Communicate effectively in healthcare

20 prompts that make you practise communicate effectively in healthcare rather than read about it — rehearse it against someone who does not fold, get told what you actually did wrong, and carry it into a situation you did not learn it in. 38 careers need this one, and it is the part of the work no software does for you. Everything here is built on 5 named sources, and on the 3 places those sources disagree.

20blueprints
38careers need it
5named sources
3real disagreements
Open it in the interactive atlas →

The blueprints

Each one is a different way in — open it up, go deeper, then carry it somewhere new.

Open it up First contact — what the skill even is, and where you already do it.

Definition stress-test

+
Define 'Teach-Back' as a clinical research coordinator so it cannot be mistaken for general patient education…
drillrehearse against a counterparty who does not let you win easily
Define 'Teach-Back' as a clinical research coordinator so it cannot be mistaken for general patient education or for Shared Decision Making. Use the Teach-Back framework by name, state exactly when in a research visit you'd use it (give minutes into a 30‑minute visit), and write three short scripts: one for informed consent key points, one for medication dosing, one for schedule adherence. Highlight one concrete linguistic tell that shows someone is substituting education for teach-back.
Grounded inTeach-BackShared Decision Making
Then sayI tried the medication dosing script but patients keep parroting the sentence verbatim — how do I rephrase so it's genuine?
If it goes shallowIf answers become generic or theoretical, force concreteness: ask the user to paste a transcript or give exact words they used. If they respond with education-only scripts, press for patient prompt language that demands recall, not affirmation.

Spectrum mapping

+
Map 'eliciting patient values in consent' across mastery levels 1–5 using the Shared Decision Making…
drillrehearse against a counterparty who does not let you win easily
Map 'eliciting patient values in consent' across mastery levels 1–5 using the Shared Decision Making framework name. For each level give 3 observable behaviors in a 15‑minute consent discussion, one metric to measure it, and one realistic upgrade task to move to the next level.
Grounded inShared Decision Making
Then sayMy team scores mostly 2s and 3s. Give a 4‑week micro‑curriculum (three 20‑minute exercises) to push them toward level 4.
If it goes shallowIf levels blur, require the assistant to produce sample transcripts or time‑based markers. If suggestions are idealistic, ask for low‑resource, high‑yield tasks.

First-principles reduction

+
Reduce 'Teach-Back' to first principles a CRC can use when patients are overwhelmed: what psychological…
drillrehearse against a counterparty who does not let you win easily
Reduce 'Teach-Back' to first principles a CRC can use when patients are overwhelmed: what psychological mechanism makes Teach-Back work (not just 'checks understanding')? Cite one cognitive or social process from the dossier or related clinical communication theory, and then give two ultra-practical rules-of-thumb (one sentence each) that reflect that mechanism and can be used at the bedside.
Grounded inTeach-Backdistinction: 'Communication is not just information transmission — it also builds trust, elicits values, and manages emotion.'
Then sayTurn one of the rules-of-thumb into a 10–12 word script the CRC can say verbatim to a fatigued patient.
If it goes shallowIf the answer is textbooky, ask: 'Which observable patient response tells me my Teach-Back succeeded? List one clear behavioral cue.'
Go deeper The real mechanics, including the parts that feel counter-intuitive.

Anti-pattern

+
Describe a Clinical Research Coordinator who believes they're excellent at 'eliciting patient perspective'…
conversationa longer back-and-forth, not a single answer
Describe a Clinical Research Coordinator who believes they're excellent at 'eliciting patient perspective' but is actually undermining enrollment and retention. Use the Four Habits Model name in your description and list five behavioral tells (concrete sentences, gestures, timing) that expose this anti‑pattern during a 20‑minute screening visit.
Grounded inFour Habits Model
Then sayGive me two short corrective micro‑behaviors the coordinator can practice next week and how to measure each one.
If it goes shallowIf the description becomes moralizing, demand observable behaviours and timing (what they say and when). If the user deflects to system causes, refocus on what the individual control is.

Scenario simulation

+
Roleplay a 12‑minute part of a study consent conversation where I must use SPIKES and Shared Decision Making…
drillrehearse against a counterparty who does not let you win easily
Roleplay a 12‑minute part of a study consent conversation where I must use SPIKES and Shared Decision Making to disclose an unexpected risk increase found in interim analysis. You play the patient: confused, anxious, and hesitant about continued participation. After each of my 3 responses, give blunt feedback (what I did well, where I failed SPIKES/SMD, and one micro-fix). Start now with my first line: 'We found new information about the risk; I want to discuss whether you should stay.'
Grounded inSPIKESShared Decision Making
Then sayI used the micro-fix but the patient got defensive. Play them again and point to the pivot where my language escalated.
If it goes shallowIf feedback is generic, require the assistant to quote the clinician's exact words it is critiquing. If the roleplay softens, instruct it to escalate patient stakes (family dependence, travel burden) to force tradeoffs.

