|
Post by Admin on Jan 24, 2016 23:17:43 GMT
When conducting a clinical trial on a new asthma drug, a nonblinded physician participating in the study notices benefits in patients who are taking the drug rather than the placebo. Knowing that the experimental drug is actually beneficial, he begins intentionally to assign newly enrolled patients with more severe asthma to the experimental group rather than the placebo group. Which of the following types of bias is this?
A. Late-look bias B. Length bias C. Recall bias D. Sampling bias E. Selection bias
|
|
|
Post by Admin on Jan 24, 2016 23:28:18 GMT
Selection bias occurs when patients in a study are not randomly assigned to a treatment group. This can occur because either the patients or the investigators select the group that an individual patient will enter. Selection bias can be avoided by randomization of participants to treatment groups in a double-blind study. Late-look bias concerns information gathered at an inappropriate time in the study. It results in selection of patients with less severe disease because those with more severe disease died before detection. An example would be enrolling patients with metastatic breast cancer in a clinical trial 6 months after diagnosis instead of immediately.
Length bias (and lead-time bias) occurs when the temporal characteristics of a disease are not considered. An example would be a new cancer screening test that detects an incurable form of cancer 3 months before the patient becomes symptomatic. The patient's survival with the disease would be "increased" by 3 months with the new test, but only because disease was detected earlier and not necessarily because the screening was beneficial.
Recall bias occurs when a patient's report of symptoms or past events is selective, either intentionally or not. An example of when this is important would be in a survey of elderly adults with shingles when they are asked detailed questions about their childhood chickenpox infections.
Sampling bias occurs when a sample group is not representative of the population from which it is taken. An example would be a study looking at obesity levels in a local town, where the surveys were distributed in the local fast food restaurants.
|
|