EBP-EAP key terms

A problem-solving approach to care that combines the best available research evidence, clinician expertise, and patient values and preferences, then evaluates the results of the decision. It is a process, not a single product.

A ranking of study designs by how well they control bias for a given question, with systematic reviews and randomized trials near the top for treatment questions and expert opinion near the bottom. Rankings shift by question type.

A graded scale assigned to individual studies or bodies of evidence to signal confidence in the findings. Scales differ by organization, so the scale used must always be named when citing a level.

A strength rating attached to a clinical recommendation, reflecting both the quality of supporting evidence and the balance of benefits, harms, and patient preferences. Strong and weak (conditional) grades guide how firmly to act.

A structured system for rating certainty of evidence and strength of recommendations across outcomes. Certainty starts high for randomized evidence and is downgraded for risk of bias, inconsistency, indirectness, imprecision, and publication bias.

A structured clinical question naming the Population, Intervention, Comparison, and Outcome. It converts a vague clinical concern into searchable concepts and clarifies what evidence would actually answer the question.

A focused, answerable question about a specific patient or population, such as the effect of a named therapy on a defined outcome. It contrasts with broad background questions about a condition's general nature.

A broad question about a condition, its pathophysiology, or general management, usually answered by textbooks and narrative reviews. It typically precedes and informs a focused foreground question.

A reproducible plan specifying databases, search terms, Boolean operators, filters, and limits used to find evidence. It should be documented well enough that another person could rerun it and retrieve similar results.

The connector words AND, OR, and NOT used to combine search concepts. AND narrows by requiring all concepts, OR broadens by accepting synonyms, and NOT excludes terms, risking loss of relevant records.

A standardized set of subject headings, such as MeSH, assigned by indexers to describe an article's content consistently. Searching these headings captures relevant records even when authors use different wording.

Free-text searching for words as they appear in titles, abstracts, or full text. It captures newly indexed or unindexed material and author-specific phrasing but misses synonym variations that controlled headings would catch.

Evidence produced outside commercial publishing, including conference abstracts, dissertations, government reports, and trial registries. It reduces publication bias in reviews by surfacing studies that conventional database searches miss.

Predefined characteristics that determine whether a study is eligible for a review or guideline, such as population, design, outcomes, and date range. Setting them in advance limits selective inclusion of favorable studies.

The staged process of reviewing titles, abstracts, then full texts against eligibility criteria to decide which records enter a review. Independent duplicate screening reduces errors and arbitrary exclusions.

The systematic judgment of a study's validity, results, and applicability to a clinical question. It asks whether the design controlled bias, whether the findings are precise, and whether the patients resemble yours.

The degree to which a study's design and conduct support the conclusion that the observed effect is real rather than caused by confounding, selection, or measurement error. It concerns the study's own logic.

The extent to which study findings apply to patients, settings, and circumstances beyond those studied. Narrow eligibility, specialized centers, and short follow-up all weaken it.

The likelihood that a study's results are distorted by flaws in design, conduct, or reporting, such as inadequate randomization or unblinded outcome assessment. It is assessed per domain, not as a single global judgment.

Distortion of an exposure-outcome relationship by a third variable that is associated with both and is not on the causal pathway. Randomization and adjustment in analysis are the main defenses against it.

Systematic error arising when the groups compared differ in ways related to the outcome, or when included participants differ from the target population. It can enter at enrollment, allocation, or follow-up.

Allocating participants to groups by chance so that known and unknown confounders are distributed similarly across arms. It supports causal inference and must be paired with concealment to succeed.

Hiding the upcoming group assignment from the person enrolling participants, for example with central randomization. It prevents recruiters from steering certain patients into a preferred arm and protects the randomization.

Keeping participants, clinicians, or outcome assessors unaware of group assignment to prevent expectations from influencing behavior or measurement. Blinding of assessors matters most for subjective outcomes.

Distortion caused by participants dropping out or being lost to follow-up in ways related to their group and outcome. High or differential loss threatens the comparability that randomization created.

Analyzing participants in the group to which they were randomized regardless of adherence or crossover. It preserves the benefits of randomization and gives a conservative estimate of effectiveness in real-world use.

Analyzing only participants who adhered to the assigned protocol, which can show the treatment's effect under ideal use. It breaks randomization and risks overestimating benefit because adherers often differ from non-adherers.

A prospective experiment in which participants are randomly assigned to an intervention or control condition and followed for outcomes. It is the strongest single design for questions of treatment effect.

A structured summary of all studies meeting prespecified eligibility criteria for a focused question, using explicit search, appraisal, and synthesis methods. It reduces bias in summarizing a body of evidence.

