| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Frederick J. Kier, PhD, VA Pittsburgh Healthcare System, Highland Drive Division (116 A-H), 7180 Highland Drive, Pittsburgh, PA 15206. E-mail: frederick.kier{at}med.va.gov
 |
Abstract |
|---|
 TOP Abstract Diagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
Recently, a home screening test for AD became available for purchase by the general public. The Early Alert Alzheimer's Home Screening Test (AHST) costs roughly $20.00 (U.S.) and is a variant of the Smell Identification Test (SIT), also known as the University of Pennsylvania Smell Identification Test (Doty, Shaman, & Dann, 1984). The SIT is considered to be an excellent test of odor identifications when administered properly by trained researchers and professionals (Thompson, Knee, & Golden, 1998), and numerous studies have discovered a connection between deficits in odor identification, as indicated by tests such as the SIT, and AD (see Thompson, Knee, & Golden, 1998, for a review). While the SIT consists of 40 items, the AHST is a briefer, 12-item test apparently derived from a shortened version of the SIT, the Cross-Cultural Smell Identification Test (CC-SIT; Doty, Marcus, & Lee, 1996), for which much less validity and reliability information is available. The AHST consists of a small box with a test booklet consisting of 12 scratch-and-sniff style microencapsulated scents, an instruction booklet, a physicians' information card (that a person with a positive result is instructed to take to their doctor), a small pencil, and an answer key at the end. The test is available at many drugstores nationwide.
There are considerable issues and considerations in regards to this test. There are general ethical questions concerning at-home testing of psychological and medical conditions with life-threatening or, more commonly in the case of AD, quality-of-life damaging effects. There are also specific concerns regarding the AHST in terms of its accuracy. First, we will discuss the specific concerns regarding the AHST, followed by a general discussion of the ethics of psychological do-it-yourself testing.
|
Diagnostic Utility of Smell Identification Tests for Diagnosing AD |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
Research also indicates that olfactory impairment is not just restricted to those with pathology. One study found that 34% of a sample consisting of nonpsychotic relatives of schizophrenics score in the impaired range on the SIT (Kopala et al., 2001). A study using a modified SIT on 510 older Chinese (Liu, Wang, Lin, & Lin, 1996) yielded an average level of performance of 46% correct. It should be noted that the sample was largely illiterate and poorly educated. This may indicate that the test has some multicultural and educational biases as well, although relatively poor test reliability and/or administration may have played a role in this particular study. A recent large sample study of olfactory impairment in older adults discovered a high incidence (24.5%) of olfactory impairment in the older general public. Age, sex, current smoking, nasal congestion or upper respiratory infection, stroke, and epilepsy were associated with olfactory impairment (Murphy et al., 2002). Ship, Pearson, Cruise, Brant, and Metter (1996) looked at SIT performance, over a 3-year period, in a sample of 161 people 19–95 years of age. They discovered a general decline of SIT scores with age. Most notably, they found that, for those aged 80 and older, a 1-point drop in SIT scores was typical, regardless of what medical problems the person had, and that women scored better than men did, indicating a possible gender bias. Structural equation modeling conducted by Dulay and Murphy (2002) supports the finding that olfactory problems are associated with the natural aging process. Doty (2001) notes that decrements in olfactory function because of smoking can take years to recover, and that olfactory function is compromised in urban residents and workers in the paper and chemical manufacturing industries. In a review of the literature on olfaction in persons with AD, Thompson, Knee, and Golden (1998) mention that respiratory illnesses and medication use can also lead to poor olfactory functioning, producing inaccurate results on odor identification tests such as the SIT.
Thus, a broad review of the literature renders the conclusion that poor performance on the SIT is not restricted to those with just AD (or those who are likely to get AD). There appear to be a large number of psychological, neuropsychological, medical, and social and/or demographic factors that can lead to severe olfactory deficits. Some lifestyle habits rather common in the older population, such as alcoholism and smoking (Adams & Cox, 1997; Lisansky, 1999), are also linked with olfactory dysfunction. The first, and possibly most significant, specific problem with the AHST is that a large number of variables, besides AD, can lead to a positive result.
