Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
118 Cards in this Set
- Front
- Back
|
What is a primary study
|
collect new data to answer a specfic question in a population
|
|
|
What are secondary studies
|
attempt to combine or synthesize results from 2 or more primary studies (existing) to generate a global/overall answer to a question
|
|
|
Example of a primary study
|
single clinical trial
|
|
|
Example of secondary study
|
meta-analysis of several clinical trials
|
|
|
What is primary literuature
|
original researach reported
|
|
|
What is secondary/tertiary literautre
|
complications of research or knowledge in a field
|
|
|
What are GOALS of descriptive studies
|
record events observations, activtes, and therby describe a pop. disease
|
|
|
Descriptive studies provide
|
an INITAL picture of some population or disease
|
|
|
Descriptive studies DO NOT provide information about
|
causes or clinical efficacy
|
|
|
Prototpyical descriptive method
|
case reports/series, surveys, porgram description
|
|
|
What are goal of explanatory studies
|
compare and explain difference between things to shed light on the etiology or prognosis of disese or intervation
|
|
|
What study is used to establish "cause and effect"
|
eplanaotry study
|
|
|
Most common types of explantory studies
|
Case control, cohort (TRYING to establish cause and effect)
|
|
|
What are observational studies
|
observe subjects in naturalistic setting, measure exposures and outcomes, makie comparison and inferenes about cause-effect
|
|
|
What arole do researches play in observational studies
|
are "observers" only-do not conrol intervetion
|
|
|
Common obersvational studies
|
cross-section, case control cohort
|
|
|
What are experimental studies
|
study subject in a controlled setting measuing exposures and outcomes, and make CONCLUSIVE info about cause-effect
|
|
|
Do reseraches control the intervention in experimental
|
YES---who gets it and who does not
|
|
|
What is the gold standary method
|
RCT
|
|
|
Case reports/series are prototypical descriptive method
|
YES
|
|
|
What are CASE REPORT goals
|
share experiences
provide inital step towards more spohistaced research hypothesis generating |
|
|
Does the complexity of case reports vaires
|
YES-single cases give little quntiative info, weher many cases can give substanial info
|
|
|
PRos of CASE reports
|
quick, easy, inexpesnive, and hypothesis generation
|
|
|
Problems of Case reports
|
few (anecdotal info)
No control of any kind |
|
|
What is problems that case report have not control
|
bias--investagors looking for cause, chance, and confounding
|
|
|
Is it hard for case reports to answer the question "DOES DRUG X cause EFFECT Y"
|
diffiuclt to establish beyond suggestion--
|
|
|
What is a causality assement
|
used in case reports to help quantify the likelihood that a drug is cuasing an observed adverse reaction
|
|
|
What are Cross-Sectional Studies AKA
|
prevalance studies or surveys
|
|
|
What does Cross-sectional studies collect information on
|
disease prevalence
prgnosis population norms satifaction, opinion |
|
|
Cross-sectional studies can generate several things at one point in time,
|
YES
|
|
|
Cross-sectional studies generate PREVALENCE date, and measure correlations between variables, but DO NOT
|
gnerate data that can be used to support cause-effect
|
|
|
Correclation does not equal causation
|
HIGH CORRELATION DOES NOT EQUAL PROOF
|
|
|
What are advantages of CROSS Sectional studies
|
Quick inexpensive to conduction
good starting pt for future research |
|
|
What are disadvantages of cross-sectional studies
|
survey-biasa
response rates are low difficult to establish time sequencing of events |
|
|
What do Case-control studies RETROSPECTIVE REVIEWS THAT start with
|
PERSONS with AN OUTCOME (cases)
|
|
|
What do case-control studies compare with
|
similar people NOT having the outcome (controls)
|
|
|
Case-control studies look at rates of expousre of each group to
|
the potential cause
Rash--exposure to drug A No rash--exposure to drug A |
|
|
What is the odds ratio
|
case control study
Rate cases/Rate controls |
