Clinical and Experimental studies

• Clinical trials are research designs in which researchers observe participants and their interventions/exposures without controlling them.

Observational Studies

• Cross-sectional, cohort, and case-control are types of observational studies.
• Observational studies are most useful for studying rare exposures or outcomes.
• It is possible to observe harmful behaviors since you cannot randomize individuals to harmful behaviors.
• Confounding of association is a weakness of observational studies.
• No random assignment to groups introduces confusing results and bias from self-selection into groups.

Experimental Studies

• Experimental studies are studies where the researcher controls the interventions that participants receive.
• Clinical trials are prospective studies on humans comparing the effect of an intervention to a control, whether it be an established standard of care or a placebo.
• The word "clinic" is derived from "clinic" (bedside caring) + "trier" (try).
• Randomized control trials (RCTs) are a type of clinical trial.

Clinical Trial Characteristics

• Prospective designs follow participants forward in time, like cohort studies.
• Participants are randomly assigned to a treatment arm or a control arm.
• Clinical equipoise means there is not a definitive answer as to whether the standard of care (SOC) or the intervention is better.
• The trial must test a new intervention, such as drugs, surgery, or lifestyle changes.

Purpose of Clinical Trials

• Clinical Trials determine if the new intervention is safe and effective.
• They compare the effectiveness of new interventions to existing standards of care.
• Clinical Trials monitor how different patients respond to interventions by biological groups, such as age or cancer stage.
• They also confirm what's effective/ineffective.
• Clinical Trials correct misconceptions in biomedical knowledge.
• It determines for whom the intervention is recommended. Note that biological group comparisons can be complicated, as race doesn't biologically change how a drug works, but marginalization can.

Strengths and Weaknesses of Clinical Trials

• Strengths of clinical trials allow for causal conclusions.
• Controls for known and unknown confounders occur with randomization.
• Clinical trials help define effectiveness and safety.
• Weaknesses include ethical concerns when randomizing to potentially inferior treatments.
• They are expensive and resource-intensive.
• Limited generalizability is another weakness, as they are often conducted in academic centers.

Phases of Clinical Trials

• The phases of clinical trials must be completed for the intervention before the FDA approves it, and it can be marketed.
• Phase I focuses on safety and dosage.
• Phase II focuses on efficacy and side effects.
• Phase III confirms effectiveness and monitors side effects in larger populations.
• Phase IV involves post-market surveillance.

Key Concepts - PICOT Framework

• Population: the target group.
• Intervention: what's being tested.
• Comparison: the control group or intervention being compared against.
• Outcome: what's being measured.
• Time: the duration of the study.

Efficacy vs Effectiveness

• Efficacy intervention effect in controlled settings and ideal compliance.

• Efficacy is always higher than effectiveness.

• Effectiveness is the intervention effect in real-world settings.

• Single-blind trial: only the researcher knows the treatment randomized to the participant.

• Double-blind trial: neither the participant nor the researcher knows the randomized treatment.

• Blinding/randomization minimizes bias and confounding.

Ethics in Clinical Trials

• Equipoise dictates that it is unethical to randomize participants to a treatment that is known to be inferior.
• Participants must be informed of the trial's purpose, procedures, benefits, risks, and use of their data/samples, and their right to withdraw, this is "Informed consent."
• Data monitoring is a continuous monitoring of safety and effectiveness during the trial.

Safety

• Adverse events are categorized by severity, such as mild, moderate, or severe.
• Informed consent and the right to withdraw are ethical requirements.

Study Protocols & Control Groups

• Standardizing study details and conducting for consistency and reliability are important in making a study protocol.
• Protocols are reviewed by regulatory bodies like the IRB and DSMB.
• Keep protocol changes minimal, with clear justifications.
• Creating windows for monthly measurements allows wiggle room, e.g., +/- 2 weeks are still counted as on that month.

Control Groups

• Control groups are used to compare against the intervention to see if the intervention is better.
• Compare the intervention to standard care or placebo.
• Minimize the Hawthorne effect by not changing behavior while being watched as well as minimizing the placebo effect where health appears to improve without therapeutic intervention.

Finding Control Groups

• Randomized controls consist of participants randomly controlled.
• Advantages include removing bias from allocation, balanced outcomes, and validity of common statistical tests.
• Disadvantages include some patients potentially being denied an effective treatment.
• Non-randomized concurrent controls consists of enrolling controls from the same time period and hospital.
• Advantages include the influence of the same factors on participants (similar infections, vaccinations, diets).
• Disadvantages include confounding variables.
• Nonconcurrent historical controls involve comparing new interventions with past treatments and their outcomes.
• Advantages include a willingness to participate due to already knowing they're receiving the intervention. It can also compare a historical threshold instead of patient by patient and is good for established prognosis and highly fatal disease.
• Disadvantages include needing access to historical data, publication bias, different guidelines based on time periods, non-completeness, and cut points for diagnosing change of time.
• Hybrid controls are some historical + some concurrent controls with a greater proportion randomized to the intervention arm.

Study Designs in Clinical Trials

• In crossover trials, each participant is their own control. Participants receive both treatments with washout periods.
• The order of treatments is randomized.
• Between each treatment is a washout period that allows the condition to return to baseline.
• Advantages include reduced sample size and being useful for stable chronic diseases.
• Disadvantages include the condition not returning to baseline and the first treatment potentially "carrying over" and affecting the second.
• Example: Measuring tooth extraction pain with two interventions. Participants might expect pain after first treatment, which could reduce their pain perception during the second treatment, complicating the effect of the 2nd intervention on pain.