It should be mentioned that many of the studies cited above used the SIT, not the AHST that is now available to the general public (nor the CC-SIT from which AHST was immediately derived). Although the SIT parent test may be useful in detecting particular olfactory deficits, there appear to be no published studies looking at the reliability and validity of the in-home AHST test, and only a few examining the CC-SIT. In the initial standardization study, Doty, Marcus, and Lee (1996) administered the CC-SIT to 198 individuals spanning the age range from young children to very old adults. They found that the pattern of test scores for the CC-SIT did not differ from the SIT's large normative sample; therefore, they used SIT norms for the development of the CC-SIT norms. Unfortunately, no details are presented on how the CC-SIT was self-administered, and no validity studies are presented to indicate whether or not the CC-SIT is comparable to the SIT in differentiating varied age or diagnostic groups. In a large epidemiological study, Graves and colleagues (1999) evaluated 1604 mostly Japanese American older adults in the community using the CC-SIT. They found that twice as many subjects who were identified as anosmic at baseline showed cognitive decline 2 years later. However, trained interviewers apparently administered the CC-SIT (it was not self-administered as in the AHST) and this study did not investigate whether the cognitive decline was associated with AD. The authors conclude that the CC-SIT may be sensitive to AD diagnosis, but recommend that such findings be supplemented with neuropsychological testing or APOE genotyping to make olfactory tests more specific for AD diagnosis.
There are various factors that may make the AHST a far less reliable and valid test than the SIT and/or the CC-SIT. Given the tendency for older adults to be more anxious in unfamiliar evaluative situations (Poon et al., 1986), aged users of the AHST may not self-administer the test correctly, or become flustered when concerned loved ones try to have them respond to such a novel task. Errors may reflect such test anxiety rather than AD, thus causing needless worry among the individual and family members. The answer key is readily available in the back of the booklet, making cheating possible. In addition, the SIT is a 40-item test, whereas the AHST has only 12 items. The CC-SIT is comparable in length to the AHST, but as aforementioned there are no validity studies differentiating AD subjects' performance on the CC-SIT from other groups. Thus, any reliability and/or validity data from previous studies may not translate to the AHST. During development of the SIT, Doty and colleagues (1984), using scores from a large standardization sample, found a 6-month test–retest reliability of.918. However, length of a test influences reliability. The smaller 12-item test may not perform at the same level as the 40-item version, although a study looking at 10-, 20-, and 30- item fractions of the SIT showed very high internal consistency (Doty, Frye, & Agrawal, 1989). The initial standardization study for the CC-SIT yielded a modest test–retest reliability quotient of.71 (Doty et al., 1996).
It is also important to keep in mind that the sample used affects reliability and validity measures, because it is the test scores, not the test itself, that is analyzed (Thompson, 1994). The original development and standardization sample of the SIT was conducted using samples of relatively young individuals (mean ages of 24.87, 35.24, 33.69, 44.13, 42.41 for each of the five research samples; Doty, Shaman, & Dann, 1984). This raises another issue relating to the predictive validity of the SIT and related tests, such as the AHST. These tests do not appear to have been standardized on a sample of those with AD, or a sample exclusively made up of older adults. Tests that are developed and/or validated on one clinical and/or community sample, such as the SIT, may not generalize to a different sample.
Devanand and colleagues (2000) evaluated the utility of the 40-item SIT in predicting AD in 77 patients experiencing mild cognitive deficits. Although the authors found that the SIT was useful in predicting AD, the question of a good cut-off score for the test was raised. A cut-off score of 34 produced a sensitivity of 88.9%, but a specificity of 50%. A cut-off score of 33 produced a sensitivity of 77.8% and specificity of 59.1%, and a 32 cut-off score yielded a 61.1% sensitivity and 68.2% specificity. Thus, although the test was somewhat predictive, its sensitivity and specificity were not outstanding, and the question of an adequate cut-off score remained.
