B5.031 - Scholarly Projects Flashcards
when is mean preferred to measure central tendency for continuous data
symmetric distributions
what is the mean sensitive to
extreme values and will be biased in presence of outliers
when is median better for central tendency
for skewed distributions or distributions with outliers
when is mode a preferred measure of central tendency
pretty much never
what are SD and variance measures of
variability
when is interquartile range the best estimate of variation of total
in the presence of outliers
what will be higher mean or median

median, mean will be skewed low bc of the outlier
what is a t test used for
to compare mean values between groups
what are chi squared tests used for
test the difference between proportions between two groups
when can correlation be evaluated
when comparing two continuous variables, (eg. height and weight are highly correlated)
what is a risk ratio
a measure of effect calculated by dividing the risk of an outcome in one group by the risk of the outcome in a different group
Researchers conducted a randomized controlled trial to test the efficacy and safety of oral BG-12 (dimethyl fumarate) at a dose of 240 mg two or three times daily as compared to placebo in patients with relapsing-remitting multiple sclerosis. Patients were randomly assigned to receive either oral BG-12 twice daily, oral BG-12 three times daily, or placebo (control). The investigators compared the average baseline Expanded Disability Status Scale (a score ranging from 0 to 10 with higher scores indicating a greater degree of disability) (EDSS) score of patients across the three treatment groups (twice daily, three times daily, placebo) and reported a p-value of 0.7. What statistical test did the authors conduct to produce this p-value?
ANOVA
what is ANOVA
An ANOVA compares the mean of a continuous variable between two or more groups.
the difference between ANOVA, Chi squared, t-test
An ANOVA compares the mean of a continuous variable between two or more groups. Though a t-test is also used to compare the mean of a continuous variable between groups, it is not appropriate for comparing more than two groups. A chi-square test compares proportions between groups.
Researchers conducted a randomized controlled trial in 500 patients to test the efficacy and safety of catheter ablation for atrial fibrillation in heart failure patients. Patients were randomly assigned to undergo either catheter ablation or medical therapy (control). The primary outcome was the relapse rate over a period of 2 years. Two patients randomized to the catheter ablation group did not receive ablation. The authors implemented the intention-to-treat (ITT) principle and the outcomes for these 2 patients were included in the catheter ablation group. What is the best statement concerning the ITT analysis?
The results of the ITT analysis should not differ substantially from the per-protocol analysis because of minimal noncompliance. In this study, a small degree of non-compliance was observed and so the analytic sets for the ITT and per-protocol analysis will contain nearly the same set of subjects, resulting in relatively small differences in results from the two analysis methods.
what is an ITT
An intent-to-treat (ITT) analysis is the most appropriate method for determining efficacy when analyzing a randomized controlled trial (RCT) because RCTs often suffer from noncompliance and missing outcome data. ITT analysis includes every subject who is randomized according to his or her randomized treatment assignment, ignoring non-compliance, protocol deviations, withdrawal, etc.
what is the primary purpose of ITT
The primary purpose is to maintain the prognostic balance generated from the original random treatment allocation; however, in the presence of protocol deviations the ITT analysis generally provides a conservative (smaller) estimate of the treatment effect.
what is a per protocol analysis
a per-protocol analysis limits the analysis to only those study subjects who were treated (and observed) according to protocol—this approach actually measures the beneficial effects of treatment in patients who completed treatment. When there is a large degree of non-compliance in a study, a per-protocol analysis will exclude a large percentage of study subjects’ data.
Researchers aimed to evaluate the association between ASD and exposure to prenatal ultrasound. They conducted a case-control study that included 107 patients with ASD and 209 controls with typical development. Researchers created exposure categories of high and low. Given their study design, what measure of association would be the most appropriate for the authors calculate during data analysis to examine the relationship between ASD and ultrasound exposure level?
Odds ratio.
When a case-control study is conducted, risk cannot be directly calculated, since individuals are not followed forward in time to look for outcome incidence (new cases of disease). Instead, the odds of exposure among cases is compared to the odds of exposure among controls using an odds ratio.
what is a hazard ratio
A hazard ratio is used when time to event data are available (from a forward-directional study like a cohort or randomized controlled trial; these study designs start with exposure and follow people forward to look for the development of an outcome of interest).
when are risk ratio and rate ratios used
Both risk ratios and rate ratios are also used when incidence can be determined.
when are prevalence ratios used
Prevalence ratios can be used in a variety of situations, but are used in special situations when a comparison of two prevalence values is needed (as the name suggests).
Progesterone is essential for the maintenance of pregnancy. However, whether progesterone supplementation in the first trimester of pregnancy would increase the rate of live births among women with a history of unexplained recurrent miscarriages is uncertain. A recent systematic review reported on four small trials that examined the risk of miscarriage among women who received progesterone compared to the risk among women who received placebo or no treatment. An overall odds ratio of 0.39 (95% confidence interval [CI], 0.21 to 0.72) was reported, but the quality of the four trials was considered to be moderate to poor. What conclusions about the relationship between progesterone and the risk of subsequent miscarriage can be drawn from this study?
Early progesterone treatment reduced the risk of subsequent miscarriage.
in meta analysis what type of ratio is usually calculated
In meta-analyses, an odds ratio is typically calculated. The four studies being used in the example are described as “trials”, suggesting randomized controlled trials. Thus, even though an odds ratio is calculated, inferences about risk can be made.