Are your CME evaluations measuring satisfaction, or demonstrating that clinicians learned something and changed their practice? An excellent session rating answers whether participants valued the experience. It doesn't establish whether they can apply a guideline, change a clinical decision, or improve patient care.
Effective CME evaluation questions separate reaction, learning, behavior, and results. Those are the categories in Kirkpatrick's evaluation model used throughout this guide. Mapping each question to a level helps an education committee see what its evidence supports, and where its evaluation form stops short.
That distinction matters because educational impact has long been difficult to document. A systematic review of formal CME found that only 2 of 32 reviewed studies addressed the relevant higher-level outcomes involving physician knowledge and attitudes, practice, and patient health.
Build your evaluation around these question types:
- Reaction: Overall satisfaction, faculty effectiveness, independence perceptions, and operational access.
- Learning: Objective achievement and pre/post knowledge or competence assessment.
- Behavior: Commitment to change, followed by evidence of implementation.
- Results: Follow-up questions about measurable effects on care.
Not every activity needs an extensive outcomes study. Every activity does need a clear distinction between what learners enjoyed, what they demonstrated, and what they did afterward.
Table of Contents
- 1. Overall Session and Program Satisfaction Rating (Likert Scale)
- 2. Learning Objective Achievement Questions
- 3. Pre- and Post-Activity Knowledge and Competence Self-Assessment
- 4. Intended Practice Change and Commitment-to-Change Questions
- 5. Faculty Effectiveness and Presentation Quality Questions
- 6. Commercial Bias and Independence Perception Questions
- 7. Logistics, Technology, Credit Claim, and Certificate Access Questions
- 8. Longitudinal Outcome Follow-Up Questions (Kirkpatrick Levels 3 and 4)
- The 8 CME Evaluation Question Types at a Glance
- Building an Evaluation Form That Proves Impact
1. Overall Session and Program Satisfaction Rating (Likert Scale)
Kirkpatrick mapping: Level 1, reaction. Overall satisfaction questions measure the learner's judgment of the educational experience. They're useful for program improvement, but they aren't evidence of learning.

A practical item is: How satisfied were you with the overall educational quality of this session? Use a consistently labeled scale from very dissatisfied to very satisfied, with a separate not-applicable option when needed.
Don't combine educational quality, faculty performance, and clinical relevance into the same statement. A learner may appreciate the speaker but find the content irrelevant to their specialty, and a combined rating won't tell the committee what to change.
Make the rating comparable, not overcomplicated
For a regional society meeting, use the same overall-quality item across tracks, then examine session results alongside attendance mode and respondent counts. Keep the wording stable between program cycles so changes in the questionnaire don't masquerade as changes in quality.
- Keep the core short: Reserve the immediate form for questions that support an actual decision.
- Show the denominator: Report completed evaluations alongside eligible participants.
- Add an explanation: Ask what should be retained or changed, rather than requesting an unrestricted comment.
A systematic review of UK course evaluations found that shorter questionnaires produced higher response rates. That's a reason to remove repetitive satisfaction items, not to remove learning assessment.
Use these post-event survey questions to distinguish experience feedback from questions about educational impact.
Practical rule: Keep satisfaction as a baseline. Never relabel it as competence, practice improvement, or patient benefit.
2. Learning Objective Achievement Questions
An oncology symposium promises that participants will identify appropriate treatment options. Its evaluation should ask about that objective directly, rather than whether the activity broadly met expectations.
Kirkpatrick mapping: Level 2, learning, when achievement is demonstrated. A participant's rating that an objective was met is only perceived achievement. A content-specific assessment provides stronger evidence.

The CME for MOC Evaluation Guide describes an evaluation component that assesses individual knowledge, competence, or skill rather than satisfaction alone. Apply that distinction when converting planning objectives into evaluation items.
Match the question to the objective's verb
If the objective says identify, ask learners to identify the correct option. If it says apply, give them a clinical scenario requiring application, not a rating of whether the presentation was clear.
- Perceived achievement: How well did this activity prepare you to identify appropriate treatment options?
- Demonstrated achievement: Which treatment option best fits the patient scenario presented?
- Remaining gap: Which part of this objective requires additional education?
For a scientific program committee, keep the objective, assessment item, answer rationale, and results connected in the activity record. Faculty can then distinguish an objective that wasn't adequately addressed from an assessment question that was poorly written.
The broader explanation of why learning outcomes matter for applicants is useful when clarifying the difference between course content and what a learner should be able to do.