There is also debate over whether olfactory deficits are the primary cause of errors when a cognitively impaired individual has trouble identifying specific smells. Nagy and Loveland (2001) note that mistakes could be caused by semantic categorization problems rather than olfactory impairment per se. If this distinction is borne out, they suggest that pure categorization tests would be more appropriate for screening. Finally, as discussed earlier in this section, the AHST can yield positive results for many other ailments, yet it is clearly marketed as an Alzheimer's screening test. This might be considered an exaggeration of the power and usefulness of the AHST. Is there really any data to suggest that it can distinguish between delirium and dementia, AD and Vascular Dementia, let alone other less common causes of cognitive impairment such as nutritional deficiencies or alcohol dementia?
|
The "Fine Print" |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
Only once the test is purchased, and the AHST kit opened, are the limitations of the test made apparent to the buyer. This is a questionable ethical practice because truly informed consent prescribes that prior to purchase the consumer must be made aware of the limitations of the product or procedure, whether it be therapy or testing. A pamphlet entitled "Read This First!" is enclosed in the box and explains that "Failing this test does not necessarily mean you have AD," and "If you fail this screening test you should consult your physician. If you do not fail this test, and are experiencing symptoms of AD (see signs and symptoms), contact your physician." This last statement strongly illustrates the limitations of the test. Most individuals probably would not buy this test if they thought that they were not having memory problems (i.e., a symptom of AD). In this case, whether they fail or pass, they should see their physician.
Even if they consult their doctor, many physicians, particularly those in primary care, may not have training in identification and differential diagnosis of dementing illnesses. In a review and guide for primary care physicians, Geldmacher (2002) states "Physicians often express a lack of clarity regarding what to do to diagnose dementia" (p. 68). In a study looking at generalists versus geriatric health care specialists, it was found that many generalists lacked even basic knowledge of dementia (Barrett, Haley, Harrell, & Powers, 1997). For example, only 40% of generalists, compared to 97% of geriatric specialists, knew that AD was the most common cause of memory loss in older adults (Barrett et al., 1997).
Clinicians trained in geriatrics appear to be a much better referral choice, with one study showing several statistically significant differences relating to quality of care between AD patients seen by their primary care physicians versus those seen by a geriatric psychiatrist (Aupperle & Coyne, 2000). Lichtenberg (1998) argues that neuropsychological evaluation should be integrated into primary care to ensure accurate diagnosis and aftercare, an argument bolstered by a recent paper by Zakzanis and Boulos (2002, August), which suggests that many neuropsychological instruments have good sensitivity and specificity when it comes to diagnosis of early AD. In any event, if the result is positive for AD, the advice of the makers of the AHST for users to see their physicians may not be the best advice in terms of getting an accurate diagnosis and aftercare in the absence of additional informed referral to trained specialists.
|
Lack of Available Counseling |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
The most critical difference, however, is the lack of available counseling, the third and most troubling aspect with the AHST. The Food and Drug Agency approval of home HIV testing was not without controversy, mainly because of the issue of counseling (Beall & Whyte, 1995). Home HIV tests incorporate a method by where the user calls a phone number for test results, and, if the result is positive, an option for counseling is given (Branson, 1998). The AHST offers no phone number to call for counseling. In surveying users of the HIV home test, Branson found that 65% of those testing HIV positive accepted referrals for medical and psychosocial services, 7% of users expressed shock and dismay at their positive result, and 5% hung up after finding out the test came back positive for HIV. One caller (out of the sample of 610) expressed suicidal ideation. Concerns over lack of counseling led to the HIV home test being banned and/or restricted for use in many countries (Schopper & Vercauteren, 1996; Karcher, 1997).
