In 1949, the English-born physician John Wild, working at the University of Minnesota, discovered that he could determine the thickness of bowels injured in the war by bouncing sound waves though the abdominal wall.
Over the next 30 years, medical ultrasound technology improved markedly, ultimately leading to the many uses we’re all familiar with. Once ultrasound became available for widespread use in the late 1970s, radiologists held a near-exclusive franchise for a while. But the last 25 years have seen the diffusion of the technology to a variety of specialties, including obstetrics, surgery, emergency medicine, and critical care. I believe that the time has come to – carefully and thoughtfully – add hospitalists to this list.
The first medical ultrasound machines “lived” in the radiology department, along with other stationary behemoths like CAT and rectilinear scanners. With the exception of the relatively uninformative portable plain radiograph, whose source was an unwieldy beast wheeled to a patient’s bedside (after which, the film was rushed back to the radiology department for interpretation), patients who needed a whiff of Roentgen’s magic needed to “go to radiology” – for both the creation of the image and its interpretation.
In the late 1980s, hewing to a medical version of Moore’s Law, ultrasound machines became smaller, faster, more powerful… and portable. Radiologists sometimes went with their machines to the patient’s bedside, but mostly they didn’t. Technicians were sent out – like modern sentries – to record the images and bring them back to Fortress Radiology, where they were read by the experts.
Since buying a fancy espresso machine a few months ago, we rarely go to Starbucks. And so it is with many disruptive innovations, in which new technologies allow lesser-trained mortals to upend previously ironclad franchises. As soon as ultrasound machines could leave radiology departments, radiologists lost the ability to fully control their use. Obstetricians began ultrasounding pregnant bellies and before long, OB-ultrasound became an essential part of an obstetrician’s toolkit. Surgeons developed the Focused Assessment with Sonography for Trauma (FAST) exam, rendering diagnostic peritoneal lavage a historical curiosity. The moat surrounding the fortress had been breached.
One can only imagine how this development must have annoyed the specialists in radiology (and cardiology, in the parallel case of echocardiography), who trained for years in their craft and are unambiguously expert in it. Predictably, they erected barriers for the use of ultrasound by non-radiologists, generally clothed in the language of quality assurance but undoubtedly influenced by issues of turf and money as well. Writing in 2007, a group of emergency physicians and intensivists offered their view of the controversy:
The paradigm of emergency ultrasound has historically been a difficult concept for traditional providers of ultrasound to understand or accept. Emergency ultrasound is not an inferior imitation of comprehensive consulting imaging services, but rather it is a focused, appropriate application of technology that provides essential diagnostic information and guidance of high-risk procedures…. Issues of physician credentialing, ownership of technology, exclusive contracts, reimbursement, and specialty society advocacy positions have overshadowed clinical evidence and the practical experience of improved patient care.
While the pushback was predictable, so was the outcome. The upstarts eventually won out, since they could buy their own equipment and develop training and certification programs to address the quality concerns. Moreover, radiology is a consultative business, and the emergency medicine doctors, surgeons, and obstetricians lived upstream, in full control of what happened to their patients. Radiology, for all its successes in the marketplace (the average US radiologist earns $350,000 annually, with one-in-five earning more than $500,000), remains a practice largely reliant on the kindness of others. This makes it challenging to engage in a full-throated effort to block those others from doing what they consider best for their patients.
Over the last 15 years, I’ve been in the privileged (and daunting) position of helping to chart the course of the young specialty of hospital medicine. One of my key beacons has been the field of emergency medicine (EM), a site-defined generalist specialty like hospital medicine but one 30 years our senior. This timeline means that one can give the illusion of being prescient by simply learning what happened in EM and extrapolating it to our field. Using that logic, I knew it was inevitable that hospitalists would eventually enter the world of – and the controversy over – bedside ultrasound.
Starting in 1988, emergency physicians began to report using ultrasound to help with procedures like line placement and thoracentesis, and then, slowly, to aid in diagnosis. Within a few years, the ER docs had produced research demonstrating that they could safely and effectively perform procedures and make accurate diagnoses with limited, question-focused (i.e., is there biliary duct dilatation? Is there an ectopic pregnancy? Is there an aortic aneurysm?) exams. Soon, ultrasound – both for procedures and diagnosis – was a core part of EM training curricula.
