EBP-EAP flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Definition of evidence-based practice

Integration of the best available research evidence with clinical expertise and patient values and preferences to guide decisions about individual patients and populations.

Five steps of EBP

Formulate an answerable question, search for the best evidence, critically appraise it, apply it with expertise and patient preferences, then evaluate the outcome.

Internal versus external validity

Internal validity is whether results are correct for the study sample; external validity is whether they apply to other patients, settings, and circumstances.

GRADE downgrading domains

Risk of bias, inconsistency, indirectness, imprecision, and publication bias each can lower certainty in a body of evidence by one or more levels.

Background versus foreground questions

Background questions ask general facts about a condition or test; foreground questions ask about a specific decision in a defined population and drive the search.

PICO components

Population or patient, Intervention or exposure, Comparison, and Outcome; together they define a searchable foreground clinical question.

Boolean AND versus OR

AND requires both concepts and narrows results; OR accepts any listed synonym and broadens results within a single concept.

Controlled vocabulary searching

Using standardized index terms such as MeSH or Emtree to capture synonyms; it is combined with free-text keyword searching for sensitivity.

Grey literature

Unpublished or non-commercially published material such as conference abstracts, dissertations, trial registry entries, and organizational reports; searching it reduces publication bias.

Citation chasing

Backward citation chasing reviews references of a key paper; forward chasing identifies later papers that cite it, extending the search without new database terms.

Search reproducibility elements

Report database and platform names, the full search strategy, coverage dates, the date searched, all limits applied, and result counts for each source.

Allocation concealment

Hiding the upcoming randomization assignment from those enrolling participants; it prevents selection bias, which randomization alone may not achieve.

Intention-to-treat analysis

All randomized participants are analyzed in their assigned groups regardless of adherence; this preserves prognostic balance and avoids overestimating treatment benefit.

Number needed to treat

One divided by the absolute risk reduction; it estimates how many patients must be treated for one additional patient to benefit.

Absolute risk reduction

The control group event rate minus the intervention group event rate; unlike relative measures, it reflects baseline risk and clinical impact.

Confidence interval interpretation

A range of plausible values for the estimate; if it excludes the null value, the result is statistically significant at the corresponding alpha level.

Sensitivity and specificity

Sensitivity is the proportion of those with disease who test positive; specificity is the proportion of those without disease who test negative.

Positive likelihood ratio

Sensitivity divided by one minus specificity; it expresses how much a positive test raises the odds of disease from pretest to posttest.

Spectrum bias

Distortion of diagnostic accuracy when study patients differ from the clinical population in severity, comorbidity, or case mix.

I-squared statistic

The percentage of variability in effect estimates attributable to between-study heterogeneity rather than chance; higher values suggest pooling may be questionable.

Random-effects meta-analysis

A pooling model assuming study true effects vary around an overall mean; it generally gives wider confidence intervals than fixed-effect models when heterogeneity exists.

Forest plot diamond

The diamond marks the pooled effect estimate, its center is the point estimate, and its width spans the confidence interval.

Funnel plot asymmetry

An asymmetric funnel plot suggests small-study effects or publication bias; interpretation is unreliable with fewer than about ten pooled studies.

GRADE strong recommendation

A recommendation that most informed patients would choose the intervention, stated as recommended, based on clear balance of benefits and harms.

Implementation outcomes

Measures of the implementation process itself: acceptability, adoption, appropriateness, feasibility, fidelity, cost, penetration, and sustainability, distinct from clinical effectiveness.

PDSA cycle

Plan a change and prediction, Do a small-scale test, Study results against prediction, Act to adopt, adapt, or abandon the change.

Special-cause variation

Signal on a control chart from an assignable cause, such as a point beyond control limits or a nonrandom pattern, requiring investigation of that event.

Donabedian quality model

Evaluation of quality through structure (resources and organization), process (care delivered), and outcomes (results experienced by patients).

Balancing measures

Measures that detect unintended consequences of an improvement change elsewhere in the system, such as increased workload or delayed care.

Belmont Report principles

Respect for persons, beneficence, and justice; their applications are informed consent, risk-benefit assessment, and fair selection of research subjects.

SBAR communication

A structured handoff format covering Situation, Background, Assessment, and Recommendation to convey clinical information clearly and reliably.

Conflict of interest management

Beyond disclosure, management may include recusal from decisions, divestment, independent oversight, or exclusion from guideline development panels.