Don't treat unanimous agreement that an objective was covered as proof that everyone mastered it. Report perceived achievement and demonstrated achievement separately.
3. Pre- and Post-Activity Knowledge and Competence Self-Assessment
Confidence can improve without demonstrated competence improving. Paired assessments are useful only when the report preserves that distinction.
Kirkpatrick mapping: Level 2, learning. Self-assessment measures perceived knowledge, skill, or confidence; scored questions and case-based exercises assess demonstrated learning or application.

For a webinar about interpreting a diagnostic guideline, ask before and after: How confident are you in applying this guideline to a patient with competing risk factors? Keep both the wording and response labels identical.
Pair that item with a clinical vignette requiring a decision and an explanation. A peer-reviewed CME evaluation study distinguishes satisfaction, learning, and competence as separate constructs, supporting a question bank that doesn't collapse them into a single score.
Build a matched comparison
Use a protected registration or login identifier to connect responses. Explain why matching is necessary, restrict access to identifiable results, and report aggregate findings to the program committee.
- Match participants: Compare responses from learners who completed both assessments.
- Separate measures: Report confidence changes apart from scored knowledge results.
- Inspect missed concepts: Review question-level responses before summarizing overall improvement.
- Provide feedback: Explain the correct answer and clinical reasoning after assessment.
The trade-off is administrative effort. A brief matched assessment requires more setup than an anonymous end-of-session survey, but it supports a more defensible interpretation.
Don't report the difference between unrelated pre-activity and post-activity groups as though every individual improved. If matching isn't possible, describe the groups clearly and acknowledge that participation differences may affect the comparison.
4. Intended Practice Change and Commitment-to-Change Questions
Ask participants what they will do differently, not whether they intend to improve their practice. A useful commitment names an observable action, a clinical setting, and an implementation trigger.
Kirkpatrick mapping: A precursor to Level 3, behavior. Intention isn't observed behavior, so label the result planned practice change until follow-up confirms implementation.

A useful prompt is: What specific action will you take in your clinical work because of this activity? Follow it with: What might prevent you from making that change?
For a medication-safety course, a participant might commit to using a medication-reconciliation checklist during eligible visits. That's more actionable than a general promise to provide safer care.
Make commitments usable later
Preserve the original statement so the follow-up survey can return it to the learner. Group responses into consistent categories such as diagnostic decisions, treatment selection, referral, communication, and workflow.
- Specify the action: Ask what the learner will start, stop, or modify.
- Allow no change: Include options for already following the practice or finding it inapplicable.
- Identify dependencies: Ask whether implementation requires approval, staffing, technology, or team agreement.
Joint Accreditation's criteria connect program evaluation with changes in team skills or strategy, performance, and patient outcomes. A commitment question helps connect the activity to that broader improvement process, but doesn't complete the measurement.
Where clinical applicability needs clarification, participants can find condition-specific clinical guidance rather than relying on an isolated presentation slide.
A detailed intention is a follow-up target, not proof of practice change.
5. Faculty Effectiveness and Presentation Quality Questions
Kirkpatrick mapping: Level 1, reaction. Faculty ratings describe how learners experienced instruction. They don't establish that the speaker improved clinical performance.
Start with observable teaching behaviors: The faculty member explained the clinical reasoning clearly. Another useful item is: The faculty member used examples relevant to the stated learning objectives.
Avoid personality-based questions about charisma, enthusiasm, or likability. Those responses won't tell a faculty-development lead whether to revise a dense slide deck, add case discussion, or explain conflicting evidence more clearly.
Give speakers feedback they can act on
Suppose a scientific committee is reviewing a first-time presenter. Strong relevance ratings paired with comments about unreadable slides call for presentation support, not automatic exclusion from the next program.
- Clarity: Ask whether explanations were understandable.
- Application: Ask whether examples helped learners use the content.
- Evidence: Ask whether recommendations were explained with supporting evidence.
- Interaction: Ask whether participants could obtain clarification, where the format allowed it.
Add Which aspect of the teaching most helped or hindered your understanding? This gives the speaker a concrete improvement target without turning the evaluation into an unrestricted personal review.
Response context matters. A study of web-based surveys found an average response rate of 57% for meeting and conference evaluations, which shouldn't be treated as a guarantee of representativeness.
Set a reporting threshold before reviewing faculty results, show respondent counts, and avoid ranking speakers when the underlying samples aren't comparable. Share individual feedback privately, while giving the committee aggregate themes and documented recommendations.