Failure autopsy

+
Walk me through a realistic failure where poor use of Teach-Back by a CRC led to a protocol deviation and…
drillrehearse against a counterparty who does not let you win easily
Walk me through a realistic failure where poor use of Teach-Back by a CRC led to a protocol deviation and patient hospitalization. Use the Teach-Back framework name, identify the earliest two warning signs (with timestamps in a 45‑minute enrollment visit), and give a sequence of five decision points where a different communication call would have prevented escalation.
Grounded inTeach-Back
Then sayWrite two short PSAs I can share with my team about the first two warning signs (one line each).
If it goes shallowIf the autopsy stays high-level, force the assistant to produce exact wording and clock times. If it blames patient literacy alone, push it to name coordinator actions that were missing.

Context shift

+
I manage a team of CRCs across a 200-person academic center and a small community hospital. Using Shared…
drillrehearse against a counterparty who does not let you win easily
I manage a team of CRCs across a 200-person academic center and a small community hospital. Using Shared Decision Making and SPIKES explicitly, outline how I should coach CRCs differently in the academic site (high regulatory oversight, long enrollment lists) versus the community site (fewer staff, closer patient relationships). Give three concrete coaching scripts (one per context) for an initial enrollment conversation where eligibility is borderline.
Grounded inShared Decision Making (SDM)SPIKES
Then sayConvert the community-site script to an asynchronous written consent prompt for remote enrollment — what must change to preserve SDM?
If it goes shallowIf responses blur contexts, ask for a one-sentence priority list for each site (top 3 coaching focuses) and to bold the differences.

Translation exercise

+
I wrote this email to a patient participant who missed two follow-up visits and I need it rewritten to…
grounded studylearn the real rules, including where the experts disagree
I wrote this email to a patient participant who missed two follow-up visits and I need it rewritten to demonstrate strong Teach-Back and Shared Decision Making (SDM). Here’s the draft: "We noticed you missed visits on 6/3 and 6/17. These visits are important for safety labs and adherence data. Please reschedule ASAP or we may withdraw you from the study." Rewrite it so it: (a) opens with the Four Habits 'invest in the beginning', (b) uses Teach-Back to confirm understanding, (c) invites preferences per SDM, and (d) keeps the regulatory constraint (possible withdrawal) clear but nonthreatening. Keep it short enough for an email subject+3 brief paragraphs.
Grounded inTeach-BackShared Decision Making (SDM)Four Habits Model
Then sayIf the patient replies saying they missed visits due to transportation issues, rewrite the email to include two realistic solutions (rideshare voucher, later clinic hours) and the exact phrasing to document consent to a solution.
If it goes shallowIf the reply becomes generic or repeats textbook steps, ask: 'Which single sentence here would a tired 70-year-old patient actually read and act on? Cut the rest.'

Culture clash

+
I coordinate a multi-site oncology trial with patients from urban US, rural Midwest, and an immigrant…
drillrehearse against a counterparty who does not let you win easily
I coordinate a multi-site oncology trial with patients from urban US, rural Midwest, and an immigrant community whose first language is Vietnamese. Using the Four Habits model and the distinction 'patient education ≠ shared decision making', list 3 concrete ways communication breakdowns are likely to differ across these cultures when discussing side-effect reporting. For each, give one exact phrase or question (in English) a CRC should use to avoid the misunderstanding.
Grounded inFour Habits Modeldistinction: 'patient education is not the same as shared decision making'
Then sayFor the immigrant community example, rewrite the suggested phrase into two versions: one for a trained interpreter present, one for bilingual staff limited in medical vocabulary.
If it goes shallowIf responses stay generic, demand: 'Show the exact words. No abstractions like "build rapport"—give me one sentence to say.'

Counterfactual

+
Take the 2010 UK 'Mid Staffordshire' care failures (a well-known breakdown of patient safety/communication).…
grounded studylearn the real rules, including where the experts disagree
Take the 2010 UK 'Mid Staffordshire' care failures (a well-known breakdown of patient safety/communication). Argue, in two paragraphs, how consistent use of the Four Habits Model and SPIKES during escalation and family updates could plausibly have prevented key harms. Then in one paragraph name two concrete obstacles CRCs would face implementing those models in a stressed ward and how they'd realistically be overcome.
Grounded inFour Habits ModelSPIKESdistinction: 'Communication is not just information transmission'
Then sayPick one obstacle you named and draft a two-sentence script a CRC could use to get a busy nurse/physician to pause for a 3-minute family update.
If it goes shallowIf the counterfactual is heroic, prompt: 'Trim one claimed causal link—leave only what a CRC (not a chief exec) could reasonably influence.'

