Respiratory Therapists

29-1126.00
Median wage $82,280/yr139,790 employed (US)Rank #741 of 923 scored · top 80% by substitution

Assess, treat, and care for patients with breathing disorders. Assume primary responsibility for all respiratory care modalities, including the supervision of respiratory therapy technicians. Initiate and conduct therapeutic procedures; maintain patient records; and select, assemble, check, and operate equipment.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution17
Exposure17
Augmentation50

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

22 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

0%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%17

panel mean rating 1.7/5 → substitution pressure 17/100

Technical feasibility todayw 20%18

panel mean rating 1.7/5 → substitution pressure 18/100

Cost vs. human wagew 15%17

panel mean rating 1.7/5 → substitution pressure 17/100

Adoption barriersw 20%inverted — strong barriers lower the score15

panel mean rating 4.4/5 (barrier strength) → substitution pressure 15/100

Sector adoption velocityw 10%21

panel mean rating 1.9/5 → substitution pressure 21/100

Task breakdown (22 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Maintain charts that contain patients' pertinent identification and therapy information.

70

CI 50–90 · exposure 75 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Healthcare systems, especially hospitals, have rapidly adopted EHR-integrated documentation and AI-assisted charting tools over the past decade; production deployment of voice-to-text and auto-population for therapy records is widespread in large clinical settings.
Sector adoption velocityclaude-sonnet-53/5Healthcare documentation is a target of active AI investment and pilots, but adoption of automated charting for allied health specialties like respiratory therapy lags behind physician-focused tools.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-powered documentation tools significantly enhance respiratory therapist productivity by auto-generating notes from voice input, auto-populating routine fields, and flagging missing data, allowing clinicians to focus on patient care rather than manual data entry.
Augmentation potentialclaude-sonnet-54/5AI scribes and templated EHR autofill meaningfully speed up chart maintenance and reduce administrative burden, letting therapists focus on patient care while retaining oversight.
Task automatabilityclaude-haiku-4-5-202510015/5Chart maintenance—entering and organizing structured patient data, therapy records, and identification information—is highly automatable with current EHR systems and AI-powered data entry tools that can achieve >50% time savings by auto-populating fields, extracting data from speech/documents, and organizing records with minimal human oversight.
Task automatabilityclaude-sonnet-53/5Structured charting of identification and therapy data is largely templated and can be significantly automated via EHR integrations and voice/ambient documentation tools, though clinical nuance and verification still require human input.'
Adoption barriersclaude-haiku-4-5-202510013/5While HIPAA, credentialing standards, and organizational IT governance create friction, there is no legal requirement that a licensed respiratory therapist personally perform data entry; charts can be maintained by staff or automated systems under the therapist's accountability, making adoption feasible but subject to policy oversight.
Adoption barriersclaude-sonnet-53/5Charts are legal medical records requiring accuracy and accountability, and clinicians typically must review or sign off, creating moderate regulatory and liability friction.
Cost vs. human wageclaude-haiku-4-5-202510015/5The cost of EHR systems and AI documentation tools amortized per patient encounter is orders of magnitude cheaper than the loaded wage of a respiratory therapist ($60k–$80k annually), especially when handling high-volume routine data entry.
Cost vs. human wageclaude-sonnet-53/5AI-assisted documentation tools reduce time spent charting but require licensing fees, integration, and human review, so savings are moderate rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510015/5Mature EHR platforms (Epic, Cerner, etc.) with integrated documentation automation and voice-to-text systems are deployed at scale in hospitals and clinics today, reliably handling chart maintenance and therapy record management in production environments.
Technical feasibility todayclaude-sonnet-53/5EHR systems and ambient clinical documentation products (e.g., Nuance DAX, Epic tools) are deployed in hospitals today, but respiratory therapy-specific charting still involves manual entry and correction, limiting full reliability.

Conduct tests, such as electrocardiograms (EKGs), stress testing, or lung capacity tests, to evaluate patients' cardiopulmonary functions.

50

CI 7–92 · exposure 50 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Hospitals and diagnostic centers are rapidly adopting AI-assisted cardiopulmonary testing, with automated ECG interpretation, spirometry, and stress-test analysis already common in production environments across major health systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside/physical diagnostic procedures, adopts automation slowly due to regulatory, safety, and infrastructure constraints compared to purely digital sectors.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments therapist productivity by automating data collection, preliminary analysis, and report drafting, allowing therapists to focus on patient positioning, quality assurance, and clinical communication rather than manual measurement and interpretation.
Augmentation potentialclaude-sonnet-54/5AI-assisted interpretation software already helps analyze EKG waveforms and flag abnormalities, improving speed and accuracy while the therapist administers and oversees the test.
Task automatabilityclaude-haiku-4-5-202510015/5Modern AI and automated systems can reliably conduct and interpret EKGs, stress tests, and spirometry with high accuracy, meeting or exceeding human performance and saving >50% of operator time through automated data capture, analysis, and reporting.
Task automatabilityclaude-sonnet-51/5This requires hands-on patient contact, equipment operation, and physical maneuvering (electrode placement, mouthpiece fitting, monitoring patient during stress) that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510012/5While some regulatory oversight exists (FDA clearance for diagnostic devices), clinical protocols increasingly permit non-physician technologists and automated systems to perform and interpret these tests; liability is often distributed and oversight is routine rather than gatekeeping.
Adoption barriersclaude-sonnet-54/5Clinical testing on patients typically requires licensed personnel for safety, liability, and quality assurance, especially with stress testing where medical emergencies can occur.
Cost vs. human wageclaude-haiku-4-5-202510015/5Once equipment is purchased, the per-test cost of AI-driven analysis (primarily hardware amortization and minimal software licensing) is orders of magnitude cheaper than the loaded wage of a respiratory therapist to conduct and interpret each test.
Cost vs. human wageclaude-sonnet-51/5AI cannot independently execute this physical task, so there is no viable AI-only cost to compare; a human respiratory therapist remains necessary for administration.
Technical feasibility todayclaude-haiku-4-5-202510015/5Deployed products exist across all three test types: ECG interpretation algorithms are FDA-cleared and in widespread hospital use, automated spirometers are standard in pulmonary labs, and stress-test protocols are increasingly AI-assisted in cardiology centers.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously conducts EKGs or lung capacity tests on patients; devices exist but require a human operator to administer the physical test.

Relay blood analysis results to a physician.

