Module 3 Flashcards
1
Q
Demographic, psychological, and social factors as variables (3)
A
Dependent variable- as outcome (ie marital status)
independent variable- impacts outcome- ie marital status and stress management
control variable- things you control for (ie severity of depression controlled for to determine impact of counseling on quality of life)
2
Q
2 ways of conceptualizing age
A
- as time since birth
- As likelihood of dying “real age”
- not good validity yet
3
Q
2 ways to measure age
A
continuous variable- if constant effect
categorical variable- if inflection point where age effect is maximized
4
Q
2 ways of conceptualizing residence (with details)
A
- as location
- indicator of access to health services
- health exposure
- rural v. urban
- indirect measure of ses
- as characteristics of dwelling or larger built environment
- ex: stairs after hip surgery
5
Q
Definition of race (historically)
A
population that lives within specified geographical area and has common gene pool
consider genotype v phenotype
6
Q
Ecological fallacy (race)
A
attributing group characteristics to the individual
7
Q
ethnicity
A
cultural factors that ID a person as part of a group, can’t ID from race
8
Q
Marital status
A
civil arrangement with legal support
consider divisions (single, married, widowed, divorced)
looks at life stresses, emotional support, or can be an outcome itsenf
9
Q
SES
A
looks at factors such as income, education, occupation (with different ways to code)
now moving from strategication and toward social capital
10
Q
Tools to measure SES (3)
A
- Hollingshead Index of Social Position
- Duncan socioeconomic index
- Nam-Powers Socioeconomic Score
11
Q
Mind-body connection
And what it includes (4)
A
includes behavioral medicine, esp placebo effect
- well being
- locus of control
- pain
- stress
12
Q
well being
A
ranges from emotional to economic well being
overlap with health related qulatiy of life
don’t measure specific health-well being relationship
more general in nature
13
Q
Measures of well being (2)
A
- General well being schedule
- index of well being
14
Q
Locus of control
A
most share internal and external aspects
outgrowth of mastery and efficacy
15
Q
Measures of locus of control:
general (2)
health (2)
A
General
- spheres of control battery
- internality, powerful others, and chance scale
health
- multidimensionality health locus of control scale
- mental health locus of control scale
16
Q
pain
A
frequently measured with VAS
difficult to compare across individuals
frequency and intensity most relevant
17
Q
measures of pain (6)
A
- brief pain inventory (location, severity, quality)
- mcgill pain questinonaire (sensory nature of pain)
- MOS Pain measure (pain and function)
- Low Back Pain disability questionnaire
- pain and distress scale
- fatigue, energy, econsciousness, adn sleepiness
18
Q
stress
A
emotional, physiologic, social, economic
perceived or experienced
consider :perception, stressful experiences, coping resources
19
Q
stress measures (4)
A
perceived
- perceived stress questionnaire
experiences
- social readjustment rating scale
- life stressors and social resources inventory
- life stressor checklist
20
Q
Other mind-body
A
nonfunctional adjustment (illness behavior questionnaire)
readiness to change
21
Q
Affect
A
construct validity very important
consider: responsiveness, diagnosis v. behavior/perception
anxiety and depression with biggest focus
22
Q
depression
A
most prevalent mental health problem in us
related to loneliness and social isolation
some tools target specific populations, others environmental factors
23
Q
depression tools (5)
A
- self rating depression scale
- center for epidemiologic studies depression scale
- geriatric depression scale
- carroll rating scale
- depressive experiences questionnaire
24
Q
anxiety
A
range of defniitions from situational and well defined to vague state experienced
recent focus on state v. trait
25
state v. trait anxiety
state=transitory experience, short term
trait=stable, persistant response to environmental factors
26
anxiety tools
* self rating anxiety scale
* hamilton anxiety scale
* state trait anxiety measure
* endler multidimensional anxiety scale
27
Other affect considerations (with tools-3)
general affect or morale or positive emotions
* affect balance scale
* memorial university of newfoundland scale of happiness
* PGC morale scale
28
cognitive function
doesn't measure IQ
addresses judment, memory, ability to perform interpretive and related tasks
29
cognitive function tools (2)
* mini mental state exam
* short portable mental status questionnaire
30
social function
ability to fulfill roles
giving and receiving help
dependent on roles held previously
31
social support
can be objective (how many hours) or subjective (level of need, feelings about it)
32
social support measures (4)
social support questionnaire
MOS social support survey
duke-unc functional social support survey
duke social support and stress scale
33
social function and adjustment
individual's status and function
long tools due to complexity
take positive or maladaptive approach
34
social function and adjustment tools (4)
social functioning schedule
social adjustment schedulue
social maladjustment scale
social dysfunction scale
35
complex organizations
looks at aspects of jobs, such as commitment and work control
36
what is a treatment?
