Red Door Telehealth Triage: Building Evidence-Based Protocols for High-Acuity Home Encounters

High-acuity telehealth encounters — chest pain, stroke symptoms, respiratory distress, suspected sepsis — push triage protocols to their limits. Without structured, evidence-based workflows, clinicians risk misjudging urgency, delaying care, or overloading emergency systems. This guide walks experienced telehealth teams through designing protocols that handle these high-stakes presentations in home settings. We cover prerequisite infrastructure, a step-by-step triage workflow, tooling considerations, adaptations for resource-constrained environments, and common failure modes. A practical FAQ and checklist help teams audit their current protocols against best practices.

Who Needs This and What Goes Wrong Without It

Telehealth triage for high-acuity conditions is a different beast from routine low-acuity care. Teams that treat every encounter with the same generic script often miss red flags. For instance, a patient reporting 'chest tightness' might be having anxiety — or a STEMI. Without a protocol that systematically evaluates cardiac risk factors, symptom characteristics, and home monitoring capabilities, the clinician may default to 'call 911' for everything, which erodes trust and overwhelms local EMS. Or worse, they may under-triage and advise watchful waiting when the patient needs immediate intervention.

Who specifically needs this? Telehealth programs that serve patients with chronic conditions (heart failure, COPD, diabetes) who are at higher baseline risk for acute events. Also, remote monitoring programs where patients have home devices like blood pressure cuffs, pulse oximeters, and glucometers — but no structured way to interpret trends during an acute call. And any virtual urgent care or tele-stroke service that wants to reduce unnecessary ED referrals while maintaining safety.

What goes wrong without a protocol? First, inconsistent assessment: one clinician may ask about chest pain radiation, another may skip it. Second, documentation gaps that create medicolegal exposure. Third, failure to escalate appropriately — either too slow or too fast. Fourth, patient confusion when they receive conflicting advice from different providers. We've seen programs where triage nurses had no decision support for differentiating Bell's palsy from stroke, leading to unnecessary CT scans or missed thrombolysis windows. The cost of these failures isn't just clinical — it's reputational and financial.

The good news: building an evidence-based protocol doesn't require a research grant. It requires a systematic approach to synthesizing existing guidelines (like those from the American Heart Association, National Stroke Association, and sepsis bundles) into a telehealth-specific workflow that accounts for the limitations of a camera and microphone. The rest of this guide shows you how.

Prerequisites and Context to Settle First

Before you write a single algorithm, you need to establish the foundation that makes a protocol work in practice. These prerequisites are often overlooked by teams eager to jump into flowcharts.

Clinical Governance and Scope

Define who can triage and under what supervision. In many jurisdictions, triage is a registered nursing function, but some protocols allow paramedics or advanced practice providers. Clarify whether your protocol assumes independent decision-making or requires real-time physician backup. For high-acuity encounters, we recommend a two-tier model: a triage nurse performs initial assessment using a structured tool, then escalates to a supervising physician for disposition decisions. Document this hierarchy explicitly.

Technology and Data Integration

Your protocol is only as good as the data feeding it. Ensure your telehealth platform can capture and display vital signs from home devices in real time. If patients use Bluetooth-enabled pulse oximeters, the system should flag SpO2 below 92% automatically. Similarly, integrate historical trends — a patient with a baseline SpO2 of 88% due to COPD may not be in acute distress, while a previously healthy patient at 92% could be deteriorating. Without this context, your protocol will generate false positives.

Training and Competency

Every clinician using the protocol must complete simulation-based training, not just a PDF review. Run scenarios for chest pain, stroke, anaphylaxis, and pediatric respiratory distress. Test their ability to follow the algorithm under time pressure. We've found that even experienced nurses skip steps when they're anxious — simulations reveal those gaps. Also, train on how to handle technical failures mid-encounter (e.g., video drops during a stroke assessment).

Legal and Regulatory Check

Review your state's telehealth practice laws regarding out-of-state patients, prescribing, and emergency protocols. Some states require that the triage clinician be licensed in the patient's location. Others have specific rules about when you can recommend against calling 911. Your protocol should include a disclaimer that it's a decision-support tool, not a substitute for clinical judgment, and that patients with concerning symptoms should seek emergency care. We'll revisit this in the FAQ.

Core Workflow: Sequential Steps in Prose

Our recommended triage workflow follows a 'S-T-E-P' framework: Screen, Triage, Evaluate, Plan. Each step has specific actions and decision points.