37

CI 11–62 · exposure 38 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare has lagging AI adoption for clinical decision-making; most respiratory therapy departments continue manual verification and relay of lab results, with only basic automated alerting for critical values in some institutions, reflecting both regulatory caution and organizational inertia.
Sector adoption velocityclaude-sonnet-53/5Healthcare adoption of automated lab alerting and EHR-integrated notifications is moderate and growing, but hospitals remain cautious and often keep human relay as a redundancy in critical care contexts.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by auto-generating result summaries, flagging critical values, and suggesting the appropriate physician to contact, but the human therapist must remain in the loop to verify clinical appropriateness and take final responsibility for communication.
Augmentation potentialclaude-sonnet-54/5AI-driven EHR alerts, dashboards, and automated flagging significantly help therapists and physicians quickly identify and act on abnormal results, improving speed and accuracy in communication.
Task automatabilityclaude-haiku-4-5-202510011/5Relaying lab results is primarily a communication task requiring human judgment about which physician to contact, timing, and escalation urgency—decisions that depend on clinical context and patient safety priorities that current AI cannot reliably determine without real-time human verification.
Task automatabilityclaude-sonnet-54/5Relaying structured lab values (e.g., ABG results) to a physician is a simple data-transfer/communication task that can largely be automated via EHR integration and automated alerting systems, though it requires reliable interfacing with clinical systems.
Adoption barriersclaude-haiku-4-5-202510015/5Legal and regulatory frameworks in healthcare require a licensed professional (respiratory therapist or physician) to review and relay critical lab results; medical-legal liability for missed or delayed communication, and the Joint Commission and state board requirements, create hard barriers to autonomous substitution.
Adoption barriersclaude-sonnet-53/5While relaying data itself isn't heavily regulated, clinical protocols often require a licensed provider to confirm interpretation and communication, and hospitals have liability concerns about relying solely on automated alerts for critical values.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI could handle routine message formatting and transmission at low cost, but healthcare liability, regulatory oversight, and the need for human clinical judgment to determine urgency mean the human therapist remains essential and the all-in automation cost approaches or exceeds the human's time burden.
Cost vs. human wageclaude-sonnet-54/5Automated result routing via existing lab/EHR infrastructure costs very little marginal compute compared to therapist time spent communicating results, though initial integration costs exist.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can parse and extract lab values from records, actual deployment of autonomous result notification remains limited; most healthcare systems require a licensed respiratory therapist or clinician to review and relay critical values, with only incremental alerting automation in some EHRs.
Technical feasibility todayclaude-sonnet-53/5EHR systems and clinical decision support tools already push lab results and flag abnormal values automatically to physicians, but full replacement of the therapist's clinical framing and verbal communication is not yet standard practice in most hospitals.

Monitor cardiac patients, using electrocardiography devices, such as a holter monitor.

31

CI 30–32 · exposure 34 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of autonomous cardiac monitoring without human oversight remains limited in healthcare settings. While ECG analysis tools are used widely, they are predominantly assistive rather than replacement systems, reflecting both regulatory constraints and conservative adoption in clinical environments.
Sector adoption velocityclaude-sonnet-53/5Healthcare has moderate AI adoption for diagnostic support tools like automated ECG analysis, but hands-on patient monitoring tasks lag behind due to regulatory and safety requirements.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-based ECG analysis significantly augments respiratory therapist productivity by automating rhythm detection, flagging anomalies, and reducing manual chart review, allowing therapists to focus on clinical interpretation and patient intervention while remaining in the monitoring loop.
Augmentation potentialclaude-sonnet-54/5AI-based ECG/Holter analysis significantly speeds up interpretation and flags abnormalities, meaningfully augmenting the therapist's ability to monitor and respond to cardiac patients.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can detect arrhythmias and anomalies in ECG data with high accuracy in controlled settings, real-time cardiac monitoring requires clinical judgment about patient context, medication interactions, and clinical trajectory that current systems cannot fully replace. The task involves continuous observation and intervention decisions that exceed 50% time-savings at equal quality.
Task automatabilityclaude-sonnet-52/5AI can assist with arrhythmia detection and signal analysis, but the full task includes device application, patient monitoring, clinical correlation, and response to changes that require hands-on presence and judgment.'
Adoption barriersclaude-haiku-4-5-202510014/5Strong regulatory barriers exist: patient monitoring for cardiac events is a licensed clinical function under state respiratory therapy boards and medical directives. Liability for missed arrhythmias is high, and patient safety regulations require qualified professionals to interpret and act on cardiac monitoring data.
Adoption barriersclaude-sonnet-54/5Cardiac monitoring involves licensed clinical staff, regulatory oversight of medical devices, and liability concerns that require a credentialed human to apply and validate monitoring.
Cost vs. human wageclaude-haiku-4-5-202510012/5While ECG analysis software is inexpensive to run, integration into clinical workflows, validation, and the respiratory therapist time required for oversight and clinical decision-making makes the all-in cost comparable to or slightly cheaper than direct human monitoring, not orders of magnitude cheaper.
Cost vs. human wageclaude-sonnet-52/5AI ECG analysis software is relatively cheap per reading, but the overall task still requires a paid respiratory therapist for physical setup and oversight, keeping blended costs comparable to human labor.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI products for ECG analysis and arrhythmia detection exist in clinical settings (e.g., FDA-cleared algorithms), but they are typically decision-support tools requiring respiratory therapist review rather than fully autonomous monitoring systems. Deployment is common in hospitals but still requires human oversight.
Technical feasibility todayclaude-sonnet-53/5FDA-cleared AI-based ECG analysis and Holter interpretation software (e.g., automated arrhythmia detection) is deployed in clinical practice, but a human therapist still applies leads and oversees monitoring.

Educate patients and their families about their conditions and teach appropriate disease management techniques, such as breathing exercises or the use of medications or respiratory equipment.

29

CI 25–34 · exposure 25 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI for clinical education remains limited and cautious; most respiratory therapy services still rely on direct clinician-patient education, with AI used only as supplementary material. No measurable displacement of respiratory therapist education roles in production settings is evident.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially allied health/respiratory therapy, has been slower than information-sector fields to deploy AI agents for patient-facing education, with most current use limited to supplementary written materials.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by drafting educational materials, generating visualizations of breathing techniques, or creating personalized summaries of disease management steps for patients to review, but the therapist must still deliver, demonstrate, and validate learning in person.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully augment therapists by generating tailored educational handouts, simplifying medical jargon, translating materials, and answering routine patient questions, freeing therapist time for hands-on instruction.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can generate educational content and instructional videos on breathing techniques and medication use, respiratory therapy education requires individualized assessment, real-time adaptation to patient comprehension, and hands-on demonstration that current systems cannot reliably deliver end-to-end. Significant human supervision and modification would be needed.
Task automatabilityclaude-sonnet-52/5AI can generate educational content and answer questions, but personalized in-person teaching of breathing techniques and equipment use requires physical demonstration, observation of patient technique, and real-time correction that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare regulations and standards of care typically require licensed respiratory therapists to directly educate and assess patient competency in critical techniques like proper inhaler use or ventilator management; legal and liability frameworks generally do not permit autonomous AI substitution for this patient-facing, safety-critical education.
Adoption barriersclaude-sonnet-53/5No strict licensing mandate requires a therapist to deliver all education, but liability concerns, need for physical demonstration, and patient safety around equipment misuse create meaningful friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5An AI-generated educational video or chatbot may be cheap per unit, but oversight, customization for individual patients, and verification that learning occurred still require trained respiratory therapists, making the all-in cost comparable to or higher than direct human teaching.
Cost vs. human wageclaude-sonnet-53/5AI-generated educational materials and chatbot support are cheap to produce, but the hands-on instruction and assessment component still requires a paid therapist, keeping overall cost comparable rather than dramatically cheaper.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI chatbots and educational platforms exist for generic patient education, but deployed systems lack the ability to assess whether a patient actually understands or can perform techniques correctly, adjust for cognitive/physical limitations, or build the therapeutic rapport essential to disease management. No production system reliably handles this full scope.
Technical feasibility todayclaude-sonnet-52/5Chatbots and patient education apps exist for general health information, but no deployed product reliably teaches hands-on respiratory equipment use or corrects breathing technique in clinical practice.