Considering that over half of those who tested positive on an HIV home test expressed interest in support services and/or counseling, as well as over 1 in 20 expressing significant psychological distress over their positive results (Branson, 1998), one would expect that the AHST would have congruent numbers in terms of those needing or wanting counseling and/or support after positive test results. The importance of providing counseling, even of an informational nature, cannot be underemphasized in chronic, terminal illness. In a sample of cancer patients, Evans (1995) found that a lack of information about their disease was the greatest cause of anxiety. Psychotherapy is often helpful, even for those with cognitive deficits from AD. Reichlin (1999) describes a successful group therapy program for persons with AD and their caregivers, with indications that such programs with educational components lead to better outcomes for those with AD. Unfortunately, the AHST does not appear to offer any kind of informational support and/or referral service, beyond asking users to see their physicians. Considering that depression, anxiety, and even homicide and/or suicide (Cohen, 2000) may be a reaction to a diagnosis of dementia, the unavailability of counseling is a critical shortfall of the AHST.
Thus, even if the AHST was more precise diagnostically, had better sensitivity and/or specificity, and so forth, this problem alone is a severe shortfall of using the AHST. The ethical issue here is that of avoiding harm, in that the results of this test could lead to negative psychological consequences for the individuals taking this test. In discussing the American Psychological Association (APA) ethics code (APA, 1992), Canter, Bennett, Jones, and Nagy (1994) note that, in avoiding harm, "it would be important for the psychologist to demonstrate that reasonable steps were taken to avoid negative consequences" (p. 45). With the AHST, the psychological consequences of a positive result could well be emotionally disastrous for an individual, and a reasonable step to counter this would be the requirement for well-documented immediate availability of pre- and post-test counseling and crisis intervention services.
With the probability of significant misclassification because of the lack of demonstrated predictive validity for the AHST, this problem becomes even more pronounced. A similar issue was raised for rapid at-home HIV testing. Sibbald (2000) discussed the issues surrounding a 15-min HIV test that could be done outside of a laboratory. She notes that even the ELISA HIV test, which is highly sensitive, can lead to a false-positive (out of 300,000 ELISA tests in Ontario, 3,000 were positive, but only 1,000 were found, after a second, confirmatory test, to actually be HIV positive). The consequences of a false-positive test, without counseling, could easily lead to adverse psychological reactions. Precounseling for HIV tests is a common standard of care in many healthcare systems as a result, and even the at-home HIV test has a means for users to get counseling services. Even questions regarding home pregnancy tests, which have been available for quite some time, have been raised. In a study comparing 200 women using home pregnancy tests at home with 200 laboratory technicians running the tests in a laboratory setting, all of the laboratory technicians results were accurate, but 10% of the women who tested at home had tests that yielded inaccurate results (Hicks & Iosefsohn, 1989). The authors of this study cite that home testing may "result in needless emotional turmoil for patients and their families" (Hicks & Iosefsohn, 1989, p. 321). A similar argument can be made on the same grounds with the AHST.
|
Social, Legal, and Ethical Issues |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
AD, like HIV infection and acquired immunodeficiency syndrome (AIDS), is a disease that the public often misperceives and misunderstands. AD, like HIV/AIDS, is often considered a death sentence by the general public, and much misinformation still persists about AD. Gatz and Pearson (1988) evaluating the literature on attitudes toward aging noted:
Alzheimer's Disease is very connected with aging in the public eye and for psychologist readers. This media attention may have an unintended consequence. Namely, older adults may become hyperaware of their own lapses and needlessly fear that they are developing the disease... We argue that one important manifestation of this process would be overestimating the prevalence of Alzheimer's Disease. (p. 186)
Indeed, Gatz and Pearson (1988) cite several studies, using samples from different populations, in which roughly half of the sample overestimated the prevalence of AD. Given this, many older adults, and those around them (i.e., relatives, caregivers) are extremely concerned about their cognitive status vis á vis AD. It is easy to see how this climate of misinformation and fear of AD could create situations where a test like the AHST, without the proper safeguards, could lead to bad outcomes.