an intervention designed to improve a health state
could be a cure or prevention, procedure, drug, behavior change, how health care is delivered
any potential modifiable factor
37
components of care
goal is to standardize intervention
consider level of specification and what you will emphasisize depending on area of focus
38
Treatment components (3)
Medication, prodedure, counseling/education
39
7 elements of tx
type
dosage
route
frequency
duration
onset/timing
provider characteristics/technical aspects
40
diagnosis v. treatment
each can be a component of tx
pay attention to diagnsosi when it has an impact on treatment (ie if diagnosing depression means that it's more severe to start with, or that someone is more likely to have a stigma against it or something like that)
41
treatment components: medications
anything taken into body tha timpacts health status
can be drug, supplement, nutrition
consider: type, dosage, route, frequency, duration, onset/timing
NOT provider
42
treatment components: procedures
anything physically done to pt
consider: type, frequency, duratino, onset/timing, technical aspects/provider
NOT: dosage or route
43
treatment components: counseling
exchange of info for therapeutic purpose
consider type, dosage, frequency, duration, onset/timing, provider
NOT route
44
clinical guidelines
help standardize tx
must be adapted to ind. pt needs
is making the guideline part of the tx?
how are they implemented?
45
isolating tx of interest
use comparison group
analyze change over time relative to when intervention implemented
utilize placebo for "attention control"
46
quality improvement
special kind of intervention
highly contextualized, so may have to treat it differently
47
challenge to compare interventions
hard to do less effective intervention (extant data?)
adherance
48
treatment summary (3)
1. understand the components of tx
2. include variation in tx- comparison
3. understand tha teffect of tx is difference in groups after everythign else is controlled for
49
risk adjustment
takes into account pt attributes for making valid inferences about effectiveness and quality of care
50
examples of risk adjustors
age
sex
gender
race
ethnicity
acute clinical stability
principal diagnosis
severity
comorbidity
functional status
hrqol
51
severity
classify pt's primary problem in terms of prognosis
ex: APACHE: Acute physiological score, age, chronic health eval
ex: repeated hospital admissions
ex: DRG
52
comorbidity with 3 reasons to measure
coexisting conditions unrelated to principal diagnosis
Not equal to complications
1. control for selection bias
2. predict outcomes
3. form basis for subgroup analysis
53
examples of diagnosis specific severity measure (3)
AHA Stroke outcome classification
Canadian neurological scale
low back classification system
54
ways to measure comorbidity
can list dummy coding
can list only salient ones
can weight them
can consider severity of comorbidities
55
why control for comorbidity? (3)
control for selection bias
improve prediction of outcomes
form a basis for subgroup analysis
56
data sources for risk adjustment (3)
medical records- not always complete, time consuming to abstract data
admin data- easier, but designed to maximize reimbursement
medications prescribed- as proxy, but flawed
57
considerations in selecting risk adjustment strategy (4)
purpose
relative importance of diagnosis to risk
data realted considerations
Role of competing risks
58
increasing reliability in data for risk adjustments (5)
clarity of definition
operationalized definition
extent of quality of data- review, completeness, etc
training of reviewers
environment of review
59
statistical performance
look for goodness of fit, predictive validity, model discrimination
60
Uses of ICD codes (7)
1. Workload and length of stay tracking
2. Quality of care
3. VA- allocate resources to medical centers
4. Study patterns of disease, care, and outcomes
5. Effectiveness of care (risk adjusted)
6. Inclusion and exclusion criteria
7. Rate reporting
61
original use of ICD codes
classify causes of death, then adopted by WHO
62
Sources of error along patient trajectory
* communication between patient and admitting clerk/clinicians
* Clinician's knowledge of best tests
* Clinician's ability to interpret results
* Recording of diagnosis
* Changes in diagnostic accuracy over time
63
sources of errof along the paper trail (ICD codes)
* synonyms used in record
* omissions in medical record
* transcription in medical record
* incomplete info when chart is reviewed
* coder training
* coder experience and attention
* miscoding (generic for specific)
* resequencing codes (primary diagnosis)
* upcoding (for reimbursement)
* errors at attestation by physician
64
5 ways to view code accuracy
1. sensitivity
2. specificity
3. positive predictive value
4. negative predictive value
5. K coefficient
65
3 things required in order for race/ethnicity measurs to be meaningful
1. produce consistent data over time
2. allow comparability across populations and surveys
3. use terms that are widely understood by the groups completing the instruments
66
race as biology?
