Anesthesiologist Assistants
29-1071.01Median wage $135,880/yr162,150 employed (US)Rank #856 of 923 scored · top 93% by substitution
Assist anesthesiologists in the administration of anesthesia for surgical and non-surgical procedures. Monitor patient status and provide patient care during surgical treatment.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution12
Exposure13
Augmentation45
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
16 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%12
panel mean rating 1.5/5 → substitution pressure 12/100
Technical feasibility todayw 20%16
panel mean rating 1.6/5 → substitution pressure 16/100
Cost vs. human wagew 15%11
panel mean rating 1.4/5 → substitution pressure 11/100
Adoption barriersw 20%inverted — strong barriers lower the score7
panel mean rating 4.7/5 (barrier strength) → substitution pressure 7/100
Sector adoption velocityw 10%14
panel mean rating 1.6/5 → substitution pressure 14/100
Task breakdown (16 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.
Participate in seminars, workshops, or other professional activities to keep abreast of developments in anesthesiology.
55CI 25–85 · exposure 62 · augmentation 88 · importance 3.8/5 · click for rater detail
Participate in seminars, workshops, or other professional activities to keep abreast of developments in anesthesiology.
55| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Medical institutions are beginning to adopt AI-assisted continuing education tools, but participation in live professional activities remains standard practice; adoption is accelerating in pockets (major medical centers) but remains middle-of-the-road across the sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare professional education is a highly regulated, slow-moving sector with limited AI-driven transformation of CME/CE requirements so far. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI powerfully augments anesthesiologists' ability to stay current by rapidly synthesizing vast literature, highlighting clinically relevant updates, and personalizing recommendations based on subspecialty or institutional protocols; this maintains the human in the loop while dramatically expanding information intake. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can effectively curate literature, summarize new findings, and personalize learning recommendations, meaningfully augmenting how professionals stay current. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Current AI systems can fully automate the task of staying abreast of anesthesiology developments by aggregating and summarizing literature, clinical guidelines, conference proceedings, and educational content with >50% time savings; an AI agent can monitor journals, extract relevant findings, and deliver curated summaries more efficiently than manual participation in seminars and workshops. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help curate and summarize research or CME content, but the task itself involves active participation, discussion, and professional engagement that isn't a document-processing job to be fully automated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Institutional culture and professional norms strongly favor live participation in seminars and continuing education for credentialing and networking; regulatory bodies (CME/CME credits) may require documented human attendance, creating adoption friction despite technical feasibility. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Continuing education and licensure maintenance typically require documented human participation in accredited activities, a regulatory/credentialing barrier that AI cannot fulfill on behalf of the professional. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | The cost of AI-driven continuous monitoring, summarization, and knowledge delivery is orders of magnitude cheaper than the loaded human cost of attending seminars, workshops, and travel; a single AI system can serve many clinicians for minimal marginal cost. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply summarize content, but the professional requirement is attendance/participation which still requires the human's time and cost, limiting cost savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Deployed products like large language models, automated literature review tools, and content aggregation systems reliably perform continuous learning and knowledge synthesis across medical domains today; academic medical centers already use AI-driven evidence surveillance platforms at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products like literature summarizers and AI-curated CME tools exist, but no deployed system substitutes for actual participation in seminars or workshops. |
Collect and document patients' pre-anesthetic health histories.
34CI 25–43 · exposure 38 · augmentation 63 · importance 4.6/5 · click for rater detail
Collect and document patients' pre-anesthetic health histories.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI-assisted pre-anesthetic questionnaires and template population is emerging but slow. Most anesthesia departments still rely on traditional paper or EHR-based forms filled by humans. Pilot programs exist, but production-scale replacement of human history-taking is not yet common in healthcare organizations. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially perioperative/anesthesia care, is a slower-adopting sector for AI due to regulatory, safety, and workflow integration constraints, despite growing interest in ambient documentation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully augment the process by pre-populating templates from prior records, flagging potential drug interactions, or standardizing questionnaire formats, thereby reducing clerical work and improving consistency. However, the core task—eliciting and interpreting patient responses—remains primarily human-driven. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribes and structured intake tools can meaningfully speed up history collection and documentation, letting the assistant focus on clinical judgment and patient interaction. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract and structure some data from existing medical records or help format a questionnaire, collecting a genuine pre-anesthetic history requires direct patient interaction, clarification of symptom details, and clinical judgment about what follow-up questions to ask. The task is not automatable end-to-end at the 50% time-saving threshold without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can transcribe, extract, and structure history data from intake forms and EHRs, but confirming accuracy, probing for red flags, and integrating with physical exam requires human judgment, limiting full automation to roughly half the workflow. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pre-anesthetic history is legally and clinically critical to patient safety and informed consent. The anesthesia provider has direct liability for the accuracy and completeness of the history, and in most jurisdictions a licensed anesthesiologist or AA must personally review and sign off. This creates a hard barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Pre-anesthetic assessment is a clinical safety-critical task typically requiring a licensed provider to verify and sign off, given liability and patient safety concerns around anesthesia. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted intake (chatbot screening, auto-population of templates) has modest cost, but the full value delivered by current systems does not justify replacement of the human collector without substantial rework and verification. The all-in cost of AI infrastructure plus human verification remains comparable to or higher than direct human collection. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted intake/documentation tools reduce clinician time meaningfully but still require licensed oversight and verification, keeping costs roughly comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably collects and documents complete pre-anesthetic histories autonomously. Some EHR systems and chatbots can assist with standardized questionnaires or pre-fill templates from existing records, but clinically adequate history-taking—especially for complex or unusual presentations—remains dependent on human clinicians in production settings. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient documentation and clinical NLP tools (e.g., DAX Copilot, EHR-integrated history intake) are deployed in some clinics, but comprehensive pre-anesthetic history collection with clinical nuance is not yet standard practice everywhere. |
Assist anesthesiologists in monitoring of patients, including electrocardiogram (EKG), direct arterial pressure, central venous pressure, arterial blood gas, hematocrit, or routine measurement of temperature, respiration, blood pressure or heart rate.
