Loading ...

As physician shortages deepen and waits grow longer, health systems are asking whether nurses can take on more tasks traditionally handled by physicians. For some physicians, however, that framing misses the harder safety question: Who recognizes the atypical presentation, the deteriorating patient, or the medication decision that needs another level of review?

That concern about complex cases and escalation ran through comments on a recent Medscape article on nurse-physician substitution. The article centered on a Cochrane review published in February that found little overall difference between physician- and nurse-delivered care on key outcomes, including mortality, quality of life, and patient safety events. But the review also cautioned that the substitution models examined were "highly heterogeneous," making broad conclusions difficult. For many readers, the differences between care models became the central issue.

The Training Gap Physicians Keep Citing

One commenter put the training concern bluntly: "Anyone can practice by clicking protocoled boxes … and 90% of the time the patient will be okay. The entire purpose of medical training is for the 10%."

Another offered a numerical contrast: Attending physicians receive 21,000 hours of clinical training, whereas a nurse practitioner receives around 500. 

Exact comparisons vary by country, specialty, and pathway, but European sources support the broader point: Irish advanced nurse practitioner registration includes 500 supervised clinical practice hours, while European Union rules require specialist medical training to meet defined minimum durations and include supervised participation in clinical care.

The argument behind both comments is that routine care is not where the greatest risk lies. It lies in the cases that do not fit the script. Recognizing those cases requires a depth of training that may not be fully captured by outcome studies.

A commenter with 27 years in healthcare pushed back against this argument by emphasizing what nurse practitioners may add. They often consider whether a patient can "afford or tolerate" a treatment plan, the reader said, not just whether it is clinically optimal.

The Cochrane review's findings point in a similar direction. Some of the better outcomes in nurse-delivered care appeared to stem from structural differences — more frequent appointments, added educational components, and greater on-demand access — rather than from clinical skill alone. 

When the Anecdotes Arrived

Some commenters focused on experience rather than on principles. One described a 4-minute urgent care visit with a physician assistant — billed at $1060 — during which they had to prompt the clinician to examine their nose and throat. Another detailed two specialist appointments with no physical examination and a medication recommendation that later appeared inconsistent with guidance from a national health institute. A third estimated that 80% of clinic visits in their area were handled by nurse practitioners or physician assistants and described "the utter lack of attention to detail" as the defining concern.

The accounts highlight a recurring, problematic pattern: brief visits, absent examinations, and clinical decisions made without adequate evaluation. 

A neurosurgery physician assistant with 18 years of experience responded to that thread, describing the skills they developed through long collaboration with nursing and physician colleagues and the important role they play in ensuring continuity within their practice. 

The exchange clarified that the real problem does not lie within a professional category but in the care model applied. A care model can be poorly designed regardless of who staffs it. In practice, "substitution" may describe very different arrangements, from structured team care with defined escalation pathways to more independent task-shifting.

How Access Changes the Question

Where many commenters were weighing the risks of substitution, a smaller group was weighing the risks of the alternative.

"In some rural areas, you can be seen by NO ONE for months," one wrote. Another addressed skeptics directly: "Would you want to be seen by no one at all? Because that is the current state of the American healthcare system."

That reframe did not settle the debate. But it did shift its terms. The question was no longer only whether nurse-delivered care is equivalent to physician-delivered care in settings where both exist. It was whether a supervised, protocol-driven model with clear escalation rules is better than a months-long wait — and whether health systems are being honest about which comparison they are making when they expand nonphysician roles.

What Clinicians Said They Would Need

The comments did not produce consensus, but they did produce a list of conditions for those designing these types of care models:

• Define the model: collaboration, supervised delegation, or replacement. Several commenters supported team care but objected to blurred boundaries.
• Build in escalation protocols so that complex or deteriorating cases reach physician review without delay.
• Address training and supervision requirements directly. Several commenters flagged the Cochrane review's warning that outcomes do not transfer automatically across settings.
• Include physician assistants in workforce planning. More than one commenter noted that physician assistants were absent from the substitution framing despite a training model closer to medicine than nursing.
• Protect time for physical examination and medication review. The anecdotal complaints centered consistently on rushed, examination-free visits, not only on professional credentials.
• Name the access trade-off. The case for expanded roles looks different when the real comparison is not physician care vs nonphysician care, but delayed care vs timely care.