An ABA risk benefit analysis worksheet should keep each option's expected benefits, risks, burdens, alternatives, unknowns, safeguards, and evidence visible to the people making a case-specific decision. It should not produce a number that authorizes treatment. The copyable record below preserves client perspective, unmet safeguards, dissent, decision ownership, and the conditions for stopping or reconsidering an option.

Clinicians & ABA Professionals / Clinical Management, Supervision and Leadership.

This worksheet is educational and cannot determine whether an assessment or intervention is appropriate. It is not informed consent, an assent determination, a human-rights review, a safety plan, an incident report, a medical decision, a treatment plan, or legal advice. Qualified professionals must follow current clinical evidence and the applicable consent, licensure, payer, organizational, regulatory, and legal processes.

Compare defined options, not optimistic labels

Start with the client's purpose and the decision to be made. Then describe every live option at comparable detail, including continuing current supports, modifying an option, gathering more information, consulting or referring, pausing, and taking no new action when that is genuinely available. “Do treatment” and “do nothing” are not adequate option definitions.

The current Ethics Code for Behavior Analysts, reached through the BACB's ethics-code page, addresses maximizing benefits, minimizing risk of harm, assessment, and intervention selection. It also addresses risks, benefits, side effects, client and stakeholder preference, less intrusive means, informed consent, data use, documentation, and continual evaluation. The code does not prescribe this worksheet or a universal scoring method.

The BCBA Test Content Outline, Sixth Edition identifies ethics, risk, collaboration, culturally responsive practice, assessment, client-informed goals, intervention selection, and data-based evaluation as entry-level knowledge areas. The BACB's test-content outline page identifies the current outline. The CASP ABA Practice Guidelines Version 3 access page identifies a licensed guideline intended to inform ABA assessment and treatment decisions for autism. This article neither reproduces that licensed material nor claims that the worksheet meets a specific review standard.

A complete table can still support the wrong decision

Benefit, risk, burden, and uncertainty are not one-dimensional. A low-frequency harm can be serious. A highly likely benefit can be trivial to the client. A procedure can appear efficient while creating intolerable distress or excluding the person's communication method. A less intrusive option may need more time or resources without being less worthy of consideration.

A review of ethical behavior analysis through evidence-based practice places benefit, harm avoidance, dignity, autonomy, justice, integrity, competence, and effective practice within a broader decision framework. A proposed functional-analysis risk-assessment decision tool illustrates one attempt to structure assessment risk, but it is specific to that context and does not validate a generic score for all ABA decisions. A paper on socially meaningful, risk-driven case conceptualization discusses risk to desired and quality-of-life outcomes. These sources support disciplined questions, not arithmetic permission.

Copyable review charter

FieldEntryClient-described purpose and desired changeDecision that must be madeCurrent plan, assessment, and evidence versionsDirect-client perspective and communication methodCaregiver and other stakeholder perspectives, separately attributedOptions that will be comparedQualified decision owner and scope of authorityRequired clinical, medical, interdisciplinary, human-rights, consent, legal, or payer reviewImmediate stop or escalation conditionsReview date, record version, and next review date

Build communication access into the review conditions. ASHA Communication Access recommends learning how the person communicates and which supports work for them. Address the person directly, allow time, use the preferred method, and verify that the exchange was understood. Record how options and uncertainty were explained, how the person responded, and when someone else is reporting from a different role.

Option register and evidence ledger

Use the ABA risk benefit analysis worksheet to name each live option before comparing it. Give current supports, modified supports, consultation, referral, a pause, and no new action enough detail to make the differences reviewable.

Option IDExact procedure or actionWhat stays unchanged?Source and versionWho may implement?Setting and durationStatusproposed / current / unavailable / rejectedproposed / current / unavailable / rejectedproposed / current / unavailable / rejectedproposed / current / unavailable / rejected

Claim or concernOptionEvidence source and dateDirect, indirect, research, policy, or opinion?Population and setting fitLimitation or conflictReviewer

“No adverse event found” is not the same as evidence that an option is harmless. A systematic review and meta-analysis of early-childhood autism interventions reported inadequate monitoring of adverse events, limiting benefit-harm interpretation. A separate review of adverse-event reporting in intervention research for young autistic children likewise describes major reporting gaps. Research evidence and clinical service monitoring are different activities, but both sources show why missing harm evidence should stay visible.