|
|
What are advantages of Case control studies
|
quick, inexpensive to conduct
simple results |
|
|
Case control studies are good for studing
|
rare/negative outcomes
mutiple possible cuases "exposures" |
|
|
What are disadvantages of case-control studies
|
dependent on historical date
biasis difficult to selcet/obtain appropriate control groups |
|
|
What are cohort studies
|
follow groups of subjects based on expousre status over time to determine risks out outcome
|
|
|
Cohort studies are often used to study
|
drug use in real world
disease pronosis 'natural course' Risk factors and relation to outcomes |
|
|
Two ways that Cohort studies can be conducted
|
prospectively
retrospectively |
|
|
What is prospectively
|
start today, and end at some pt in future (allow for control over measurement
|
|
|
WHat is retrospecitvely
|
Uses existing/historical records--allow for quick completion of study
|
|
|
What are advatnages of Cohort studies
|
more relistic than clinical trials, can control measurement and many biases, establishes time sequence of events, and good for studing rate exposures
|
|
|
What are disadvantages of Cohort studies
|
costly
significant degree of Loss to follow up difficult to establish cause-effect |
|
|
Randomized Controlled trials AKA
|
Clinical trials, or controlled clinical trials
|
|
|
What are 2 key compents of RCT
|
random allocation to treatments
investigator control of investigation |
|
|
What are types of RCT
|
parallel
Crossover |
|
|
What are is parallel RCT
|
in which each pts receive on e of the two treatments being compared
|
|
|
What is the most common design of RCT and fewest problems
|
parallel
|
|
|
What are parallel subtypes
|
Factorial
Large Simple Trials |
|
|
What is a factorial design
|
randomizing 2 treatment (when we are intersted in more than one factor, factorial design allow the simulatensou analysis of any number of factors
|
|
|
What are large simple trials
|
large studies with very limited includsino and exclusion criteria, better efectiveness info
|
|
|
What is Cross-over RCT or 2 Period Cross-over
|
Sujects are intially assigned to treatment A or B, then after a period of time "crossed-over"
|
|
|
When is a subject its "own control"
|
cross-over RCT
|
|
|
What type of study design more realistic based on EFFECTIVNESS
|
cohort studies
|
|
|
What type of study design is based on efficacy
|
RCT
|
|
|
Case-Control Studies caclculate what
|
"ODDS RATIO"
Rate cases/Rate controls |
|
|
Cohort Studies calculate what
|
"Relative Risk"
Rate of PPI/Rate H2RA |
|
|
Cross over RCT have limited number of siutations where acceptable such as
|
chronic stable disease
drugs with short-half life short-treatment periods baseline and washout peroids are feasible |
|
|
Probelsm with Cross-over design
|
inadeqatue washoutperiods
high dropout rates inadeate sample size |
|
|
What are problems with RCT
|
CRABS I
Controls Run-in periods Allocation Blinding Sample Size Intention to Treat |
|
|
What are RCT issuse with CONTROLS
|
active (comparative), placebo or historical
|
|
|
What are RCT issue with Run-in periods
|
used to establish definitive diagnosis of disease by washing current past treatments from body
|
|
|
What are RCT issues with Allocation
|
randomication must be true NOT SYSTEMATIC
|
|
|
RCT Issues with BLINDING
|
single, double triple, placebo vs dummy
|
|
|
RCT issues with SAMPLE SIZE
|
should be determined before to assure study have power to find a differance
|
|
|
RCT issues with Intetntion to TREAT (FDA requires)
|
all studies have some dropouts, results should reflect all subjects
|
|
|
What are advantages of RCTs
|
best cause effect design
Gold standard for FDA approval |
|
|
What are disadvantages of RCTS
|
costly, time consuming
"articifical" may be unethical/unfeabsilbe |
|
|
Some biasis still remain in RCTs such as
|
volunteer biasas, referral bias, Hawthrone effect, and placebo
|
|
|
What is Hawthorne effect
|
people act different when their being watch
|
|
|
If one RCT is good, combinding several is even better
|
YES
|
|
|
What are 3 major types of review articles exist
|
1. Review article