Step 1: Screen for Immediacy

Within the first 30 seconds of the encounter, determine if the patient is stable enough for a structured assessment. Ask: 'Are you having trouble breathing? Do you have chest pain or pressure? Are you bleeding heavily? Do you feel like you might pass out?' If yes to any, initiate emergency protocol immediately — instruct the patient or bystander to call 911, stay on the line, and provide first aid instructions if trained. Document the time and reason for escalation.

Step 2: Triage Using a Validated Tool

For stable patients, use a condition-specific triage tool. For chest pain, we adapt the HEART score (History, ECG, Age, Risk factors, Troponin) — but since ECG and troponin aren't available at home, we use a modified version: symptom characteristics (crushing vs. sharp), radiation, associated symptoms (nausea, diaphoresis), age >65, and known CAD. For stroke, use the FAST-ED or CPSS (Cincinnati Prehospital Stroke Scale) via video. For respiratory distress, use the modified Borg Dyspnea Scale and accessory muscle use observation.

Step 3: Evaluate with Home Data

Collect vital signs from home devices: blood pressure, heart rate, SpO2, temperature, and if available, blood glucose. Compare against the patient's baseline if known. For example, a heart rate of 110 in a patient with baseline 70 is more concerning than in a patient with chronic sinus tachycardia. Also assess mental status: ask orientation questions, observe speech clarity, and check for facial droop via video.

Step 4: Plan Disposition

Based on the triage score and evaluation, assign one of three dispositions: Emergency (call 911 or go to ED immediately), Urgent (see primary care or urgent care within 24 hours), or Routine (manage at home with follow-up). Document the rationale and provide written aftercare instructions. For emergency dispositions, confirm that the patient or caregiver understands and agrees to seek care — if they refuse, document the refusal and escalate to a supervisor.

This workflow is iterative. If new symptoms develop during the call (e.g., patient starts slurring speech), re-screen and re-triage immediately.

Tools, Setup, and Environment Realities

Even the best protocol fails if the tools aren't right. Here's what we've found works in practice.

Decision Support Software

Embed your triage algorithm into the telehealth platform as a branching logic tool. This can be as simple as a structured form in the EHR or as advanced as a clinical decision support (CDS) system that auto-calculates scores. We recommend using a tool that forces completion of critical fields before disposition — for example, cannot mark 'Routine' without documenting SpO2 and symptom duration. This reduces omission errors.

Home Device Integration

If your program provides home monitoring kits, ensure they include devices that transmit data automatically to the EHR. Manual entry is error-prone and delays triage. For patients using their own devices, train them to show the reading on camera, and have a protocol for validating device accuracy (e.g., compare with a second device if available).

Video Quality and Lighting

High-acuity assessments rely on visual cues: skin color, accessory muscle use, facial symmetry, pupil reactivity. Ensure your platform supports HD video and that clinicians can instruct patients to adjust lighting (e.g., 'turn on a lamp and face the window'). For stroke assessments, the ability to zoom on the face is critical. Test your platform's performance on low-bandwidth connections — many rural patients have limited internet.

Documentation Templates

Create condition-specific templates that capture all elements of your triage tool. For chest pain, include fields for onset, quality, radiation, associated symptoms, risk factors, and home ECG if available. For stroke, include time of symptom onset (last known well), FAST-ED score, and blood glucose. These templates serve as both clinical records and medicolegal protection. We recommend using structured data (dropdowns, checkboxes) rather than free text for key decisions.

Backup Communication

Have a backup plan for when video fails. A telephone-only protocol should be available, though it reduces assessment capabilities. For example, without video, you can still assess speech clarity, respiratory effort by listening, and mental status by questions. But you lose visual cues for facial droop and skin color. Document the mode of communication and any limitations.

Variations for Different Constraints

Not every telehealth program operates in ideal conditions. Here are adaptations for common constraints.

Low-Resource Settings (No Home Devices)

If patients lack home monitoring equipment, rely more on symptom assessment and visual observation. For respiratory distress, ask the patient to count breaths per minute (you can time it) and observe for retractions or nasal flaring. For chest pain, use the PQRST mnemonic (Provocation, Quality, Radiation, Severity, Time) and assess for associated symptoms like diaphoresis. Without vitals, your threshold for escalation should be lower — any concerning symptom pattern should trigger emergency referral.

Pediatric Encounters

Children present unique challenges. Triage tools like the Pediatric Assessment Triangle (appearance, work of breathing, circulation) can be adapted for video. Assess the child's interactivity, respiratory rate, and skin color. For fever, ask about activity level and fluid intake. Use age-appropriate vital sign thresholds. Remember that children can deteriorate quickly — if the parent reports lethargy or poor feeding, escalate even if vitals are normal.

