Section 1
Preview this deck
What is the gold standard for a study
Front
| 5 | 0 | ||
| 4 | 0 | ||
| 3 | 0 | ||
| 2 | 0 | ||
| 1 | 0 |
Active users
0
All-time users
0
Favorites
0
Last updated
6 years ago
Date created
Mar 1, 2020
Cards (73)
Section 1
(50 cards)
What is the gold standard for a study
Randomized, double blinded clinical trial
How does the positive predictive value change with disease prevalence
It increases as the prevalence of a disease increases If the prevalence of a disease is low, the PPV is also low (PPV is the same as the direction of the prevalence)
When can the odds ratio be used to approximate the relative risk?
When the disease being evaluated has a very low prevalence. The Rare disease assumption says that when the prevalence is very low, the odds ratio from a case control study can be effectively used to estimate the Relative risk
What are cohort studies
They evaluate people without a disease and categorize them into exposure vs not exposed. They are then followed over time, and the incidence of the disease is determined later in each group to evaluate the exposure risk
What is the purpose of using Matching during a case control studies
Decrease the likelihood of a confounding variable
What kind of study evaluates a population to determine disease prevelance
cross sectional
What do cross sectional studies measure
Prevalence of disease
How can you use the sensitivity to calculate the false negative ratio
1 - sensitivity
Do you use sensitivity or specificity to screen
sensitivity
What are case control studies
A study to compare the frequency of expsure to a possible risk factor between people with a disease, and people without a disease Evaluates people with lung cancer and people without lung cancer and compares exposure to asbestos between both groups
What does a positive likelihood ratio indicate
Shows how much the odds of disease are increased if the test result is positive If a very positive LR, then a positive test means that it is very likely the disease is present (unlikely to be false positive)
What do cohort studies measure
relative risk and incidence of disease
How are groups in case control studies selected
Selected as either having the disease or not, then evaluating those people for the risk factor in question
Are cohort studies retrospective or prospective?
Can be either - the cohorts can be assembled now and watched for a prospective study, or gathered from past records and followed to current day for a retrospective study
What is selection bias
When participants in a study are selected in a way that causes them to be more or less likely to have a disease, exposure, etc, than the general population
How do you calculate the odds ratio, given the disease vs risk table values
AD/BC
If a test has a NPV of 97 percent, and a patient's test is negative, what is the likelihood that the patient actually does have the disease
3 percent
When can you use a case control study to determine the relative risk of an exposure to a disease
when the prevalence of that particular disease is very low
What is the attributable risk percentage
The excess risk a population exposed to something has of developing a disease Calculate: (Risk of smokers developing lung cancer minus the risk of non-smokers developing lung cancer)/risk of smokers developing lung cancer Also can be calculated with Relative risk. (RR - 1)/RR
What kind of study measures odds ratio
case control
How do you calculate the false positive ratio using specificity
1 - specificity
What is the power of a study
probability that the study will find a statistically significant difference when one is present How likely is the study to find the difference that exists
What is a likelihood ratio?
