PPAP I: Intro to EBM

Evidence-Based Medicine (EBM)

How do?

• Formulate clinical question

• Assessing the literature

“the important thing to learn is how to learn on your own”

5 Steps of Evidence-Based Medicine

Clinical Question

Evidence Acquisition

Evaluate the Evidence

Apply the Evidence

Assess the Outcome (REPEAT)

helps you get the answer you are looking for quick! from reliable source

Formulating a Clinical Question

PICO!!: assessment of need and problems INTO structure clinical Q (adequate, focused information w/o too m detail)

• Patient/Population: disease/condition you are investigating, and the particular demographic you are wanted to learn about

• Intervention: primary treatment option

• Comparison/Control: secondary treatment option or control groups (if appicable)

• Outcome: desired outcome

additional timeframe component (PICOT)

PICO: Patient/Population/Problem

Patient/Patient Group

• gender/race/sex/age

How described to a medical colleague?

• disease/condition?

• Stage of illness?

• setting of __?

PICO: Intervention

What you plan to do

• consider…

  • type of treatment

  • intervention level (dosage/freq not necessarily known)

  • stage of intervention (preventative, early, advanced)

  • Delivery method (hospital admin, at home, self)

PICO: Comparison

Alternative to intervention

• standard of care, placebo, other approved therapies

Only assess one alternative at a time

• optimal when in reviewing studies on prognosis or diagnosis

PICO: Outcome

result you want to review for the patient

• when possible: measurable terms (BP mmHg, A1C..)

• others: SE, Quality of Life, Cost Effectiveness, Improvement of Symptoms

Types of PICO

Diagnosis: How to select and evaluate diagnostic tests?

• types: cross-sectional study/ prospective blind comparison to a gold standard

Therapy: Is a treatment effective?

• Types: systematic review, randomized controlled trial (RCT) or cohort study

Prognosis: What is the patient’s clinal course over time?

• Types: Cohort studies, case studies, or case series

Harm/Etiology: What cause this disease/condition?

• Types: Cohort studies, case studies, or case series can answer question of harm

Acquiring the Evidence

Skills:

• ask a focused clinical question

• searching medical literature

• retrieving the medical literature

Appraising the Evidence

Directness → will i read?

Validity → will i believe?

Results → does it work?

Applying the Evidence

Applicability → can i use?

Individualizing results → how will it affect my patient?

Sharing decision → what does my pt think?

Randomized Controlled Trials (RCT)

High Validity

• top choice for all types of questions

Low Feasibility

• big sample size

• ethical outcomes

• requires long term followup

Case-Control

Observational Retrospective study

• assignment

  • popu w/ disease VS popu AT RISK w/o disease

• measure predictor variables

  • exposure to certain drug/ condition

LOW validity and HIGH feasibility

• start at outcome and assess exposure status

• Strength

  • overcomes temporal delay (waiting time to find out)

  • may only need small sample size

  • can assess mult exposures

• Weakness

  • only assess one condition/disease

  • susceptible to bias → esp selection bias and recall bias (self-reporting)

  • limited validity (certain popu)

Cohort Studies

Observations Prospective study

• can be retrospectives (via claims databases)

• select sample → measure predictor variables (risk factors) → follow-up the cohort → measure outcome (incidence/chance of dev disease/condition)

Start! exposure status to assess outcome event (measuring variable)

moderate validity and feasibility

Strengths:

• feasible whenrandomizaiaont of exposure is not possible or unethical

• assess multiple outcomes

Weaknesses

• susceptible to bias

• limited validity (potential to lose subjects to follow up)

• longer to conduct

• more expensive than case control

Case Reports

manuscript reporting sev cases on particular incident or therapy in specific situation

• new ideas/discussion can emerge

Strengths

• fast to report outcomes to med community

• observationsal to potentially lead to larger study

• good for unique, rare clinical scenarios

Weaknesses:

• not generalizable (external validity)

• difficult to compare cases and replicate

• uncontrolled

• can include bias

Bias

Recall Bias: retrospective studies

• pt recalls exposure

• want obj!! if subj → collab w/ medical record

Selection Bias: consider random sequence generation and allocation concealment methods of RCT

Performance Bias: difference b/tw study groups as result of systematic differences in performance outside study treatment (MUST blind subjects/caregivers)

Detection Bias: possibility for differences b/tw comparison groups w/ regard to how outcomes measured/assessed (BLIND assessors)

Attrition Bias: result of systematic cause of pt withdrawals in a study that disproportionately affect a certain subset of pt

• could imbalance/impact results/conclusions

• asses study analysis and look further into results

Reporting Bias: concerns w/ regard to outcomes reported w/in results of study

• report only some measured outcomes (deemed not important)

Internal Validity

how well a study’s results reflect the true cause-effect relationship w/in the study

• confidence that study’s results are conclusive and NOT influences by other factors/bias

Threats:

• Selection Bias

• Performance biase

• Detection Bias

• Attrition Bias

External Validity

Extent the study’s results can be generalized to other settings or populations

• real world application? what degree?

Threats:

• studies excluding certain popu

• restrictions on medial comorbidities

• restrictions on concurrent medications

• short study length

Assessing Applicability

Sex: physiological, hormonal, biochemical differences

Comorbidities: coexisting conditions effecting outcome risk for given exposure?

Race: racial/ethnic differences that might alter risk for outcome?

Age

Pathology: more extensive infection, more catastrophic course (malignant histology causes earlier metastasis)

Socioeconomic factors: Location, lifestyle

Sharing Decision

come up w/ estimate of prognosis specific to individual patient based on characteristics

• helps pt make treatment decisions, diagnosis, and harm

• help pt make general plans for future