Absolute Risk, ch. 11
Notes- Probability of the outcome within both groups; The incidence of disease in a population; indicates the magnitude of risk in a group of people with a certain exposure, does not take into consideration the risk of disease in nonexposed individuals.
how to measure Excess Risk
ratio of risks or of incidence rates
Disease risk in exposed/Disease risk in nonexposed
or by difference in incidence rates- Incidence in exposed- inicidence in nonexposed
Ratio of risk of disease in exposed ppl to the risk of disease in nonexposed ppl; RR= Risk in exposed/Risk in nonexposed
Randomized Trial, ch. 7
The ideal design for evaluating the effectiveness and side effects of new forms of interventions; begins with a defined population that is randomized to receive either new treatment or current treatment, and follow subjects in each group to see how many improved in the new treatment group versus the current treatment group.
Historical controls, ch. 7
using a comparison group from the past, past medical records from a group of patients; flaws associated with this includes: that the data obtained from study groups may not be similar in quality, if there is a difference in outcome between two groups we are unsure if it is due to the therapy or due to other events/factors that occurred over the time period.
Stratified Randomization, Ch. 7
stratify study population by each variable considered important (gender, age), and thenrandomize participants to treatment groups within each stratum.
Masking (Blinding), ch. 7
ensuring that subjects do not know which groups they are assigned to, one way to do this: Placebo-
Double Blinding, ch. 7
In addition to blinding the subjects, also want to mask or blind the data collectors in regard to which group the patient is in.
Crossover, (Planned & Unplanned), ch. 7
Planned Crossover- subjects randomized to therapy A or B, and after being observed for a certain period of time on 1 therapy, they are switched to the other therapy.
Carryover, Ch. 7
If subject is changed from therapy A to therapy B and observed under each therapy, observations under therapy B will be valid if there is no residual carryover from therapy A.
Potential Biases in Cohort Studies, ch. 9
Bias in assessment of the outcome- problem can be addressed by blinding (masking) the person who is making disease assessment, Information Bias- if quality & extent of infor is different for exposed vs. nonexposed ppl (more likely to occur in historical cohort studies), Biases from nonresponse and follow-up (loss-to-follow up- incidence rates in exposed vs. unxposed groups will be difficult to intepret, Analytic Bias- via the researchers that might unintentionally introduce their biases
When is a cohort study warranted?
When good evidence suggests an association of disease with a certain exposure or exposures- rare exposures, particularly useful when attritition (losses to follow-up) can be minimized, easy to conduct when interval b/t exposure and development of disease is short, long-term studies of childhood health & disease
Prospective cohort study
Is concurrent, because investigator identifies the original population(a well defined population) at the beginning of the study and accompanies the subjects concurrently through calendar time until the point at which the disease develops or does not develop. Permits investigator to study multiple exposures (i.e.- The framingham study)
Retrospective Cohort or Historical Cohort Study, AKA Noncurrent Prospective Study
uses historical data from past; short duration time; still looks at exposed vs. unexposed groups; only difference between this and prospective cohort study is calendar time. Exposure is det from past records and outcome (disease or no disease) is ascertained at the study began.
In cohort studies of the role of a suspected factor in the etilogy of disease, it is essential that:
The exposed and nonexposed groups under study be as similar as possible with regard to confounding factors
Which of the following is not an advantage of a prospective cohort study?
It usually costs less than a case control study
Retrospective cohort studies are characterized by all of the following except:
The required sample size is smaller than that needed of a prospective cohort study
A major problem resulting from lack of randomization in a cohort study is:
The possibility that a factor that led to the exposure, rather than the exposure itself, might have caused the disease
In a cohort study, the advantage of starting by selecting a defined population for study before any of its members becomes exposed, rather than by selecting exposed and nonexposed individuals, is that:
A number of exposures can be studied simutaneously
Background Risk, ch. 12
Both exposed and nonexposed ppl have this risk; i.e.- smoking & lung cancer- nonsmokers have some risk due to environmental carcinogens
Total Risk of Disease in Question, Ch. 12
Sum of the background risk that any person has and the additional risk due to the exposure in question. To find how much of total risk in exposed ppl are due to exposure, subtract background risk from total risk.
