1. Choose a paper describing the
development or validation of a measure of relevance in health disparities
research (please give the full citation and/or upload the paper if that's
possible).
Julayanont P, Tangwongchai S, Hemrungrojn S, et al. The Montreal
Cognitive Assessment-Basic: A Screening Tool for Mild Cognitive Impairment in
Illiterate and Low-Educated Elderly Adults. Journal of the American Geriatrics
Society. 2015;63(12):2550-2554. doi:10.1111/jgs.13820
2. What was the definition of the construct?
The construct of focus for this
publication is the effect of an individual’s level of education on the
appropriate cutoff for diagnosing dementia. Many of the current cognitive
functional exams were designed and validated amongst individuals with high school
or greater education, and the cutoff scores for a diagnosis of dementia reflect
the relevant cognitive effects of such an education. This has led to a
disparity of appropriate diagnoses of individuals with less education, especially
those with less than five years of schooling. Although “years of schooling” is
not the only way to measure education, it is one of the most common, and can
even be a component of inclusion/exclusion criteria for studies. Because of
this, those with less education represent an understudied portion of the
population that suffers from dementia, and the result is a disparity of
appropriate diagnosis tools, let alone interventions and medical care.
This publication presents the MoCA-Basic, a version of the Montreal Cognitive Assessment
tool that is designed to better distinguish between adults with normal or mild
cognitive impairments (MCI) who have less than five years of education. The
MoCA-B version excludes components of the test that requires literacy and formal
education, such as alphabet-based tests, clock-reading, and cube-drawing.
3. How did the authors provide
evidence on the validity of the measure? Could you think of additional
approaches to validating the measure?
Evidence provided in support of the measure’s validity: A pilot
study using the MoCA-B test was conducted in cognitively normal and impaired
subjects with 12 years of schooling or less. The larger study was conducted in
cognitively normal and impaired subjects with less than 5 years of schooling. All
participants in the study were given a clinical interview that included the
individual and a caregiver along with the standard Clinical Dementia Rating
Scale (The MoCA-B is intended to be a screening tool). Additionally,
individuals with medical, neurological, or psychological conditions having possible
cognitive repercussions were excluded. Individuals taking medications that could
affect cognition were also excluded. The physician administering the CDR was
blinded to the result of the MoCA-B score. Lastly, the cognitively normal
control group all had a CDR score of 0 out of 3.
Additional approaches: The study group could publish a follow-up that
includes longer term results of the participants. Since the argument in favor
of accurately diagnosing MCI in patients, and MCI typically leads to more
progressed dementia, it would be worth showing that the MCI diagnosis obtained from
the MoCA-B reflected cases that went on to develop dementia.
4. How did the authors provide
evidence on the reliability of the measure? Could you think of additional
approaches to evaluating the reliability of the measure?
Evidence provided in support of measurement reliability: A subset
of the participants were randomly selected to repeat the MoCA-B within 2 months
of the first test.
Additional approaches to evaluate reliability: Although it would be resource intensive, I think it would be worthwhile to repeat the standard that the MoCA-B was compared with (CDR Score). This could then suggest that it was reliably valid (there is variation in CDR due to variation in cognition depending on factors like time of day, mood, health, etc) rather than just reliably consistent. Additionally, the individual who administers the test can affect the result, so a measure of inter-rater reliability would be worthwhile as well.
5. Describe the implications of a lack of measurement validity or reliability for future research applications.
If the MoCA-B is going to be used for future research, the results of the test will affect the conclusions made within new studies. For example, if the exam was invalid such that someone with MCI could still score well on the test, then it’s possible that such measurements as biomarkers, brain scans, or functional skills that should be associated with cognitive impairment might be overlooked. This could translate into less care available to the individual. On the other side, if the MoCA-B was invalid such that it was still too difficult, and individuals with lower education who were cognitively normal were scoring low, then they might receive a diagnosis of MCI, and this could affect their social lives, insurance access, health care costs, etc. If the MoCA-B was not reliable, then additional misdiagnosis would result.
Part 2:
1. Find a paper describing a health disparity (please give the full citation or, even better, upload the paper so everyone else can look at it too)
Gianattasio KZ, Prather C, Glymour MM, Ciarleglio A, Power MC. Racial disparities and temporal trends in dementia misdiagnosis risk in the United States. Alzheimers Dement (N Y). 2019;5:891-898. doi:10.1016/j.trci.2019.11.008
2. Summarize the construct and measurement of the dimension of disparity (e.g., racial inequalities?, SES inequalities?) and the outcome measured (e.g., self-rated health).
This paper is focused on
disparities in misdiagnosis of dementia across racial/ethnic groups. The
outcome for the study is likelihood of a dementia diagnosis. An algorithm based
on a logistic regression was used to assign dementia status to individuals
based on various sociodemographic, health, social engagement, and cognition
variables. The study uses data from the HRS study, which is a longitudinal,
nationally representative study of US adults aged 50 and older (beginning in
1992). The diagnosis of dementia is important for individuals to be able to
make plans for their future to sustain a higher quality of life. A correct
and early diagnosis allows for more potential interventions, time to make
financial plans, communication with social support either through family,
community, or both, and more. A misdiagnosis can greatly diminish the prospects
for sustained high quality of life.
3. What is the evidence for the
validity and reliability of the measures?
The algorithm was developed using data from the substudy from
HRS, the Aging, Demographics, and Memory Study, in which the subjects were
given formal in-person dementia.
4. What is the reference category
used for the disparity measure (ie, who is the comparison group)? Why does this
reference category make sense (or not) for this research question?
The reference category was non-Hispanic whites. In the context
of this paper, the goal is to determine whether or not there are disparities in
proper diagnosis. The algorithm used suggests that non-Hispanic whites have the
highest percentage of correct diagnoses, so lower percentages relative the
highest group would represent the proxy measurement for disparities in dementia
diagnosis.
5. How is the disparity quantified
or measured? Is this an absolute or relative measure or are both
provided? Describe which type of measure you would prefer for this
research area, or, if both, why.
The paper determines whether each individual has a dementia
diagnosis according to a dementia ICD-9-CM code in the medical record. This is
then compared to the algorithmic prediction of diagnosis status, and the
prediction is actually taken to be the true diagnosis. Finally, the measurement of
misdiagnosis is expressed as either over- or under-diagnosis by presence of
ICD-9 w/o prediction and absence of ICD-9 with prediction, respectively. This
is an absolute measure because it is looking at the exact percentage of
cases that are over and under diagnosed rather than the relative amount of over
or underdiagnoses above or below the reference group. I think both relative and
absolute measures are important. A relative measure could show that there is a
fold-difference between groups, whereas an absolute measure could give a sense
of how many are affected in each group.