Evidence-based practice

Evidence-based practice is an integration

The practice question is broader than “Did a study find an effect?” A clinician asks whether credible research addresses a similar problem, whether the methods and outcomes are trustworthy, whether the person wants the target and approach, and whether the setting can deliver it safely.

Slocum and colleagues describe evidence-based ABA as integrating best available evidence, clinical expertise, and client values and context. These parts interact rather than forming a ranked checklist.

Start with an answerable question

Define the person or population, activity or concern, proposed approach, comparison or alternative, meaningful outcomes, risks, setting, and time. A narrow question produces a more useful search than “Does ABA work?”

Include outcomes the person values, such as communication access, autonomy, participation, pain, time burden, relationships, or quality of life. A proxy measure can miss the actual priority.

Appraise the body of evidence

Look at research design, measurement quality, replication, sample, setting, comparison, attrition, missing outcomes, follow-up, treatment fidelity, adverse effects, and conflicts. One positive paper can be weaker than a consistent body of transparent studies.

Distinguish efficacy under research conditions from effectiveness in ordinary practice. Check whether evidence includes people with similar communication, health, culture, age, support needs, and goals.

Clinical expertise is bounded

Expertise includes assessment, technical skill, knowledge of alternatives, cultural responsiveness, communication, measurement, coordination, and recognition of limits. It also includes referring when the question requires another profession.

Experience can produce useful judgment and bias. Record the rationale, invite peer review, use data, and explain uncertainty. Certification or seniority does not establish competence for every condition, method, person, or setting.

Client values change the decision

Ask what the person wants more or less of, which outcomes matter, what feels acceptable, and how burden fits daily life. Offer accessible explanations of options, likely benefits, uncertainties, risks, and alternatives.

Honor refusal and withdrawal through the governing consent and assent process. A plan can have strong research support and still be a poor fit for one person’s goal, identity, sensory needs, schedule, culture, or risk tolerance.

A fictional decision example

Fictional adult Mei wants to travel independently to a weekly ceramics class. The clinician finds research on route instruction, reviews Mei’s prior travel skills and vision needs, and discusses options with Mei. Mei chooses picture-based route checks and practice during quiet travel times.

Across five baseline trips, Mei independently completes 9 of 20 defined route checks. During six supported trips, she completes 27 of 30. Mei rates five of those trips acceptable and asks to keep a human backup for transfers.

The team retains the backup, checks for prompt dependence and safety, and tests the plan on a second familiar route. The change is compatible with improvement, but the small uncontrolled sequence does not isolate a cause.

Use ongoing evidence from the actual case

Define the target, opportunity, data source, decision rule, unwanted effects, review date, and client feedback before implementation. Keep raw data and meaningful events visible.

When progress is weak, examine measurement, health, communication, access, implementation, teaching, environment, and goal fit. Increasing intensity is only one possible response and needs its own evidence and consent.

Implementation fit is part of practice

A plan requiring unavailable supervision, inaccessible materials, or unsustainable family time will not operate as studied. Assess staff competence, setting, technology, travel, privacy, cost, coordination, and maintenance.

Adaptation can improve fit, yet a large change may move beyond the evidence. Document what changed and why, then evaluate the adapted plan rather than citing the original study as proof.

When evidence points in different directions

Research, client preference, clinician judgment, and setting constraints may conflict. Name the conflict and its uncertainty. Compare feasible alternatives, seek consultation, and agree on a reversible next step with monitoring and stop criteria.

Avoid presenting the clinician’s preferred option as the only evidence-based choice. Record why the selected option fits now and what new evidence would trigger reconsideration. Shared decisions can change as the person’s priorities or circumstances change.

When no option has strong support, say so plainly. A smaller, reversible trial with agreed measures may be more responsible than a confident recommendation built on weak evidence.

Keep the BACB source in scope

The BACB BCBA Test Content Outline, 6th edition covers identifying research-based interventions, evaluating evidence, client preferences, contextual variables, and data-based decisions. It is training and examination content. It does not endorse an individual plan or establish authority to practice.