Hospitalization Patterns for Inpatient Pediatric Surgery and Procedures in California: The Ground Is Shifting

See Article, p 1070 Comprehensive assessment of a physician workforce can help identify emerging challenges in the marketplace and strategically prepare an organization for change. Effective allocation of a physician workforce can also increase efficiency, productivity, and effectiveness by ensuring that there is an adequate supply of physicians with the skills, knowledge, and experience required to achieve their strategic objectives. Workforce assessment is a particularly challenging exercise in subspecialty areas of medicine such as pediatric anesthesiology. Muffly et al1,2 have embraced this challenge in previous publications, and in this issue of Anesthesia & Analgesia, they add to our understanding of the pediatric anesthesiology workforce with an analysis of the shifting landscape of inpatient perioperative pediatric care in California over a 16-year period. In the United States, pediatric care occurs in a variety of settings that are associated with general hospitals (GH) and children’s hospitals (tertiary care children’s hospitals, children’s units within [GH], general [adult] hospitals, and ambulatory surgery centers). The authors sought to describe inpatient surgery demographics in children 5 years old or younger. Historically, low complexity procedures in children were as likely to occur in adult GH as they were in pediatric tertiary care centers.3 This pattern appears to be evolving. The authors documented that for 20 of the most common inpatient pediatric surgical procedures, volume appears to be shifting to pediatric tertiary care facilities for both complex and simple procedures. Importantly, while procedural volume is increasing in the tertiary care centers, overall inpatient pediatric procedural volume is declining statewide. This shift is occurring ahead of the American College of Surgeons Children’s Surgical Verification (ACS CSV) Quality Improvement Program which defines the resources required of a medical center to provide a specific complexity of pediatric care. The information in this study has potentially important implications for the pediatric anesthesiology workforce and for pediatric anesthesiology education. To put the authors’ current findings into context, it is important to appreciate their previous study on the determinants of pediatric anesthesiology workforce supply and demand. The primary supply of pediatric anesthesiologists is the number of fellow graduates produced from US fellowships. The current rate is approximately 220 fellows per year. This has increased significantly over the past 20 years. The authors predicted that there will be a 50% increase in the number of pediatric anesthesiologists by 20354 if we continue on our current path of increasing the number of fellows trained every year. Even if the number of graduating fellows is held constant, growth of pediatric anesthesiology workforce will continue because the per annum number of retiring anesthesiologists is smaller than the number of new graduates added to the workforce. As it stands now, these data suggest that we may be training too many pediatric anesthesiologists. However, there are other important determinants of supply that complicate the picture. Full- or part-time practice patterns of pediatric anesthesiologists will affect the supply of physicians that provide care for children. Graduates who choose to work in either children’s hospitals embedded in an adult hospital or in adult hospitals will have a mixed practice, and pediatric care may constitute a fraction of the total anesthesia care they provide. Finally, the interest of residents in the specialty will influence the size of the pipeline into the specialty. If residents perceive a tighter job market, they may choose another anesthesia subspecialty or forego subspecialty training altogether to quickly enter the workforce. In an ideal scenario, the supply of pediatric subspecialists would be balanced by the demand. There are several factors that affect the demand for pediatric anesthesiologists. The primary determinants include the pediatric population and pediatric surgical and nonsurgical procedural volume. There are other factors that have a yet undefined impact on demand; these include the concerns over neurotoxicity, staffing ratios, pediatric anesthesiology participation in pediatric pain and critical care, and capitation models of insurance. If we look at the demand for pediatric anesthesiologists based on the pediatric population and their surgical and nonsurgical procedural volumes, it appears to be flat or decreasing. Fertility rates in the United States have been steadily declining for almost 3 decades (https://www.cdc.gov/nchs/data/vsrr/vsrr-007-508.pdf), and the pediatric population is projected to be flat or only slightly increasing over the next 2 decades.4 Case numbers reported to several large, national, multi-institutional procedural databases (Kid’s Inpatient Database [KID], Healthcare Cost and Utilization Project, National Health Interview Survey) demonstrate a plateau or a reduction in inpatient pediatric procedural volume. In this