Untitled

• The objectives are to recall the tasks of Evidence-Based Medicine, understand the components of creating a research question, define different types of research design, and identify what specific research questions can be answered by a variety of research designs.
• Evidence-Based Medicine tasks include asking the clinical question, searching for evidence, assessing evidence quality, incorporating evidence into clinical decisions, and evaluating the process.

Asking a Clinical Question

• Background questions give general information on a topic
• As an example, what are the risk factors for osteoporosis?
• Foreground questions give guidance for clinical care of particular patients or groups
• As an example, in children less than 8 years of age, what antibiotic regimen is most effective for recovery of Lyme Disease with less side effect?

Elements of a Research Question: PICO(T)

• Population describes the patient or population and ensures it is sufficiently similar to the reader's patient population.
• Intervention/Exposure defines what will be done for the patient
• As an example, this could be medication, procedure, or diagnostic test
• Comparison is what the intervention is compared against, like a placebo, standard intervention, or non-exposed population.
• More than one comparison group is possible
• Outcome signifies what the clinician hopes to measure, achieve, or affect.
• Timing indicates when the outcome will be measured
• Two trials with different follow-up periods can yield very different results

PICO Question Types

• PICO questions can address a variety of clinical conditions
• These include diagnosis, etiology or harm, prognosis, and treatment

Important Terms

• Prevalence is the proportion of persons affected with a particular disease/condition at a specific time.
• Incidence is the number of new cases of a disease during a specified period, as a proportion of the number of people at risk.
• Association uses statistical techniques to quantify the strength of the relationship between two or more variables.
• Causation means the exposure produces the effect
• This includes the presence of an adverse exposure or absence of preventative exposure
• Correlation measures the magnitude of the association between two variables, sometimes expressed as a coefficient.

Searching the Evidence

• Filtered Evidence is gathered and synthesized by experts into a format readily usable by clinicians
• This includes Clinical Guidelines, Evidence Based Summaries, Systematic Reviews, and Databases
• Unfiltered Evidence consists of original research articles published in peer-reviewed journals
• The PICO(T) framework allows for a systematic literature search
• Use databases like PubMed, Ovid

Research Design

• Study design is an important feature of research studies that affects the validity of that research
• Validity is how well the results among the study participants represent true findings among similar individuals outside of the study

Types of Study Design

• Study design is divided into all studies, descriptive, and analytic
• Descriptive includes survey and qualitative
• Analytic studies can be experimental, observational

Descriptive Studies

• These studies describe a population and outcomes (PO)
• They do not compare two groups
• Instead, they focus on describing the characteristics, presentation, and outcomes of a specific patient or small cohort of similar patients
• Examples consist of case reports, case series, qualitative studies, and surveys.
• A disadvantage to Descriptive Studies, is without a comparison group, you can only hypothesize why certain outcomes/characteristics are seen
• They also have the lowest strength of evidence

Analytic Studies

• These studies quantify the relationship between two factors
• Analytic studies look at the effect of an intervention (I) or exposure (E) on an outcome (O) in comparison (C) groups
• Actively change the factor or impose an intervention means experimental
• As an example, this is a randomized control trial
• Passive researcher involvement indicates observational

Observational Studies

• Retrospective studies' outcome already occurred prior to start of the study
• The researcher is looking back in time
• Prospective studies' outcome has yet to occur prior to the start of the study
• As an example, there are case control and cohort studies

Experimental: Randomized Control Trial

• Participants are divided into treatment/intervention or control/placebo groups using a random mechanism
• They are best for studying the effect of an intervention or exposure

Advantages of Randomized Control Trial

• Unbiased distribution of confounders
• Blinding is more likely
• Randomization assists statistical analysis

Disadvantages of Randomized Control Trial

• Expensive and time consuming
• Volunteer bias is a factor
• Ethically problematic at times

Experimental: Randomized Crossover

• A control trial where each study participant has each treatment
• All subjects serve as their own controls
• Error variance is lessened, which reduces the sample size needed
• All subjects receive treatment
• Blinding can be maintained

Disadvantages to Randomized Crossover

• All subjects receive placebo or alternative treatment at some point
• Washout period between treatment A and treatment B can be lengthy
• These cannot be used for treatments with permanent effects

Observational: Cohort Study

• A group of people with a common characteristic are gathered
• Participants are divided into two or more groups based on their level of exposure to the independent variable of interest
• Participants are followed over time to see who develops the outcome of interest
• These are typically prospective

Advantages to Cohort Studies

• Likely strongest study type if randomization is not possible
• Ethically safe
• Can measure multiple outcomes
• Easier and cheaper than RCT

Disadvantages to Cohort Studies

• Confounding may occur due to lack of randomization
• No causation, only correlation
• Blinding can be difficult
• Large sample size or long follow-up necessary

Observational: Case-Control Studies

• Look at population with a certain outcome or disease (cases) vs similar population (control) without the outcome or disease
• Information is collected on whether subjects have been exposed to the factor under investigation
• They are typically retrospective

Advantages of Case-Control Studies

• Quick and inexpensive
• Good for rare disease/Disorder

Disadvantages of Case-Control Studies

• Confounders
• Potential bias: recall, selection
• No causal relationship can be established

Observational: Cross-Sectional Study

• A "snap-shot" or "slice in time" study
• Examines the relationship between diseases (or other health-related characteristics) and other variables of interest as they exist in a defined population at one particular time
• Best for quantifying the prevalence (not incidence) of a disease or risk factor, and for quantifying the accuracy of a diagnostic test

Advantages of Cross-Sectional Study

• Inexpensive, ethically safe

Disadvantages of Cross-Sectional Study

• No causal relationship able to be drawn
• Association is the most that can be established
• Recall bias
• Confounders may be unequally distributed
• Group sizes may be unequal

Hierarchy of Study Design

• In order from high to low value: Meta-analysis, experimental, observational (cohort > case control), descriptive studies