Relative Risk vs. Attributable Risk, Ch. 12
Relative risk is a measure of the strenth of the association and possibiluty of a casual relationship; whereas, attributable risk indicates potential for prevention if the exposure could be eliminated.
Absolute Risk reduction (notes)
difference in the absolute risk between the two
groups; Absolute Risk (Control Group)= 60/100= 60% or .6
Absolute Risk (Zinc Lozenge Group)= 20/100=20% or .2
Relative risk reduction or relative risk (RRR)
ratio of two absolute risks; RRR = (.6- .2/.6) x 100 = .4/.6 x 100 = 66.6 or 67%
Preventative & Therapeutic Trials, Lecture 6
also known as interventional studies, may or may not be randomized, but the clear gold standard is the RCT or randomized clinical trial, highest quality evidence for therapeutic efficacy, best way to assess cause and effect
if ethically and practically possible
Internal Validity, ch. 8
If study is properly done without major methodologic problems, and takes into account all the issues
External Validity, ch. 8
Being able to generalize study findings to all patients with the disease, therefore, you must know the extent to which patients being studied are representative of all patients with the disease in question.
Randomized trial, ch. 7
Considered the ideal design for evaluating both the effectiveness and side effects of new forms of intervention, begins with a defined population that is randomized to receive either new treatment or current treatment, and follow subjects in each group to see how many are improved in new treatment group compared with current treatment group.
Case Study or Case Series, ch. 7
No comparison is made with an untreated group or with a group receiving some other treatment.
Problems associated with randomized trial, ch. 7
Subjective biases of the investigstors, either overt or covert may be involved in selecting patients for 1 treatment group vs. another; if subjects are not randomized properly, the groups may be different in terms of many demographic, social, psychological, and cultural variables that may have important roles in determining outcome.
Masking (Blinding), ch. 7
To ensure that the subjects do not know which group they are assigned to, usually done through "placebo".- substance that looks/acts like the active substance.
Placebo, ch. 7
Plays a major role in identifying both the real benefits of an agent and its side effects.
Double Blinding, ch. 7
In addition to blinding subjects, also blind the data collectors in regard to which group the patient is in. Purpose is to: avoid observer & subject bias
Crossover, Ch. 7
"Planned Crossover"- Subjects randomized to therapy A or B, after being being observed for a certain pd of time are switched over to other therapy, "Unplanned Crossover'- Patient unexpectedly has to switch groups because they chose not to receive the designated intervention or because they have to do due to some medical reason.
Carryover, ch. 7
If subject is changed from 1 therapy to the other, the observations under second therapy will be valid only if there is no residual carryover from the 1st therapy.
All of the following are potential benefits of a randomized clinical trial, except:
c. The external validity of the study is increased
- a. The liklihood that the study groups will be compared is increased.
-b. Self-selection for a particular treatment is eliminated.
-d. Assignment of the next subject cannot be predicted.
-e. The therapy that a subject receives is not influenced by either conscious or unconscious bias of the investigator.
The major purpose of random assignment in a clinical trial is to:
Reduce selection bias in allocation of treatment
Cohort Study, ch. 9
investigator selects group of exposed ppl and a group of nonexposed ppl and follows both groups to compare icidence of disease in the two groups
Two-by-two table/ design of a cohort study
Disease No Disease Incidence Rates
Exposed a b a/a+b
Not Exposed c d c/c+d
incidence among exposed, cohort study
incidence among nonexposed, cohort study
How is a cohort study similar to randomized clinical trial?
In both types of designs, an exposed group is compared with a nonexposed group or to a group with another exposure; cohort studies begin with a defined population
Problems with a nonrandomized study
If we observe an association of an exposure with a disease, we are left with uncertainty as to whether the association may be a result of the fact that ppl were not randomized to the exposure, or whether it was factors that led ppl to be exposed, that are associated with the disease
Problems with a prospective cohort study
Takes a long time to complete, sometimes 20 years, research grant funding generally limited to a few years, risk that study subjects will outlive the investigator, or that investigator may not survive to the end of the study.
Main difference between a randomized trial & cohort study
Cohort study, subjects are not randomly assigned to be exposed, because randomization to exposure to possibly toxic agents would not be ethical. Case- control & cohort studies address questions of etiology.