Read prescription, measure arterial blood gases, and review patient information to assess patient condition.

25

CI 25–25 · exposure 25 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI assistance for documentation and record review is increasing in healthcare, but automation of the hands-on measurement and clinical assessment aspects remains limited to pilot or research settings rather than widespread production deployment.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially respiratory therapy, has slower and more cautious AI adoption due to regulatory, safety, and workflow integration constraints compared to sectors like finance or IT.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist by rapidly summarizing patient history, flagging relevant prior results, and suggesting data-driven interpretation patterns, substantially augmenting a therapist's efficiency while they retain responsibility for the measurement and final clinical judgment.
Augmentation potentialclaude-sonnet-54/5AI can significantly assist by flagging abnormal ABG values, summarizing EHR data, and suggesting differential considerations, meaningfully speeding up the therapist's assessment process while they retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with reading prescriptions and reviewing patient information from records, measuring arterial blood gases requires hands-on clinical skill and equipment operation that cannot be fully automated today. AI cannot perform the essential phlebotomy or blood-gas analyzer operation component, limiting overall task automation to perhaps 40% time savings at best.
Task automatabilityclaude-sonnet-52/5Interpreting blood gas values and correlating with patient records requires clinical judgment and hands-on assessment that current AI cannot fully replace, though data aggregation and flagging abnormal values can be automated.'
Adoption barriersclaude-haiku-4-5-202510014/5Respiratory therapists must be licensed and credentialed to perform arterial blood draws and interpret blood gases under clinical and hospital regulations; liability for misinterpretation falls on licensed personnel. Clinical judgment and legal/regulatory requirements create substantial barriers to full automation.
Adoption barriersclaude-sonnet-54/5Assessing patient condition and acting on blood gas results is a licensed clinical function with direct patient safety and liability implications, requiring a credentialed practitioner to sign off.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI costs for document processing and data extraction are modest, but the manual measurement of blood gases and clinical assessment remain human-dependent, making total cost-per-task comparable to or potentially higher than direct human performance for the complete workflow.
Cost vs. human wageclaude-sonnet-52/5AI-assisted data analysis is cheap, but the overall task still requires a licensed therapist's judgment and physical interaction, so total cost savings are limited relative to human wages.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI products can extract and summarize prescription and patient record data reliably, but no deployed system independently performs arterial blood gas measurement or clinical assessment that meets clinical reliability standards. Clinical decision-support tools exist but require human respiratory therapist validation and oversight.
Technical feasibility todayclaude-sonnet-52/5Clinical decision support tools exist that flag abnormal ABG values and summarize patient charts, but no deployed product independently performs full patient condition assessment in respiratory therapy workflows.

Enforce safety rules and ensure careful adherence to physicians' orders.

24

CI 3–45 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Hospitals are adopting clinical decision support and monitoring systems at moderate pace, with many pilots and some production deployment, but full enforcement automation remains uncommon due to regulatory and liability concerns.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside respiratory care, is a slower-adopting sector for autonomous AI decision-making due to regulatory and safety constraints, though documentation-support tools are spreading.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-driven monitoring dashboards, real-time alerts on protocol deviations, and automated documentation of compliance checks substantially enhance a therapist's ability to track multiple patients and catch safety lapses faster than manual review alone.
Augmentation potentialclaude-sonnet-53/5AI can assist by flagging order discrepancies, alerting to safety threshold breaches, or supporting checklist compliance, but the human must still verify and enforce compliance directly.
Task automatabilityclaude-haiku-4-5-202510013/5AI systems can monitor physician orders, flag deviations, and generate real-time safety alerts against encoded protocols, reducing manual review work by 50%+ on routine cases. However, complex clinical judgment calls and mediation with physicians when orders may be unsafe still require human oversight.
Task automatabilityclaude-sonnet-51/5Enforcing safety rules and ensuring adherence to physician orders requires real-time clinical judgment, physical presence, and accountability that current AI cannot replicate end-to-end at the bedside.
Adoption barriersclaude-haiku-4-5-202510014/5Hospital liability, regulatory oversight by state boards and OSHA, and the legal requirement for licensed respiratory therapists to validate and sign off on safety compliance create substantial barriers to full automation of enforcement and adherence verification.
Adoption barriersclaude-sonnet-55/5This task involves licensed clinical responsibility, direct liability for patient safety, and regulatory requirements mandating a credentialed human to enforce orders and safety protocols.
Cost vs. human wageclaude-haiku-4-5-202510013/5Implementation of AI monitoring and alerting systems requires significant infrastructure and integration costs that approach or equal the cost of an additional respiratory therapist FTE, making the economics roughly neutral or modestly favorable depending on scale.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human clinician who is legally and practically required to do it.
Technical feasibility todayclaude-haiku-4-5-202510013/5Electronic medical record systems and monitoring software exist in hospitals that check orders and alert clinicians to safety issues, but current AI systems lack authority to enforce compliance and cannot reliably detect all safety violations in real-time clinical settings without human supervision.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously enforces clinical safety compliance or oversees order adherence in respiratory therapy settings; this remains a human clinical oversight function.

Monitor patient's physiological responses to therapy, such as vital signs, arterial blood gases, or blood chemistry changes, and consult with physician if adverse reactions occur.

