Week 6 Flashcards
Once the disease diagnosed, the next questions are…
What can be done about it?
Is there a treatment that improves the outcome of disease?
Very tough questions that clinicians have to face
Why should treatment be given?
Treatments should be given “not because they ought to work, but because they do work.”
Voir slide 5
What are the 2 types of studies of treatment?
Observational studies
Experimental studies
What is an observational study of treatment?
- Frequency of disease in exposed vs. unexposed (we only observed what actually happened for a certain pop at a given time)
- Frequency of exposure in disease vs. non-disease
- We cannot observe the counterfactual
What are the types of experimental study of treatment?
- Randomized controlled trials
- Community trial
- Natural experiments
What are natural experiments?
are unplanned situations in nature that resemble planned experiments
What is a Randomized controlled trials (RCT)?
Randomly allocate individuals to experimental and control groups to test the effects of an intervention.
A prospective study using randomization (participants) to compare the effect and value of an intervention against control in human beings
RCT increases what?
Increases comparability between exposed and unexposed groups on both known and unknown factors that cause the outcome
What is the research purpose of RCT?
Usually it is believed that the experimental intervention might be better than the control, but that has not been conclusively established by strong research.
The primary outcome must be benefit; treatments cannot be randomly allocated to discover whether one is more harmful than the other.
Voir slide 11
What are the steps to do an RCT?
- Developing a research hypothesis
- Selecting the sample
- Determining sample size requirements
- Allocating subjects into experimental and control groups
- Applying the intervention
- Assessing the outcomes during follow-up
What are the criterias that are found in RCT?
- Ethics
- Sampling
- Comparison groups
- Intervention
- Allocating treatments
- Differences arising after randomization
- Compliance
- Blinding
- Assessment of outcomes
What is ethics?
Under what circumstances is it ethical to assign treatment at random, rather than as decided by the patient and physician?
- Equipoise: no compelling reason to believe that either of the randomly allocated treatments is better than other. (Equipoise not only applies when RCT is initially designed, but throughout the trial as well)
- Patients fully understand the consequences of participating in the study, know that they can withdraw at any time without comprising their health care, and freely give their consent to participate. (They need to know what kinda study the will be part of – but not know what group they will be in (placebo or control))
What is the Data Safety Monitoring Board (DSMB) and what do they do?
DSMB is a group of individuals who have no conflict of interest regularly monitor trial data
- Ensure safety of participants
- Ensure scientific integrity of the trial
- Stop trial for early evidence of harm or benefit
- If theres more harm than benefit to the study they will force stop the study
- If no difference or results will be observed they stop the study
What are the sampling levels?
Reference pop (practical representation)
Experimental pop
Study sample
The study sample includes what?
Eligibility criteria
- Inclusion
- Exclusion
The eligibility criteria selected for an RCT should involve what?
- the study objectives,
- the possible effects on internal and external validity of the study,
- the potential benefit or harm to the subjects,
- issues related to convenience and efficiency.
What are common Eligibility inclusion Criteria in RCT?
- Certain age, sex, or racial characteristics
- Absence of certain diseases or conditions
- Not currently receiving the proposed intervention
- Located in certain communities
- Meets eligibility criteria during certain time periods
What are common Eligibility exclusion Criteria in RCT?
- Potential harm to participants, e.g. pregnancy
- Intervention unlikely to be effective
- Potential for poor compliance with the study protocol
- Practical difficulties with participants
- Patients with contradictions to the treatment cannot be randomized.
Sampling example: A RCT conducted at a hospital in Philadelphia. The study was designed to evaluate the efficacy of lithium in treating aggressive behavior in children and adolescents.
What would be the inclusion and exclusion criterias?
Inclusion criteria: male and female patients with a history of severe aggressive behaviour and a diagnosis of conduct disorder, who were 10-17 yrs of age and residing in an acute-care child and adolescent psychiatric inpatient ward of a hospital.
Exclusion criteria: mental retardation, pervasive developmental disorder, recent substance dependence, and a number of other psychiatric disorders, in addition to pregnancy, certain major medical problems, recent prescriptions for psychoactive drugs, and previous inclusion in a lithium trial.
What are things used to determine sample size?
- The proportion of outcome expected in the control group;
- The proportion of outcome expected in the experimental group;
- P-value
- Power
What is Power?
- The smaller sample size requires a greater the effect size
- If the sample is not very representative of the experimental or reference population, it would not affect the internal validity, but the external validity.
- Losses to follow-up and noncompliance, will affect the internal validity.
Voir slide 21
Voir slide 22
Voir slide 23
What are Differences arising after randomization?
- Patients may not have the disease being studied
- Compliance/adherence
- Cross-over
What are 2 Measurement bias?
- Placebo effect: People in placebo group tend to report the efficacy of treatment even there is no intervention.
- Noncompliance rate
What is the difference between single, double and triple blinding?
- Single blinding: subjects
- Double blinding: subjects and investigators
- Triple blinding: subjects, investigators, and analysts
T or F: A well-designed placebo-controlled trial with randomization and double- blinding (or triple-blinding if necessary) is the ideal RCT.
True
A placebo-controlled trial requires what?
the experimental intervention demonstrate a level of superiority that exceeds any placebo effect found in the control group
Randomization has the potential of eliminating what?
selection bias and confounding as alternate explanations for any association between exposure and outcome;
Double blinding has the effect of minimizing what?
measurements bias that can occur due to knowledge of subject assignments to experimental or control groups.
What are the 4 Variations in RCT?
- Factorial design
- Cross-over design
- Large sample trials
- Practical clinical trials (pragmatic CT)
What is the factorial design?
- Intend to answer two or more questions in an RCT
- Cost savings (cheap to do)
What is the cross-over design?