more variation within races than between
race exists as socio-political construct rather than biological one
67
race and racism?
race is social construct that INCLUDES the effects of racism on an individual
racism includes institutionalized racism and internalized racism
68
institutionalized racism
having differential levels of access based on race to societal goods, services, and opportunities
educational and access racism
69
internalized racism
feelings of resignation, helpflessness, and hopelessness
can lead to risky health behaviors
70
ethnicity
cultural identification
can be fluid over time
can influence health beliefs
71
ways of measuring race and ethnicity:
revised directive number 15
5 categories of race, can pick more than 1
hispanic/nonhispanic
72
ways to measure race/ethnicity:
census 2000
allow multiple responses to categories, some confusion over specificying hispanic ethnicity
73
ways to measure race/ethnicity:
sillitoe survey
added religion item to get at ethnicity- did not chagne responses
74
ways to measure race/ethnicity:
stephenson multigroup acculturation scale
looks at degree of acculturation
not designed to measure cultural change among acculturating individuals
factors: ethnic society immersion, dominant society immersion
75
5 research goals related to race/ethnicity
1. look at outcomes related to race ethnicity
2. disentangle ses from race/ethnicity
3. effects of racism on different groups realted to outcomesrationale for collection of race/ethnicity data
4. manner in which data was collected could be reported
76
treatment fidelity
methodological strategies used to monitor and enhance the reliability and validity of behavioral interventions
goal is to increase confidence that changes in dependent variable are due to changes in independent variable
77
best practice for treatment fidelity (5 categories)
design
training
delivery
receipt
enactment
78
treatment fidelity: design
factors that should be considered when designing a trial and factors that should be reported in order to evalute and replicate the trial
example: content and dose for tx and comparison conditions, provider credentials, theoretical framework
79
treatment fidelity: training
if using human providers, consider capability of delivering intervention , training, measurement of provider skill acquiasition and how skills are maintained over time
80
treatment fidelity: treatment delivery
processes of monitoring and improving the delivery of the intervention
mechanism to assess how treatment is delivered
look for impact of provider
81
treatment receipt
ensuring that the participants understand ino provided
(ie literacy, cognitive level, English language, etc)
82
treatment fidelity: treatment enactment
processes to monitor and improve teh ability of patients to perform treatment realted cognitive strategies and behabioral skills in their daily lives
83
importance of treatment fidelity study
give consumers a way to evaluate
give researchers a plan
give more guidelines for RCTs and funding
84
Comorbidity measures
Charlson- assigns weights related to hospital mortality, have caution with ICD9 codes, weight different for different populations, no reason to conlude that it's comprehensive
## Footnote
Elixhauser comorbidity index- broader applications
The list created in the study we read:
allows DRG filter for comorbidities (to exclude complications), comorbidities predict resources utilized better than mortality
85
5 concepts that represent burden of illness
1. primary diagnosis
2. severity of primary diagnsois
3. complications that arise from process of care
4. unimportant comorbodities
5. important comorbidities (that are unreatled to main reason for hospitalization) but increase likelihood of poor outcome
86
what is attribution bias?
falsely attributing an aspect of an outcome to one cause, when it may have been due to another
87
how can you lessen attribution bias in outcomes research?
make sure that comparison groups are very well matched and defined
RISK ADJUST!
88
what is sampling bias?
In statistics, sampling bias is when a sample is collected in such a way that some members of the intended population are less likely to be included than others.
89
What is recruitment bias?
Give one source
Recruitment bias is a cause of sampling error due to methods used to recruit participants into a study.
Possible sources: how the study is advertised (phone calls v. mailers v. posters, convenience studies, stopping people on the street, from a clinic, etc)
90
What is assignment bias?
Give one source
Assignment bias would occur anytime something is done that results in people systematically being assigned to a treatment.
For example, if you're assigning people based on treatment received, maybe it is impacted by setting (ie academic v. nonacademic med setting)
91
What is intention to treat?
How can it be used in studies in which noncompliance is believed to be a problem?
an analysis based on the initial treatment intent, not on the treatment eventually administered. ITT analysis is intended to avoid various misleading artifacts that can arise in intervention research
Help estimate what will happen "in real world," though it may understimate actual efficacy of treatment
92
what is selection bias
Selection bias is a statistical bias in which there is an error in choosing the individuals or groups to take part in a scientific study
93
what is a selectivity corrected model?