23

CI 20–25 · exposure 25 · augmentation 75 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of autonomous clinical decision-making remains slow and cautious; most deployed systems are advisory (assisting clinicians) rather than autonomous. Regulatory, liability, and cultural factors limit production deployment of automated monitoring-and-consult workflows.
Sector adoption velocityclaude-sonnet-52/5Healthcare overall lags in AI deployment for clinical decision-making due to regulation, liability, and interoperability challenges, though monitoring/alert tools are slowly being piloted.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-powered dashboards, alerting systems, and real-time physiological data aggregation can substantially assist respiratory therapists by surfacing trends and flagging critical values, allowing clinicians to focus on intervention and physician communication rather than manual monitoring.
Augmentation potentialclaude-sonnet-54/5AI-based monitoring systems, trend alerts, and predictive analytics can meaningfully help therapists detect adverse changes earlier and prioritize attention, while the human remains responsible for interpretation and physician consultation.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can analyze physiological data and flag abnormalities, the task requires real-time clinical judgment, integration of multiple data streams, and immediate intervention decisions that current systems cannot reliably perform end-to-end without human oversight. Monitoring alone is automatable; the consultative decision-making and response component is not.
Task automatabilityclaude-sonnet-52/5AI can help track and flag vital signs and lab trends, but the physical monitoring, patient assessment, and clinical judgment to consult a physician require a licensed human present in real time.
Adoption barriersclaude-haiku-4-5-202510014/5Respiratory therapists operate under clinical protocols and state licensure; monitoring patient responses and consulting physicians are legally delegated acts. Patient safety liability, the requirement for a licensed clinician to make consultative decisions, and regulatory oversight of clinical algorithms create substantial barriers to full automation.
Adoption barriersclaude-sonnet-55/5This is a licensed clinical task with direct patient care and legal liability; only a certified respiratory therapist or physician may interpret findings and decide on escalation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Continuous monitoring infrastructure (sensors, integration, review systems) combined with necessary physician oversight integration keeps costs comparable to or above the loaded wage of a respiratory therapist, especially considering liability and quality assurance overhead.
Cost vs. human wageclaude-sonnet-52/5Monitoring software is cheap to run, but the task bundles hands-on patient assessment and clinical decision-making that still requires a paid licensed professional, so overall cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5Clinical decision support systems exist to flag abnormalities in vital signs and blood gas values, but no deployed product reliably performs the full task of monitoring *and* deciding when/how to consult a physician without substantial human validation. Error rates in clinical recommendations remain material.
Technical feasibility todayclaude-sonnet-52/5Deployed clinical decision support and monitoring alert systems exist in ICUs, but they augment rather than replace the respiratory therapist's continuous bedside assessment and escalation decisions.

Determine requirements for treatment, such as type, method and duration of therapy, precautions to be taken, or medication and dosages, compatible with physicians' orders.

17

CI 14–20 · exposure 25 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare adoption of AI for autonomous clinical decision-making in hospital respiratory care is minimal; the sector remains heavily human-supervised and risk-averse. Regulatory scrutiny and patient safety criticality slow deployment.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside respiratory care, adopts AI slowly due to regulatory, safety, and workflow integration constraints compared to purely digital sectors.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could usefully assist by cross-referencing protocols, flagging drug interactions, or organizing patient history, helping a therapist work faster. However, the core judgment function—determining compatibility with orders and patient needs—remains fundamentally the therapist's responsibility.
Augmentation potentialclaude-sonnet-53/5AI can assist by summarizing physician orders, cross-referencing dosing guidelines, and flagging inconsistencies, improving efficiency while the therapist retains final clinical decision-making.
Task automatabilityclaude-haiku-4-5-202510012/5This task requires interpreting physician orders, assessing patient-specific clinical factors, and applying professional judgment to determine therapy parameters. While AI could assist in organizing information or suggesting protocols, the integration of individualized medical decision-making with legal accountability means no current system can perform this end-to-end with the required reliability at 50% time savings.
Task automatabilityclaude-sonnet-52/5This requires integrating physician orders, patient-specific clinical status, and real-time judgment about respiratory therapy protocols; current AI can suggest options but cannot reliably finalize treatment parameters end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Strong regulatory and liability barriers exist: respiratory therapists are licensed professionals whose judgment on treatment parameters directly affects patient safety, and scope-of-practice regulations require a qualified human to determine therapy compatibility with physician orders. Malpractice and clinical negligence liability remain human-anchored.
Adoption barriersclaude-sonnet-55/5Determining treatment type, dosage, and duration compatible with physician orders is a licensed clinical function with direct patient safety and liability implications, requiring a credentialed therapist's judgment and sign-off.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of a clinical decision-support system, including validation oversight, licensing, and integration, exceeds the marginal cost of a respiratory therapist performing the task themselves. The therapist's expertise is the value-add and cannot be substantially undercut.
Cost vs. human wageclaude-sonnet-52/5AI decision-support tools have real integration, validation, and oversight costs comparable to therapist time saved, since a licensed professional must still verify all outputs.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs clinical decision-making for respiratory therapy treatment determination in production settings. Decision-support tools exist, but they operate as adjuncts requiring respiratory therapist validation, not autonomous performers of the task.
Technical feasibility todayclaude-sonnet-52/5Clinical decision support tools exist that flag dosing ranges or suggest ventilator settings, but no deployed product autonomously determines full respiratory treatment plans in production without therapist oversight.

Perform pulmonary function and adjust equipment to obtain optimum results in therapy.

14

CI 3–25 · exposure 13 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI for clinical decision support is advancing slowly in practice, particularly in specialized respiratory care. Most deployments remain pilot or advisory; hospitals have not yet shifted to AI-driven automation of core therapeutic tasks due to regulatory, safety, and liability constraints.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside respiratory care, has been slow to adopt AI for hands-on physical tasks despite adoption of AI in diagnostics and documentation elsewhere in the sector.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist respiratory therapists by providing real-time data interpretation, trend analysis, and equipment optimization recommendations, helping them make faster and more informed adjustments. However, augmentation is limited to decision support; the therapist remains the agent executing and accountable for all clinical actions.
Augmentation potentialclaude-sonnet-53/5AI can assist with interpreting pulmonary function test data, flagging abnormal results, and suggesting equipment settings, improving decision support even though the physical task remains human-performed.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI cannot end-to-end automate patient pulmonary function assessment and real-time equipment adjustment. While AI can assist in interpreting spirometry data, the task requires live patient interaction, clinical judgment, and adaptive equipment calibration that demands human expertise and continuous real-time decision-making.
Task automatabilityclaude-sonnet-51/5This requires hands-on physical manipulation of patients and equipment, real-time clinical judgment, and adaptive adjustments based on physiological feedback that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Respiratory therapy is a licensed healthcare profession where a certified therapist must perform patient assessment, adjust medical equipment, and sign off on results. Regulatory and liability requirements create strong barriers, as patient safety and clinical responsibility cannot be fully delegated to automated systems.
Adoption barriersclaude-sonnet-55/5Respiratory therapy involves licensed clinical practice, direct patient contact, and regulatory/liability requirements mandating a credentialed human perform and adjust equipment.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI infrastructure and oversight costs for partial automation (data analysis, recommendations) do not offset respiratory therapist labor costs substantially. The remaining human work and required clinical oversight make the all-in cost comparable to or higher than current workflows.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so AI cost cannot be compared favorably; a human therapist is required regardless of cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs complete pulmonary function testing and equipment optimization independently. AI tools exist for data analysis and interpretation support, but the critical components—patient assessment, physical equipment adjustment, and safety oversight—remain primarily human-performed tasks in production settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs hands-on pulmonary function testing or equipment adjustment in clinical settings; this remains a physical, in-person clinical task.

Use a variety of testing techniques to assist doctors in cardiac or pulmonary research or to diagnose disorders.