Keep faculty reaction scores separate from objective-achievement results. An engaging presentation can still leave a learning gap, and a demanding session can produce useful learning despite mixed satisfaction.
6. Commercial Bias and Independence Perception Questions
A learner's concern about commercial influence deserves review, but a perception survey isn't a substitute for disclosure collection, relationship resolution, and content review.
Kirkpatrick mapping: Level 1, reaction, with a separate independence-monitoring purpose. These questions measure perceived balance and promotional influence, not knowledge acquisition.
Use a neutral prompt: Did you perceive commercial bias in this activity? If the learner answers yes, ask them to identify the session, content, and reason for concern.
Collect enough detail to investigate
A comment that an activity felt promotional is difficult to assess without context. Request the product or topic involved, the relevant recommendation, and whether alternatives appeared to be omitted or treated unevenly.
- Use conditional detail: Display the follow-up field when a concern is reported.
- Assign an owner: Route flagged responses to the person responsible for independence review.
- Connect records: Examine the concern alongside disclosures, presentation files, and content-review notes.
- Document disposition: Record what was reviewed, the conclusion, and any corrective action.
For example, a symposium participant may flag repeated brand references and a lack of discussion of alternatives. The education team should examine the actual slides and clinical context before deciding whether faculty coaching, content revision, or another response is warranted.
Include the item across the program rather than limiting it to activities with commercial support. Use financial disclosure management as a connected preparation process, not a replacement for participant feedback.
Review standard: A flagged comment is a reason to investigate. An absence of comments isn't proof that an activity was free from influence.
Keep the wording stable so the committee can monitor concerns over time. Don't bury independence findings inside an overall satisfaction average.
7. Logistics, Technology, Credit Claim, and Certificate Access Questions
Could a participant complete the activity but still miss a handout, lose access to the virtual session, or fail to retrieve a certificate? These are different operational problems, so ask about the task that failed rather than folding every issue into a general satisfaction rating.
Kirkpatrick mapping: Level 1, reaction and participation conditions. Logistics questions identify barriers to taking part. They do not measure learning.
Ask whether participants could access the presentation materials and whether credit requirements were clear. If someone reports a problem, follow up with the specific session, access method, or step they could not complete.
Separate access failures from educational judgments
For a hybrid meeting, use attendance mode to show relevant questions. Ask virtual attendees about audio and platform access; ask onsite attendees about venue access. A participant who could not hear a clinical discussion had an access problem that may have prevented participation, not necessarily a content complaint.
- Materials: Could the participant open the relevant handouts or presentations?
- Technology: Which session and platform problem affected access?
- Credit requirements: Did the participant understand what was required to claim credit?
- Certificates: Could an eligible participant find and open the certificate?
A national GP CME program analysis identified regional resource differences and limitations in evaluating implementation. Survey responses alone may not capture the experience of people who could not participate or complete the evaluation.
Give non-attendees and people who abandoned registration a separate way to report access barriers. If international participation matters, ask about the activity's actual access requirements rather than assuming that general information about postgraduate opportunities for foreign-trained doctors explains those barriers.
Use a consistent participation certificate format. Then check evaluation completion, credit eligibility, and certificate availability against the activity's rules. Staff can verify an administrative mismatch directly, so learners should not have to diagnose it.
8. Longitudinal Outcome Follow-Up Questions (Kirkpatrick Levels 3 and 4)
Kirkpatrick mapping: Level 3, behavior, and Level 4, results. Follow-up questions ask whether participants implemented learning and whether the change affected care. Keep those outcomes separate.
A virtual gynecologic-oncology CPD evaluation assessed reaction and learning but not later behavior, clinical performance, or patient outcomes. That boundary is important: immediate feedback can't establish sustained workplace change.
For a screening-protocol activity, return the learner's original commitment and ask: Have you implemented this change? Offer responses that distinguish full implementation, partial implementation, no opportunity, and a decision not to proceed.
Ask for action, then evidence
Follow with What did you change? and What prevented or limited implementation? For results, ask which measure was examined and what evidence is available, without requesting identifiable patient information.
- Behavior evidence: A revised workflow, changed clinical decision, or adopted team process.
- Results evidence: An appropriate quality, safety, efficiency, or equity measure.
- Attribution context: Other training, policy changes, staffing changes, or technology introduced during the same period.
ACCME reported that 8% of accredited CME activities in its system measured community or population health outcomes in 2024, according to this published discussion of CME outcomes. Such measurement requires deliberate planning, not an extra satisfaction item.