- Reduce sample size
- Decrease potential for confounding
What is large sample trials?
usually randomized, are characterized by large sample sizes, broad entry criteria consistent with the intended target population, streamlined data collection, objectively measured outcomes (e.g., death, hospitalization), and follow-up that mimics normal clinical practice.
What is practical clinical trials (pragmatic CT)?
real-world questions in the actual care of patients by including the kinds of patients and interventions found in ordinary patient care settings.
What are the strengths of RCT?
- Demonstrate causal relationships with a high level of confidence, due to tightly controlled conditions not possible in observational studies.
- Allow investigators to control the exposure levels as needed.
What are the weaknesses of RCT?
- Limit applicability due to ethnical considerations and sometimes due to the artificial setting of experiment.
- Noncompliance, withdrawals, loss follow-up
- Cost, time consuming
- Difficult to achieve adequate sample size
- Limit external validity
How does Analysis of RCT work?
Planning
- Intention-to-treat analysis
- real sample analysis/ explanatory
* focusing on compliant subjects
Procedure
Voir slide 33
Voir slide 34
What is the Intent-to-treat analysis?
- Analyze based on randomized group membership
- Randomization preserved
- Internally valid estimate: comparability of groups
What is the As-treated analysis?
- Analyze based on final group membership
- Randomization not preserved
- Estimate may be affected by non-comparability of groups
How do you calculate the Intent-to-treat and As-treated analysis?
Risk ratio = (a/a+b)/(c/c+d)
What is the Number needed to treat?
The number needed to treat (NNT) is an epidemiological measure used in assessing the effectiveness of a health-care intervention, typically a treatment with medication. The NNT is the average number of patients who need to be treated to prevent one additional bad outcome (i.e. the number of patients that need to be treated for one to benefit compared with a control in a clinical trial). It is defined as the inverse of the absolute risk reduction. It was described in 1988.
What is the ideal NNT?
1, where everyone improves with treatment and no one improves with control. The higher the NNT, the less effective is the treatment
Variants are sometimes used for more specialized purposes. One example is number needed to vaccinate
Voir slide 39
What are the steps to conduct a Community trials?
- Developing a research hypothesis
- Selecting the communities
- Collecting baseline data on study outcomes
- Allocating communities
- Applying the interventions
- Assessing the outcomes during follow-up
What is ‘‘Developing a research hypothesis’’ for a community trial?
- Test etiologic hypotheses about health-related exposures and outcomes;
- Evaluate the impact of health-related programs, projects, or campaigns on communities.
What is ‘‘Selecting the communities’’ for community trials?
- Similarities among the selected communities
- Relatively stable population
- Willingness of communities to participate in an experimental study
- Sufficient evidence of experiencing high enough levels of exposures and outcomes
What is ‘‘Collecting baseline data on study outcomes’’ for community trials?
- Determine if an intervention is effective, it is crucial to be able to measure changes in study outcomes over the course of the trial.
- Use consistent methods to collect.
What is ‘‘Allocating communities’’ for community trials?
- Ideally, the allocation of communities into experimental and control groups should be random.
- Technically, community trials that use randomization to allocate communities to experimental and control groups are classified as true experiments, while those that do not are referred to as quasi- experimental studies.
What is ‘‘Applying the interventions’’ for community trials?
- Test an epidemiologic hypothesis about exposure and outcome
- Evaluate the impact of a program on community health status
What is ‘‘Assessing the outcomes during follow-up’’ for community trials?
Community trial may involve collection, measurement, and analysis of data on individuals within the communities being compared, but the ultimate interest is in comparing the results for the communities as a whole.
What are the strengths of community trials?
- Provide evidence in support of cause-effect relationships
- may be the most appropriate epidemographic design in certain situations (e.g. when risk factors for the outcome are very common and when the social context influences compliance with the intervention)
WHat are the weaknesses for community trials?
- Differences in the study outcomes may be due to causes other than the intervention (this is especially true for studies that involve only one intervention and control group)
- Expensive, time consuming
- Selection bias when randomization is not feasible
- Ecological fallacy can occur if inferences based on group data are made about individuals in the communities
What is a Natural or quasi experiment?
assignment to exposed and unexposed groups determined by nature
Chat says: A quasi-experiment in epidemiology is a study design that evaluates the effects of an intervention or exposure without randomization. Unlike randomized controlled trials (RCTs), quasi-experiments rely on naturally occurring or administratively assigned interventions, making them useful in real-world public health and policy research where randomization may not be feasible or ethical.
What were the problems with the Dutch Famine Natural Study (1944-1945)?
- Unethical to randomly assign famine
- Prenatal exposure to famine and health outcomes in adulthood?
- Areas of German-occupied Holland were malnourished due to blockage of food and fuel
- Areas under blockage vs areas not under blockage
What is the best clinical evidence?
SR of RCTs or RCTs
RCTs can not answer the following conditions:
- where there are good reasons to think the intervention is not likely to be beneficial or is likely to be harmful;
- where the outcome is very rare (e.g. a 1/10000 fatal adverse reaction);
- where the condition is very rare;
- where very long follow up is required (e.g. does drinking milk in adolescence prevent fractures in old age?);
- where the evidence of benefit from observational studies is overwhelming (e.g. oxygen for acute asthma attacks);
- when applying the evidence to real clinical situations (external validity) (results cannot be applied to any pop);
- where current practice is very resistant to change and/or patients would not be willing to take the control or active treatment;
- where the unit of randomization would have to be too large (e.g. a nationwide public health campaign); and
- where the condition is acute and requires immediate treatment. Of these, only the first case is categorical. For the rest the cut off point when an RCT is not appropriate is not precisely defined. If RCTs would not be appropriate we search and include the best appropriate form of evidence