A model that has either been risk adjusted or used propensity scores to equate groups.
94
Treatment setting as proxy
Potentially:
Insurance status (inpt v no)
Severity (locked psych unit v. outpatient)
SES (transportation)
95
Should all elements of treatments be measured in a single study
You should measure each of them that are relevant so as to idolate the treatment of interest as much as possible.
Consider #s involved in study.
96
distinguish among the different components of severity
general physiological severity (regardless of diagnosis)- likelihood of mortality
functional component- degree of impairment
relation of baseline condition to condition at different measured times
97
3 types of severity measures
APACHE
CNS
AHA Stroke Outcome Classification
98
2 most common time points at which severity is measured (with disadvantages)
admission- can be at different points in illness (ie on the way up v. on the way down)
discharge- often impacted b things like insurance
99
How can one identify components of severity to avoid selection bias?
Look at the if the severity of illness is associated with outcome
AND
if severity of illness is associateed with treatment but is not a conseuence of the treatment
100
5 criteria to choose severity measure
covers appropriate domains
appropriate for time point in which it's being utilized
specific enough to allow for risk adjustment
valid in the population in which it's being utilized
differentiates between complications and comorbidities
101
4 severity of illness measures and one example of when each would be used
APACHE-II, if you're determining risk of death in ICU patients
Probability of repeated hospital admissions- to identify elderly patients who might benefit from a comprehensive evaluation
AHA Stroke Outcome Classification- to impairment, severity, and functional classification through the assessment of 6 domains after a stroke
Canadian Neurological Scale= to predict patient outcomes and ID symptoms specific to acute stroke.
102
determining comorbidity v. complication
literature review
DRG
timing- comorbidities present at onset, compications arise during the episode
103
why do a subgroup analysis when examing comorbidity?
looks at influence of effect of comorbodity on response to treatment (interaction effect)
104
why are disease of primary interest and prognostic end points factors in choosing and measuring comorbidity?
help you decide how to consider comorbidity- all of them, salient ones, weighted, etc, based on likelihood that they will impact the outcome of interest
105
1 advantage of a weighted comorbidity approach
lets you control more for the comorbidities that are likely to ahve the biggest impact on outcome
106
Charleson and Duke- explain and one example of when it might be good to use as comorbidity tool
107
gender v. sex
Sex refers to biological differences; chromosomes, hormonal profiles, internal and external sex organs.
## Footnote
Gender describes the characteristics that a society or culture delineates as masculine or feminine.
Which one used should depend on why you're measuring the construct at all.
108
advantage of composite ses index?
the use of a composite ses index can help to ensure that you are actually measuring the intended underlying contstruct v. just proxies of that construct.
109
list at least 3 depression scales and give an advantage
Geriatric depression scale- speciically for administration in geriatric populations in the clinical setting
carroll rating scale- assesses severity of depression by specifically identifying 17 symptoms associated with depression; self-rating, easy- yes/no questions
beck depression inventory- allows for measurement of change over time because it evaluates depth of the depression
110
list at least 2 psychological well being scales and give an advantage and disadvantage of each
general well-being schedule- developed for the national center for health statistics, this is intended to measure well being in the general population
The GWB Schedule is good for group studies of subjective well-being but less is known regarding its adequacy as a case-detection instrument.
index of well being- composed of eight semantic differential items and one life satisfaction item
looks at health states and rates their importance (societal view)
each can miss situational issues
111
comorbidity measures (2)
Charlson Comorbidity Index- weighte dcomorbidity score to predict hospital mortality (relative risk)., also available in ambulatory care; disadvantage- narrow application
## Footnote
elixhauser comorbidity index- broader application across a range of conditions and outcomes
112
Duke Severity of Illness Checklist
four parameters of a health problem: symptoms, complications, 6-months prognosis without treatment, treatment potential. T
he following three types of severity score (from 0 [lowest degree of severity] to 100 [highest degree of severity]) can be calculated: (1) the DUSOI diagnosis score for each diagnosis stated, (2) the DUSOI overall score for the set of all health problems stated for a patient, and (3) the DUSOI comorbidity score, i.e., all problems except for any one problem of principal interest.
The DUSOI is suitable for patients from the entire chain of medical and rehabilitative care, although it had initially been developed for the ambulatory sector
113
