NURSE: Empathic responses to emotion
Back, Arnold, Baile, Tulsky, Fryer-Edwards, and VitalTalk
Naming, understanding, respecting, supporting, and exploring. The rubric scores how a clinician responds when a patient or family member shows emotion.
Ready to use scoring catalog
Each item below has a published rubric attached to at least one current catalog case. Choose the conversation first, then review its scoring source and limits before a pilot begins.
Serious news, emotion, and goals of care
SPIKES: Six step protocol for delivering bad news
Baile and colleagues, The Oncologist, 2000
Six steps cover setting, perception, invitation, knowledge, emotion, and strategy and summary.
REMAP: Goals of care conversation framework
Childers, Back, Tulsky, and Arnold, 2017
Reframe the situation, expect emotion, map the future, align with values, and plan treatment around those values.
Serious Illness Conversation Guide
Ariadne Labs
Five steps cover setup, understanding and information preferences, prognosis, goals and tradeoffs, and a recommendation with follow up.
Error disclosure
AHRQ CANDOR: Disclosure communication
Agency for Healthcare Research and Quality
The disclosure checklist covers preparation, listening, an explanation without blame or speculation, apology, next steps, and follow up.
Information, decisions, and consent
Informed consent: Consent discussion
AMA Code of Medical Ethics and StatPearls
Six elements cover capacity, the proposed intervention, risks and benefits, alternatives, understanding, and a voluntary decision.
Ask, tell, ask and teach back
VitalTalk and the Agency for Healthcare Research and Quality
Ask what the patient understands, explain in small pieces, ask what they heard, and use teach back to check the explanation.
Motivational Interviewing: MITI global ratings
Miller and Rollnick; MITI 4.2.1 by Moyers, Manuel, and Ernst
The rubric uses the MITI global ratings for change talk, sustain talk, partnership, and empathy, with a fifth rating for OARS skills.
Clinical interview and assessment
Calgary-Cambridge Guide to the Medical Interview
Kurtz, Silverman, and Draper
Five stages organize the visit, with providing structure and building the relationship scored throughout the encounter.
KEECC-A: Kalamazoo communication checklist
Makoul, 2001; adapted by Joyce and colleagues, 2010
Seven tasks cover rapport, the opening, information gathering, the patient's perspective, information sharing, agreement, and closure.
SEGUE Framework: Communication skills
Makoul, 2001
Five task sets follow the encounter from setting the stage through eliciting and giving information, understanding the patient, and closing.
HEEADSSS: Adolescent psychosocial screening
Goldenring and Cohen; expanded by Goldenring and Rosen
Eight domains structure a private adolescent interview: home, education, eating, activities, drugs, sexuality, suicide and depression, and safety.
Team communication
SBAR: Structured communication
Agency for Healthcare Research and Quality, TeamSTEPPS
Situation, background, assessment, and recommendation structure a concise clinical escalation or handoff.
I-PASS: Handoff communication
I-PASS Study Group and Agency for Healthcare Research and Quality
Illness severity, patient summary, action list, situation awareness and contingency planning, and synthesis by the receiver structure a handoff.
Interprofessional Collaborator Assessment Rubric
Curran and colleagues
Six categories cover communication, collaboration, roles, patient and family centered work, team function, and conflict management.
TeamSTEPPS: Teamwork and communication competencies
Agency for Healthcare Research and Quality and the US Department of Defense
The rubric scores communication, team leadership, situation monitoring, and mutual support when the case gives one speaker a chance to show them.
Feedback and coaching
R2C2: Feedback and coaching conversation
Sargeant and colleagues
Relationship, reactions, content, and coaching structure a feedback conversation between a supervisor and learner.
Institution supplied standards
Bring an approved policy, service model, script, or rubric. We use its element names and definitions, then identify which elements a spoken encounter can show.
1. Use the approved definitions
You send the document your institution uses. We keep its element names and definitions, and we ask when an element is unclear.
2. One score per element
A clinician speaks with an AI patient, and the report scores the encounter against each element rather than a general impression.
3. The clinician's own words under every score
Each element score carries the transcript excerpt behind it, so a reviewer can inspect the evidence and disagree with the rating.
4. Out of scope, not marked down
An element is scored only when the case gives the clinician a chance to show it. The report states which elements were excluded.
The scope rule
An element is scored only where the conversation gave the clinician a chance to show it. Elements the case never raised are marked out of scope and excluded from the total, and the report says how many.
Without this rule, a low score can reflect a case that never asked for the behavior. Faculty need to see which elements the case tested before they interpret the total.
What one element looks like
From a published example on this site: a pediatric critical care fellow obtaining consent from a mother for a central line in the PICU, scored against a six-element consent framework.
Assess decision-making capacity
2 out of 5The line that earned the score
"Absolutely. I'm really glad you asked."
Why
The fellow answers the question and moves straight into explaining the procedure. In a surrogate consent conversation the element asks for something earlier than that: confirming the mother's role as the decision maker, finding out what she already understands, and asking how much detail she wants. The fellow adapted well to her questions, which is why this is not a failing score, but the assessment happened reactively rather than up front.
Same report, same conversation: risks and benefits scored 5 out of 5, alternatives 3 out of 5, and no elements were excluded, because the case gave the fellow a chance at all six.
Frequently asked questions
What do you need from us to score against our own framework?
The element definitions as you have written them, in whatever document you already use: a consent policy, an escalation policy, a preceptor rubric, a disclosure standard, a badge card. We do not rewrite them, shorten them, or map them onto a rubric of ours. If an element is ambiguous we ask you what you meant rather than deciding for you.
What is the scope rule?
An element is scored only where the case gave the clinician a chance to show it. If a scenario never raises a decision about alternatives, the alternatives element is marked out of scope and left out of the total rather than scored as a failure. Every report states how many elements were excluded and why.
How is this different from an AI simulation platform that scores against its own rubric?
ClinicalSim can score the element definitions your program already uses instead of limiting every case to a vendor written rubric. The report keeps the original element name beside the score and transcript evidence, so the program can compare the result with its own standard.
Can a framework be scored if it was written for teams rather than individuals?
Only for the elements a single speaker can demonstrate. TeamSTEPPS names speech acts an individual either performed or did not, such as a check-back or a two-challenge escalation, and those score cleanly. A team-level item about shared mental models does not, and we say so rather than inventing a proxy for it.
Are the scores validated against expert human raters?
Not yet. Consistency between model runs is not the same thing as agreement with faculty raters, and we say that before anyone asks. Measuring agreement on a customer's own rubric is the work a first pilot should do, and it is the gate before any high-stakes use.
What if our framework is not on this list?
Send it with the element definitions you use. This list names ready to use rubrics already attached to current catalog cases. It does not limit what an institution can add for a pilot.
What this page does not claim
ClinicalSim is not endorsed by, affiliated with, or approved by the ACGME, AHRQ, VitalTalk, Ariadne Labs, or the framework authors named above. Scores map to published frameworks; they do not establish compliance with any standard, which a surveyor determines against the institution's own policy.
ClinicalSim output is formative. It is evidence that informs program judgment, and it is not intended for employment, credentialing, privileging, licensure, or other high-stakes decisions.
Consistency between model runs is not the same thing as agreement with expert human raters. Faculty-rater validation on a customer's own rubric is the gate before any high-stakes use, and that work has not been done.
A published rubric in ClinicalSim does not mean the framework's owner reviewed or approved ClinicalSim, and published human rater reliability does not validate an AI score.