lower-than-desired response rate largely reflects difficulties making initial contacts with persons, rather than outright refusals to participate, a common problem in random-digit-dialed surveys.18
The mean age of respondents was 43.3 years. Respondents were predominantly female; white, non-Hispanic; and in excellent or very good health (Exhibit 1
Seventy-nine percent of respondents stated that they would take the genetic test for Alzheimers disease (Exhibit 2
Inclination to take either test was fairly constant across sex, education, race, and health status categories (Exhibit 3
Mean willingness to pay was $324 for the fully predictive test and $170 for the partially predictive one; those not willing to take the test were assumed to have zero willingness to pay. Mean willingness to pay for the 116 respondents willing to take both tests was $871 for the perfect test and $461 for the imperfect test.21 Respondents stating that they would not want to take the test cited living with the burden of the disease (68.5 percent) and the fact that there were no preventive treatments (61.1 percent) as their main concerns.22 Only 30.4 percent of negative respondents noted privacy/confidentiality as an important concern.
When asked what they would do with information from a positive test, a majority of respondents (who were allowed to choose one or more options from a list) stated that they would sign advance directives, spend more time with family, get their finances in order, and/or buy long-term care insurance (Exhibit 4
Most respondents to our survey were prepared to learn whether they carried a gene for Alzheimers disease, even though there are no preventive or curative treatments now available. Inclination to take the completely predictive test was generally strong across age, income, and racial categories. Almost half wanted the partially predictive test, despite a one in ten chance that test results would be incorrect. Mean willingness to pay for the entire sample for the completely and partially predictive tests was substantial. The data suggest that people would value this knowledge for both personal and economic reasons. Respondents stated, for example, that they would spend more time with family and organize their finances. Moreover, a majority stated that they would obtain advance directives or purchase long-term care insurance. Importantly, though, many fewer people wanted the partially predictive test. The results have implications for a number of key decision makers. Implications for individuals and families. For individuals, the availability of genetic testing will inevitably create new choices and dilemmas. Our data suggest that people may have strong personal preferences for testing information, although not necessarily for immediate health or medical reasons. Inclination to take the genetic test was strongest among those stating a family history of Alzheimers disease or experience caring for someone with the illness; this suggests that more knowledge may increase demand for testing. Relatively few expressed concerns over others obtaining access to their test information. This may indicate a perception that test information about a geriatric disease is less likely to affect employment or insurance. But the data also suggest the value of counseling potential test recipients. While a large majority of respondents wanted the completely predictive test, many more declined testing after being given information about the possibility of inaccurate results. Research on preferences for genetic testing for Huntingtons disease yielded similar findings.23 The results highlight the importance of having a genetic counselor involved in the process before and after testing. Implications for clinicians. For health professionals, the data raise questions about patients access to information. Clinicians have actively debated whether to inform patients about a diagnosis of Alzheimers disease, especially given that patients may not be able to comprehend or remember the information.24 Current guidelines recommend that patients be told about an Alzheimers diagnosis, although many clinicians do not follow this practice.25 Telling people about their genetic predisposition for Alzheimers disease raises additional questions. Individuals rights to know about their genetic composition must be weighed against concerns that they will misinterpret the data or be unduly burdened or stigmatized, given the low sensitivity and specificity of existing tests and the lack of preventive or curative measures. As noted, leaders in the field currently recommend against APOE genotype testing for these reasons.26 Our data generally support these recommendations, particularly since the predictive values of the hypothetical tests in our survey are higher than would be expected with currently available tests. But the results also suggest that more work is needed in this area. Importantly, our data suggest that genetic testing could lead more people to obtain advance directives and long-term care insurance. Testing also could encourage people to engage in behavior that promotes cognitive vitality in aging, including exercise, social engagement, stress reduction, and proper nutrition.27 Cost and access. Widespread testing for Alzheimers and other diseases could raise health costs. Questions will increasingly emerge about