EVIDENCE BASED FINAL

false

T/F most epidemiologist would conclude that case-control study reliably estimates the odds of disease risk when properly designed

all free from cancer

all participant in cohort study in smoking and oral cancer have the following characteristic at base line:

false

T/F patients are randomly selected from a population in most randomized controlled trials

residual confounding

bias which is caused by imperfect adjustment

false

T/F it is simple to account for confounding bias

overdiagnosis

increase of disease with increased screening but mortality rate remains the same

better control over confounding

why does mendelian randomization hold great promise for the future in providing more reliable evidence

a business man

Claude hopkins had significant impact on the hypothesis that plaque removal prevents carie. claude hopkins was:

a person with interproximal dental caries

in a case control study on interproximal dental caries and dental floss habits, a case is

identifying changeable causes of ill health

clinical practice and public health is mostly focused on

false

T/F George Davey Smith in the YouTube movie on Mendelian randomization talks about how a disease and the identification of a disease (e.g., high blood pressure) can lead to a change in behaviors (e.g., reduced alcohol drinking).  George Davey Smith subsequently states that such biases can be controlled for in an observational study by careful history taking of the participating subjects

false

T/F George Davey Smith reports that genotyping study participants is expensive making Mendelian randomization studies very expensive when compared to randomly assigning study participants to treatments (i.e., randomized controlled trials).

true

True or false: George Davey Smith reports that mothers who drink during pregnancy in the studies had a higher education than those who did not.  

genetic variant are unrelated to confounders

What is the reason Mendelian randomization is superior to observational studies according to George Davey Smith

confounding

George Davey Smith describes which bias when pointing to the lack of trustworthiness of epidemiological findings (which fall within the zone of bias)

false

True or False:  George Davey Smith states that cohort studies provide the best way of getting the most reliable evidence that a changeable risk factor suspected of being beneficial might influence disease.

true

True or false: George Davey Smith states that there are genetic variants which relate to drinking more or less alcohol.

true

T/F George Davey Smith reports that Mendelian randomization shows that for "mothers who drink more alcohol during pregnancy, their offspring actually do worse on IQ tests and in school performance rather than do better as the observational studies suggested".

retrospective

registering trial AFTER data collection has begun

2000

When approximately did clinical trial registration become required

time

what variable does the x-axis represent

size (extent) of disease

what variable does the y axis represent

size at which disease causes death

what does the red circle indicate?

size at which disease of interest causes morbidity

what does the red circle indicate?

start of disease progression

what does the red circle indicate?

death from other causes

what does the red circle indicate?

fast and slow

which arrow leads to morbidity during a person’s lifetime?

very slow and non-progressive

For which of the arrows can screening only cause harm.

fast

Which of the arrows is the least likely to be captured with regular screening programs

non-progreessive and very slow

which arrows lead to overdiagnosis

overdiagnosis

the identification of a disease that will not cause symptoms or death during a patient's lifetime, often leading to unnecessary treatment

true

T/F As a general rule, randomized controlled trials can be used to reliably test small risks or benefits, epidemiology cannot.

turkey shoots

the consistent identification of large risk increases for diseases (i.e., large relative risks for smoking and lung cancer, huge relative risk for hep B and liver cancer)

true

True or false: Success stories in epidemiology are often the result of "turkey shoots"

true

True or false: Compliance with the dentist recommendation to use dental floss is one of the many markers for healthy user bias. 

unexpected harm

which result of epidemiological study is most reliable

hard measures (ex:bone biopsy)

most reliable exposures to measure in epidemiological studies

towards null (ie. 1)

The crude odds ratio between periodontitis and a marker of infection is 3.3. When the odds ratio is adjusted for BMI it equals  2.2.
The change in direction from the crude to the adjusted odds ratio is ...

true

True or False: Randomized controlled trials typically do not adjust for confounding variables (magic!), epidemiological studies typically have to adjust for confounding variables and the ability to do so is not perfect

prescriber bias

the influence of a prescriber's personal beliefs or preferences on their prescribing practices, potentially leading to inappropriate treatment choices.

false

True or false:  Case-control studies are associated with a smaller zone of bias than cohort studies

decreases odds of outcome

a odds ratio below 1 indicates that the exposure:

statistically insignificant

a confidence interval that includes 1 is:

zone of bias

The range in which results are not clearly conclusive and are more likely to be affected by bias, confounding, or random error.

true

T/F a statistically insignificant finding fall within the zone of bias

false

T/F An odds ratio moving from 0.61 to 0.66 is a move away from the null hypothesis

true

T/F when a model has more confounders, trustworthiness increases as the odds ratio moves away from the null hypothesis

confounding

Which bias has been referred to as the Achilles heal of epidemiology?

selection bia

distortion in measure of association to to sample selection that doesn’t reflect target population

14%

percent of clincical trials that are fabricated

surrogate dont always relate to real world health benfits

importance of true endpoints

cohort study

Follows a group over time to assess exposure-outcome relationships

case-control study

Starts with outcome (cases) and compares past exposures to controls

.5 - 2.0

zone of bias for cohort studies ?

Statistical significance

confidence interval does not include 1.0

Harm (increased risk)

relative risk > 1

Benefit (decreased risk)

relative risk < 1

200% risk increase

what does a relative risk of 3 mean

endpoint switching

Changing primary outcomes after a trial has started or ended — can mislead results and inflate findings

confounding

major source of bias in observational studies

associated with exposure and independently affects outcome

what two criteria need to be satisfied for a variable to be a confounder

reverse causality

The situation in which the outcome influences the exposure, rather than the other way around, leading to misleading conclusions in studies.

evenly distributes confounders

Why is randomization is magic?

Null hypothesis

relative risk and odds ratio = 1

confounding inflated original result

if an odds ratio move towards null (closer to 1) after adjusting for confounding variable what does that suggest?

stronger real association

If a value moves away null (farther from 1) after controlling for confounders what that suggest?

placebo

positive effects from inert treatment due to expectation of benefit

nocebo

negative effect from inert treatment due to expectation of harm

overdiagnosis

detecting disease that would not cause symptoms or harm in a person's lifetime

reduce frequency of screening, limit screening to high risk, raise diagnostic threshold, RCTs

What actions did the National Cancer Institute suggest to minimize overdiagnosis?

earlier detection of condition that will cause death and early treatment more beneficial

Two criteria needed for screening to provide tangible patient benefits.

lead time bias

refers to the perceived increase in survival time due to earlier detection of a disease, despite no actual increase in lifespan.

length time bias

occurs when slower-growing diseases are more likely to be detected by screening, giving an illusion of improved survival rates.

incidentalomas

Accidental findings that may not be clinically significant but can lead to unnecessary treatment

backed by high-quality randomized evidence

what is a requirement of preventative interventions for healthy individuals

 linear-no-threshold

Radiation doses greater than zero will increase risk of excess cancer or herritable disease in proportionate manner in low dose range

zombie trial

Clinical trials that continue to be cited or used in practice despite being outdated, flawed, or refuted

.33 - 3.0

zone of bias for case control study