A statistical technique that pools effect estimates from multiple studies to produce a combined estimate with greater precision. Pooling is only defensible when studies are clinically and methodologically similar enough.

Variation in results across studies beyond what chance alone would produce, arising from differences in populations, interventions, or methods. High heterogeneity argues against simple pooling and calls for explanation.

A graph displaying each study's effect estimate with its confidence interval alongside a pooled summary estimate. It lets readers see the spread of results, precision, and the influence of individual studies at a glance.

A range of values likely to contain the true effect, with the width reflecting precision. A 95% interval means that repeated sampling would capture the true value in about 95% of cases.

The probability of observing a result at least as extreme as the one found if the null hypothesis were true. It indicates statistical compatibility, not the size or clinical importance of an effect.

Whether an effect is large enough to matter to patients and practice, judged against a minimal important difference rather than a threshold for statistical significance. Small but real effects may be clinically trivial.

The smallest change in an outcome that patients would perceive as beneficial or that would justify a change in care. It anchors interpretation of effect sizes and sample size planning.

The number of patients who must receive an intervention for one additional patient to benefit, calculated as the reciprocal of the absolute risk reduction. Lower values indicate a more efficient treatment.

The arithmetic difference in event rates between control and treatment groups, expressed in percentage points. Unlike relative measures, it conveys the actual size of benefit and depends on baseline risk.

The ratio of the odds of an outcome in one group to the odds in another, often reported in case-control and logistic analyses. It approximates relative risk only when the outcome is rare.

Distortion of the evidence base because studies with positive or significant results are more likely to be published, and faster, than null ones. Funnel plot asymmetry and registry searches help detect it.

The judgment of whether evidence from a study or guideline fits a particular patient, setting, and resource context. It weighs patient characteristics, baseline risk, and feasibility alongside the strength of findings.

A collaborative conversation in which clinicians present evidence about options and the patient's values and preferences shape the choice. It is most useful when options carry different trade-offs.

The values, goals, and priorities a patient brings to a care decision, including attitudes toward risk, inconvenience, and outcomes. They are one of the three core elements of evidence-based practice.

A tool that presents balanced information about options, benefits, and harms in plain language to support shared decisions. Well-designed aids improve knowledge and reduce decisional conflict.

A person's capacity to obtain, understand, and use health information to make decisions. Low health literacy changes how evidence should be communicated and is a common barrier to informed choice.

The study of methods that promote the systematic uptake of evidence-based practices into routine care. It addresses context, barriers, and strategies rather than assuming that good evidence spreads on its own.

The process of moving research findings into practice through dissemination, adaptation, and support for change. It spans from summarizing evidence to sustaining its use at the point of care.

Systematically developed statements that recommend care for specific clinical circumstances, based on a review of evidence and an explicit assessment of benefits and harms. Trustworthy guidelines disclose conflicts and date their search.

A small-scale, iterative improvement method in which a change is planned, tested, observed, and then adapted or adopted. Repeated rapid cycles build local evidence before wider spread.

A systematic, data-guided effort to make measurable improvements in care processes and outcomes within a specific setting. It uses local data and repeated tests of change rather than generalizable research designs.

An outcome tracked during improvement work to detect unintended harm caused by the change, such as longer waits after a new triage process. It guards against improving one result by damaging another.

A graph of a measure over time used in improvement work to distinguish real change from ordinary variation. It shows trends, shifts, and patterns without requiring statistical control limits.

A method using control charts with calculated limits to separate common-cause variation inherent in a process from special-cause variation signaling an actual change. It guides whether to act on a process.

The extent to which an improvement or evidence-based practice continues after initial project support ends. It depends on embedded workflows, ownership, and ongoing measurement rather than enthusiasm alone.

The process of giving a capable person the information needed to voluntarily agree to or refuse care or research participation. It requires disclosure, comprehension, voluntariness, and capacity.

A situation where financial or personal interests could improperly influence professional judgment or the presentation of evidence. Disclosure and management, not merely awareness, are the expected responses.

Working with practitioners from different professions toward shared goals through mutual respect, clear roles, and coordinated communication. It improves care and is essential for translating evidence into practice.

Structured exchange of information among team members using tools such as closed-loop confirmation and standardized handoffs. Clear communication reduces errors and supports consistent application of evidence.

The principles governing studies involving humans or animals, including respect for persons, beneficence, and justice. Oversight by review boards and honest reporting are core obligations.

A person's ability to understand relevant information, appreciate consequences, reason about options, and express a choice. It is decision-specific and assessed rather than assumed from a diagnosis.