Schedule follow-up when implementation could reasonably have occurred, and set the date before launching the activity. Report the invited population, respondents, and evidence type.
Self-reported change is useful evidence of reported behavior. It isn't equivalent to an independently verified improvement in patient outcomes.
If participants haven't changed practice, retain the finding. Barriers may reveal a need for team-based education, implementation support, or a better-targeted activity.
The 8 CME Evaluation Question Types at a Glance
Use the Kirkpatrick level to see what each question actually measures. Satisfaction and faculty ratings capture reaction; objective-linked assessments address learning; follow-up questions examine behavior and results. Commitment to change records an intention, not behavior itself. Commercial-bias perception is a Level 1 reaction measure with a separate independence-monitoring purpose, while logistics questions assess event operations rather than learning outcomes.
| Question Type | Kirkpatrick Level | Question Format | Best Used For |
|---|---|---|---|
| Overall session and program satisfaction rating | Level 1: Reaction | 5- or 7-point Likert scale | Comparing learner satisfaction across sessions |
| Learning objective achievement | Level 2: Learning | Objective-linked agreement or confidence scale | Checking whether stated objectives were addressed |
| Pre- and post-activity knowledge and competence self-assessment | Level 2: Learning | Matched pre- and post-activity rating items | Measuring perceived knowledge or confidence gains |
| Intended practice change and commitment to change | Precursor to Level 3: Behavior | Open-ended commitment or likelihood rating | Capturing planned application and setting up follow-up |
| Faculty effectiveness and presentation quality | Level 1: Reaction | Speaker-specific ratings or yes/no items | Improving presentation quality and guiding faculty feedback |
| Commercial bias and independence perception | Level 1: Reaction, with a separate independence-monitoring purpose | Yes/no, agreement scale, or open-ended response | Detecting perceived promotional influence |
| Logistics, technology, credit claim, and certificate access | Not a Kirkpatrick learning outcome measure | Rating scales, yes/no items, or open-ended response | Improving event operations and credit access |
| Longitudinal outcome follow-up | Levels 3 and 4: Behavior and results | Follow-up survey with specific actions and outcomes | Assessing reported practice change and longer-term impact |
Building an Evaluation Form That Proves Impact
An effective evaluation form doesn't ask every possible question. It collects the evidence needed to judge whether the activity achieved its educational purpose, identifies barriers, and supports the next program decision.
Start by labeling every existing question reaction, learning, behavior, results, or administration. Be honest about ambiguous items: confidence is perceived learning, commitment is intended behavior, and satisfaction remains reaction regardless of how favorable the score looks.
Audit the form before the next activity
- Keep a satisfaction baseline: Use consistent wording and a clearly labeled scale, then request actionable feedback.
- Connect objectives to assessment: Match each objective with an appropriate question or demonstration, not just a general rating.
- Separate confidence from competence: Preserve self-assessment, but add scored knowledge or application questions where feasible.
- Capture a specific commitment: Ask what participants will change and what support they'll need.
- Review teaching and independence separately: Give faculty useful feedback and investigate commercial-bias concerns through a documented process.
- Check operational access: Ask about technology, materials, credit requirements, and certificates without blending those answers into content scores.
- Schedule follow-up: Return to the commitment, ask what happened, and request an appropriate measure when results can be assessed.
- Report limitations: Include response rates, subgroup context, missing data, and whether outcomes were self-reported or independently documented.
Keep a stable core question bank, then add specialty-specific items only when they serve a defined purpose. Store question wording, objective mappings, response options, and report definitions together so committees can interpret changes between activities.
The administrative side matters just as much as question design. Planstone's evaluation and credit modules collect participant feedback, support program-specific credit rules, maintain credit records, and make certificates available to attendees, allowing education teams to focus on interpreting results and improving the next program.
Assign responsibility before the activity begins: faculty validate assessment content, education staff manage evaluation delivery, and a named reviewer owns follow-up analysis. Protect identifiable learner results and separate them from anonymous activity feedback where the workflow requires different handling.
Finally, decide what finding would trigger a change. A missed assessment concept might require revised teaching; an implementation barrier might require a team-focused follow-up activity. Audit your current form against all eight question types before the next activity cycle, then close the Kirkpatrick-level gaps your program is equipped to measure.
Planstone offers configurable conference and continuing education management software connecting participant evaluations, learning assessments, credit administration, and certificate access. Visit Planstone to discuss your current evaluation forms, program-specific credit rules, and reporting requirements, and identify where a coordinated workflow could support your next activity.
Let’s talk 