Scribd
Upload
36 views

Journal Article Cheat Sheet-1

Uploaded by

George Wong
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
36 views

Journal Article Cheat Sheet-1

Uploaded by

George Wong
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Study design comparison table

STUDY TYPE WHAT HAPPENS CALCULATE CRITIQUE QUESTIONS


Descriptive
Case Series Measure exposure(s) Calculate descriptive –Is the sample representative? (Best if
in a series of subjects statistics but cannot test a recruited consecutively or sample includes all
with a specific disease hypothesis about an cases in a population)
(outcome) association because no
non-cases*
Observational
Cross-Sectional Exposure and Prevalence –Did the exposure precede the outcome?
outcome assessed at (temporal relationship?)
ONE time –Prevalence (survival) bias?
–Selection bias?
Case-Control Cases and non-cases OR –Were cases and controls similar with respect
assessed for PAST to reasons for exposure? (selection bias for
exposure controls?)
–Were circumstances and methods for
determining exposure similar for cases and
control? (measurement bias of exposure?)

Cohort Known exposure RR –Sample representative? (referral bias?)


follow FORWARD –Group homogeneous with respect to
until outcome prognosis? (self-selection bias?)
–Was follow-up complete? (selective drop out?)
–Were outcome criteria objective and
unbiased? (measurement bias of outcome?)
Experimental
Randomized Strength of association –Do groups start with same prognosis?
Controlled Trial between intervention (randomized? allocation concealed? similar
(exposure) and outcome risk?)
–Was study blinded? (all unaware of group)
–Do groups end with same prognosis?
(complete follow-up? intent to treat analysis?,
per protocol?)
*Descriptive statistics examples- measures of morbidity (incidence rate, prevalence rate, quality-adjusted life years),
measures of mortality (mortality rate, case-fatality rate, life expectancy, age-adjusted rates)
Internal Validity VS External Validity/Generalizability

(Internal) Validity External Validity/Generalizability


Definition/Question Is the statistically significant association “TRUE” Is the statistically significant association “TRUE”
for the sample group that was studied? Can we for the general population? Can we believe that
believe that the DIFFERENCE found between the DIFFERENCE found between the test group
the test group and control group really exists? and control group APPLIES TO OTHER SIMILAR
PEOPLE NOT in the STUDY?
Threats (things that Random measurement error/Non-differential Narrowly selected sample
make us question misclassification
whether the
difference and its Selection bias Volunteer bias
applicability)
Measurement bias/Differential misclassification Survival bias

Confounding Cohort effect

Effect Modification (Interaction)

Common Pitfalls and Advice

Factors affecting Selection Bias are frequently incorrectly listed as reasons for poor External Validity/Generalizability and the
threats to External Validity/Generalizability are listed under Internal Validity (See chart above, these are in separate
categories).

Sample size- Just because a sample “seems” small does not mean it is inadequate
–Were appropriate calculations done and what parameters were chosen (Paired/unpaired, Alpha (type I or false-
positive) error level, Is Beta error considered at all, Size of variance in data anticipated, What size difference do you want to
detect)

If patients are randomized, then selection bias (the groups being significantly different) is unlikely because the whole point of
randomization is to make sure the groups are different only with regard to the thing that you want to expose them to (the drug
or counseling, etc).

You might also like