Feedback rehearsal

+
Last week I had a clinic visit where a patient considering joining our oncology trial seemed hesitant but…
referencea plain fact, asked directly
Last week I had a clinic visit where a patient considering joining our oncology trial seemed hesitant but never voiced a clear reason. I used Teach-Back at the end and they could repeat logistics, but later declined enrollment. Coach me: how should the Four Habits Model have shaped my responses in that encounter? Include concrete lines I could have said in the moment and a 60–90 second rehearsal script I can practice aloud.
Grounded inFour Habits ModelTeach-Back
Then sayListen to my 60–90s rehearsal (I’ll paste a transcript) and mark 3 places where my wording weakens elicitation or empathy.
If it goes shallowIf responses get theoretical, demand specific verbatim lines and ask for the immediate next sentence to follow each line.

Junior-to-senior delta

+
Explain concretely what excellent use of SPIKES looks like for a junior CRC, a senior CRC, and a CRC manager…
drillrehearse against a counterparty who does not let you win easily
Explain concretely what excellent use of SPIKES looks like for a junior CRC, a senior CRC, and a CRC manager delivering trial-risk information to a patient. For each level give one specific sentence they'd say at 'Perception' and one behavioral difference in how they prepare the consultation.
Grounded inSPIKES
Then sayGive me two small coaching prompts a manager could use to push a junior from scripted SPIKES to cue-driven SPIKES.
If it goes shallowIf answers stay generic, require a concrete prep checklist (time needed, interpreter arranged, family present, room setup) for each level.

Conflict pairing

+
Simulate a realistic disagreement between a study PI who prioritizes recruitment targets and a CRC who…
conversationa longer back-and-forth, not a single answer
Simulate a realistic disagreement between a study PI who prioritizes recruitment targets and a CRC who prioritizes patient-centered shared decision making (SDM) about trial risks. Use SDM and the SBI feedback model: one speaks first (PI), the other responds (CRC). End by listing where SDM and SBI pull them apart and a quick adjudication—who should concede which point and why.
Grounded inShared Decision Making (SDM)SBI (Situation-Behavior-Impact)
Then sayNow rewrite the CRC lines so they use SBI to give feedback to the PI without inflaming defensiveness.
If it goes shallowIf the simulation is overly polite, instruct the assistant to add specific tensions (percent recruitment shortfall, enrollment target date) and repeat the scene.
Test it elsewhere Carry it into a situation it was not learned in.

Trade-off probe

+
I coordinate oncology trials and I worry our consent conversations swing between two extremes: long,…
drillrehearse against a counterparty who does not let you win easily
I coordinate oncology trials and I worry our consent conversations swing between two extremes: long, exhaustive explanations (protocol, risks, logistics) that overwhelm patients, and very brief checklists that speed enrollment but leave confusion. Using the Four Habits Model explicitly, tell me when 'too much information' in a consent visit is harmful and what it trades off against (e.g., trust, comprehension, autonomy, enrollment rates). Roleplay a 7–10 minute inner debate I can rehearse where I weigh stopping vs continuing the explanation for a confused 68-year-old with limited health literacy and a caregiver rushing out the door.
Grounded inFour Habits ModelTeach-Back
Then sayRewrite the inner debate to favor Teach-Back after the caregiver presses to finish; show the exact two teach-back prompts I'd use and what patient cues would make me stop.
If it goes shallowIf the roleplay becomes generic, ask for timestamps or word counts per segment (e.g., first 90s: invest in the beginning), forcing concrete pacing.

Measurement challenge

+
I need to assess a candidate's skill at communicating trial risks in a 45-minute interview without directly…
grounded studylearn the real rules, including where the experts disagree
I need to assess a candidate's skill at communicating trial risks in a 45-minute interview without directly asking 'are you good at consent?'. Using Teach-Back and the SBI model, design a 5-step interview station (with timing) that reveals practical ability, including 3 indirect tasks or prompts, the behavioral anchors you’ll watch for, and how you'll score them quickly.
Grounded inTeach-BackSBI
Then sayTurn one of your tasks into a standardized prompt for a roleplayer (exact patient vignette and one-line backstory).
If it goes shallowIf the station lacks measurable anchors, demand three observable behaviors per task and a pass/fail threshold.

Self-diagnosis

+
Assess my current competence at 'eliciting patients' values during enrollment' by asking me 10 questions, one…
drillrehearse against a counterparty who does not let you win easily
Assess my current competence at 'eliciting patients' values during enrollment' by asking me 10 questions, one at a time. After each answer, gently challenge or request an example. After all ten, give a candid readout with one clear developmental priority and two specific actions I can use this week. Use the Four Habits Model and SDM to shape the questions.
Grounded inFour Habits ModelShared Decision Making (SDM)
Then sayBased on your readout, produce the exact wording for the two actions so I can practice them tomorrow on my next enrollment.
If it goes shallowIf questions become yes/no, instruct the assistant to always follow each with 'Give a concrete recent example or say "no example"'.

