Statistics

Question 1

Which of the following statistical metrics is used to measure the extent to which an observational study dataset derived from patient care registries can be generalized to broad patient populations?

A. External Validity
B. Selection bias
C. Random selection
D. Internal Validity
E. Blinding

Answer 1

External Validity

The answer is external validity. Internal and external validity are concepts that reflect whether or not the results of a study are trustworthy and meaningful. While internal validity relates to how well a study is conducted (its structure), external validity relates to how applicable the findings are to the real world. Factors That Threaten External Validity include pre- and post-test effects, Sample features, selection bias, and situational factors.

Question 2

Institution X introduced a change in practice in 2010, where every patient undergoing spinal fusion received pulse irrigation prior to closure of their incision. A neurosurgeon designs a study to investigate the effect of pulse irrigation on postoperative infection using patients who had surgery prior to 2010 as the control group. This study design demonstrates which of the following types of bias?

A. Selection bias
B. Publication bias
C. Observation bias
D. Confirmation bias
E. Recall bias

Answer 2

Confirmation bias

The correct answer is confirmation bias. Confirmation bias occurs when researchers, consciously or unconsciously, look for information or patterns in their data that confirm the ideas or opinions that they already hold. Observation bias occurs when participants in a study are aware that they are being observed by scientists and, either consciously or unconsciously, alter the way they act or the answers they give. Selection bias occurs when certain groups are under-represented. Recall bias occurs when different types of events are more likely to be remembered than others, causing respondents to report those types of experiences more readily. Publishing bias occurs when negative findings, for example when no significant results are found, are less likely to be submitted by scientists or published by scientific journals because they are perceived as less interesting.

Question 3

A diagnostic test has a sensitivity of 0.99 and a specificity of 0.99. It is administered as a screening test to a population with a disease prevalence of 1:10,000. Which of the following is the most appropriate counsel for a patient with a positive test result?

A. The probability that this patient has the disease is low
B. The probability that this patient has the disease is high
C. More information is needed to counsel the patient
D. None of the above
E. A positive test indicates that this patient definitely has the disease

Answer 3

The probability that this patient has the disease is low

The correct answer is A. Sensitivity is the percentage of true positives (e.g. 90% sensitivity = 90% of people who have the target disease will test positive). Sensitivity = (true positive) / (true positive + false negative) = Probability of being test positive when disease present.
Specificity is the percentage of true negatives (e.g. 90% specificity = 90% of people who do not have the target disease will test negative). Specificity = (true negative) / (true negative + false positive) = Probability of being test negative when disease absent.
Positive predictive value is the percentage of patients with a positive test who have the disease, which is what this question is asking for. PPV = (true positive) / (true positive + false positive) = Probability (patient having disease when test is positive). In this case, as the specificity and sensitivity of the test are high but the prevalence is low, the probability that this patient has the disease is low.

Question 4

A clinical trial shows that the risk of death in the control group is 50%, while the risk of death in the intervention group is 40%. Which of the following is the number needed to treat (the number of patients who need to be treated to prevent one additional death)?

A. 40
B. 30
C. 10
D. 50
E. 20

Answer 4

10

The correct answer is 10. The number needed to treat (NNT) is the inverse of the absolute risk reduction (ARR). The ARR is the absolute difference in the rates of events between a given activity or treatment relative to a control activity or treatment, i.e. control event rate (CER) minus the experimental event rate (EER), or ARR = CER - EER. If a drug reduces the risk of a bad outcome from 50% to 40%, the ARR = 0.5 - 0.4 = 0.1. Therefore, the NNT = 1/ARR = 10.

Question 5

Which of the following aspects of a prospective, multicenter, double-blind, randomized controlled trial is most critical in addressing the problem of error from unknown confounders?

A. Calculation of Power
B. Blinding
C. Randomization
D. Appropriate patient selection
E. Study setting

Answer 5

Randomization

The correct answer is randomization. Confounding variables or confounders are often defined as the variables correlate (positively or negatively) with both the dependent variable and the independent variable. A confounder is an extraneous variable whose presence affects the variables being studied so that the results do not reflect the actual relationship between the variables under study. The researchers therefore need to account for these variables - either through experimental design and before the data gathering, or through statistical analysis after the data gathering process. In randomization, the random assignment of study subjects to exposure categories to breaking any links between exposure and confounders. This reduces potential for confounding by generating groups that are comparable with respect to known and unknown confounding variables.

Question 6

A neurosurgeon designs a study comparing spinal cord stimulation to surgical decompression and fusion for patients with mechanical back pain. When evaluating five-year postoperative outcomes, the neurosurgeon notices that more patients are lost to follow up in the decompression and fusion group. The neurosurgeon attributes this finding to the fact that spinal cord stimulator patients routinely follow up for reprogramming sessions and generator replacement. This study demonstrates which of the following types of bias?

A. Observation bias
B. Transfer bias
C. Recall bias
D. Confirmation bias
E. Selection bias

Answer 6

Transfer bias

The correct answer is transfer bias. In almost all clinical studies, subjects are lost to follow-up. In these instances, investigators must consider whether these patients are fundamentally different than those retained in the study. Researchers must also consider how to treat patients lost to follow-up in their analysis. Well-designed trials usually have protocols in place to attempt telephone or mail contact for patients who miss clinic appointments. Transfer bias can occur when study cohorts have unequal losses to follow-up. This is particularly relevant in surgical trials when study cohorts are expected to require different follow-up regimens, such as this one, in which the spinal cord stimulator patients may be expected to follow up for reprogramming visits.
Selection bias occurs when certain groups are underrepresented. Recall bias occurs when
different types of events are more likely to be remembered than others. Observation bias occurs when participants are aware that they are being observed and alter the way they act based on this knowledge. Confirmation bias occurs when data is interpreted in a way that supports one’s prior beliefs.