29CI 25–32 · exposure 34 · augmentation 75 · importance 4.8/5 · click for rater detail
Assist anesthesiologists in monitoring of patients, including electrocardiogram (EKG), direct arterial pressure, central venous pressure, arterial blood gas, hematocrit, or routine measurement of temperature, respiration, blood pressure or heart rate.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI-assisted monitoring dashboards and alerting systems is slow in clinical practice despite decades of availability; regulatory conservatism, risk aversion, and the critical nature of errors in anesthesia limit rapid substitution of human monitoring roles. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare has been steadily adopting smart monitoring and clinical decision-support tools, but adoption in the OR for anesthesia-critical functions remains cautious and augmentative rather than substitutive. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly augments human assistants by automating data collection, generating alerts for out-of-range values, trending parameters, and detecting arrhythmias, reducing cognitive load and allowing the assistant to focus on intervention and communication with the anesthesiologist. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled monitors and predictive alerting systems significantly enhance the assistant's ability to track multiple physiological parameters simultaneously and catch early warning signs, improving safety and efficiency while the human remains in control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can process and interpret physiological signals (EKG, blood pressure, temperature), the task requires real-time integrated monitoring, clinical judgment on parameter changes, and immediate intervention readiness—currently still requiring human oversight. Partial automation of signal collection and preliminary flagging is feasible, but end-to-end autonomous monitoring without human involvement does not meet the 50% time-saving bar at equal safety. |
| Task automatability | claude-sonnet-5 | 2/5 | AI/software can process and flag abnormalities in monitored physiological data, but the task requires hands-on patient monitoring, physical intervention capability, and real-time clinical judgment during anesthesia that current systems cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory and legal barriers exist: the anesthesiologist or assistant must legally be present and responsible for patient monitoring during procedures, and liability for adverse events rests on credentialed personnel. Automation cannot substitute for the legal and professional requirement for a qualified human to monitor and respond. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Intraoperative anesthesia monitoring is subject to strict licensure, liability, and patient-safety regulations requiring a qualified anesthesia provider physically present and legally accountable. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Monitoring equipment and AI-assisted analysis systems exist but integrate into existing infrastructure; the loaded cost of a trained anesthesiologist assistant still undercuts the total system cost when factoring in hardware, software licenses, integration, and mandatory human oversight. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Monitoring hardware/software is already embedded in OR costs, but replacing the human oversight role would still require a licensed clinician present, so there is no meaningful cost reduction from AI substitution alone. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed monitoring systems exist that automatically capture and display vital signs, and some AI tools can assist in arrhythmia detection or trend analysis, but production-grade systems still have material error rates in critical contexts and require anesthesiologist review. No fully autonomous monitoring product reliably replaces trained personnel in surgical settings today. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Clinical monitoring systems already exist in production (smart alarms, EKG analysis, arterial line monitors with automated alerts) and are widely deployed in ORs, but they augment rather than replace the human performing continuous assessment and intervention. |
Verify availability of operating room supplies, medications, and gases.