Separate benefit, risk, burden, and alternative panels

OptionExpected benefit to the clientOutcome and time horizonProbability or uncertainty basisClient importanceEvidence gap

OptionRisk or unwanted effectSeverity and reversibilityExposure or triggerDetection methodImmediate responseResidual uncertainty

OptionTime, effort, distress, opportunity, financial, or access burdenWho bears it?Client viewMitigationBurden after mitigation

AlternativeHow it differsExpected benefitRisk and burdenWhy feasible, pending, or unavailableEvidence or consultation neededNo new action for now

Do not hide risks of continuing the current plan or delaying action. Also do not manufacture a “no treatment” comparison when ordinary supports, environmental changes, medical evaluation, communication access, consultation, or a different service remain possible.

Safeguard-readiness log

Safeguard or preconditionWhy requiredVerified, pending, not met, or not applicableEvidence and dateOwnerBlocking?Recheck dateyes / noyes / noyes / noyes / noyes / noyes / noyes / noyes / no

A contemporary review of the right to effective behavioral treatment discusses welfare, effective treatment, collaboration, and less intrusive procedures in relation to current ethical expectations. The ABAI position statement on restraint and seclusion is an older, narrow statement about restrictive practices and should not be generalized to ordinary treatment or treated as current law. When restrictive, punishment-based, emergency, or high-risk procedures are at issue, the applicable review, training, authorization, monitoring, rights, medical, and legal requirements need case-specific verification.

An unmet blocking safeguard remains unmet even when many other rows are complete. Use “not applicable” only when the reviewer records why the safeguard does not apply to that exact option and version. Do not change “not met” to “pending” to improve a readiness percentage. If the safeguard cannot be met, revise or reject the option, obtain the required help, or use the authorized escalation path.

Decision and monitoring record

Decision fieldEntryOptions actually comparedClient preference, questions, dissent, and accessCaregiver and other stakeholder views, separately attributedBenefits with supporting evidence and uncertaintyRisks, unwanted effects, burdens, and no-action risksLess intrusive and other reasonable alternativesUnmet, pending, and verified safeguardsRequired consent, authorization, consultation, or reviewDecision: proceed, revise, hold, reject, gather evidence, consult, or referRationale and unresolved uncertaintyMonitoring measures and source recordsStop, pause, escalation, and reconsideration conditionsQualified reviewer, signature, date, and next review

The record should point to the approved plan and consent documentation rather than becoming either one. If authority is shared or disputed, state that plainly. Consultation can inform the decision without transferring the decision owner's responsibility.

Fictional worked example

This example is fictional. It does not describe a real client, recommend a procedure, define a high-risk intervention, or establish a standard of care.

A team is considering a short, structured transition-support trial for a fictional client-selected library activity. The option remains intentionally generic. Before any start, the qualified reviewer lists eight case-specific safeguards: accessible explanation, current assessment basis, direct-client willingness check, caregiver communication, staff competence verification, agreed stop signal, monitoring method, and required supervisory approval.

At the review, six safeguards are verified. The monitoring method is pending final setup. Staff competence verification is not met and is marked as blocking. Verified safeguard coverage is therefore 6 / 8 × 100 = 75.0%.

Safeguard stateCountInterpretationVerified6Evidence is present for this version and datePending1Monitoring setup is unfinishedNot met1Required competence evidence is absent; blockingTotal8All planned rows remain visible

The fictional decision is “hold and revise,” not “75% ready.” The essential unmet competence condition blocks the trial, and the pending monitoring system also requires closure. Six verified rows cannot average away either fact. If the team later supplies the missing evidence, it must update the record, recheck client willingness and context, and obtain whatever current approval is required. The worksheet still would not prove that expected benefits outweigh risks.

Reopen the review when the option or evidence changes

Reconsider when the goal, procedure, dose, setting, implementer, client preference, assent or consent status, health information, or communication access changes. Reopen the record as well when evidence, an adverse-effect signal, a burden, an alternative, a safeguard, the law, a payer term, or the responsible reviewer changes. Preserve the prior version and explain the transition.

Treat unexpected deterioration, pain, distress, injury, coercion, loss of communication access, withdrawn participation, suspected abuse, medical concern, or other urgent welfare signal according to the applicable immediate response process. Do not wait for a scheduled worksheet review. A useful record slows down unsupported certainty without slowing down necessary protection.

Related resources

Sources