Clinical Reasoning/EBP/Scientific Method Flashcards
What is deductive reasoning?
“top down” logic
-creates a linear relationship between pt’s restrictions/impairments etc and their pathology of body structures, personal factors, environment, etc
What is narrative reasoning?
using specific observations to draw conclusions
-achieved throughout open ended questions and active listening to gather info about the pt’s personal and environmental factors
What is Bayesian reasoning?
basically involves application of probability as data is gathered to identify if something is more or less likely or the probability that something is true or not true
What is the cluster of findings for nociplastic pain(central sensitization)
characterized by pain that is disproportionate, non-mechanical, unpredictable, and diffuse
-in the absence of red flag findings
Sensitivity and Specificity of nociplastic pain cluster?
sensitivity 91.8%
specificity 97.7%
What is the cluster of findings for peripheral neuropathic pain(radicular or referred)
-symptoms that are referred in a dermatomal (radicular) or cutaneous (referred) distribution
-history of nerve injury, pathology, or mechanical compromise of the nerve w/ symptom provocation via mechanical testing
Sensitivity and speficifity of peripheral neuropathic pain cluster?
sensitivity 86.3%
specificity 96%
What is the cluster of findings for nociceptive pain?
-symptoms localized to an area of injury or dysfunction
-provocation and/or alleviation are clearly identifiable and proportionate
-match known mechanical and anatomical distributions
-symptoms usually intermittent and start w/ onset movement or mechanical provocation
-quality of symptoms may be a constant dull ache or a throb at rest
Sensitivity
the ability to identify a condition when it truly exists
-find true positives
-higher sensitivity means you will catch it if it’s there
-most people with the condition will test positive
-lower sensitivity means you will get some false positives
SnNOUT
-negative result rules it out
-only people without the condition get ruled out, but you may have false positives
Specificity
the ability to identify when a condition is not present
-find true negatives
-higher specificity means that when the test is negative, you are confident that the condition is not present
-most people with the condition will test positive, but anyone who DOES NOT have it will test negative
-lower specificity means that when test is negative, you are less confident it is a true negative. might be a false negative
SpPIN
-positive result rules it in
-only people WITH the condition will test positive
Type 1 error
“backing a loser”
FALSE POSITIVE
concluding that there IS a significant difference when there actually ISNT
-set a stricter alpha level for statistical significant difference to reduce likelihood
-larger sample size also helps
-well defined hypothesis
-replication studies with different samples can help validate findings
Type 2 error
“missing a winner”
FALSE NEGATIVE
concluding that there ISNT a significant difference when there actually IS
-increase # of subjects in the study to reduce likelihood of this
Large effect size
0.8 and up
Moderate effect size
0.5-0.7999
Small effect size
0.2-0.4999
Trivial effect size
<0.2
Cohen’s kappa (k)
describes how reliable something is
0 = no reliability(no better than chance)
1= perfect reliability
0 no better than chance
<0.4 poor
0.4-0.6 fair
0.6-0.75 good
>0.75 excellent
1 perfect reliability
positive likelihood ratio
these will be >1
> 10 large shift in probaiblity
5-10 moderate shift
<5 small shift
1 = no change
negative likelihood ratio
these will be <1
<0.1 large shift
0.1-0.2 moderate shift
>0.2 small shift
1.0 no change
Placebo effect
individual believes in positive effect, even though no therapeutic value
good research should have placebo group so that test group improves MORE than placebo group
good research should have placebos that look like the real treatment
-placebo group needs to have belief that they actually had real treatment
placebo effect most powerful on outcomes mediated by the brain
Nocebo effect
negative beliefs about a treatment
-exaggerated when subjects expect negative side effects or negative experiences with treatments in the past
research groups example
-good to have exclusion criteria
-in a research study on spinal manipulation this would exclude those who have positive or negative expectations about it
-randomization also helps reduce this effect
Hawthorne effect
subjects who know they are observed as part of a research study tend to work harder than they would otherwise
-we find that people change their behavior(positive or negative) due to being observed
Observer effect
people work harder when being watched
-more improvement in response to more attention
-control for this by making sure all treatment groups get about the same amount of attention from clinician
group that receives more attention likely to improve more
example:
group 1: 30 mins manual
group 2: 30 mins exercise
group 3: 30 mins manual + 30 mins exercise
-60 total minutes, they are likely to improve the most
John Henry effect
“legend of john henry outcompeting the machine”
control group perceives that they are disadvantaged compared to experimental group
-they seek out additional treatment/help because they think they aren/t getting enough care otherwise
best way to reduce
-blind subjects so that they don’t know if they are in control or experimental group
Pygmalion effect
AKA “rosenthal effect”
expectations of those in authority shape results of their subjects
-high expectations lead to improved performance
-low expectations lead to poor performance
best way to reduce
-blind clinicians who treat the group and those who assess the group
Level 1 evidence
high quality diagnostic studies, prospective studies, RCTs, or systematic reviews
Level 2 evidence
comes from lesser quality diagnostic studies, prospective studies, RCTs, or systematic reviews
-weaker diagnostic criteria
-weaker reference standards
-improper randomization
-no blinding
- <80% follow up
Level 3 evidence
case control studies or retrospective studies
Level 4 evidence
case series
-has no control compared to case control study
Level 5 evidence
expert opinion
Grade A recommendation
STRONG
-preponderance of level 1 and/or level 2 studies support the recommendation
-includes AT LEAST 1 level 1 study
Obligation
“Should”
Grade B recommendation
MODERATE
-a single high quality RCT or a preponderance of level 2 studies support the recommendation
-includes studies w/ only short term follow up
(less than 3 months) or small sample size (less 100 subjects)
Obligation
“May”
Grade C recommendation
WEAK
-single level 2 study supports the recommendation
Obligation
“Can”
Grade D recommendation
CONFLICTING OR NO EVIDENCE
-level 1 and level 2 studies disagree with respect to their conclusions or provide no evidence of benefit
Obligation
“Should NOT”
Incidence vs prevalence
incidence = # of NEW cases during a period of time
“10 of 100” people develop LBP during 2024
prevalence = # of cases at a specific point in time
“10 of 100” people have LBP on 12/10/2024
-prevalence includes ALL cases, including new and pre-existing
MDC
smallest change detected that is not attributed to measurement error
example: moving from 2/10 to 1/10 on VAS, however MCID may be 2 full points
MCID
smallest change in condition that either pt or clinician may consider important
