So Long to Solo Practice Medicine

In a very recent issue of JAMA, I was reading about the general problem of the decline in small medical practices, “Can Small Physician Practices Survive? Sharing Services as a Path to Viability.”(1) Although the scope of this article is more concerned about the use of new resource models that would help small practices stay afloat, the beginning sections of the commentary speak about the very important topic of quality of care. After describing briefly a few of the examples the authors provide, I will illustrate how these ideas directly affect small vs larger sleep medicine practices.

The first data point is the decline in solo practitioners from 1983 to 2014, a drop from 44% to 19%. In contrast, in 2015 34% of all doctors were in practices involving 100 or more physicians. These changes are dramatic to say the least, and the trends certainly do not bode well for solo docs. I worked in an academic medical setting from 1979 to 2000, and then I switched to solo practice 18 years ago. All around me, I see a dwindling supply of solo practitioners in various fields of medicine. In New Mexico, I am not certain of the actual count, but I believe there may only be a handful or fewer solo sleep medicine specialists. The largest proportion of sleep patients are served by large hospital centers, academic, or community-based facilities. All these sites have three or more sleep doctors, and most have several mid-level professionals in the form of physician assistants and nurse practitioners. By definition, some of these practices are still considered small, but then again they are still influenced by larger administrative bureaucracies that impact quality and delivery of care.

The second and main data points address quality of care in ways that may or may not be surprising. In one study, a comparison looked at practices of only 1 to 2 docs versus practices with 10 to 19 docs and found that the smaller groups demonstrated a 33% lower rate of preventable hospital admissions. Small practices of only 3 to 9 docs showed a 27% lower rate of preventable hospital admissions compared to larger practice groups. Presumably, “preventable” simply meant various factors in the examined caseloads that could have been identified and managed to prevent the hospitalizations, and the smaller practices were far better at this form of management. Another study looked at a very wide spectrum of practices measured from 5 to 750 doctors. Not only were the smaller practices better again with lower rates of hospitalizations but also better in keeping costs lower in managing diabetic patients.

Can you guess why?

The commentary does not address this issue other than to indirectly suggest that the “personal touch” had something to do with these better and cheaper outcomes. There are probably a lot of different reasons, but one that jumps immediately to mind might be as simple as the type of person who stays in solo private practice in this day and age of aggregate data, if not aggregate doctors. One of the big ideas about BIG DATA is the belief that with so much information, the probability is increased of finding out so many more things about entire populations that eventually the insights filter down into specific patient situations. However, the conclusions drawn from aggregated data may sometimes end up serving a purpose that places financial or expedient aspects over quality of care.

Consider the inception of home sleep testing. As HST cannot provide the same amount of information as a polysomnogram, why would insurers promote this model of care? The answer is not just a straightforward cost-analysis. Rather, entrepreneurs who invented HST devices looked at the big picture and realized that an enormous number of patients could be diagnosed with sleep apnea without ever having to know what was transpiring in their sleep patterns. Yes, there is tremendous irony in the fact that a home sleep test usually is not measuring sleep at all. Instead, it measures a sufficient amount of information regarding breathing and oxygenation to reliably detect the presence of a sleep breathing disorder. And, even though my colleagues and I want to see more sleep specific information (EEG brain waves) to actually measure the important variables of sleep stages and sleep fragmentation, we cannot deny the HST model provides a clear-cut tool for diagnosis OSA in specific cases. Some of these devices purport to test for UARS as well.

Thus, this BIG DATA idea that you can detect and measure the severity of the diagnosis of OSA without knowing what’s going on with the patient’s sleep has had a major impact on the field of sleep medicine. Nowadays, HST provokes a lot of controversy not to mention a great deal of struggle between sleep centers and insurers, whenever the necessity for more in-depth sleep information is in play. No doubt, technology will solve this problem as well, probably sooner than later, as HST devices become more sophisticated, but that’s another story.

Think of the potential impact of HST on solo practitioners. On the one hand, this model of care has actually driven some private practice doctors into retirement (and closed some academic centers as well), because these sleep professionals were situated in an environment where their particular insurance companies rapidly instituted policies promoting and then demanding HST’s. If a solo practitioner or any other sleep doc for that matter had to rely on some proportion of cases tested in the sleep lab, then as soon as the number drops below that percentage, the practice may no longer have been financially viable.

