Physician assistants

Research the latest treatments to ensure quality of patient care

AI excels at literature search, synthesizing recent studies, and identifying relevant treatment updates from vast medical databases. Systems can already monitor journals, extract key findings, and summarize evidence-based recommendations with minimal human involvement beyond selecting which summaries to review.

BLS evidence: The final duty listed is 'Research the latest treatments to ensure quality of patient care.'

Assess and record patients' progress during treatment

Recording patient progress involves documenting objective measurements, symptoms, and treatment responses—largely structured data entry and pattern recognition. AI can auto-generate progress notes from vitals, labs, and brief inputs, with human review for accuracy, substantially reducing labor content.

BLS evidence: The duties list includes 'Assess and record a patient's progress' as a routine documentation task.

Order and interpret diagnostic tests such as x-rays or blood tests

AI systems already interpret x-rays, CT scans, and lab values at or above human performance in many contexts. Ordering tests based on clinical presentation can be protocol-driven and AI-assisted. The main human role is accountability and handling edge cases, making this highly automatable with light oversight.

BLS evidence: The duties list states 'Order and interpret diagnostic tests, such as x rays or blood tests.'

Obtain and review patients' medical histories

AI can extract, summarize, and flag relevant information from electronic health records efficiently. Reviewing histories is largely pattern recognition and information synthesis, tasks where AI excels. Human review is needed mainly for clarifying ambiguities and patient communication, not the core review work.

BLS evidence: The first duty listed is 'Obtain and review patients' medical histories.'

Prescribe medication to patients as appropriate for their condition

AI can recommend evidence-based prescriptions based on diagnosis, patient factors, and drug interactions with high accuracy. However, prescribing carries legal liability, requires nuanced patient-specific judgment about adherence and preferences, and involves regulatory constraints that keep humans in the loop for final decisions.

BLS evidence: The duties section explicitly includes 'Prescribe medication' as a physician assistant responsibility.

Diagnose patients' injuries or illnesses based on examination and test results

AI diagnostic systems can analyze test results and suggest differential diagnoses with high accuracy, but clinical diagnosis requires integrating physical exam findings, patient history, and contextual factors with high-stakes accountability that currently requires physician oversight. AI assists substantially but doesn't replace the clinician.

BLS evidence: The duties explicitly list 'Diagnose a patient's injury or illness' as a key responsibility.

Educate and counsel patients and their families on treatment and self-care

Patient education requires adapting explanations to individual health literacy, cultural context, and emotional state, plus building trust through empathetic communication. AI can generate educational content but cannot yet replicate the interpersonal dynamics and real-time adaptation that make counseling effective.

BLS evidence: The duties include 'Educate and counsel patients and their families on a variety of issues, such as treatment and self-care for asthma.'

Examine patients to assess their medical condition

Physical examination requires hands-on palpation, auscultation with a stethoscope, visual inspection of physical findings, and real-time clinical judgment in an unpredictable patient interaction. AI lacks the physical embodiment and tactile sensing capabilities needed for competent examination.

BLS evidence: The duties section states 'Examine patients' as a primary task, and the opening summary identifies 'examine, diagnose, and treat patients' as the core work.

Provide surgical care including closing incisions and pre- and post-operative care

Surgical procedures including wound closure require precise manual dexterity, real-time adaptation to tissue characteristics and bleeding, and handling of unexpected intraoperative findings. Surgical robotics exist but require human operators; autonomous surgical AI is far from clinical deployment.

BLS evidence: The page states 'a physician assistant working in surgery may close incisions and provide care before, during, and after an operation.'

Provide medical treatment such as setting broken bones, stitching wounds, and administering immunizations

Setting bones requires precise manual manipulation and tactile feedback, suturing demands fine motor control in varied tissue conditions, and administering injections involves physical patient interaction. Current robotics cannot reliably perform these procedures across diverse patient presentations and clinical settings.

BLS evidence: The duties section specifies 'Provide treatment, such as setting broken bones, stitching wounds, and immunizing patients.'