This, of course, led to food fights in hospital credentials committees around the country, as emergency physicians sought to add ultrasound to their list of privileges, and radiologists resisted. Ultimately, though, this was a tide that could not be held back, particularly once studies demonstrated competence and utility, and EM residents began to be trained in the practice. And radiologists’ economic concerns have been muted by the fact that, at least in some radiology departments (including my own), ED ultrasound has been associated with increases, not decreases, in radiology ultrasound volume. It turns out that while some ED ultrasounds obviate the need for a gold standard radiology department study, others lead to the identification of abnormalities that need further, more comprehensive investigation.
After about a ten-year time lag, a similar trend has played out in intensive care medicine. Many intensivists now use ICU ultrasound and echocardiography to assess fluid status, diagnose pericardial tamponade and cholecystitis, and aid in central line placement. As in EM, the generally positive experiences have led to calls from the field to make this a core part of the critical care medicine curriculum and intensivists’ scope of practice. The outcome is likely to mirror that of EM.
Which brings us to hospitalists. Starting a few years ago, some hospitalists trained in the use of bedside ultrasound for procedures, and some hospitalist programs – including my own – built stand-alone procedure services to help perform (and, in training programs, teach residents how to perform) procedures such as central line placement, paracentesis, and thoracentesis. Studies have shown that such services lead to superb outcomes and high levels of patient satisfaction. At UCSF, our radiology department – after some initial hesitation – has been supportive, recognizing that it is better for a patient to have a procedure done safely at the bedside than be transported to and from radiology for a site marking and then returned to the floor – a journey that is both wildly inefficient, unpleasant for the patient, and periodically risky.
And, as in emergency and critical care medicine, hospitalists are now turning their attention to diagnostic uses of ultrasound for things like fluid status, gall bladder disease, pericardial and pleural effusions, and venous thrombosis. Some of this work will be in the ICU, where, in many community hospitals, hospitalists make up a significant fraction of the physician workforce (given the tremendous shortage of trained intensivists). But some of the uses are certain to be on the wards as well.
Last month, I had the occasion to catch up with an old friend and former UCSF resident, Dr. Liz Turner, now the director of critical care medicine at UC Irvine and an expert in training clinicians in bedside ultrasound. With support from the new University of California Center for Health Quality and Innovation, Liz recently completed a pilot training program, with impressive results. Over three months, using both web-based and hands-on sessions (with competency testing after each segment), critical care physicians learned to use ultrasound for general cardiac and thoracic indications, and achieved high levels of competency and confidence. There’s no reason to believe that such a curriculum wouldn’t be equally effective with hospitalists.
For hospitalist programs that are seeking to dip their feet into the ultrasound pool, there are lots of obstacles to overcome, including financial and logistical ones like purchasing the equipment and archiving the images in the electronic record. But emergency medicine programs have overcome these, and they don’t seem too daunting. The most important challenge is to negotiate our relationship with our radiology and cardiology colleagues, the true experts in ultrasound and echocardiography. We need to convince them that we are able to manage this technology competently, that we understand its – and our – limitations, and that we are committed to a collegial relationship.
We also need to sort out how widely distributed we want this competency to be within the hospitalist universe. For example, should all hospitalist faculty be trained in the use of ultrasound? Probably not, at least for now. At a place like UCSF, for example, I’d aim for a core group of faculty to become our local go-to bedside ultrasonographers. When we built our procedure service, we staffed it with a core group of about six to eight of our (50) faculty who committed to receive rigorous training and certification and who could maintain a sufficient yearly volume to remain skilled. In our setting, a similar model may be the right one for diagnostic ultrasound, at least for the next several years.
On the other hand, community programs, in which a hospitalist may be the only physician in the building at night, may ultimately need all the hospitalists to be trained and certified. The growing group of American hospitalists choosing to spend some of their professional time in resource-poor countries, where ultrasound is emerging as an essential tool, will also need this skill.
But these are details. The larger issue now is to decide whether we believe that building competency in ultrasound among generalist physicians – in this case hospitalists – will enhance patient safety, quality, and value. Personally, I do. In fact, I’d wager that the hospitalist of 2022 is more likely to be carrying an ultrasound probe than a stethoscope. If this is right, it’s time to begin the hard work of developing the partnerships, training programs, competency assessments, quality assurance methods, and economic models to support bedside ultrasound by hospitalists.