optimal strategies for testing, given the costs and predictive values of tests, the availability of treatments, and individual preferences. Formal cost-effectiveness analyses of testing and treatment options would be helpful.28 The high rates of caregiving in our sample also highlight the societal impact of Alzheimers disease on individuals, families, and employers.29 Discrimination and stigma. The results also have implications for ongoing debates about genetic discrimination. Conceivably, ones decision about whether to obtain a genetic test for Alzheimers disease could be influenced by concerns about the general stigma associated with the disease or by anxiety over ones ability to obtain health insurance, life insurance, or employment.30 Recent legislation and administrative rulings have begun to address the issue. At the federal level, for example, the Health Insurance Portability and Accountability Act (HIPAA) of 1996 prohibits insurers from considering a genetic trait without the manifestation of disease as a preexisting condition for coverage purposes.31 A year 2000 executive order signed by President Bill Clinton prohibits federal agencies from using genetic information in hiring or promotion of employees.32 Most states have passed legislation that limits genetic-discrimination in employment, and several have passed genetic privacy laws that restrict disclosure of, and access to, genetic information without informed consent.33 However, enforcement is uneven and the extent to which the laws influence behavior is not clear. Study limitations. The results presented here should be interpreted with caution for a number of reasons. First, the relatively low response rate to our survey (47 percent) raises questions about representativeness. In particular, persons who chose to participate may be more familiar with Alzheimers disease than is the general population. If true, this could have inflated our estimates regarding respondents inclination to take the genetic test, as well as their willingness to pay for test information, and could have affected their conclusions about what they would do with test information. The fact that 24 percent said that they had cared for someone with Alzheimers disease may suggest a possible response bias, because the number seems high, given the prevalence of the disease in the population (68 percent of persons over age sixty-five). But the estimate may simply reveal heightened awareness of and experience with the disease and may itself be inflated because people may mistake other forms of dementia for Alzheimers disease.34 In addition, the caregiver estimate in our sample may include caregiving by respondents who are health professionals. Furthermore, it is important to note that even if knowledge about the disease in our sample is somewhat higher than average, inclination to receive testing was not substantially higher among those with family history or caregiving experience, compared with those without, and differences were not significant. Another concern relates to the hypothetical nature of the questions. While people may claim on surveys that they would obtain a genetic test, in reality they may not actually do so; in similar fashion, respondents who stated that they would purchase long-term care insurance might not follow through with this action. There are also concerns about the effect of the framing of questions, particularly with respect to how responses might vary with alternative wording about the accuracy of the partially predictive test, and with alternative sequences of questions. Finally, the predictive values of the tests stated in our survey were higher than would be expected with existing genetic testing. For the sake of simplicity, we told respondents to assume that the partially predictive test had a one in ten chance of being incorrect. Given the low sensitivity and specificity of currently available APOE tests, actual predictive rates would be much lower. We thus recommend that caution be exercised in interpreting our results and that this study be viewed as exploratory in nature. As noted, little is known about public preferences toward genetic testing for Alzheimers disease. This study contributes new information to the debate, but more research is needed in many areas.
Peter Neumann is assistant professor of policy and decision sciences and deputy director of the Program on the Economic Evaluation of Medical Technology at the Harvard School of Public Health in Boston, Massachusetts. James Hammitt is associate professor of economics and decision sciences, Center for Risk Analysis, Harvard School of Public Health. Curt Mueller is a senior research director at the Project HOPE Center for Health Affairs in Bethesda, Maryland. Howard Fillit is executive director of the Institute for the Study of Aging in New York City. Jerrold Hill is director of outcomes research there. Nii Tetteh is a student at Harvard Medical School, in the Harvard/MIT Division of Health Sciences and Technology. Ken Kosik is professor of neurology and neuroscience, Brigham and Womens Hospital and Harvard Medical School. This project was funded by a grant from the Institute for the Study of Aging. The authors are grateful to Richard Chapman and Ying Zhou for statistical and programming assistance.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||