14

CI 3–25 · exposure 13 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While hospitals and clinics use AI-assisted analysis of test results, the actual administration and technique selection remain manual; adoption of AI for the end-to-end task is minimal because regulatory and clinical practice requirements keep the human therapist in the loop.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside clinical procedures, adopts AI slowly due to regulatory, safety, and workflow integration challenges, though diagnostic-assist software is growing in narrow niches.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist respiratory therapists by providing rapid interpretation of test results, flagging abnormalities, and suggesting relevant follow-up tests, improving their diagnostic accuracy and workflow efficiency while they retain clinical decision-making.
Augmentation potentialclaude-sonnet-53/5AI can assist in analyzing test results, flagging abnormalities, and supporting research data analysis, improving efficiency in the diagnostic and research support components of the task.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in analyzing test data and flagging patterns, the task requires administering varied testing techniques to patients, requiring clinical judgment, real-time adaptation, and hands-on interaction that current systems cannot perform end-to-end with 50% time savings at equal quality.
Task automatabilityclaude-sonnet-51/5This involves hands-on administration of pulmonary/cardiac diagnostic tests (spirometry, blood gas sampling, stress tests) requiring physical patient interaction and real-time clinical judgment that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Legal and regulatory requirements mandate that licensed respiratory therapists perform diagnostic testing procedures; medical boards restrict which healthcare workers can administer certain cardiac and pulmonary tests, creating hard licensing barriers to full automation.
Adoption barriersclaude-sonnet-55/5Diagnostic testing and clinical decision support in respiratory therapy require licensed practitioners, regulatory oversight, and direct patient contact, creating strong legal and safety barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI analysis tools for test interpretation are relatively inexpensive, but they only automate part of the task; the labor cost of skilled respiratory therapists for hands-on testing and clinical judgment remains substantial and not offset by available AI.
Cost vs. human wageclaude-sonnet-51/5Physical test administration requires human hands and presence, so there is no viable AI substitute whose cost could be compared favorably.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI tools exist for analyzing respiratory and cardiac test data (spirometry interpretation, ECG analysis), but no deployed product performs the full task of selecting appropriate tests, administering them, and adapting based on patient response in production clinical settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously conducts cardiopulmonary testing procedures on patients; AI is at most used for interpreting resulting data, not performing the testing itself.

Teach, train, supervise, or use the assistance of students, respiratory therapy technicians, or assistants.

12

CI 7–16 · exposure 5 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare remains a laggard sector for autonomous AI; while e-learning and AI-assisted modules are emerging, direct substitution of human supervisors and trainers is rare in production respiratory therapy departments due to regulatory and quality-assurance requirements.
Sector adoption velocityclaude-sonnet-52/5Healthcare training and supervision functions see slow AI adoption due to clinical liability concerns and the hands-on nature of respiratory therapy education.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by generating training modules, providing on-demand reference materials, and helping with documentation and competency tracking, but the core supervisory and interpersonal functions remain human-dependent.
Augmentation potentialclaude-sonnet-53/5AI-based simulation tools, e-learning modules, and virtual patient scenarios can meaningfully support training content and knowledge assessment even though supervision itself stays human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Teaching, training, and supervising human staff requires judgment, personalized feedback, real-time adaptation, and relationship-building that current AI systems cannot perform end-to-end. While AI can generate training materials or assist with documentation, it cannot autonomously supervise or train technicians in a clinical setting.
Task automatabilityclaude-sonnet-51/5Teaching, training, and supervising staff in a clinical setting requires hands-on demonstration, real-time judgment, and interpersonal mentorship that current AI cannot replicate end-to-end., particularly for procedural respiratory care skills.
Adoption barriersclaude-haiku-4-5-202510014/5Respiratory therapist licensure and accreditation standards typically require qualified humans to directly supervise and train technicians and students; regulatory bodies and hospital credentialing often mandate human accountability for trainee competency and patient safety.
Adoption barriersclaude-sonnet-54/5Supervision of clinical trainees and technicians typically requires credentialed professionals per institutional and accreditation standards, creating strong role-based barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-generated training content and virtual assistants have some cost advantage, but the overhead of oversight, verification, and personalized mentoring still requires substantial human involvement, making the combined cost comparable to or exceeding traditional human instruction.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for supervisory/training duties, so cost comparison favors the human entirely; any AI tool would only add cost as a supplement, not a replacement.
Technical feasibility todayclaude-haiku-4-5-202510012/5Limited commercial products exist for autonomous training or supervision; AI can support via chatbots or e-learning modules, but deployed systems do not reliably handle the interpersonal, adaptive, and accountability aspects of clinical supervision at scale.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs clinical supervision or hands-on training of respiratory therapy staff; at most AI provides supplementary educational content.

Set up and operate devices, such as mechanical ventilators, therapeutic gas administration apparatus, environmental control systems, or aerosol generators, following specified parameters of treatment.

8

CI 0–16 · exposure 13 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for AI automation due to regulatory stringency, high error-cost asymmetry, and entrenched clinical workflows. No evidence shows displacement of respiratory therapist device-operation tasks in production hospital systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially bedside respiratory care, has been slow to adopt autonomous AI systems for physical device operation, though software-assisted ventilator management is growing modestly.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by monitoring parameters, alerting therapists to deviations, logging data, and recommending parameter adjustments—useful support functions that enhance therapist productivity without removing the human from decision-making or physical setup.
Augmentation potentialclaude-sonnet-53/5AI-enabled smart ventilators and decision-support algorithms can suggest parameter adjustments and monitor trends, helping therapists titrate treatment more efficiently, but the human still performs the physical operation.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with parameter monitoring and logging, the task requires real-time physical setup, adjustment of hardware, and clinical decision-making in response to patient physiological state—functions that current AI systems cannot execute end-to-end without human intervention. Setup and safe operation of life-critical equipment falls below the 50% time-saving threshold for full automation.
Task automatabilityclaude-sonnet-51/5This is a hands-on clinical task requiring physical setup and manipulation of medical devices on patients, which current AI cannot perform end-to-end; no software-only system can substitute for the physical act of connecting and adjusting ventilators or aerosol equipment.
Adoption barriersclaude-haiku-4-5-202510015/5Respiratory therapist licensure and scope of practice are legally defined; clinical judgment on ventilator settings and device operation typically requires a licensed healthcare professional. Regulatory requirements (FDA oversight of medical devices, clinical liability) and the patient-contact requirement create hard barriers to automation.
Adoption barriersclaude-sonnet-55/5Operating life-support and therapeutic gas equipment on patients requires a licensed respiratory therapist under clinical protocols, with strict liability, safety regulation, and physical presence requirements.
Cost vs. human wageclaude-haiku-4-5-202510011/5The inference cost of AI monitoring or recommendation systems does not offset the need for a trained respiratory therapist to physically set up equipment and make real-time clinical adjustments. The loaded wage of a respiratory therapist remains lower than the combined cost of AI infrastructure, maintenance, and required human oversight.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so cost comparison favors the human therapist entirely; robotic alternatives are not commercially deployed for this purpose.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs end-to-end setup and operation of mechanical ventilators or similar clinical devices without human oversight. Monitoring systems and decision-support tools exist, but physical setup and parameter adjustment remain dependent on human respiratory therapists in all production healthcare settings.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product autonomously sets up or operates ventilators or gas delivery apparatus on patients; smart ventilators exist but require a human therapist to physically configure and monitor them.

