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;
40 Cards in this Set
- Front
- Back
|
What is Epidemiological Research?
|
-focuses on large groups of people
-describes the distribution of health and illness -studies risk factors -can be descriptive, analysis of relationships or analysis of differences -usually NONEXPERIMENTAL (documents existing illness rather than attempting to change the state of illness) |
|
|
Ratio
|
epidemiological study
-a/b ex) what is the ratio of inversion to eversion ankle sprains? |
|
|
Proportion
|
epidemiological study
=a/(a+b) ex) what proportion of ankle sprains are inversion sprains? |
|
|
Rate
|
epidemiological study
-a proportion expressed over time --> # of people w/inversion ankle sprains in one year divided by total population at risk a)crude rates= based on entire pop at risk b)specific rates= based on a subgroup of pop (like athletes) c)adjusted rates= used to compare two pops with unequal proportions |
|
|
Prevalence
|
epidemiological study (a proportion)
=# of existing cases in a population at a given point in time *prevalence will go up as time goes on (ie: study about ppl w/diabetes) -calc this by taking # of cases tht exist divided by total pop at risk |
|
|
Incidence
|
epidemiological study
-gives rate of development of new cases in a population at risk =new cases divided by pop at risk during that time period ex)how many ankle sprains existed over one year? |
|
|
Relationship btw Incidence and Prevalence
|
Incidence= New Cases (INCidence)
PRevalence= Existing cases (PREvalence) |
|
|
Relative Risk
|
epidemiological study
-to compare the probability of disease in groups with different risk profiles =[a/(a+b)] / [c/(c+d)] *compares the incidence rate of one group with the incidence rate of another over same period of time *consider table horizontally *must reference entire population at risk |
|
|
Odds
|
epidemiological study
a/c or b/d -only looking for odds of ONE situation *consider table vertically ex) what are the odds that a person with lateral epicondylitis will be a tennis player vs. not a tennis player? |
|
|
Odds Ratio
|
epidemiological study
(a/c) / (b/d) *consider table vertically (comparing the two vertical columns) -comparing the odds of having and not having the disease from ppl of different subgroups ex)How much higher are the odds of getting lateral epicondylitis for tennis players vs. non-tennis players? |
|
|
When to use Risk Ratios vs. Odds Ratios
|
Use Risk Ratios when...
*you can measure representative samples from an entire pop at risk *you want to know the probability of having a condition btw subgroups Use Odds Ratios when... *you don't have access to the entire pop at risk *you want to understand if a risk factor (athlete vs non-athlete) increases the likelihood of a condition |
|
|
3 common epidemiological research designs
|
1. Cross-Sectional Designs
*used to document health status at a SINGLE POINT in time for each participant in study 2. Case-Control Designs *start with effects, look for causes *allows researchers to study things that may not otherwise be ethical 3. Cohort Studies *researchers work forwards from cause to effect *subjects are selected bc they do NOT have the condition of interest but may have risk factors that could cause it *they are followed for a certain period of time to compare relative risks |
|
|
What is non-experimental research?
