Workforce in hepatology: Update and a critical need for more information

In 1950, 12 individuals attended the first meeting of the American Association for the Study of Liver Diseases (AASLD).1 Sixty-five years later over 10,000 individuals from all over the world attended the 65th annual AASLD meeting. In that time, the field of hepatology has rapidly evolved with the widespread success of liver transplantation and the development of effective new therapies for many liver diseases. Cure is now possible for most patients with hepatitis C, one of the most common diseases hepatologists treat. Arguably, no other medical discipline has seen these kinds of advances in such a short time. Training in hepatology has also transformed considerably. In the past, gastroenterology fellows with an interest in liver disease sought additional training in hepatology at a few specialized centers. Now, the primary pathway to become a hepatologist is through a formal, accredited fellowship year dedicated to advanced/transplant hepatology.

In 2001, the Lewin group presented its findings of a 2-year workforce study commissioned by the AASLD2 (data on file AASLD). At the time, the study defined hepatologists as gastroenterologists for whom 50% or more of the patient mix consisted of patients with liver disease. Transplant hepatologists were defined as those hepatologists who specialized in patients who require a liver transplant. The findings of the Lewin group were instrumental in demonstrating that transplant hepatology is a distinct discipline that requires “cognitive expertise over and above that provided by the hepatology training in a standard gastroenterology training program.”3, 4 This was formalized in 2006 with the establishment of a certificate of added qualification in transplant hepatology that eventually became a subspecialty certification from the American Board of Internal Medicine (ABIM).5 The Lewin group study also concluded, primarily based on waiting times for initial appointments, that transplant hepatologists were in excess demand. This led to substantial efforts to increase the hepatology workforce through novel training pathways and improved compensation models.

With the aforementioned dramatic changes in the field of hepatology, it is important to reassess previous assumptions and reevaluate the current and future workforce. Workforce evaluations in medical subspecialties are not unique to hepatology or to the United States.6 Workforce issues in hepatology have been the topic of several articles that tended to be qualitative because adequate data were not available to accurately estimate the number of hepatologists needed to meet demand in the United States.4, 7 The purposes of this article are to provide an update on the current landscape of the issues facing the hepatology workforce, identify challenges in determining accurate workforce estimates, and propose the next steps to address these issues. Table 1 lists the key issues, proposed next steps, and sources of information needed to evaluate the hepatology workforce.

Current Status and Future Demand of the Hepatology Workforce

An assessment of the workforce in hepatology should include the current number of hepatologists in the workforce, the number of hepatologists entering the workforce, and the number of hepatologists leaving the workforce. Estimates of the number of practicing hepatologists are difficult to ascertain, in part because of the heterogeneity in defining a hepatologist. A hepatologist may be better defined now than the working definition used by the Lewin group in 2001 and may include those who formally trained in an accredited transplant hepatology fellowship and passed a certification exam or those who participate in continuing education in hepatology, in addition to those for whom 50% or more of their practice consists of patients with liver disease or are “self-identified” as hepatologists. Recently, it was estimated that there are “only a thousand or so hepatologists to serve our communities” or an average of one hepatologist per 330,000 individuals.1 In contrast, there is one hepatologist per 524,000 individuals in the United Kingdom (acknowledging that the health care systems are very different).4, 7-9 In 2006, 214 physicians sat for the first ABIM certification exam in transplant hepatology. This number increased to 430 physicians certified in transplant hepatology in 2013.7 In 2013, fewer than 2,000 AASLD members indicated that more than 50% of their time was spent practicing hepatology.7

There is a perceived, unmet need for transplant hepatologists, a perception that was formalized by the Lewin group study and continues, in part due to the rising incidence rates of nonalcoholic fatty liver disease (NAFLD) and hepatocellular carcinoma and the growing number of liver transplant recipients in need of posttransplant care.10-12 The results from a survey of medical directors of US liver transplant programs published in 2008 suggested that the demand for transplant hepatologists is high.11 The authors reported an average of 3.8 full-time hepatologists at transplant centers and that 81% of transplant centers were recruiting or would be recruiting one or more hepatologists in the following 3 years, amounting to a demand of approximately 130-150 additional hepatologists. This does not account for demand for hepatologists in nontransplant settings such as private practice, nontransplant hospitals, or Veterans Affairs hospitals and does not account for the demand to care for the increasing rate of older patients hospitalized for complications of cirrhosis.13 Thus, this is likely an underestimate of the true demand. At the current rate of training 40-50 transplant hepatology fellows per year, it is conceivable that the demand could be met within a few years. Furthermore, a hepatologist can remain in the workforce for 30 or more years. The annual rate of retiring hepatologists or those leaving practice is likely to be lower than the rate of those entering the workforce. A workforce evaluation should analyze the age distribution of hepatologists to determine if the number of hepatologists entering the workforce will replace the number of hepatologists retiring or approaching retirement, or alternatively reducing their clinical workload, in some cases due to physician burnout. These estimates should be considered when determining workforce needs and the number of training positions.

A workforce assessment needs to consider all potential practice settings. At the time the Lewin group study was done, more than 80% of hepatologists (with or without a transplant focus) practiced at large institutions such as medical schools and other academic settings. Since then, hepatologists have expanded into new practice settings such as nontransplant hospital-based single or multispecialty groups or private practice “liver institutes” that are supported by extensive clinical care and research programs that may not be affiliated with a university. These “nontraditional” practice settings need to be accounted for in any workforce projections. Furthermore, if a significant number of self-identified hepatologists primarily engage in research rather than clinical care, this would impact workforce needs and projections.