Demonstrate respiratory care procedures to trainees or other healthcare personnel.

7

CI 3–13 · exposure 5 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare education, particularly clinical skills training, remains heavily dependent on human instruction and hands-on mentoring; while some simulation and video aids are adopted, meaningful AI-driven autonomous demonstration remains rare in production respiratory therapy programs.
Sector adoption velocityclaude-sonnet-52/5Healthcare training environments adopt AI slowly for hands-on clinical skill instruction, though simulation-based digital tools are growing modestly.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist respiratory therapists by generating supplementary instructional videos, creating procedural guides, or supporting simulation platforms, but the core teaching relationship and real-time feedback remain fundamentally human-centered and difficult for AI to enhance substantially.
Augmentation potentialclaude-sonnet-53/5AI-based simulations, videos, and virtual training modules can supplement demonstrations and reinforce learning, though the core hands-on teaching remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Demonstrating procedures to trainees requires live modeling, physical manipulation of equipment, real-time interaction with learners, and responsiveness to questions—capabilities that current AI systems cannot perform end-to-end in a clinical training setting.
Task automatabilityclaude-sonnet-51/5This requires physical, hands-on demonstration of clinical procedures (e.g., ventilator setup, airway management) on real or simulated patients, which current AI cannot physically perform.
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory requirements and clinical credentialing mandate that licensed respiratory therapists directly supervise and validate trainee competency in respiratory care procedures; this legal and professional requirement creates hard barriers to AI substitution.
Adoption barriersclaude-sonnet-54/5Clinical training and competency verification typically require credentialed respiratory therapists, with institutional and accreditation requirements around who can certify hands-on skills.
Cost vs. human wageclaude-haiku-4-5-202510011/5Creating effective instructional content or simulations requires significant human curation and clinical oversight; the all-in cost of deploying AI-assisted training infrastructure currently exceeds the cost of a respiratory therapist conducting a live demonstration.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing physical demonstrations, so cost comparison favors the human entirely.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate instructional videos or static demonstrations, no deployed system reliably performs live, adaptive procedural teaching that responds to trainee comprehension and adjusts in real-time. Existing products fall far short of replacing in-person clinical demonstration.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product physically demonstrates respiratory care procedures; this remains a hands-on clinical training activity performed by licensed personnel.

Inspect, clean, test, and maintain respiratory therapy equipment to ensure equipment is functioning safely and efficiently, ordering repairs when necessary.

7

CI 0–14 · exposure 8 · augmentation 38 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare equipment maintenance remains highly regulated and localized to clinical staff; adoption of autonomous AI for this task is negligible, with hospitals relying on certified technicians and manufacturer service contracts rather than algorithmic solutions.
Sector adoption velocityclaude-sonnet-51/5Healthcare equipment maintenance is a physical, safety-critical domain with very low AI/robotic adoption; this is a laggard use case even within a moderately digitizing healthcare sector.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could usefully assist by automating maintenance scheduling, flagging equipment-usage anomalies from sensor logs, or providing diagnostic decision support, thereby raising therapist productivity without removing human oversight of critical safety judgments.
Augmentation potentialclaude-sonnet-52/5AI could assist with maintenance scheduling, logging, or predictive failure alerts based on sensor data, but this offers only modest support to the core physical task.
Task automatabilityclaude-haiku-4-5-202510012/5While some aspects like data logging and scheduling maintenance could be partially automated, the core tasks—physical inspection, cleaning, hands-on testing, and judgment about whether equipment is functioning safely—require human expertise and direct equipment interaction that current AI cannot perform end-to-end at scale.
Task automatabilityclaude-sonnet-51/5This is a hands-on physical task requiring manual inspection, cleaning, and testing of medical equipment; current AI cannot physically manipulate or clean devices, and only marginal software-based diagnostic support exists.
Adoption barriersclaude-haiku-4-5-202510015/5Healthcare regulatory requirements (FDA equipment handling, compliance verification, infection control protocols) and implicit licensing of respiratory therapists to validate equipment safety create hard barriers; human sign-off and accountability are legally and clinically mandated.
Adoption barriersclaude-sonnet-54/5Patient safety regulations, equipment certification standards, and biomedical engineering protocols require qualified personnel to inspect and maintain life-support equipment, creating strong institutional and regulatory barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5The specialized domain knowledge, liability exposure, and need for physical presence and hands-on work mean that human respiratory therapists remain substantially cheaper than any current AI-enabled solution that could legally and safely perform these duties.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical labor involved, so there is no viable AI cost comparison; a human technician remains the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI systems reliably perform equipment maintenance tasks requiring direct manipulation, sensory assessment (visual inspection for wear, listening for abnormal sounds), and real-time troubleshooting of respiratory devices in production healthcare settings.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical inspection, cleaning, or maintenance of respiratory equipment; this remains entirely a manual biomedical/clinical engineering task.

Explain treatment procedures to patients to gain cooperation and allay fears.

4

CI 0–7 · exposure 0 · augmentation 25 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains highly regulated and resistant to substituting human-patient communication with AI, especially in acute care settings; adoption of AI for this interpersonal function is negligible and unlikely to accelerate near-term.
Sector adoption velocityclaude-sonnet-52/5Healthcare bedside patient communication is a low-digitization, high-touch domain with slow AI adoption for direct patient interaction, despite AI scribes and documentation tools gaining traction elsewhere.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could generate drafted explanations or talking points for a therapist to review and personalize, but the core value—real-time patient reassurance and adaptive communication—remains overwhelmingly human-centered; assistance is marginal.
Augmentation potentialclaude-sonnet-52/5AI can help therapists prepare patient education materials or scripts in advance, but it offers limited real-time assistance during the actual empathetic, adaptive conversation with a frightened patient.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires empathetic human engagement, trust-building, and real-time responsiveness to patient anxiety and individual concerns—capabilities that current AI systems cannot reliably deliver. While an AI could generate generic explanations, gaining genuine patient cooperation and alleviating fears demands human presence and emotional attunement.
Task automatabilityclaude-sonnet-51/5This requires real-time, in-person emotional reassurance and physical presence with a patient undergoing a medical procedure, which current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory and professional standards require a licensed respiratory therapist to directly communicate with and counsel patients on treatment procedures; liability for misalignment between explanation and clinical care, and patient preference for human contact, create hard barriers to automation.
Adoption barriersclaude-sonnet-54/5Respiratory therapy involves licensed clinical practice with patient communication as a core competency tied to clinical judgment, liability, and often institutional protocols requiring a credentialed provider.
Cost vs. human wageclaude-haiku-4-5-202510011/5The labor cost of a respiratory therapist performing this communication task is lower than the combined cost of AI systems, clinical integration, oversight, and liability coverage required to handle patient anxiety and medical explanations safely.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot substitute for the in-person reassurance and trust-building required, the human clinician remains necessary, making AI not a viable cost alternative for this task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed healthcare product reliably performs this task autonomously in clinical settings; patient communication around treatment procedures remains a human responsibility. AI chatbots exist but are not substitutes for the clinical reassurance a licensed respiratory therapist must provide.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously explains respiratory treatments to allay patient fears in a clinical bedside setting; this remains a human clinical interaction task.