28CI 25–30 · exposure 30 · augmentation 50 · importance 4.5/5 · click for rater detail
Verify availability of operating room supplies, medications, and gases.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While hospitals adopt general inventory management tech, automated verification of OR supplies remains a laggard use case; most systems still require manual spot-checks by anesthesia staff to meet liability and compliance standards. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially perioperative safety checks, is a slow-adopting sector for full automation due to regulatory and safety constraints, though digital inventory systems are gradually being adopted. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Automated inventory dashboards and alerts can assist anesthesiologist assistants by highlighting low stock or expiration dates, reducing time spent on manual scanning, though the final verification step remains human-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled inventory management and alert systems can help track supply levels and flag shortages, improving efficiency of the verification process even though physical checks remain human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Checking inventory lists and logs could be partially automated via database queries or RFID scanning, but verifying actual physical availability, handling edge cases (expired stock, damage), and ensuring clinical-grade compliance still requires human judgment and physical inspection. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical verification of supplies, medications, and gas lines requires in-person inspection and manipulation of equipment, which current AI cannot perform without robotic embodiment; only inventory tracking software components could be automated.: |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory barriers exist: federal and state pharmacy laws require licensed personnel to verify controlled substances and critical medications; Joint Commission and hospital credentialing policies typically mandate human accountability for operating room readiness before surgery. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Patient safety regulations and anesthesia practice standards require a qualified clinician to personally verify equipment and drug availability before administering anesthesia, creating strong liability and licensing barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current RFID and automated inventory systems require significant upfront infrastructure investment and maintenance; the labor cost of a single anesthesiologist assistant verification is relatively low, making the ROI marginal for this specific task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Software-based inventory tracking is cheap, but the physical checking and safety-critical verification still requires trained personnel, so overall cost savings versus the human are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Hospital inventory management systems exist and can flag stock levels, but integrating reliable real-time visibility across distributed operating rooms with medication-specific verification (potency, integrity) remains inconsistently deployed in practice. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Inventory management systems exist and are deployed in hospitals, but they support rather than replace the physical verification step done by the clinician before a case. |
Monitor and document patients' progress during post-anesthesia period.
14CI 7–20 · exposure 17 · augmentation 63 · importance 4.3/5 · click for rater detail
Monitor and document patients' progress during post-anesthesia period.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of AI in clinical monitoring is slow and cautious; most deployed systems are assistive rather than autonomous, regulatory oversight is strict, and professional standards strongly prefer human accountability in post-anesthesia care despite digitization of some documentation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially perioperative care, is a slow-adopting sector for autonomous AI due to safety regulation, though ambient documentation tools are spreading gradually. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by auto-trending vital signs, flagging abnormal patterns, and auto-populating documentation templates, moderately raising the efficiency of the anesthesiologist assistant's monitoring workflow while the human retains full clinical authority and judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted vital sign trend analysis, early warning alerts, and automated documentation meaningfully augment the assistant's ability to track and record patient status during recovery. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Monitoring and documenting post-anesthesia patient progress requires real-time clinical judgment, vital sign interpretation, adverse event detection, and immediate intervention—tasks deeply embedded in human cognition and physical presence that current AI cannot perform end-to-end with 50% time savings at equal safety. |
| Task automatability | claude-sonnet-5 | 2/5 | Continuous physiological monitoring and clinical judgment about complications during emergence from anesthesia requires real-time hands-on assessment and intervention capacity that current AI cannot fully replicate, though documentation portions are automatable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Regulatory and liability barriers are extremely high: anesthesia oversight requires a licensed anesthesia provider present or immediately available, regulatory bodies mandate human clinical responsibility, and malpractice liability for patient harm falls on the responsible clinician, making full automation legally and professionally impossible. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Post-anesthesia monitoring is a regulated clinical safety function requiring a licensed anesthesia provider to be present and legally responsible for patient recovery, representing a hard licensing and liability barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current AI monitoring systems require significant infrastructure, integration, oversight, and clinical validation; their total cost per patient monitoring episode exceeds the marginal cost of anesthesiologist assistant labor for this critical safety function. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI documentation tools can reduce charting time cheaply, but the clinical monitoring component still requires a costly licensed provider physically present, keeping overall cost comparable to human-only staffing. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with vital sign trending and documentation flagging, no deployed product reliably performs independent post-anesthesia monitoring. Production systems lack the real-time clinical decision-making, patient interaction, and liability tolerance required for this high-acuity task. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some products (smart monitors, automated vital-sign charting, EHR-integrated documentation assistants) exist and are deployed in PACUs, but no product independently monitors and manages post-anesthesia recovery without a clinician present. |
Pretest and calibrate anesthesia delivery systems and monitors.