|
-does not involve manipulating variables
-independent variable cannot be controlled by researcher -results do not show cause and effect |
|
|
Examples of Non-Experimental Study Designs
|
Case-Control (start with effects, look for causes- compares retrospective data)
Cohort Design (cause to effect) Cross-Sectional Design (look for inter-relationships btw observable variables) Epidemiological Research (cross-sectional design) Cost-Analysis Research (cost of care and relationship to quality of care) |
|
|
Types of Group Designs
|
(experimental)
1. Single Factor Design 2. Multi Factor Design 3. Repeated Measures Design 4. Mixed Design 5. Randomized vs. Randomized-Block Designs |
|
|
Single Factor Design
|
Experimental Group Design
One Independent Variable (may have one or more levels) ie: ind var= walking aid, levels= 3 diff types of walking aids |
|
|
Multi Factor Design
|
Experimental Group Design
2 or more Independent Variables (with 2 ind variable, multiply numbers of levels in each variable to get total number of groups needed for study) ex: 2 ind var, i with 2 levels, the other with 3 2 x 3 study- needs 6 groups |
|
|
Repeated Measures
|
Experimental Group Design
all subjects receive all conditions being tested |
|
|
Mixed Design
|
Experimental Group Design
-2 Independent Variable -1 is between-groups factor (different groups for each level of ind var) -1 is within-groups factor (same as repeated measures- everyone in this group received all levels of independent variable) |
|
|
Randomized vs. Randomized-Block Designs
|
Experimental Group Designs
Randomized= active variables (subjects can be randomly assigned) Randomized-Block= attribute variables (subjects cannot be randomly assigned- ie: gender, race) |
|
|
Types of Non-Experimental Research
|
1. Descriptive
2. Analysis of Relationships 3. Analysis of Differences |
|
|
Descriptive Research
|
Non-experimental
-purpose is to document a condition through systematic collection of data -no cause or effect -retrospective of prospective -uses observation, examination, interview or questionnaire to collect data |
|
|
Analysis of Relationships
|
Non-experimental
-analyzes relationships among variables -single group is tested on several different variables -retrospective (use medical records) or prospective |
|
|
Analysis of Differences
|
Non-experimental
-looks for differences btw groups or treatments (usually 2 groups) -retrospective or prospective |
|
|
Single-System Design
|
Experimental
-involves one group with repeated measures -usually involves baseline, treatment and withdrawal (or modification) of treatment 5 basic variations: 1) A-B Designs 2) Withdrawal Designs 3) Multiple baseline designs 4) Alternating Treatment Designs 5) Interaction Designs |
|
|
A-B Design
|
Experimental
-variation of Single-System Design -baseline taken, followed by treatment phase -only one group |
|
|
Withdrawal Design
|
Experimental
-variation of Single-System Design -aka A-B-A design -baseline taken, followed by treatment, then another baseline taken |
|
|
Multiple-Baseline Design
|
Experimental
-variation of Single-System Design -conduct several single-system design studies with baselines at different time or for different durations |
|
|
Alternating-Treatment Design
|
Experimental
-variation of Single-System Design -multiple treatments administered independently of each other -usually used with treatments that have short-lived effects |
|
|
Interaction Designs
|
Experimental
-variation of single-system design -used to evaluate the effect of different combinations of treatments |
|
|
Limitations of Single-Systems Design
|
-ethical dilemma of withdrawing a treatment that may be helping
-weaker single-systems designs are subject to internal validity threats- not enough control for extraneous factors that may affect outcomes -designs are difficult to replicate (bc they're so individualized) -single-systems designs are in their infancy |
|
|
What is Diagnosis?
|
the process and end result of evaluating information from examination
-help determine most appropriate intervention |
|
|
What is Screening?
|
the examination or testing that is used to separate those who are well from those who have an undiagnosed disease or conditon
|
|
|
Specificity
|
when a screening test identifies ppl without the disease as negative
-if a test has good specificity it won't make a lot of false positives *bc of this, it is safe to assume a positive result from a test with high specificity is a TRUE positive SPPIN -high SPecificity -Positive test -rule it IN (person HAS disorder) |
|
|
Sensitivity
|
when a screening test identifies ppl with the disease as positive
-if a test has good sensitivity it will not miss anyone who has the disease (however it may have false- positives) SNOUT -high Sensitivity -Negative test -rule it OUT (person DOES NOT have disorder) |
|
|
Likelihood Ratios (LRs)
|
ratio of likelihood that someone has a condition (used in sensitivity/specificity)
LR+ (positive LR) =likelihood that a pos test result was obtained in a person w/the condition vs. a person w/o the conditon *LR+ should be high in a useful test LR- (negative likelihood ratio) =likelihood that a neg test result was obtained for a person w/the condition vs. a person without the condition *LR- should be LOW in a useful test |
|
|
3 ways to synthesize literature
|
Narrative Review
Systematic Review w/meta analysis Systematic Review w/o meta analysis |
|
|
Narrative Review
|
a method for synthesizing literature
-usually summarizes reviewed articles in a series -doesn't provide details about methods of review -doesn't include formal evaluational scoring -usually by one reviewer |
|
|
Systematic Review w/o meta analysis
|
method for synthesizing literature
-must include documented research strategies -must include explicit inclusion/exclusion criteria (of studies reviewed) -must have at least 2 reviewers -formal ratings of study quality -NO meta analysis |
|
|
Systematic Review WITH meta analysis
|
method of synthesizing literature
-most desirable type of review -includes meta analysis= statistical pooling of results across studies |