study, Muffly et al5 also report a reduction in the overall inpatient pediatric procedural volume in California from 2000 to 2016. A critically important point to highlight is that ambulatory surgery data are not included in this study. Outpatient ambulatory surgery makes up a significant percentage of pediatric anesthetizing volume. There are approximately 5–6 million pediatric anesthetics performed in the United States every year and inpatient procedures requiring anesthesia contribute approximately 1 million to the total anesthetizing volume.6 This means 4–5 million pediatric anesthetics (up to 80%) are ambulatory. Far less is known about the details of pediatric ambulatory procedural volume, and this limits the impact of Muffly et al’s5 inpatient data on a pediatric workforce analysis. However, the same database used by Muffly et al5 (California Office of Statewide Health Planning and Development [OSHPD]) does provide both inpatient and ambulatory information for children 0–5 years old from 2005 to 2018. This database demonstrates that while there has been an increase in the percentage of some pediatric cases being performed on an ambulatory basis (repair of fracture, esophagogastroduodenoscopy [EGD], and cleft repair), the ambulatory surgical volume for 6 of the top 9 most common ambulatory procedures has decreased. Likewise, total procedural volume, including both inpatient and ambulatory, is flat or decreased in 9 of the 10 most common pediatric procedures from 2005 to 2018 in children 0–5 years old in California.7 This appears to be mirroring the trends seen in the national inpatient databases. Rabbitts et al8 in 2010 described the growth in outpatient surgery for children <15 years old. They examined the ambulatory anesthetizing volume nationally in children at 2 different time points and 10 years apart. From 1996 to 2006, outpatient ambulatory procedures nearly doubled in children <15 years old likely describing a shift in anesthetizing practice pattern from inpatient care to outpatient care. The ambulatory data from OSHPD suggest that this shift slowed after 2006 in California and there has been a decline in ambulatory surgical volume in children <5 years old in 6 of the 9 most common ambulatory procedures. It would appear that on balance the data suggest that the volume of inpatient procedures, and possibly ambulatory procedures, is not increasing. This may have implications for workforce, maintenance of clinical skills, and education of pediatric anesthesiologists. In a previously published workforce analysis, Muffly et al4 estimated demand by assuming that 60% of pediatric patients are cared for by pediatric anesthesiologists and 40% are cared for by adult anesthesiologists. This assumption is based on 2013 data that 40% of pediatric patients obtained their care at adult hospitals.3 The current publication helps to clarify this assumption. It would appear that more pediatric care is being provided at tertiary care pediatric centers by pediatric anesthesiologists. This is good news for pediatric patients receiving care at these tertiary care centers. The complication rates for surgeons, anesthesiologists, and medical centers decrease as case volume increases.9–12 Even for low complexity procedures such as appendectomies, the odds ratio of complications is lowest for tertiary care pediatric centers.10 This would appear to be an attractive benefit of this shifting health care landscape. However, if this shift is representative of a larger national pattern, there are likely other consequences (ie, reduced access to pediatric care in rural environments, pediatric skill regression for physicians at nontertiary care centers, and the need to modify fellowship training models). In the United States, there already exists significant heterogeneity in the geographic distribution of pediatric anesthesiologists.1 The vast majority of pediatric anesthesiology consultants and tertiary care centers are located in urban environments, and 70% of the pediatric population (0–17 years) live within 25 miles of a pediatric anesthesiologist. Unfortunately, this is not the case for the remaining proportion of the pediatric population. Lorenz curves—originally designed to describe the uniformity of wealth distribution—can also be applied to define the degree of uniformity of health care distribution of medical specialties. In 2016, Muffly et al2 used Lorenz curves and the Gini coefficient to demonstrate health care disparities. Gini coefficients have a value between 0 and 1. The closer the coefficient approaches “0,” the greater uniformity there is in the distribution of a resource (less disparity). Likewise, as the coefficient approaches “1,” the more nonuniform the distribution of the resource (more disparity) is. The authors demonstrated significantly greater nonuniformity for both pediatric anesthesiologists and surgeons (Gini coefficients 0.32 and 0.33, respectively) when compared to all anesthesiologists and pediatricians (0.19 and 0.18, respectively). This analysis demonstrates that access to perioperative pediatric specialists (anesthesiologists and surgeons) is difficult for a large portion of the pediatric population. What is the driver