13CI 0–25 · exposure 13 · augmentation 38 · importance 4.5/5 · click for rater detail
Pretest and calibrate anesthesia delivery systems and monitors.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare settings, particularly operating rooms, are conservative adopters of autonomous automation for safety-critical equipment checks. Pilot projects exist, but production adoption of AI-driven calibration remains minimal due to regulatory and trust constraints. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare equipment pretesting is a highly manual, physically embedded task in a low-digitization procedural context with minimal AI agent deployment for this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by automating data logging, flagging drift in sensor readings, and prompting calibration checks, reducing manual recording burden and improving consistency. However, the human assistant remains essential for decision-making and physical execution. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Smart monitors and built-in self-check software in modern anesthesia machines can flag calibration issues, offering some assistance, but the human must still perform and verify the physical pretest process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist with data collection and some monitoring alerts, the task requires physical manipulation of equipment, hands-on testing, and real-time calibration decisions that demand human expertise and accountability. Current systems cannot independently perform the full pretesting and calibration workflow. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical task requiring manual inspection, connection, and testing of medical hardware in a physical operating room, which current AI systems cannot perform end-to-end without robotic embodiment.<br>No off-the-shelf AI system can physically pretest and calibrate anesthesia machines today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory and liability barriers exist: clinical engineering standards and FDA oversight require documented human sign-off on equipment safety verification, and errors carry high patient-safety consequences. A licensed healthcare professional must legally validate calibration. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a patient-safety-critical pre-anesthesia check governed by strict clinical protocols and often regulatory/accreditation standards requiring qualified personnel to verify equipment before use. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems for equipment monitoring carry high integration and oversight costs, while anesthesia assistant labor is relatively efficient for this safety-critical task. The cost difference does not favor automation at current pricing and reliability. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since no AI system can perform this physical calibration task, there is no viable AI cost basis to compare against human labor for this specific task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform end-to-end equipment pretesting and calibration without human oversight. Some monitoring systems offer alerts, but actual calibration and troubleshooting remain human-dependent in clinical practice. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | There are no deployed AI products that autonomously pretest and calibrate anesthesia delivery systems; this remains a manual clinical safety check performed by trained personnel. |
Assist in the application of monitoring techniques, such as pulmonary artery catheterization, electroencephalographic spectral analysis, echocardiography, or evoked potentials.
9CI 3–16 · exposure 13 · augmentation 50 · importance 3.9/5 · click for rater detail
Assist in the application of monitoring techniques, such as pulmonary artery catheterization, electroencephalographic spectral analysis, echocardiography, or evoked potentials.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare and anesthesia are highly regulated sectors with slow AI adoption in clinical workflows. While monitoring interpretation tools are emerging, actual automation of the application of monitoring techniques has seen minimal real-world deployment due to safety, liability, and regulatory constraints. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare/anesthesia is a highly regulated, physically-grounded field with slow AI adoption for hands-on procedures, though software-based monitoring analytics see more use. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted interpretation of monitoring data (automated EEG analysis, echocardiography quality assessment) can help anesthesiologists and assistants make faster decisions, but assistance is limited to analysis rather than the physical application task itself, offering moderate productivity gain. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based signal processing (e.g., EEG spectral analysis, echocardiography image analysis) can assist interpretation and flag abnormalities, aiding clinician decision-making during monitoring. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with interpretation of monitoring data (e.g., ECG or EEG analysis), the task of physically applying monitoring techniques—positioning catheters, placing electrodes, operating ultrasound probes—requires direct manual intervention and real-time decision-making that current AI cannot perform end-to-end. AI might help analyze resulting data but cannot replace the hands-on application itself. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on invasive/technical clinical procedure requiring physical placement of catheters and probes plus real-time interpretation; no current AI system can perform the physical assistance component. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Significant legal and regulatory barriers exist: anesthesia monitoring is legally required to be performed or directly supervised by a licensed provider (anesthesiologist or CRNA), patient safety liability is severe, and clinical standards mandate human judgment in real-time monitoring adjustments. A licensed healthcare provider must remain accountable. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Invasive monitoring procedures require licensed, credentialed clinical personnel under strict medical practice regulations and liability standards, making autonomous AI performance legally and clinically prohibited. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of specialized monitoring hardware, real-time AI systems, regulatory compliance, and human oversight in a clinical setting far exceeds the labor cost of a trained anesthesiology assistant for the application and adjustment of monitoring techniques. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical task, so there is no viable cost comparison; a human anesthesiologist assistant remains the only option for hands-on assistance. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Current AI systems can interpret some monitoring outputs (spectral EEG analysis, echocardiography image recognition) but no deployed product reliably performs the full task of applying these techniques in real operating-room conditions. Clinical-grade real-time monitoring integration exists, but autonomous application of catheters or electrode placement remains research-stage. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs invasive monitoring catheter placement or physical echocardiography probe manipulation; AI exists only for signal interpretation in narrow research/clinical decision support contexts. |
Control anesthesia levels during procedures.
8CI 0–16 · exposure 13 · augmentation 63 · importance 4.9/5 · click for rater detail
Control anesthesia levels during procedures.
8| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI decision-support for anesthesia monitoring occurs in some high-resource hospitals but remains slow due to clinical conservatism, integration complexity with legacy OR systems, and the high stakes of patient safety. Production deployment of autonomous features is minimal; most adoption is assistive alerts only. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare, especially high-stakes intraoperative care, is a slow-adopting sector for autonomous AI control due to safety, regulatory, and liability constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-driven monitoring systems, depth-of-anesthesia indices, and predictive dashboards meaningfully assist anesthesiologists and assistants by reducing cognitive load, flagging deviations, and improving situational awareness, allowing them to manage multiple parameters and patients more efficiently while remaining in full control. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based decision-support and monitoring tools (e.g., depth-of-anesthesia monitors, predictive alerts) can assist clinicians in tracking patient state and flagging anomalies, improving situational awareness without replacing human control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can monitor vital signs and suggest adjustments, anesthesia management requires continuous real-time decision-making, physiological interpretation, and immediate manual interventions (drug dosing, airway management) that cannot be fully automated today. No current system can reliably manage anesthesia end-to-end with 50% time savings at equal safety. |
| Task automatability | claude-sonnet-5 | 1/5 | Real-time titration of anesthesia requires continuous physical monitoring, physiological judgment, and immediate response to patient crises; no current AI system can perform this end-to-end without a human clinician physically present and in control. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong regulatory and legal barriers exist: anesthesia administration is a licensed, credentialed function requiring physician or certified assistant oversight under law. Liability for adverse outcomes (awareness, overdose, hypoxia) rests on the responsible clinician, creating a hard requirement for human sign-off and accountability. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Anesthesia administration is tightly regulated and requires licensed, credentialed personnel (anesthesiologists/CRNAs/AAs) with legal responsibility for patient safety, creating hard licensing and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Implementing AI monitoring systems requires significant capital investment, infrastructure integration, and ongoing maintenance, while the primary cost of anesthesia control remains the trained clinician's wage. AI currently reduces workload rather than replacing the clinician entirely, making full-task cost per outcome unfavorable compared to human labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task independently, so cost comparison favors the human entirely; any AI-assisted monitoring adds cost on top of required clinical staff rather than replacing them. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Anesthesia monitoring AI products exist (e.g., depth-of-anesthesia monitors, predictive alerts) but they function only as decision-support tools, not autonomous controllers. The task requires direct human oversight and manual adjustment; no deployed system performs autonomous anesthesia management in production. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously controls anesthesia dosing during live procedures; closed-loop anesthesia delivery systems remain largely experimental/research-stage and not standard clinical practice. |
Administer anesthetic, adjuvant, or accessory drugs under the direction of an anesthesiologist.
1CI 0–3 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail
Administer anesthetic, adjuvant, or accessory drugs under the direction of an anesthesiologist.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Medical anesthesia administration remains in heavily regulated, human-centric sectors with minimal AI adoption for drug administration. Hospitals are slow to adopt automation for clinical drug administration due to liability, regulatory requirements, and patient safety criticality. No measurable displacement or production-level AI adoption exists in this domain. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially perioperative/surgical settings, adopts AI slowly for direct clinical intervention tasks due to safety, liability, and regulatory constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could potentially assist with monitoring data aggregation, alerting on drug interactions, or documenting procedures, but the core task of physically administering drugs and managing the patient's anesthetic state remains entirely human-dependent. Limited augmentation potential exists within the narrow scope of decision support rather than task execution. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based monitoring systems, dosage calculators, and predictive alerts can assist anesthesiologist assistants in tracking patient vitals and drug interactions, improving safety and efficiency without replacing the human role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time clinical decision-making, physical drug administration via IV/epidural lines, and continuous patient monitoring under a physician's direction. Current AI cannot physically administer medications, cannot operate in the dynamic operating room environment, and cannot provide the dynamic supervision that an anesthesiologist requires. The task is fundamentally tied to human clinical judgment and physical action. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires hands-on, real-time physical administration of drugs, continuous patient monitoring, and split-second clinical judgment during surgery, which current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is legally restricted to licensed medical professionals (Anesthesiologist Assistants or Anesthesiologists) under state medical boards and federal regulations. Administration of controlled substances requires DEA licensure and pharmacy oversight. Legal liability, patient safety requirements, and direct supervision mandates create hard regulatory barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Administering anesthetic drugs is a tightly regulated, licensed medical act requiring certified human oversight and legal accountability, making substitution essentially barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | There is no current AI system capable of performing this task, so cost comparison is not applicable. The human anesthesiologist assistant wage far exceeds any relevant AI infrastructure cost, but only because the task cannot be automated at all. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so cost comparison favors the human by default; any AI-assisted monitoring adds cost rather than replacing labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can administer anesthetic drugs or perform the clinical supervision required for drug administration. This remains a task that requires licensed human personnel with specialized training and requires direct physical interaction with patients in a controlled medical setting. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product administers anesthesia autonomously; clinical decision-support tools exist but do not perform the physical/regulatory act itself. |
Provide clinical instruction, supervision or training to staff in areas such as anesthesia practices.