The standards related to evaluation and assessment set down in the Ethical Principles of Psychologists and Code of Conduct (APA, 1992) are an example of a structured ethical code whereby the AHST certainly appears to be in conflict. Use of the AHST conflicts with (a) the requirement that evaluations be performed within the context of a defined professional relationship, (b) the mandate that diagnostic statements be based on information that can be substantiated and that is appropriate based on a proper application of testing techniques, (c) the standard to recognize limits of diagnostic judgments about specific individuals, and (d) the obligation that psychologists take reasonable steps to assure that the person tested be given appropriate explanation of results. An analogous situation to the AHST would be to develop a version of the California Verbal Learning Test–II (CVLT-II) for home use. In a recent meta-analysis, Zakzanis and Boulos (2002, August) found that the CVLT-II was found to be one of the most sensitive neuropsychological measures for detecting preclinical AD. Even though the CVLT-II appears to be highly useful in terms of detecting early AD, it would be clearly unethical to have a take-home version, as it would be used and scored by untrained individuals.
A major flaw with the AHST, and any at-home medical or psychological test, is an ethical one, stated best by McQueen (1993):
Even a simple test requires a sophisticated understanding of false-positive and false-negative results. The humane diagnosis of a disease or disorder requires the application of professional skill, knowledge, and training, and the consideration of many factors. The test result is frequently a minor and even insignificant part of the total process. Statistically, there is no perfect test, so home testing can generate "illness." In the history of quackery, fear has been used to create a market. As testing moves more into the marketplace are profit motives more influential than the interests of patients? Is it ethically and economically acceptable to create a marketplace for inappropriate or misleading tests that are easily manipulated from outside the health care system? (p. 1655)
|
Summary |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
Much like other tools, tests of olfaction, such as the SIT, in concert with other tests, can be useful in diagnosing dementias when placed in the hands of those skilled in administration and trained in dealing with the psychological ramifications of a difficult diagnosis. However, when placed in the hands of an untrained public, with no safety mechanisms integrated into the test procedure such as counseling and/or the guidance and knowledge of a well-trained clinician, the chance of causing harm to its users is a distinct possibility. For these reasons, we believe that public use of the AHST is inappropriate and may be unethical. We encourage dialogue between ethicists, geriatricians, mental health professionals, and the lay public to explore related issues such as (a) the way in which the pressure to screen may be motivated by commercial interests, (b) the liability issues related to routine screening, (c) the potential link between a positive test and automatic referral for pharmacological therapy that may not have much clinical utility, (d) the limitations of biological markers when memory impairment is viewed on a continuum rather than as a discrete diagnostic entity, (e) the thorny problem of professional publications advertising assessment instruments of dubious value, and (f) the formal procedure by which gerontological organizations should address these concerns.
|
Footnotes |
|---|
2 Louis de la Parte Florida Mental Health Institute, University of South Florida, Tampa.
Decision Editor: Linda S. Noelker, PhD
Received for publication November 14, 2002. Accepted for publication February 24, 2003.
|
References |
|---|
|
TOPAbstractDiagnostic Utility of Smell...The "Fine Print"Lack of Available CounselingSocial, Legal, and Ethical...SummaryReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Ryan, S. Wilson, S. Greenfield, S. Clifford, R. J. McManus, and H. M. Pattison Range of self-tests available to buy in the United Kingdom: an Internet survey J. Public Health Med., December 1, 2006; 28(4): 370 - 374. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lock, J. Freeman, R. Sharples, and S. Lloyd When it runs in the family: putting susceptibility genes in perspective Public Understanding of Science, July 1, 2006; 15(3): 277 - 300. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Kobayashi The Odor Stick Identification Test for the Japanese (OSIT-J): Clinical Suitability for Patients Suffering from Olfactory Disturbance Chem Senses, January 1, 2005; 30(suppl_1): i216 - i217. [Full Text] [PDF]
|
|
|
|
 |
|
 M. B. Kapp Should Home Screening Tests for Alzheimer's Disease Be Regulated? Gerontologist, June 1, 2003; 43(3): 292 - 294. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
