How Should Canceled or Shortened Sessions Appear in ABA Data?

Use separate states for every session

Record scheduled, canceled before start, started late, ended early, interrupted, completed, and rescheduled as distinct states. Add the responsible source only when known and use neutral reasons such as illness, client choice, staff absence, weather, transport, authorization hold, inaccessible setting, or technology failure.

Avoid marking an undelivered visit as a zero-response clinical session. No exposure and observed nonresponse mean different things.

Canceled ABA session data need the original denominator

If 20 sessions were scheduled, 15 completed, 2 ended early, and 3 were canceled before service, report those counts. A completion rate is 15 of 20 scheduled sessions, while delivered-session analysis may use 17 sessions with actual service. State the denominator beside each result.

Shortened sessions should retain actual minutes and opportunities rather than being scaled automatically to a full-session estimate.

Use only valid clinical evidence

A partial session can contribute data when the operational definition and observation conditions remain valid. Record the observation window, opportunities, prompts, supports, setting change, and reason for ending. A clinician should decide whether the evidence is comparable to full sessions.

The BCBA Test Content Outline covers measurement, graphing, interpretation, validity, and treatment integrity as examination content.

Keep administrative and financial records separate

The RBT Ethics Code requires accurate client data and billing records. The BACB Ethics Code addresses accurate reporting and billing, documentation, timeliness, and correction for covered behavior analysts.

A cancellation policy, employee pay rule, authorization unit rule, and payer claim rule can produce different outcomes. Reconcile them without copying one status into another.

A practical example

Zoe has eight scheduled sessions. Five are completed, one ends after 30 of 120 planned minutes because of illness, one is canceled by the family before start, and one is canceled for staff absence. Report 5 of 8 completed, 6 of 8 with delivered service, and the partial session's actual opportunities. Do not record the two undelivered sessions as zero skill performance.

Use a session-status ledger

Create one row for every scheduled visit. Record scheduled time, actual start and end, service minutes, status, neutral reason, source, notice time, setting, staff, valid clinical observation window, and reschedule or follow-up. Keep client choice and safety interruption visible when relevant.

This ledger supports several reports without forcing them to share one status. Attendance can use scheduled visits, clinical analysis can use valid observations, payroll can use compensable time under governing rules, and claims can use supported services under the payer route.

Compare partial sessions using actual exposure

A shortened visit may create fewer opportunities or less observation time. Show raw counts and actual denominators. For frequency measures, a rate may be useful when the observation conditions remain comparable. For opportunity measures, report responses over the opportunities that actually occurred.

Avoid multiplying a 30-minute count by four to estimate a planned two-hour session. The unobserved period may have contained different activities, fatigue, transitions, or opportunities. A projection is not observed clinical data.

Know when partial data are invalid

Data may become invalid if the session ends during a trial, a safety event changes conditions, the person's AAC is unavailable, the observer stops collecting, or the definition requires a longer period. Mark the exact point and reason rather than discarding every earlier observation automatically.

The qualified clinician can decide whether the valid portion supports the intended question. The family can ask which points were retained, which were excluded, and how the graph marks the shorter exposure.

Cancellations can bias the visible pattern

If cancellations cluster during illness, difficult settings, staffing gaps, or authorization holds, completed-session data may not represent the full period. Report cancellation reasons and timing separately from performance while protecting unnecessary personal detail.

For example, a graph may show strong clinic performance across six visits while four planned community visits were canceled for transportation. The evidence describes clinic performance. It should not support a claim across both settings.

Keep client choice neutral

A person may decline, pause, or end a session. Record the accessible communication, immediate response, safety context, and follow-up under the applicable plan. Avoid labeling every client-ended visit as noncompliance or converting it into zero performance.

The clinician can review whether the activity, setting, duration, or goal remains appropriate. Immediate safety or legal duties still apply when relevant.

Reconcile authorization and claim states carefully

Canceled service usually differs from delivered service for authorization use and claims, while payer rules vary by product, contract, service, and date. A partial visit may support only actual delivered units and required documentation. Qualified billing staff should verify the current rule.

If a claim, timesheet, or authorization tracker conflicts with the session record, preserve the evidence and route the correction. Do not edit the clinical duration solely to make another system reconcile.

A second example with several denominators

During one month, 20 sessions are scheduled. Four are canceled before start, three end early, and 13 are completed. Delivered-service rate is 16 of 20, or 80%. Completion rate is 13 of 20, or 65%. Clinical-data completeness might be 15 of 16 delivered sessions if one partial visit lacks a valid target measure.

Each percentage is correct for its named question. Reporting only 13 of 13 completed records would hide the seven other scheduled visits and the missing delivered-session measure.

Review patterns and corrective action

Look at cancellation and shortening by neutral reason, setting, day, staff coverage, access, and notice timing. Use the pattern to address scheduling, transportation, authorization, staffing, technology, or setting barriers. Avoid causal conclusions from small counts.

At plan review, ask whether repeated shortened visits make the current dosage or setting unrealistic. A qualified clinician can revise the clinical plan with the person and family; operations can repair the delivery barrier.

Set family-facing expectations

The provider can explain how to cancel, how client-requested stops are handled, how partial observations appear on graphs, when rescheduling is offered, and how billing questions are resolved. Families should receive a contact for discrepancies and an updated record when a correction affects them.

The clearest report accounts for every scheduled session and keeps attendance, clinical evidence, authorization, payroll, and claim status aligned without pretending they are the same record.

Questions families can use

Ask what was scheduled, what was delivered, which reason was recorded, whether the client chose to stop, how partial data are interpreted, which denominator appears in reports, whether authorization units changed, what was billed, and how corrections or rescheduling are tracked.