Why Hospitalists Should Not Routinely
Perform Bedside Procedures
After six months of critical care rotations during my internal medicine residency, I felt well prepared to perform most of the procedures that were core competencies. But after a few years in practice, I’d done only a handful of paracenteses and thoracenteses, and that confidence was gone. I don’t think my experience is unusual. Individual hospitalists simply can’t generate the volume of procedures needed to stay competent, and lower volumes translate into lower skill and potentially a higher complication rate.1 At most hospitals, a practitioner is credentialed and proctored once and never reassessed despite the fact that procedural skills can decline over time.2 An additional barrier for hospitalists performing these procedures bedside is the fact that most hospitalists haven’t received adequate training in POCUS. It’s in our patients’ best interest to have these procedures performed by people who can demonstrate genuine proficiency.
Beyond competency, asking hospitalists to perform procedures pulls them away from already very busy days, contributing to burnout and interruptions to clinical care. A single bedside procedure can take an hour or more, during which the hospitalist is unavailable to nurses, patients, consultants, and families. It isn’t worthwhile for hospitalists to carve out that time when others can do the same procedures faster and more safely.
There are two situations where it makes more sense for hospitalists to perform procedures. First, hospitalists who spend a substantial portion of their time providing critical care may do enough procedures to maintain ongoing competency. Second, larger hospitals with the volume to support a dedicated procedure service can have a subset of hospitalists maintain their procedural and POCUS skills and staff that service. This would concentrate the work in a few hands that actually do enough to stay sharp, while at the same time boosting hospital throughput without detracting from everyone else’s patient care.3
It’s important to note the differences in perspective between Dr. Migliore and me—his, coming from an academic center, and mine, from community hospitals. While his vision is compelling, it doesn’t reflect the current realities at most hospitals in the country.
Dr. Gershfield
Dr. Gershfield is a hospitalist at Sequoia Hospital in Redwood City, Calif., and director of quality and performance for hospital medicine at Vituity.
References
- Cool JA, et al. Procedural competency among hospitalists: a literature review and future considerations. J HospMed. 2021;16(4):230-235. doi:10.12788/jhm.3590.
- Hale C, et al. Cohort study of hospitalists’ procedural skills: baseline competence and durability after simulation-based training. BMJ Open. 2021;11:e045600. doi:10.1136/bmjopen-2020-045600.
- Gentile N, et al. Evaluating the impact of a newly established hospitalist medicine procedure service on length of stay and procedural outcomes. J HospMed. 2026;1-8. doi:10.1002/jhm.70311.
Why Hospitalists Should Routinely Perform Bedside Procedures
The case against hospitalist procedures is not a case against hospitalist procedures. It concedes that hospitalists in critical care should perform them. It concedes that a dedicated procedure service, staffed by hospitalists, is the correct model at any hospital large enough to run one. It’s more a scheduling preference: some hospitalists, organized deliberately, should do procedures, which is my position.
Competency is not a trait issued at graduation, which is then slowly misplaced, but rather a state function. It is built, maintained, and measured, or it is not. The other essay describes a physician credentialed once, proctored once, and never reassessed, drifting down to a handful of paracenteses a year. This happens, and it is a real failure, but more a failure of the system, not an inherent property of hospital medicine. The correct response to a decayed skill is to rebuild it, not to declare the skill unbecoming. In the same way we do not respond to a hospitalist who has gotten rusty on diabetic ketoacidosis by removing diabetic ketoacidosis from the job; we educate. Procedures deserve the same approach.
The volume argument also proves too much. If low individual frequency disqualifies competency, it disqualifies most of what makes a hospitalist a physician. Massive gastrointestinal bleeds are infrequent. Status epilepticus is infrequent. The genuinely crashing patient at three in the morning is, thankfully, infrequent. We do not outsource these because they are rare. We concentrate on exposure, we simulate, we build reps, and we hold people to a standard. The same machinery that keeps a hospitalist sharp on a dying patient keeps a hospitalist sharp on the use of ultrasound or the placement of a needle. Volume is an input we can engineer.
I also don’t agree that someone else will always do the procedure faster and more safely. Faster than whom, and when? The patient with a tense abdomen and a fever does not care that interventional radiology is excellent. He cares that it is Saturday, that the queue is four deep, and that his paracentesis has just become a Monday problem. Delay is not neutral. In spontaneous bacterial peritonitis, each hour to paracentesis carries measurable mortality.1 Outsourcing the procedure does not stop that clock; rather, it relocates the needle and lets the clock keep running. A hospitalist who can drain that abdomen at the bedside tonight is not a luxury but is the difference between a diagnosis made and a diagnosis postponed.