Micro-habit design

+
Give me one 5-minute daily micro-habit that will measurably improve my ability to 'demonstrate empathy'…
drillrehearse against a counterparty who does not let you win easily
Give me one 5-minute daily micro-habit that will measurably improve my ability to 'demonstrate empathy' during busy enrollment visits. Explain the mechanism in behavioral terms and name which part of the Four Habits Model it trains. Include an exact script (20–30 words) I can say at the start of a visit and a single 1-minute reflection prompt to use after each day.
Grounded inFour Habits Model
Then sayNow shorten that 20–30 word script to 10 words without losing its empathic function; show both versions.
If it goes shallowIf the routine is nebulous, demand timing broken into seconds and specify where in my workflow to do it.

Devil's advocate

+
Argue that Teach-Back and SPIKES are overrated for Clinical Research Coordinators running high-enrollment…
drillrehearse against a counterparty who does not let you win easily
Argue that Teach-Back and SPIKES are overrated for Clinical Research Coordinators running high-enrollment trials (first make the strongest case). Then rebut that argument, showing where those approaches actually matter most in CRC work. Finish by saying where the practical compromise should land—what CRCs should do routinely and what can be reserved for exceptions.
Grounded inTeach-BackSPIKESdistinction: 'Patient education is not the same as shared decision making'
Then sayGive three short decision rules (yes/no) CRCs can use to decide when to apply full SPIKES or a quick Teach-Back.
If it goes shallowIf the attack is weak, ask: 'Show one data point or clinic metric that would justify not using Teach-Back routinely.'

Teaching test

+
Help me design a 30-minute workshop to teach CRCs Teach-Back in a clinic. Include: a 7-minute intro framing…
conversationa longer back-and-forth, not a single answer
Help me design a 30-minute workshop to teach CRCs Teach-Back in a clinic. Include: a 7-minute intro framing (why Teach-Back vs. simple education, invoking the dossier distinction), one 12-minute paired exercise (roles, prompts, scoring rubric), a 6-minute group debrief question, and one 5-minute take-home practice assignment. Keep it runnable for 8–12 CRCs.
Grounded inTeach-Backdistinction: Patient education vs. Shared Decision Making
Then sayGive me the exact wording of the facilitator script for the 7-minute intro (150–180 words).
If it goes shallowIf the exercise is passive, demand active role constraints (e.g., patient must avoid saying 'I understand').

Retrospective lens

+
Give me five concise, behavior-focused questions I can ask myself after any patient-facing meeting to judge…
diagnosticdescribe what went wrong; get the likely causes ranked
Give me five concise, behavior-focused questions I can ask myself after any patient-facing meeting to judge whether I applied Shared Decision Making (SDM) effectively. Each question should include what to look for (a one-sentence clue) and a quick 1–2 word scoring rubric (e.g., 'Yes/No' or '0–2').
Grounded inShared Decision Making (SDM)
Then sayNow convert those five questions into a single 60-second self-check script I can say to myself immediately after leaving a room.
If it goes shallowIf questions are abstract, require a concrete observable for each (words the patient said, or actions like 'offered at least two options').

The canon behind these

Where these came from — and where the experts disagree.

On Communication in Medicine — Michael BalintSkills for Communicating with Patients — Jonathan Silverman, Suzanne Kurtz, Juliet DraperCrucial Conversations: Tools for Talking When Stakes Are High — Kerry Patterson, Joseph Grenny, Ron McMillan, Al SwitzlerThe Four Habits Model — Victor Montori / Richard M. Frankel (originators associated with habit-based consultation models)Patient-Centered Medicine — Moira Stewart, Thomas R. Roter (editors/authors in the field)

Where they disagree

Degree of clinician-directed versus patient-directed consultation

Where they disagree

Use of scripted protocols versus conversational flexibility

Where they disagree

Prioritizing efficiency (short, focused encounters) versus relational depth

What people get wrong

The confident version of the mistake.

Assuming that more information equals better communication; in healthcare, tailored, prioritized information and checking comprehension matter more.Believing technical accuracy alone suffices; communication quality depends equally on relational and emotional skills.Thinking teach-back is patronizing — when used respectfully it reduces misunderstandings and improves adherence.Assuming conflict-avoidance is best; some high-stakes situations require explicit, structured conversations to resolve tensions safely.
Soft-skill blueprints in the LLOS Work Atlas are built from the real books and named methods working professionals use — and deliberately from the places those experts contradict each other. Depth is not authority: use these to prepare for a hard conversation, never to replace the person you need to have it with.
Copyright © LLOS.ai · 2026 — original pedagogy, voice, and design — all rights reserved.

The rest of the map

Same library, five ways in.