To estimate the hepatology needs of the population, it is important to consider that some liver disorders are managed by nonhepatology providers, such as infectious disease clinicians treating hepatitis C and internists or endocrinologists treating NAFLD. Furthermore, limiting the provider workforce to physicians alone would neglect the fact that nonphysicians such as nurse practitioners or physician assistants provide care for patients with chronic liver disease in many health care delivery systems.

It is uncertain how the recent introduction of direct-acting antiviral agents for treatment of hepatitis C will affect the demand for hepatologists. In 2010, Davis et al. projected that there would be 150,000 cases of decompensated cirrhosis from hepatitis C in 2020.14 However, a model published in 2014 projected a peak of 62,700 cases of decompensated cirrhosis from hepatitis C in 2020.15 The difference in estimates is largely due to significant improvements in virologic cure with direct-acting antiviral agents. At the same time the prevalence of NAFLD is increasing, with an estimated 28.8 million individuals with NAFLD in the United States.16, 17 Whether the eventual void left by curing most patients with hepatitis C will be filled or exceeded by NAFLD remains uncertain. The prevalence of other chronic liver diseases in the United States, including those attributed to alcohol and hepatitis B, has remained stable,16 yet the acuity and complexity of care for patients hospitalized for complications from liver disease have increased.18, 19 Dramatic changes in hepatology-related disease prevalence and treatment will have a substantial impact on the population needs for hepatology providers, and accurate estimates are needed for realistic workforce projections.

Hepatologists Entering the Workforce: Trainees

The number of hepatologists who are trained each year should be determined by the burden of liver disease and the needs of the population. The number of hepatologists entering the workforce represents perhaps the most accurate data available. In the relatively recent past, available training in hepatology was only found within traditional gastroenterology fellowships, as well as some unaccredited and nonstandardized fellowships in advanced or transplant hepatology that do not currently provide eligibility for transplant hepatology board certification. There remains some demand for these unaccredited training programs but to a lesser extent. Since the last workforce study transplant hepatology has become recognized as a distinct discipline with 48 currently accredited transplant hepatology fellowship programs in the United States (https://apps.acgme.org/ads/Public/Reports/ReportRun?ReportId=1&CurrentYear=2016&SpecialtyId=149&IncludePreAccreditation=false). When addressing the hepatology workforce deficit, the length of training was identified as a potential barrier to growing the hepatology workforce.7 To mitigate this barrier, the AASLD and the ABIM began piloting a new training pathway in 2012 that is testing a competency-based medical education framework while simultaneously reducing the length of training in gastroenterology and transplant hepatology from 4 years to 3 years.20, 21 The result has been a gradual increase in the number of graduating transplant hepatology fellows each year (Table 2).

An assessment of the hepatology workforce should include the experiences of recent fellowship graduates seeking their first job. The perception of a shortage of hepatologists may ultimately suffer from similar inconsistencies in the widespread predictions of a physician shortage. That is, the predicted deficit may not be a national issue but one that affects some local markets and not others, although the impact of maldistribution on the physician shortage is controversial.22-24 This “maldistribution” may be why recent graduates of transplant hepatology fellowships have difficulty finding a job in the city of their choosing at a time when data suggest that many transplant centers are recruiting new hepatologists. Furthermore, maldistribution may disproportionately affect individuals living in rural setttings, limiting access to care in these areas. Electronic care and telemedicine are relatively new modalities that have the potential to improve access to hepatologists by underserved areas but may also increase the demands on already existing hepatology resources. This is an area that warrants further study.

Compensation Models

Although compensation is not a primary focus of this article, it is a significant driver of workforce fluctuations. Therefore, a workforce study should seize upon the opportunity to collect information on compensation. Compensation may vary by practice setting or type of patients seen in a hepatology practice. Hepatologists can practice in a variety of settings including university-based practice, hospital-based practice, private practice, or the Veterans Affairs system. Hepatologists may practice transplant hepatology and/or general hepatology. In many centers the ability to care for transplant patients is the feature that distinguishes hepatologists and is used to determine compensation. However, in many centers, the hepatologist is often compensated in a traditional production-based model that may not account for the additional expertise or time required for the management of complex liver patients. As a result, hepatologists may devote significant time to endoscopic procedures out of necessity to generate income.

Production-based compensation may not accurately represent the value of a transplant hepatologist. Production-based compensation models commonly use the work component of the relative value unit (wRVU). The Association of American Medical Colleges and the Medical Group Management Association provide estimates for annual wRVUs for a number of specialties and subspecialties, but there is no specific metric for transplant hepatology. The lack of estimates for transplant hepatologists may lead some centers to apply the gastroenterologist wRVU goal to hepatologists, yet hepatologists may generate more revenue with fewer patient encounters despite performing a lower volume of procedures than their gastroenterology counterparts.25

Transplant hepatology is associated with significant nonbillable activities and downstream revenue; therefore, non-RVU compensation models may be an alternative method to apply to compensation.26 Alternatives to wRVU-based compensation models incorporate elements based on program growth (e.g., increasing transplant volume), waiting list additions, program building (outreach), program quality metrics (e.g., length of stay, posttransplant survival), or a combination of these elements. Compensation models can include a combination of wRVUs and non-RVU metrics. Non-RVU metrics may become increasingly used as health care shifts to accountable care and quality-based metrics. A formal workforce study should collect information on compensation and models used to evaluate performance and incentive pay.

Conclusion

It is timely to address workforce issues given the changes that have occurred in health care and hepatology since the last workforce study 15 years ago. Some of the more notable changes have occurred in training pathways of hepatologists, advances in hepatitis C therapy, and the growing obesity and NAFLD epidemic. Whether there will continue to be a workforce deficit is currently unknown. An updated workforce study is needed to obtain data for optimal workforce estimates and to properly design training pathways. A comprehensive hepatology workforce analysis is thus timely and urgently needed to estimate the needs of society.