Language and Hearing Barriers

Your protocol must include steps for interpreter services (video remote interpreting preferred) and for patients with hearing loss who rely on lip-reading or sign language. Ensure that triage questions are translated and culturally adapted — for example, the phrase 'chest pressure' may not translate directly in some languages. Document the interpreter's presence and the patient's understanding.

Behavioral Health Co-morbidity

Patients with anxiety or panic disorder may present with symptoms that mimic high-acuity conditions (chest pain, shortness of breath, palpitations). Your protocol should include a differential diagnosis step: after ruling out organic causes using the triage tool, consider a behavioral health assessment. But never attribute symptoms to anxiety without first excluding cardiac or respiratory causes, especially in patients with risk factors. Document your reasoning.

Pitfalls, Debugging, and What to Check When It Fails

Even well-designed protocols fail. Here are the most common failure modes and how to address them.

False Reassurance from Normal Vitals

Normal vital signs do not rule out serious conditions. A patient with a pulmonary embolism can have normal SpO2 at rest, and a patient with a STEMI can have normal blood pressure. The protocol must emphasize symptom patterns over vital signs alone. If the patient reports sudden-onset chest pain that radiates to the jaw or arm, escalate regardless of vitals. We've seen cases where clinicians relied on a normal SpO2 to discharge a patient with pneumonia who later decompensated.

Over-Reliance on Patient Self-Report

Patients may minimize or exaggerate symptoms. A stoic patient might say 'just a little discomfort' while having a heart attack. An anxious patient might describe severe pain that is non-cardiac. The protocol should include objective measures where possible: observe the patient's ability to speak in full sentences, their facial expression, and their movement. For pain, use a numeric scale but also assess functional impact (e.g., 'can you take a deep breath?').

Technical Failures Mid-Assessment

If video drops during a stroke assessment, you lose critical information. Have a contingency: ask the patient to call back immediately, or switch to phone and document the limitation. If you cannot complete the FAST-ED score, err on the side of caution and recommend evaluation. Also, train clinicians to recognize when a patient is not visible (e.g., camera off) and to insist on video for high-acuity complaints.

Protocol Drift Over Time

Teams that don't audit their protocol compliance tend to drift. Clinicians develop shortcuts or adapt the algorithm based on past experiences, which may not be evidence-based. Schedule quarterly audits: randomly select 50 triage encounters, compare the documented assessment against the protocol, and identify deviations. Provide feedback and retraining. Update the protocol annually based on new guidelines or device capabilities.

FAQ and Checklist in Prose

We've compiled the most common questions from teams implementing high-acuity triage protocols, followed by a practical checklist.

Frequently Asked Questions

Q: Should we use a commercial triage algorithm (e.g., Schmitt-Thompson) or build our own?
A: Commercial algorithms are well-validated for telephone triage but may not account for video capabilities or home device data. We recommend starting with a commercial tool as a base and customizing it for your specific patient population and technology. Document your modifications and rationale.

Q: How do we handle patients who refuse to go to the ED when our protocol says emergency?
A: Document the patient's decision clearly, including that you explained the risks. Offer to call 911 for them. If they still refuse, escalate to a supervising physician who may speak with the patient. In some cases, involve a family member. Never force transport, but ensure the refusal is informed and documented.

Q: What about medicolegal liability if we under-triage?
A: The best protection is a documented, evidence-based protocol followed consistently. Ensure your protocol includes a disclaimer that it is a decision-support tool and that clinicians should use their judgment. Regular audits and training also demonstrate due diligence. Consult your legal team for jurisdiction-specific advice.

Q: How do we train new clinicians on the protocol?
A: Use a combination of e-learning (covering the algorithm and rationale), simulation (role-play scenarios), and supervised shadowing. Require a competency assessment before independent practice. Refresh training annually or whenever the protocol changes.

Quick Audit Checklist

• Does your protocol include condition-specific triage tools (e.g., modified HEART, FAST-ED)?
• Are vital sign thresholds defined and integrated into decision points?
• Do you have a contingency for technical failures (video loss, device malfunction)?
• Is there a clear escalation pathway to a supervising physician?
• Are documentation templates structured to capture all protocol elements?
• Do you conduct regular compliance audits and feedback sessions?
• Is the protocol reviewed and updated at least annually based on new evidence?

By addressing these questions and checklist items, your team can build a triage protocol that is both safe and practical for high-acuity home encounters. Remember that no protocol replaces clinical judgment — but a good protocol makes good judgment more consistent.