This is the part the efficiency argument doesn’t price in. The comparison is never stated as hospitalist-now versus radiology-now. It is hospitalist-now versus radiology-sometime. When framed that way, the safety calculus inverts. The complication rate of a competent bedside procedure is a real number. The complication rate of a delayed procedure is also a real number, and frequently the worst one. When the studies do put them side by side, the patient-centered numbers point in the same direction: hospitalist procedure services shorten time to procedure and shorten length of stay against the radiology alternative.2,3 Speed, when speed is the treatment, is safety.
In my opinion, there is also a deeper cost, and it is generational. A specialty that stops doing procedures cannot teach them.4 Skills that go unpracticed do not idle; they die, and they take the next cohort down with them. You cannot conjure a procedure service out of a workforce that never learned to do them. The dedicated service model depends entirely on a pipeline of hospitalists who kept the skill alive long enough to staff it. Abandon routine procedures broadly, and you will saw off the branch the exception is sitting on. Deskilling is not a plateau, but more of a slope. It points in one direction only.
POCUS makes this point rather than undermining it. Yes, most hospitalists were trained in an era that treated ultrasound as someone else’s instrument. That is a training gap, and training gaps close. SHM has already published the framework for closing it: applications, training, assessment, and program management.5 The technology that was supposed to make bedside procedures too complex for hospitalists is, in practice, the technology that makes them safer in hospitalist hands. The claim that we lack the skill is an argument for teaching the skill, on a timeline, with standards. It is not an argument for permanent retreat from it.
Which brings me to burnout. Ask any exhausted hospitalist what hollows out the day. It is not the hour spent doing a thoracentesis and then watching a patient breathe easier. It is the 90 minutes on hold with a consultant, the eleventh reconciliation of the same medication list, and the slow conversion of a physician into a dispatcher who orders, documents, and coordinates, but no longer does. The procedure is often the most physician-like hour of the shift. It is the part where you use your hands and your judgment and see the result in real time. Stripping that out in the name of wellness misdiagnoses the disease. The disease is not doing medicine. The disease is being kept from it.
So the real question is not whether an individual hospitalist can generate the volume. It is what we want a hospitalist to be. I say build the procedure service and concentrate the reps. Credential seriously and reassess honestly. Measure complications and time to procedure as the quality metrics they are. Teach POCUS with the same rigor we bring to antibiotic stewardship. Make competency something the department owns rather than something the individual is assumed to have quietly maintained on his own.
Do that, and hospitalists will perform bedside procedures routinely. Not everyone, and not carelessly, but routinely, deliberately, and well, as a core function of a physician who is present, who is fast when speed is itself the treatment, and who does not flinch when the answer is a needle rather than a consult. The bedside is not borrowed territory. It is our jurisdiction. Ceding the procedures that happen there is one more step in a long drift from physician to coordinator, and that drift, not the occasional difficult paracentesis, is the pathology.
The other side wants to protect patients from undertrained hands. I agree, completely. We disagree about the remedy. They would remove the hands; I would train them. One of those answers continues to build a specialty. The other explains why the specialty is no longer needed.
Dr. Migliore is an assistant professor of medicine at Columbia University College of Physicians and Surgeons, and director of general medicine consult and perioperative services as well as a medicine attending physician, at Columbia University Medical Center, both in New York.
References
- Kim JJ, et al. Delayed paracentesis is associated with increased in-hospital mortality in patients with spontaneous bacterial peritonitis. Am J Gastroenterol. 2014;109(9):1436-1442.
- Ritter E, et al. Impact of a hospitalist-run procedure service on time to paracentesis and length of stay. J Hosp Med. 2021;16(8):476-479.
- Nandan A, et al. Characteristics and impact of bedside procedure services in the United States: A systematic review. J Hosp Med. 2022;17(8):644-652.
- Tukey MH, Wiener RS. The impact of a medical procedure service on patient safety, procedure quality and resident training opportunities. J Gen Intern Med. 2014;29(3):485-490.
- Soni NJ, et al. Point-of-care ultrasound for hospitalists: a position statement of the Society of Hospital Medicine. J Hosp Med. 2019;14(1):E1-E6.