2. Meta Analysis 3. Systmatic Review |
|
|
What is a review article
|
>1 experts review topic make recommendations
|
|
|
What is a meta-analysis
|
results of several RCTS are combined to create a single ANSWER
|
|
|
What is a systematic review
|
reviews the atricle relavnex, and methodologic quality of revelvant articles
|
|
|
What are advantages of Meta-Analysis
|
produce singel result
may help to resovle controversies relativelt quick |
|
|
Disadvantages of meta-anyalysis
|
publication bias
literature retreival biasis study pop differance different study methods |
|
|
What is publication biasis
|
some studies are not published
|
|
|
What ia literature retrival bias
|
english only studies
|
|
|
In clinical research to genrate and or test hypothesis of causes and effect we must measure things so we created
|
variables
|
|
|
Variable may be more of less precise (examples
|
BP is 120/80 or simply HIGH
|
|
|
Variables are used for the classification of
|
exposure
outcome covariates |
|
|
Exposure is
|
IDEPENDENT variable (predictor)
|
|
|
Outcome is
|
DEPEDNENT vaiable aka effect one we are trying to exaplin
|
|
|
What are covariates
|
control variables
|
|
|
Surrogate or final outcome/effects are indicators of outcome
What are surrogate markers |
systolbic BP labs
|
|
|
What are final outcomes
|
MI, pateint important
|
|
|
What are the scales of measurement
|
NOIR
nominal ordinal interval ratio |
|
|
What is nominal
|
putting people in cateroges either
dichotmous YES/NO--cured/not categotrical--relgion |
|
|
What is ordinal
|
caterioes rank-order
consumer reposts 1.best product 2. But number 1 product is not 2x better |
|
|
What is interval
|
units of equal magitude and rank order (temp) (w/o absolute zero--0 means somthing
|
|
|
What is ratio
|
interval scale WITH absolute zero zero means nothing
|
|
|
What is difference between discrete and continous
|
discete (integer) whole number
continsous--any values |
|
|
Are many cotinous variable reported in discrete form
|
YES
|
|
|
Most statistical tests require you to classify as
|
Nominal
Ordinal Numerical (interval, ratios, disete or continous |
|
|
What does validty refer to
|
degree to which a measure represents a TRUE value
|
|
|
What is reliability
|
how reproducilbity is your data
|
|
|
What are types of validity
|
internal validity
external validty |
|
|
What is internal validity
|
how well the study was done internally (research design)
|
|
|
What is external validty
|
how well the study result apply to other unstudied subject---statistic is NOT invloved (ONLY JUDGMENT)
|
|
|
What is unsystematic variablity
|
obsevers assesing the same subject the same symtpom get different results can be subject issue or oversver issue
|
|
|
What is good news about insustematic varialbity
|
errors tend to even out
Law of averages--so overestimate and some underestimate |
|
|
What are systematic error more serious than unsystematic (random)
|
predictable BIAS
investigator subject biases |
|
|
How can we contorl measurement error
|
Blinding
Clear stands for measurements trained observers or multriple obersver |
|
|
What are the 3 major threats to study validity is
|
chance bias, and confoudning
|
|
|
What are 2 genral strategies to deal with threats to validity
|
Address in reasearch design phase
adress in analysis phase |
|
|
How does one deal with chance in design phase
|
Use largest possible sample size, and standarize instructemtns, and
|
|
|
How does one deal with chance in analysis phase
|
use statistic to evaluate likehood of observed result
|
|
|
How do you deal with bias in design phase
|
use best possible study design question and minimize preventable biases and use RCT (CRABS)
|
|
|
How do you deal with biasis in analysis phase
|
collecting additoan info
|
|
|
How do you deal with confound in design phase
|
use RCT design when possible and practilcating, and matching
|
|
|
How do you deal with coundfound in analysis phase
|
control table
mutivratie stsitical analysis |
|
|
What are 3 questions you should asssk after reading any study
|
What was measured
How was it measured Relability and validity |