expresses how much more or less likely a given test result is in diseased as opposed to nondiseased people
What study do you use to measure the incidence of a disease
cohort study
What type of error states there is a relationship when there really isn't one
Type 1 error Alpha error
What type of error states there is not a relationship when there really is
Type II error Beta error
What is the negative predictive value
Probablity that a patient with a negative does not have the disease If the pregnancy is negative, a test with a high NPV means they likely are not pregnant
What are cross sectional studies
A study that is a snapshot in time to assess risk factors and outcomes of people in a particular population
How to calculate False negative ratio using the sensitivity
1 - sensitivity
How do you calculate Negative predictive value
D/(C+D) Number of disease free people with a test divided by the total number of those with negative tests
What does a negative likelihood ratio indicate
Shows how much the odds of a disease are decreased if the test is negative A very negative LR shows that if the test is negative, it is very unlikely the disease is present (false negative is unlikely)
How do you interpret relative risk
A relative risk over 1 is indicates an increased risk of disease because of an exposure, with the higher number indicating a more increased risk (smokers getting cancer, RR of 4 indicates a higher connection and risk than a RR of 3) A relative risk below one indicates that there is a decreased risk of getting a disease following an exposure to something (smoking having a RR of 0.5 for endometrial cancer, indicates smoking is protective against endometrial cancer). Again, the farther from 1, the more protective an exposure is to the disease
What kind of study determines whether or not someone has an exposure to something, then follows them to assess for the presence of a disease
cohort study
How do you calculate relative risk, given the disease vs risk table values
(A/A+B) / (C/C+D)
What is Matching during a case control study
When researchers match participant characteristics. They match females to females, old people to old people, etc
If you decrease the lab cutoff that indicates a positive value, what happens to the PPV, NPV, specificity, and sensitivity? Ex, if you change the D-Dimer level from 0.5 to 0.3, what are you doing to the PPV, NPV, specificity, and sensitivity
The NPV and sensitivity will increase, while the PPV and specificity decrease By decreasing the level, you are saying I'm not taking any chances that the elevated level isn't just incidental, and will consider it for any level. Because now the value for a positive test is incredibly low, you know that if the test is negative, that clot is not there
What is the positive predictive value
Probability that a patient with a positive test has the disease If the pregnancy test is positive, a test with high positive predictive value means they are likely pregnant
How do you calculate the number needed to treat
1/absolute risk reduction (the differences in percentage of disease occurrence) If 20 percent of people get the flu, and only 10 percent who are vaccinated get the flu, then ARR = 0.20 - 0.10 = 0.10 (subtract the percent, not the whole number) In this case, the NNT is now 1/0.10 = 10
What is relative risk?
Expresses how much more likely a disease is to happen in a person who is exposed to something (how much more likely is someone to get mesothelioma from asbestos). It indicates a relative strength of connection between the disease and exposure RR = Incidence in exposed/incidence in unexposed
How do you calculate the positive predictive value
A/(A+B) Number of Diseased w/ positive test divided by the total number of those with positive tests
What is an odds ratio
It compares the odds of an exposure in people with a disease compared to those without a disease. It is an estimate of relative risk that is used in Case control studies. Case control studies are cheap and easy, but can't calculate RR, only the odds ratio.
If you increase the lab cutoff that indicates a positive value, what happens to the PPV, NPV, specificity, and sensitivity? Ex, if the D-Dimer significant positive result is moved from 0.5 to 0.7, what happens to the PPV, NPV, specificity, and sensitvity?
If the cutoff point is increased, the PPV and specificity will increase, because it is a more significant value, decreasing the likelihood that the value is incidentally increased NPV and sensitivity will decrease, because you are raising the level that is required to have the disease (they could have a clot with a D-dimer of 6, but now because the test is 7, this will be ruled insignificant). By increasing the level, you are making sure that the test shows they have the disease and that it isnt incidental
What is the number needed to treat
How many people need to be treated to benefit one person Give antibiotics to 7 people for one person to actually benefit from them
What is the Absolute risk reduction
Difference in risk attributable to an intervention. (absolute, not a proportion). If 20 percent of the population get the flu, and only 10 percent who are vaccinated get the flu, the Absolute risk reduction is 10 percent.
What value is calculated using a case control study
The Odds ratio
Do you use a sensitive or specific test to confirm a diagnosis
specific
What kind of study gathers people into diseased vs undiseased then assesses for risk factors
case control
Are case control studies prospective or retrospective?
Retrospective
How does the negative predictive value change with changes in disease prevalence
The lower the disease prevalence, the higher the negative predictive value. If there is a high prevalence, the NPV is very low (NPV is opposite the direction of the prevalence)
How is the power of a study measured
1 - Beta (type II error) (aka probability of determining there is no relationship when there really is one)
Section 2
(23 cards)