Biostats, Behavioral & MJP Flashcards
What are the chances I really do NOT have cancer?
NPV - free of disease with a (-) test result
High pre-test probability –> low NPV/likely to have disease
What are the chances that I DO have cancer?
PPV
- proportion of (+) results that are true (+)
Appearance of new cases
Incidence
Those without the disease at a certain time
Prevalence
Incidence x time
Inc prevalence but stable incidence
Something that prolongs the disease = improved quality of care
Prior knowledge of exposure status is what type of bias?
Observer
Longer survival in patients who underwent screening that allowed for earlier diagnosis is what type of bias?
Lead-time
Subjects selected for a study or selected from a group of w/ losses during follow-up is what type of bias?
Selection bias
120 out of 400 taking an ACE developed an MI while 100 out of 300 not taking a ACE experienced an MI - what is the relative risk of MI taking and ACE-I?
RR = 120/400 / (100/300)
Incidence in exposed / incidence in unexposed
Odd ratio
ad/bc
What type of study selects patients with a disease and without and then determines their previous exposure status?
Case-control
What type of study tracks pts with a known condition to document hx or response to tx?
Case Series
What type of study simultaneously measures exposure & outcome using surveys?
Cross-sectional (prevalence)
Positive skew
Smaller #s predominate - tail in + direction
Mean > median
Negative skew
Large #s predominate - tail extends in - direction
Mean < median
Babies delivered with neural tube defects and healthy babies - moms asked about medicine use. Type of study?
Case-Control
Odds ratio = odds of exposure w/ disease odds of exposure w/o disease
A risk factor leads to extensive dx testing and increases the probability the disease has been identified (smoking) is what type of bias?
Detection bias
Moro reflex
0-5mo
- falling –> spread arms –> clasp