Make emergency visits to resolve equipment problems.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task occurs in hospitals and care facilities with significant regulatory oversight and human-contact requirements. Adoption of AI for emergency clinical response remains extremely limited; remote diagnostic support is pilot-stage at best.
Sector adoption velocityclaude-sonnet-51/5Home healthcare and emergency medical equipment servicing are physical, low-digitization sectors with minimal AI-driven displacement of hands-on emergency response.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially help with dispatch optimization or pre-visit troubleshooting guides, but offers limited real-time assistance once the therapist is on-site addressing physical equipment failure in an emergency context.
Augmentation potentialclaude-sonnet-52/5AI could potentially assist with remote diagnostics or triage guidance beforehand, but offers little assistance for the actual on-site emergency repair task itself.
Task automatabilityclaude-haiku-4-5-202510011/5Emergency equipment troubleshooting requires physical presence at patient bedside, hands-on diagnosis of hardware failures, and real-time clinical decision-making in high-stakes settings. Current AI cannot physically repair equipment or safely intervene in respiratory emergencies.
Task automatabilityclaude-sonnet-51/5This requires physically traveling to a patient location, diagnosing and fixing malfunctioning respiratory equipment on-site, and ensuring patient safety in real time—none of which current AI can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Respiratory therapy is a licensed profession; emergency equipment intervention on patients requires a credentialed human to assess the clinical situation and authorize repairs. Liability and patient safety regulations mandate human clinical oversight, creating legal barriers to substitution.
Adoption barriersclaude-sonnet-54/5Emergency medical equipment troubleshooting on a patient requires clinical judgment and often licensed respiratory therapist credentials, plus liability concerns around patient safety create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Emergency respiratory technician visits command high labor costs due to on-call premiums, rapid response requirements, and specialized credentials. AI cannot replace this service today, making cost comparison moot; human dispatch remains the only option.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical labor and mobility required, so human cost is the only viable option, making AI effectively infinitely more 'expensive' as a non-solution.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can autonomously respond to emergency calls, navigate to patient locations, diagnose equipment malfunctions through physical inspection, or perform repairs. This task inherently requires human physical presence and clinical judgment in the field.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical emergency equipment repair visits; this remains purely a human physical-response task.

Work as part of a team of physicians, nurses, or other healthcare professionals to manage patient care by assisting with medical procedures or related duties.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare settings lag in automation adoption due to regulatory constraints, patient safety requirements, and the necessity of credentialed human professionals. AI adoption in this context remains extremely limited.
Sector adoption velocityclaude-sonnet-52/5Healthcare delivery, especially bedside physical care, has been slow to adopt AI compared to information-sector work, with adoption concentrated in documentation and diagnostics rather than hands-on procedural teamwork.
Augmentation potentialclaude-haiku-4-5-202510012/5While AI might assist with documentation or data aggregation in the background, the core collaborative and procedural elements of this task depend fundamentally on human clinical expertise and presence, limiting augmentation potential.
Augmentation potentialclaude-sonnet-53/5AI can assist with documentation, decision-support alerts, and data synthesis that support the broader care team, indirectly aiding coordination even though it cannot perform the physical procedural assistance.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time clinical collaboration, judgment, and hands-on assistance with medical procedures in high-stakes settings. Current AI systems cannot participate meaningfully in team-based clinical decision-making or perform procedural support at the bedside.
Task automatabilityclaude-sonnet-51/5This task requires hands-on physical presence, real-time collaboration, and manual assistance during medical procedures, none of which current AI systems can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Healthcare licensing laws require licensed respiratory therapists to perform scope-of-practice duties, medical procedures require human accountability, liability for patient harm falls on credentialed professionals, and patient safety regulations mandate human clinical judgment and presence.
Adoption barriersclaude-sonnet-55/5Respiratory therapy involves licensed clinical practice, direct patient contact, and legal/regulatory requirements for credentialed professionals to perform and be accountable for procedures.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of integrating AI into clinical team workflows, with necessary oversight and verification, far exceeds the marginal contribution an AI system could make compared to trained respiratory therapists.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical, collaborative clinical task, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs collaborative team-based clinical care or assists with medical procedures. This requires human presence, judgment, and coordination that existing systems cannot achieve.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product physically assists in medical procedures or substitutes for a respiratory therapist's hands-on team role; this remains firmly outside current product capabilities.

Perform bronchopulmonary drainage and assist or instruct patients in performance of breathing exercises.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare settings, particularly respiratory care, remain highly conservative in automation and heavily depend on licensed human practitioners for direct patient care. Physical and interactive clinical tasks see minimal AI displacement.
Sector adoption velocityclaude-sonnet-52/5Healthcare has moderate AI adoption in diagnostics and documentation, but hands-on respiratory therapy procedures remain largely untouched by automation.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with educational content or video demonstrations of breathing exercises, but meaningful augmentation is limited because the core task—hands-on drainage and real-time patient instruction—requires human presence and tactile feedback.
Augmentation potentialclaude-sonnet-53/5AI can help by generating personalized breathing exercise instructions, tracking patient adherence, or providing coaching content, augmenting but not replacing the therapist's hands-on role.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires hands-on physical manipulation of patients (postural drainage, percussion) and real-time assessment of patient response and breathing patterns. Current AI systems cannot perform physical therapy on patients or autonomously instruct breathing exercises in an interactive, adaptive manner.
Task automatabilityclaude-sonnet-51/5This is a hands-on physical intervention requiring manual chest percussion, positioning of patients, and real-time tactile/visual assessment that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Respiratory therapists must be licensed professionals, and bronchopulmonary drainage is a regulated clinical procedure that requires direct human oversight and patient contact. Legal and regulatory frameworks mandate human licensure and liability for patient outcomes.
Adoption barriersclaude-sonnet-55/5This is a licensed clinical procedure requiring direct physical patient contact and a credentialed respiratory therapist, with high liability for improper technique.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI has no role in directly performing this task, so cost comparison is not applicable; a human respiratory therapist is necessary and AI offers no cost displacement.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing the physical therapy, so any comparison favors the human clinician entirely; AI cannot deliver this output at any cost.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can independently perform bronchopulmonary drainage or instruct patients in breathing exercises. This requires licensed respiratory therapist judgment, hands-on clinical skills, and direct patient interaction that current systems cannot replicate.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs bronchopulmonary drainage or physically administers breathing exercise therapy; this remains purely a human clinical function today.