When I was growing up in the 1950s and 1960s, a common expression of the day was, “You can’t fight city hall.” It’s just as relevant today as government in many places continues to expand and create more and more rules and regulations about everything in our lives. The same paradigm seems to have taken hold of the insurance companies who often dictate what actually occurs in sleep medicine practices in many parts of the USA. Again, how long could a solo practice physician fight with insurers to maintain their necessary financial status as well as maintain a level of quality of care perceived as necessary and appropriate? Take the specific example of the prior or pre-authorization process, which wastes enormous amounts of time, energy, resources and money in attempting to schedule patients in a sleep center. The expanding prior authorization paradigm is already curtailing practices patterns in numerous ways, including increases costs while decreasing revenue.

Return to the specific effects on solo practice, keep in mind that many solo practice physicians are probably one of two overlapping types. They might be more curious and creative about looking at new or advanced ways to provide care to their patients, and they might be less comfortable with other doctors let alone administrators breathing down their necks telling them how to manage patient care. I suspect both perspectives are common in physicians who end up in solo practice. I do recall both factors influencing my decision to go into private practice by myself following several waxing and waning struggles with the bureaucracy of university environments. And, because I had the good fortune to learn so much about research in the university environment, I found I could eventually bring that mindset into my private clinical practice of sleep medicine to enrich my work experience and more importantly create new ways of evaluating and treating my patients. What I learned greatly expanded how I treated several of the most common sleep disorders, primarily OSA/UARS, insomnia and nightmares. 

Indeed, one of the greatest joys in my medical career has been to involve myself in researching both the patients who seek care at our center in New Mexico, along with the occasional research cohorts we recruited for smaller projects. This combination of activity, which generally falls into the retrospective (that is, simple chart reviews) and prospective (that is, a formal project where we seek Institutional Review Board approval) designs and has enabled us to learn so many things to apply in helping our patients sleep better. For instance, we had the opportunity to work with so many mental health patients with sleep disorders, we were inundated with the problem of CPAP failure, which in turn pressured us into looking for another solution. Once we latched onto the bilevel devices, we immediately found ourselves developing a new paradigm to monitor and resolve the problem of expiratory pressure intolerance. Simultaneous to these experiences, we realized contrary to the conventional wisdom that many insomniacs could in fact use PAP therapy, just not CPAP. Instead, by switching them to bilevel devices not only could they adapt and adhere to pressurized air flow, it actually decreased their insomnia more so than standard CPAP or APAP devices.

These two pieces of information radically changed the way in which we practiced sleep medicine, and then a third concept arose between 2008 and 2010 when we starting using even more advanced bilevel devices like ABPAP and ASV. We realized we could manually titrate these modes in the sleep lab to provide an even better response than simply using their auto-adjusting capabilities without the added human factor of a sleep technologist over-riding the device.

We were also fortunate to have the good sense to hire individuals with more creative bents to work in our front offices and in our sleep lab. And, all told there were four individuals who worked as sleep technologists and who at various points put in major efforts in sleep research at our non-profit Sleep & Human Health Institute. In total, their efforts comprised somewhere around 30 years of combined work. Several other sleep technologists put in part-time efforts totaling another 10 years of research experience.

All these process unfolded in the context of having some degree of free rein a private practice system and which further enabled us to open up and fund our non-profit institute. This creative component in myself and in the personalities of more than half of my staff greatly influenced our desire and motivation to remain in solo practice.

Other solo practitioners may operate with different aspects of creativity. For example, I have noticed many primary care physicians develop niche practices where they can provide competitive services in managing complex disease processes like diabetes such that their patients rarely if ever need consultations with endocrinologists. The same has proven clearly to be the case with many primary care physicians practicing elements of psychiatry in the management of psychotropic medications. None of these efforts guarantee that all these solo practices are the best they could possibly be. Likewise, I cannot say that I would not have benefited by engaging with other sleep doctors had they been more closely available to my practice setting. Though I meet with many sleep physicians in numerous venues throughout any given year, such encounters are far different than working in the trenches with another physician. For these reasons, some doctors forego solo practice, but are quite happy in groups of 2 to 3 physicians. 

It appears solo practices are a dying breed. My sense is this trend will be a bad thing, because it will likely lead to more groupthink as naturally occurs in larger institutional settings, such as seen in hospital practices. But, however things evolve, healthcare in general and sleep medicine in particular are certainly in for a great many fluctuations in how to deliver healthcare.



(1) Khullar D, Burke GC, Casalino LP. Can Small Physician Practices Survive? Sharing Services as a Path to Viability. JAMA. 2018 Apr 3;319(13):1321-1322. doi: 10.1001/jama.2017.21704.

Barry Krakow MD


Dr Krakow’s 27 years of sleep research have focused on the complex relationship between physiological and psychological sleep disorders. Dr Krakow currently operates private sleep medical center, Maimonides Sleep Arts & Sciences, Ltd., and serves as Classic SleepCare’s paid Medical Director.

Leave a comment

Comments have to be approved before showing up.