N of 1 Trials and Withdrawal trials

• N of 1 trial: where a single participant is treated with multiple individualized interventions.
• Similar to a crossover, but for one participant (instead of a group of participants).
• It is a double-blind design for each treatment phase and is often used in rare diseases.
• Withdrawal trials determine if participants can maintain benefits after stopping treatment.
• First, everyone is assigned to the drug and then randomized to a withdrawal type (stop meds & start nutritional changes OR stop meds and add no nutritional changes).

Factorial Design Example

• Multiple interventions are tested at different levels (combos).
• Benefits include testing the effectiveness of multiple interventions at once and seeing if the interventions work better together than alone.
• Disadvantages include the potential interaction between Drug A depending on Drug B being present.
• Example of a 2x2 factorial for Intervention A & Intervention B: A+B, A+Placebo for B, B+ Placebo for A, Placebo + Placebo.

Large Simple/Pragmatic Design

• Large sample size and a simpler design.
• Fewer eligibility criteria and more flexible data collection.
• Advantages include being easier to implement and reduces costs.
• Disadvantages include lower precision, potential for bias, limited data monitoring.

Monitoring & Adherence

• Monitoring involves continual assessment of safety outcomes.
• Statistical analysis determines if the trial should continue based on participant safety.
• Adherence ensures participants follow the prescribed treatment regimen.
• It is important to track side effects and interactions.

Superiority

• Demonstrates that a new intervention is better than the standard of care.
• Used when it is known the intervention is superior to SOC or placebo. Goal: Show there’s a difference between 2 groups where Intervention > SOC/placebo
• null hypothesis: no difference (μ₁ - μ2 = 0)
• alt hypothesis: difference between groups (μ₁ - μ2 ≠ 0)
• Ex trial tests if 12 months of hospital-based cardiac rehabilitation intervention increases exercise capacity in cardiac surgery patients compared to no rehabilitation placebo
• null: exercise capacity is the same between groups
• alt: exercise capacity is increased in intervention group.

Equivalency

• Intervention outcome is the same as, no better or worse than the SOC/placebo outcome.
• when:
• goal: show that clinical impact is the same when treatments are totally different OR that secondary endpoints are improved(better adherence, improved cost, better safety)
• null: difference(A)larger than clinically unimportant margin
• alt: difference is within the margin
• Ex Trial tests if 12 months of hospital based cardiac is the same or different to home rehabilitation with exercise capacity.
• null: exercise capacity is the same .
• alt: exercise capacity is different. Equivalence Margin (a) Difference between two treatments with what is clinically acceptable p value influenced by sample size → large size can show different so we’re looking at the difference with there isn’t one. Can’t use p value.

Non-Inferiority

• Intervention is not worse than the standard of care.
• Use when at least want it at possible.
• Goal is to see a secondary end point can improve better adherence and cost.
• Null hypothesis, intervention worse than standard of care.
• Alternate hypothesis : worse than standard the test again if the capacity is increased. ex trial testing 12 months intervention capacity the same capacity increase.
• Cl within equivalence margin
• If the difference is less than margin intervention is non inferior.
• RCT acetaminophen ankle

Literature review

• Evaluation of scholarly writings on a topic based on research objectives and thesis.
• Perspective and context for research and knowledge

background section

• brief intro for justification and trial
• find key terms.
• Relevant literature question proposal.

Narrative Review

• Lit review that gives broad summary of a topic rather than answering a research question
• Broad biased
• Sometimes evidence based

Systematic Review

• Lit review that collects critiques & synthesizes to estimating a treatment effect. Clear research topic And rigorous to comprehensive it. Evidence based conclusions. Different from meta analysis, can’t summarize it.

Meta-Analysis in Literature Review

Statistical analysis that data addresses. Systemically combined power and decision. That can combine results that can’t individual studies. Can’t fix flows it pulls from so meta analysis resutls Differences in patient populations

Funnel Plot

assess publication bias

• that is a plot estimate against the study measured error.
• Symmetrical, top
• Severe potential

Fixed effects assuming and estimate different Random different, effects

Heterogeneity

caused by Study Quality random variable difference. Different test is a test if the odds vary

• Test for the variation from chance size is small Influence analysis

Practice guidelines,

Recommendations or how to diagnose treatment Graded

Levels

• I low type
• RT II
• non random study
• III concurrent group Historical study Cases with no cases

Recommendations

• with the right that we feel that
• The recommendations the health settings being implemented
• Policy makers and for formal and quality criteria
• supported by credible research change one

Endpoint selection

Used trials to pick sample, data analysis and intervention effect

• single out to study increase the false result
• reduce the study to
• A secondary with primary
• Continue time frame for a phase it might life Composite outcome
• events and outcomes effectiveness by multiple outcomes different frequency and response

Surrogate

Indirect measure to predict the real outcome of the rate of heart
Better measure

Fda history

Promote health safety. 1907

• immigrants rights
• Nasty 1906 Law Interstate Revises

Drug Development

Chemically Synthesizing Derived Affects effect toxic

• predicted Toxicity

Tests that degree and toxicity

Drug development

• 10-15 years
• 2.6 billion
• Safety animals
• humans

New Drug (Ind)

The drugs Physician Can be used

Study development- Phase one testing humans

• the of this safety and this disease
• historical Sizes 20-64 From animal studies is slow and not for low. Acceptable

Curve to show relationship the dose Ex : that

Phase Testing To Toation

Goal Tests 2.5 to confirm effectiveness side effects

Randomization Assumed will be used. Bias will make sure for that

In the to will be 50/50 The probability patient characteristics of response

Randomization

Chance W/ characteristics control characteristics the .55 for treatment to a

Randomized Assignment

B B a B B a Blocks of a mixed size and the community

The Bias

Treatment based the underrepresented

With patients assignment equals Want to study balance , the assignment based now, assign on .

If assign draw , now by assign patients.