1CI 0–3 · exposure 0 · augmentation 38 · importance 4.0/5 · click for rater detail
Provide clinical instruction, supervision or training to staff in areas such as anesthesia practices.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare training and clinical supervision remain highly regulated and tradition-bound; institutions depend on credentialed humans to validate staff competency. No measurable adoption of AI-led clinical instruction exists in anesthesia training environments. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare training uses some AI-assisted simulation and e-learning tools, but adoption of AI for actual clinical supervision remains minimal and slow due to regulatory and safety concerns. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with scheduling, documentation review, or supplementary educational content delivery, but the core teaching and supervisory relationship requires human judgment, presence, and accountability. Modest augmentation of administrative and information tasks is possible; transformation of the instructional core is not. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can support training via simulations, case-based learning modules, and reference materials, aiding instructors but not replacing the supervisory relationship. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Providing clinical instruction, supervision, and training requires real-time assessment of learner competency, adaptive teaching methods, and direct observation of clinical performance—capabilities that current AI systems cannot perform reliably in clinical settings. The task involves judgment about individual learning needs and safety-critical feedback that demands human expertise and presence. |
| Task automatability | claude-sonnet-5 | 1/5 | Live clinical teaching, mentoring, and supervision require in-person judgment, adaptive feedback, and hands-on demonstration that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Clinical supervision and training of medical staff are legally required to be performed by licensed, credentialed healthcare professionals who bear liability for patient safety. Regulatory and accreditation bodies mandate human oversight, and institutional liability frameworks require a licensed individual to sign off on staff competency. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Clinical supervision and training of anesthesia staff is governed by licensure, credentialing, and institutional/legal requirements mandating qualified human oversight. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems cannot replace the credentialed anesthesiologist assistant needed to supervise and instruct clinical staff; a human supervisor remains mandatory. The cost of AI assistance would add to rather than substitute for the human instructor's salary. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot substitute for the human supervisory function, there is no viable cost comparison—human presence is required regardless of AI cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product currently performs clinical supervision or direct instruction of medical staff in anesthesia practices in production settings. This requires licensed clinical authority, real-time mentoring, and accountability that exceeds what any available AI system can deliver. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs clinical supervision or hands-on anesthesia training of staff; AI is at best used for supplementary educational content, not the supervisory role itself. |
Respond to emergency situations by providing cardiopulmonary resuscitation (CPR), basic cardiac life support (BLS), advanced cardiac life support (ACLS), or pediatric advanced life support (PALS).
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail
Respond to emergency situations by providing cardiopulmonary resuscitation (CPR), basic cardiac life support (BLS), advanced cardiac life support (ACLS), or pediatric advanced life support (PALS).
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Emergency medicine and anesthesia remain highly conservative, human-intensive fields with minimal AI agent adoption in production critical-care settings. The stakes of error and the regulatory environment have prevented meaningful displacement of resuscitation tasks by autonomous systems. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Acute emergency response in anesthesiology is a highly physical, high-stakes clinical domain with essentially no AI adoption for autonomous execution. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide decision support (protocol reminders, drug dosing calculators, post-event analysis), but meaningful augmentation is limited in the heat of resuscitation, where human expertise and physical action dominate. Current tools offer only marginal real-time assistance. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-enabled monitors and decision-support alerts can flag deteriorating vitals or suggest protocol steps, offering minor assistance, but the core hands-on response remains fully human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | While AI can provide guidance on CPR protocols, the task requires real-time physical intervention (chest compressions, airway management, drug administration) and continuous real-world judgment in chaotic, unpredictable emergency settings that current AI systems cannot perform autonomously. Current AI has no capability to physically execute life-saving interventions or reliably navigate the dynamic sensory and motor demands of actual resuscitation. |
| Task automatability | claude-sonnet-5 | 1/5 | Emergency resuscitation requires immediate physical intervention (chest compressions, airway management, drug administration) that current AI cannot physically perform or safely direct end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Hard regulatory and legal barriers exist: emergency resuscitation requires a licensed healthcare provider (anesthesiologist assistant or equivalent) to direct or perform life-saving interventions; standards of care, liability, and malpractice law vest responsibility in credentialed humans. Substitution is legally and ethically constrained. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a licensed clinical task with strict certification requirements (BLS/ACLS/PALS), direct liability, and mandatory human physical presence, making substitution legally and practically barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI deployment in emergency resuscitation would require expensive specialized hardware, integration, and continuous oversight, far exceeding the marginal cost of a trained anesthesiologist assistant performing the task. The liability and safety infrastructure costs are prohibitively high. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical labor and split-second clinical judgment required, so no meaningful cost comparison favors AI; human presence is mandatory. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs emergency resuscitation autonomously in clinical settings. Robotic systems exist for narrow, controlled contexts (CPR-only manikins), but production systems do not reliably conduct ACLS or PALS protocols in real patient emergencies with acceptable safety margins. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously executes CPR/ACLS/PALS in clinical emergencies; at most decision-support tools exist in research or narrow advisory forms. |
Provide airway management interventions including tracheal intubation, fiber optics, or ventilary support.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail
Provide airway management interventions including tracheal intubation, fiber optics, or ventilary support.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Medical procedure automation in critical care lags far behind other sectors; hospitals have shown minimal adoption of autonomous airway interventions, remaining heavily reliant on human specialists. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Physical, high-stakes clinical procedures in operating rooms show minimal AI adoption for direct execution; even robotic-assisted surgery adoption for airway tasks specifically is negligible and confined to trials. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI-enhanced visualization (e.g., enhanced ultrasound or video guidance) could marginally assist airway assessment, current systems offer limited real-time cognitive support during the actual manual intervention itself. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-assisted video laryngoscopy and decision-support tools exist to aid visualization and technique, offering some incremental help, but the core physical intervention remains almost entirely manual. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Tracheal intubation and fiber-optic airway management require real-time tactile feedback, precise hand-eye coordination, and immediate response to patient anatomy variation that current AI systems cannot perform end-to-end. These are fundamentally manual, embodied procedures demanding human physical dexterity and judgment. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical medical procedure requiring manual dexterity, real-time tactile feedback, and split-second judgment on a live patient; no AI system today can perform tracheal intubation or physically manage an airway. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Airway management is legally and clinically the exclusive domain of licensed anesthesiologists, certified registered nurse anesthetists, or anesthesiologist assistants under strict regulatory authority; no substitute provider can legally perform intubation without licensure. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Airway management is a licensed clinical procedure requiring certified anesthesia personnel, with major liability exposure and direct patient safety risk, making unauthorized automation legally and ethically prohibited. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current AI has zero capability to perform these tasks, making cost comparison moot; any comparative cost is infinite versus the human performer. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so any comparison is moot—human clinicians remain the only means of delivery, making AI effectively infinitely costlier or simply inapplicable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs live tracheal intubation, fiber-optic airway manipulation, or emergency ventilatory support autonomously. Research prototypes exist for airway visualization, but no production system reliably executes these critical interventions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical airway management; AI involvement in this space is limited to research-stage robotic intubation prototypes that have not reached clinical production. |
Administer blood, blood products, or supportive fluids.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail
Administer blood, blood products, or supportive fluids.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare remains heavily regulated and slow to adopt automation for direct patient care tasks. No sector data shows adoption of AI for blood/fluid administration; clinical judgment and human accountability remain non-negotiable. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare procedural tasks involving direct physical patient intervention show minimal AI adoption; this is a laggard area dominated by hands-on clinical practice. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with protocol verification, drug interaction checking, or vital sign monitoring displays, but the core physical task of administering fluids and the critical clinical judgment required offer limited scope for meaningful AI augmentation beyond basic decision support. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with monitoring vitals, flagging transfusion reactions, or dosage calculations, but offers limited direct assistance to the physical administration task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Administering blood and blood products requires direct physical intervention at the patient's bedside, sterile technique, and real-time assessment of patient response. Current AI systems cannot perform this physical manipulation or handle the critical safety requirements of transfusion medicine. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical clinical procedure requiring vascular access, monitoring, and real-time adjustment; no AI system can perform the physical administration of blood products.5 |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal and state regulations explicitly require a licensed healthcare provider (physician, nurse, or anesthesiologist assistant) to administer blood products and manage transfusion protocols. Liability for transfusion reactions and errors is legally tied to the responsible clinician. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Administering blood products requires licensed clinical personnel under strict medical, legal, and safety regulations, with direct liability for adverse reactions—an absolute barrier to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires a licensed clinical professional on-site; automation would require robotics and supervised learning infrastructure far more expensive than a trained anesthesiologist assistant performing the work directly. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical task, so cost comparison favors the human entirely; AI cannot replace the physical labor and clinical presence required. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs the physical act of administering blood or fluids. This task fundamentally requires a trained human to establish vascular access, monitor vital signs, and respond immediately to adverse reactions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical administration of blood or fluids; this remains purely a manual clinical task performed by trained personnel. |
Assist in the provision of advanced life support techniques including those procedures using high frequency ventilation or intra-arterial cardiovascular assistance devices.
0CI 0–0 · exposure 0 · augmentation 38 · importance 4.6/5 · click for rater detail
Assist in the provision of advanced life support techniques including those procedures using high frequency ventilation or intra-arterial cardiovascular assistance devices.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare systems continue to hire and deploy anesthesiologist assistants for these critical roles; there is no measurable displacement or substitution by AI agents in production. Regulatory constraints and the life-critical nature of the work severely limit adoption velocity. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Direct physical medical procedures in anesthesia care are among the least digitized/automated clinical tasks, with essentially no production AI deployment for this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can provide useful assistance through continuous physiological monitoring, alerting systems, and decision-support tools that flag abnormalities or suggest parameter adjustments, thereby enhancing the assistant's awareness and response time in managing complex devices. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can support monitoring, data interpretation, or alerting during these procedures, but offers minimal assistance with the actual physical execution of high frequency ventilation or intra-arterial device management. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Providing advanced life support techniques—particularly high-frequency ventilation and managing intra-arterial cardiovascular assistance devices—requires continuous real-time clinical decision-making, physiological monitoring, and manual intervention in life-critical situations. Current AI systems cannot autonomously manage these interventions or replace the bedside presence and judgment required. |