Perform endotracheal intubation to maintain open airways for patients who are unable to breathe on their own.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains among the slowest sectors to adopt autonomous clinical procedures, and intubation specifically—a high-risk, life-critical task—shows virtually no production AI automation despite decades of opportunity.
Sector adoption velocityclaude-sonnet-51/5Physical, safety-critical bedside procedures in healthcare show minimal AI/robotic adoption compared to administrative or diagnostic tasks in the same sector.
Augmentation potentialclaude-haiku-4-5-202510012/5Current AI offers minimal real-time assistance during the intubation procedure itself; imaging or pre-procedure planning tools may provide marginal benefit, but AI does not meaningfully augment the therapist's capability during the manual intubation maneuver.
Augmentation potentialclaude-sonnet-52/5AI can assist with pre-procedure decision support (e.g., predicting difficult airways from imaging) but offers little direct assistance during the physical act of intubation itself.
Task automatabilityclaude-haiku-4-5-202510011/5Endotracheal intubation is a manual procedural task requiring real-time physical manipulation of airway anatomy, precise tactile feedback, and immediate clinical judgment in dynamic emergency situations. Current AI systems cannot perform the physical insertion of a tube, assess airway patency by direct visualization, or adjust in real-time based on patient response.
Task automatabilityclaude-sonnet-51/5Endotracheal intubation is a hands-on, high-stakes physical procedure requiring manual dexterity, real-time tactile feedback, and adaptive judgment under crisis conditions—no current AI system can perform this end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory bodies (FDA, state licensing boards) require a licensed healthcare provider (respiratory therapist or physician) to perform intubation; liability, patient safety standards, and legal authorization create hard barriers to substitution by any non-licensed entity or fully autonomous system.
Adoption barriersclaude-sonnet-55/5This is an invasive medical procedure legally restricted to licensed, trained personnel, with severe liability and patient-safety consequences for error, representing a hard regulatory and licensing barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of developing, validating, and deploying an autonomous or robotic system for intubation would vastly exceed the labor cost of a respiratory therapist performing the procedure, even accounting for years of amortization.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so any hypothetical AI-robotic system would require far more investment (specialized hardware, safety validation) than employing a trained therapist.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs endotracheal intubation end-to-end; this remains a human-performed clinical skill with no robotic or autonomous system in mainstream medical practice for this specific task. Research prototypes exist but are not in production use.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs endotracheal intubation autonomously; robotic intubation devices remain experimental/research-stage with limited real-world use.

Provide emergency care, such as artificial respiration, external cardiac massage, or assistance with cardiopulmonary resuscitation.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5No meaningful AI adoption exists for emergency resuscitation delivery itself. Healthcare settings remain dependent on human respiratory therapists for immediate bedside emergency response.
Sector adoption velocityclaude-sonnet-51/5Healthcare emergency response is a low-digitization, physically-embedded task with no meaningful movement toward AI/robotic replacement of hands-on resuscitation.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could provide minor assistance through protocol prompts or real-time checklist reminders during resuscitation, but the primary work—physical intervention and clinical judgment under pressure—remains with the human therapist and sees limited productivity lift from current tools.
Augmentation potentialclaude-sonnet-52/5AI can support via decision protocols, monitoring alerts, or training simulations, but offers minimal real-time assistance during the actual physical emergency procedure.
Task automatabilityclaude-haiku-4-5-202510011/5Emergency resuscitation requires real-time physical intervention (manual chest compressions, airway management, drug administration) that current AI cannot perform. No autonomous system can reliably execute the coordinated motor control and on-site presence needed for CPR or artificial respiration.
Task automatabilityclaude-sonnet-51/5This requires hands-on physical emergency intervention on a patient's body in real time, which current AI systems cannot perform at all.
Adoption barriersclaude-haiku-4-5-202510015/5Legal and regulatory frameworks require a licensed healthcare provider to perform or directly supervise emergency resuscitation. Liability, patient safety, and licensure create hard barriers preventing automation of this life-critical task.
Adoption barriersclaude-sonnet-55/5Emergency clinical care requires licensed personnel, direct physical intervention, and carries severe liability; regulation and physical necessity make this a hard barrier to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot physically deliver resuscitation, so direct cost comparison is moot. The human therapist's presence and manual skills remain irreplaceable and far cheaper than any speculative robotic system that might theoretically perform this task.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical task, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs actual emergency resuscitation. AI can assist with protocol reminders or post-event analysis, but the core task—delivering physical emergency care—remains outside the scope of any production system.
Technical feasibility todayclaude-sonnet-51/5No deployed AI or robotic product performs artificial respiration, cardiac massage, or CPR assistance in clinical practice; this remains purely a human physical task.

Transport patients to the hospital or within the hospital.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare adoption of autonomous patient transport is minimal; pilots exist but production deployment is rare, and regulatory/cultural resistance to unsupervised robotic patient handling remains very high.
Sector adoption velocityclaude-sonnet-51/5Healthcare, especially physical patient-handling tasks, is a laggard sector for AI/robotic adoption due to safety, regulatory, and physical infrastructure constraints.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist by optimizing route planning or summoning transport equipment, but the core physical task of safe, dignified patient transport remains fundamentally human-dependent and offers limited augmentation scope.
Augmentation potentialclaude-sonnet-52/5AI can help with logistics scheduling or route optimization for transport requests, but offers minimal direct assistance to the physical act of moving patients.
Task automatabilityclaude-haiku-4-5-202510011/5Transporting patients physically through hospital corridors or to/from external locations requires mobile manipulation in uncontrolled environments with real people, unpredictable obstacles, and safety-critical handling. Current AI systems lack the embodied robotics capability and real-time adaptation needed to perform this task end-to-end.
Task automatabilityclaude-sonnet-51/5Physical transport of patients requires manipulation of medical equipment, mobility assistance, and physical presence; no current AI/robotic system can perform this end-to-end in a hospital setting.
Adoption barriersclaude-haiku-4-5-202510015/5Patient safety, liability, and regulatory requirements create substantial barriers: manual handling is often mandated by occupational health standards, patient dignity and comfort require human judgment, and hospitals face legal accountability for autonomous patient movement. Human oversight is effectively required.
Adoption barriersclaude-sonnet-55/5Patient safety, liability, and often clinical judgment during transport (monitoring vitals, managing lines/equipment) require a trained, licensed human professional present at all times.
Cost vs. human wageclaude-haiku-4-5-202510011/5Mobile robots capable of safe patient handling are highly specialized and expensive to deploy, integrate, and maintain. The operational and capital costs far exceed the loaded wage of respiratory therapists performing this task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI-driven substitute for this physical task, so any AI-based approach (e.g., robotic transport) would be far more expensive than a human aide given current technology costs and integration needs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed commercial product reliably performs full patient transport autonomously in hospital settings. While some facilities pilot autonomous delivery robots, these handle objects only and avoid patient contact, leaving the core task unmet.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously transports patients with medical equipment in production hospital environments; this remains outside current AI product capability entirely.

Related occupations — Healthcare Practitioners & Technical

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.