Research Design II, week 4

prognosis

what is the likely outocome or course of a disease (injury)

what is my chance of recovery

prognostic factors

• demographic

• individual pt. behaviors

• disease/injury specific factors

• comorbid considerations

not every study showing ____ is a prognostic prediction study

future risk

prediction studies use multiple variables aka _____ to estimate the probability of an outcome

predictors

two types of prediction study designs

diagnostic and prognostic

diagnostic prediction study design

prediction models - determine the presence of a particular (as yet undiagnosed) disease or injury

prognostic prediction study design

prediction models estimate whether a particular event/injury will occur in the future

prognostic studies:

•Predict future outcomes

•Develop or validate prediction models or identify prognostic factors

•Incorporate predictive performance (e.g., calibration, discrimination, accuracy)

•Answer: What is likely to happen?

•May include predictors that are or are not  biologically associated with the outcome

prognostic studies are not:

•Not descriptive – they do not simply describe disease frequency or characteristics

•Not causal – they do not estimate whether changing a factor changes the outcome

•Not evaluating interventions or treatment effects

•Do not answer: “What caused this?” or “What is the effect of treatment?”

a good prognostic (prediction) model study should answe four questions:

•Can it distinguish high-risk from low-risk patients? (Discrimination)

•Are the predicted risks accurate? (Calibration)

•Does it work in other patients? (Validation)

•Does it improve clinical decisions? (Clinical Utility)

discrimination

Can it distinguish high-risk from low-risk patients?

Calibration

Are the predicted risks accurate?

Validation

Does it work in other patients?

Clinical Utility

Does it improve clinical decisions?

goals doe prognostic studies

1. is appropriately conducted

2. has all the statistic components/model evaluation needed

3. know if it applies to your patients

research principles

1. beneficence

2. justice

3. respect for persons

beneficence

“do no harm - do benefit”

assessment of rusks (harms) and benefits

• minimize harm to pts

• use least harmful methods to achieve the scientific end

• favorable benefit-to-risk comparison

• maximize benefit to risk ratio and benefits to participant and society

justice

treat individuals fairly

selection of subjects/participants

• equitable distribution of research harms and benefits

• equitable selection of subjects/participants within a population

• equitable selection of population

respect for persons

individual autonomy

informed consent

• full information

• full comprehension

• voluntary, without coercion, with ability to end participation without penalty

• protect individuals with reduced capacity to exercise autonomy

priorities of the IRB

• subject

• institution

• invesigator

IRB responsibilities

• prospective subject population

• identification of potential risks

• identification of potential benefits

• risk/benefit analysis

• investigator qualifications

• informed consent

prospective subject population

• appropriate to the nature and goals of the project

• vulnerable subjects

• equitable selection

types of risk

physical, psychological, social, legal, economic

potential benefits

• to subject

• to society

• financial compensation is not a beneft

• “maximize benefits”

principle I of nuremburg code

• legal capacity

• sufficient knowledge and comprehension

• enlightened decision

the consent process

• prior to participation

• moral right to know what is being done/what are the risks

• who seeks consent?

consent form

• signed by subject, witness, investigator

• 6th grade reading level

• written in second person

elements of informed consent

• Overall purpose

• Explanation of procedures

• Potential risks & discomforts

• Potential benefits

• Alternatives to participation

• Financial obligations/compensation

• Assurance of confidentiality (data security and privacy)

• Option to withdrawal

• Questions

levels of IRB review

• full board

• expedited

• exempt

full board

more than “minimal risk”

expedited

not greater than minimal risk

exempt

less than “minimal risk”

minimal risk

probability and magnitude of physical or psychological harm that is normally encountered in the daily lives or routine (dental medical psychological exam) of healthy persons

general considerations

+All submissions and reviews are electronic via IRB Mentor

+Only Faculty & Staff can serve as Principal Investigators

+All PI’s must complete CITI Training prior to review of protocols (good for 4 years)

+All DPT students must also complete CITI Training – FOR STUDENTS , good for 3 years.