| Task automatability | claude-sonnet-5 | 1/5 | This is hands-on, high-stakes physical medical intervention requiring real-time tactile skill, judgment, and manual dexterity that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal and state regulations mandate that licensed anesthesiologist assistants—not autonomous systems—must directly provide and oversee advanced life support. Liability, patient safety mandates, and scope-of-practice laws create hard barriers to automation of these high-acuity procedures. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This requires licensed medical personnel performing invasive, high-risk procedures under strict regulatory, credentialing, and liability frameworks that legally mandate human execution and oversight. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of current AI systems capable of monitoring and decision support in critical care is not offset by labor savings when a licensed anesthesiologist assistant must remain present and responsible for all clinical decisions and manual interventions. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical procedure at all, so any comparison is moot—human clinicians remain the only viable option, making AI effectively infinitely costlier for full task substitution. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably performs the end-to-end provision of advanced life support techniques. While AI can assist with monitoring and alerting, the actual administration, troubleshooting, and adjustment of mechanical support devices remain within the exclusive domain of trained clinical personnel. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs invasive advanced life support procedures like high-frequency ventilation or intra-arterial device management; this remains far outside current AI product capability. |
Assist anesthesiologists in performing anesthetic procedures, such as epidural or spinal injections.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail
Assist anesthesiologists in performing anesthetic procedures, such as epidural or spinal injections.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare adoption of autonomous clinical procedures remains minimal and highly constrained. Anesthetic procedures involve direct patient contact and high-acuity decision-making; no systematic industry migration toward automation is evident in practice. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare procedural tasks involving direct physical intervention on patients show minimal AI adoption due to physical, regulatory, and safety constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Ultrasound guidance and imaging AI can assist with anatomical visualization before needle insertion, but current systems offer limited real-time assistance during the injection itself. The task remains largely dependent on human tactile feedback and judgment. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with pre-procedure planning, imaging guidance, or documentation, but offers little augmentation to the hands-on physical assistance itself during the procedure. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time physical manipulation (needle insertion, drug administration) and continuous patient safety monitoring in a surgical environment—capabilities no current AI system possesses. The task cannot be decomposed into a meaningful automated workflow without removing the core physical execution component. |
| Task automatability | claude-sonnet-5 | 1/5 | Physically assisting in invasive anesthetic procedures requires hands-on manipulation, sterile technique, and real-time physical dexterity that current AI systems cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is heavily protected by medical licensing requirements, malpractice liability, FDA oversight of medical devices, and the legal requirement that licensed healthcare providers perform or directly supervise anesthetic procedures. Regulatory and professional barriers to automation are maximal. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a licensed medical procedure requiring certified human providers, with strict liability, patient safety regulations, and legal requirements for credentialed personnel to perform or directly supervise it. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The hardware, software, regulatory validation, and human oversight required for any automated injection system would far exceed the salary cost of a trained anesthesiologist assistant for the foreseeable future. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so cost comparison favors the human by default since AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can autonomously perform epidural or spinal injections. While image-guided systems exist for needle placement, they require human operators and do not constitute independent task performance. This remains a research domain, not a production capability. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical assistance in epidural or spinal injections; robotic surgery assistance remains research-stage for this specific application. |
Collect samples or specimens for diagnostic testing.
0CI 0–0 · exposure 0 · augmentation 25 · importance 3.9/5 · click for rater detail
Collect samples or specimens for diagnostic testing.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare automation adoption remains slow for direct patient contact tasks. Specimen collection remains a human-performed function across all surveyed anesthesia departments and clinical settings, with no measurable displacement by automated systems. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare procedural tasks involving direct patient contact are among the slowest to adopt AI/automation due to physical and regulatory constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI might assist with sample tracking, labeling, or documentation post-collection, it provides minimal assistance during the actual specimen collection procedure itself. The core task remains fundamentally manual and procedural with limited augmentation opportunity. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI may assist with documentation, specimen tracking, or test ordering around this task, but offers minimal help with the physical collection act itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Collecting physical samples or specimens requires manual dexterity, precise positioning, and direct physical contact with patients in clinical settings. Current AI systems cannot perform the hands-on procedural work needed to extract blood, tissue, or other specimens safely and accurately. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical, hands-on clinical procedure (e.g., drawing blood or collecting specimens from a patient) requiring manual dexterity and direct patient contact that current AI systems cannot perform.atability. AI has no embodiment to execute this task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Specimen collection must be performed or directly supervised by licensed clinical staff; regulatory bodies (FDA, state medical boards) require human qualification and liability rests with a licensed provider. The legal requirement for credentialed personnel to perform or sign off on the procedure creates a hard barrier to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Sample collection from patients is a licensed clinical activity with strict safety, sterility, and liability requirements, mandating a credentialed human provider. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Manual specimen collection by trained anesthesiologist assistants remains far cheaper and faster than any robotic or AI-driven alternative currently available. The infrastructure and oversight costs for automated collection would exceed the human labor cost by a wide margin. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical task, so cost comparison favors the human by default; robotic phlebotomy devices exist only in narrow pilot contexts and are not general AI systems. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI products perform specimen collection in clinical practice today. While some robotics research exists for procedural assistance, no mature production systems reliably and independently collect diagnostic specimens at the standard of care. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product physically collects diagnostic samples from patients; this remains a manual clinical task performed by trained personnel. |
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.