Home > Preview
The flashcards below were created by user
on FreezingBlue Flashcards.
What is internal validity?
The degree to which the results of a study are attributable to the treatments rather than the flaws in the research design.
What's external validity?
The extent to which the conclusions of the study are correct for persons beyond those who were investigated
What is reliability? What is reproducibility?
Reliability = repeatability, tightness of the range of data values in the same laboratory
Reproducibility = method yields same result accross laboratories
Describe case-series (descriptive)
Case series = Description about a group of UNIQUE people, leads to the generation of hypotheses that may be subsequently investigated in case control, prospective cohort, cross section or experiential studies
What are the pros and cons regarding case series?
Pros: simple, no protocol, clinician only, inexpensive
Cons: investigator bias possible, no hypothesis testing, no masking, no hypothesis testing
Describe Case control (retrospective)
Retrospective backward looking study that analyzes previous events to find causes, association and risk factors of the dz. Longitudinal.
- Examine CASES and CONTROLS
- How did they get sick/stay well?
Pros and cons of case control?
Pros: quick, inexpensive, best for rare dz, best for identifying the cause of sudden outbreaks
Cons: recall bias/missing information
Prospective Cohort Study?
aim at identifying risk factors, natural history, incidence of disease
Study looks at incidence of dz and risk factors
What are the pros and cons of a prospective cohort study?
- Advantages: Good for multiple outcomes
- Disadvantages: bad for rare dzes, timely bias due to follow up, expensive
Looks at past at a group that was treated similarly to determine outcome
Pros/Cons of retrospective cohort?
Pros: good for determining risk factors/associations, easy, inexpensive
Cons: biased from incomplete records, confounders
What's the difference between case control and retrospective cohort?
- Retrospective Cohort:
- 1) looking forward in time from time of tx
- 2) pop = cohort receiving different treatments
- 3) how did they progress?
- Case control
- 1) looking backwards
- 2) pop not a cohort
- 3) how did they get sick/stay well?
What's the difference between a controlled and uncontrolled case study?
Controlled: one new drug tx compared to another drug/procedure that's either a placebo or an accepted tx (gold standard tx)
Uncontrolled: new tx is described but not compared to anything else
Experimental Study Groups: what is a concurrent control group?
- A group treated similarly to the test group in the same
- time frame (sham surgery)
Hawthorne effect is ________
- Hawthorne effect: when subjects change
- their behavior just bc they are in a study
What are 3 ethical concerns with clinical trials in research?
Should not compare to an inferior tx
Should not w/hold treatment known to be effective
Patients must sign informed consent form stating risks of participation
What is a Meta-Analysis?
hint: Meta means later and more highly organized
- A Meta-Analysis uses published data from several studies to produce one grand overall conclusion
- --> pooling different sources
What is a sample?
A group of people representative of the population
Statistical integrity is dependent on the obtainment/collection of a __________
Truly random sample
Random samples provide you with results with ________ probability of which you _____make inferences concerning the population of interest.
Non-random samples provide you with results with _______ probability of which you _____ make inferences concerning the population of interest.
Random: known, can
Non-random: unknown, cannot
T/F Method of sampling is vital to the determination of probabilities associated w/ various outcomes
Describe the process of randomization in the selection of subjects for a clinical study.
- Population of interest -->
- random selection of subjects -->
- sample population -->
- inclusion/exclusion criteria -->
- informed concent obtained -->
- random assignment -->
- subjects enrolled in study
Describe the 4 different types of samples and whether they're random or not.
Simple Random = selection of one unit does not affect the selection of another, each unit has an identifier number, each unit has an equal chance of being selected
Sequential/systemic random sample= selection of one unit does influence the selection of another, i.e. the first one is chosen randomly and then after that every 17th unit is chosen.
Randomized block study: subjects/units organized into blocks, once a block is randomly chosen, all those units/subjects are in the study, this is NOT a truly random sample
Convenience sample: non-random sample, no element of random sampling, external validity seriously questioned (SCCO student research is usually convenience sample)
What elementary school grades are required by lawmust to have vision screenings?
1st, 3rd, 6th
Describe the goal of the Orinda study and what was developed through it.
o Population: screened 1163 children in grades 1–8
o Goal: design the least expensive, least technical, most effective school screening program
o Modified clinical technique = VA, r’scopy, CT, color vision, ophthalmoscopy (VRCCO)
Effective in identifying more than 90% of kids w/ RE/binocular vision problems
90% sensitivity (test + = disease +)
Describe the VIP STUDY: vision in preschoolers
Goal: determine best methods to screen for major eye conditions by nurses + lay personnel
- Part I--> screening vs comprehensive evaluation
- 2 screening tests used = 2/3 sensitivity for RE, but for more complicated problems (high RE, amblyopia, aniso) it was 90% sensitivity
Part II --> Preschoolers had comprehensive evaluations and then screenings done by nurses and lay people. Outcome = nurses and lay people had about the same 85% sensitivity. Results were similar to the results from comprehensive exam. CONCLUSION: VISION SCREENINGS FOR CHILDREN DO NOT NEED TO BE CONDUCTED BY LISCENSED PROFESSIONALS
What is specificity and what is sensitivity?
Specificity = ability to detect true negatives
Sensitivity = ability to detect true positives
Describe an ROC curve
ROC curves: maximize sensitivity + specificity
- ROC = receiver operator characteristic curve
- x = false positives (1 – specificity)
- y = true positives (sensitivity)
Use different tests or combinations of several tests to provide largest area under curve
Maximum area under the curve (1.00) yields maximum sensitivity + specificity
Incidence = _______
what is incidence as a ratio?
probability of developing a certain condition over a period of time
new cases/population at risk (standardized number i.e. 100 ,1000, 10000)
Prevalence = ______________
Probability of people having a particular dz at a given time
existing case of dz/total population at risk (standardized number i.e. 100, 1000...)
What is point prevalance?
What is period prevalance
prevalence at a point in time
prevalence at a point and new cases recurrence within a period of time
Prevalence vs Incidence
1) Which one has a time component?
2) Which one is preferred when discussing dz etiology? When studying health services utilization?
- 1) Incidence
- 2) Incidence, Prevalence
CLEER study = ?
SEE study = ?
CLEER = study of refractive erro over time in specific ethnic grous
SEE = study of dry eye