for this geographic disparity? Specialists tend to work in tertiary care centers, and this may preordain them to more urban environments. That said, this does not seem to be the case for neonatologists (Gini coefficient = 0.13). Regardless of the etiology, the most recent data from California suggest that access will become even more challenging as care continues to move to tertiary care facilities which are predominantly located in urban environments. The impact of the ACS CSV Quality Improvement Program may also contribute to this disparity. Many hospitals in more rural locations may not be able to meet the ACS CSV guideline requirements and this may result in a further shift away from nontertiary care centers. For a given case volume, as pediatric procedural volume increases at tertiary care centers, it declines in the nontertiary referral centers. This has implications for maintenance of clinical skills. Skill regression for physicians at nonreferral care centers is a very real possibility if their pediatric anesthetizing volume declines. Case volume is strongly associated with the incidence of adverse outcomes. Anesthesia complication rates have been shown to be higher for pediatric patients cared for by physicians with reduced pediatric case volume.11,12 It is not clear what minimum volume is required to maintain pediatric clinical skills, but the ACS CSV Quality Improvement Program states that an anesthesiologist “must demonstrate continuous experience with children younger than 24 months, defined as 25 patients per anesthesiologist per year.”13 This metric will be more difficult to achieve for some physicians as patients move to tertiary care centers. What implications this shifting landscape has on fellowship education is also unclear. Is there a way to augment fellowship training to address the needs of the fellow and the changing landscape of pediatric health care? Can the disparity of access to pediatric anesthesiologists be addressed in fellowship training? Simply training more pediatric anesthesiologists will not address the geographic disparity. Geographic location is a strong driver for choosing a practice location and it may be stronger than financial incentives. Ideally, a fellowship prepares the graduate to be a consultant, so they have the skill sets to be leaders in their field to advance their career and the specialty. The changing landscape of pediatric anesthesia care strongly suggests that we need pediatric specialists that can work in academic tertiary care environments, and we also need pediatric specialists that can work in a community mixed practice. The needs of these trainees might be quite disparate, and building some flexibility into training programs to individualize the training experience to satisfy the needs of the graduates will be important. Having a fellowship that addresses clinical skills, scholarship, and “individualization” might address some of these needs. If we look at other pediatric medicine subspecialties, we see that the concept of “individualization” is built into their training programs. In fact, when the pediatric hospital medicine fellowship was being created, they intentionally created an “individualized” component to address the needs and aspirations of the fellow.14 As we consider the workforce landscape, we need to consider the factors that affect supply and demand. We commend Muffly et al5 for enriching our understanding with this most recent analysis. Their data demonstrate that inpatient pediatric procedural volume in a large state in the United States is flat or decreasing over the past 2 decades. It also tells us that this volume is moving to tertiary care centers. However, the findings highlight the complexity and dynamic nature of this study. Putting the whole workforce picture together is complicated by the inability to analyze all the pertinent variables. Namely, we still do not have a good appreciation of how the pediatric ambulatory volume has changed nationally over the past 20 years. This information is either missing or incomplete in most procedural databases and may contribute to a considerable volume of pediatric care. It is imperative that ambulatory data be added to this workforce analysis. Without it, our conclusions are limited. If Muffly et al5 choose to investigate these trends nationally, it will be imperative that they try to identify ambulatory data. Given the degree of complexity and the changing nature of this study, periodic reviews of the supply and demand of the pediatric workforce must take place in a structured and organized fashion. It will have important implications for our specialty, and this information will be vital for several stakeholders in our society (eg, the Pediatric Anesthesia Program Directors Association, Pediatric Anesthesia Leadership Council, and the American Society of Anesthesiologists). DISCLOSURES Name: Franklyn P. Cladis, MD, FAAP. Contribution: This author helped write and edit the manuscript. Name: Peter J. Davis, MD. Contribution: This author helped write and edit the manuscript. Name: James A. DiNardo, MD, FAAP. Contribution: This author helped edit the manuscript. This manuscript was handled by: Thomas R. Vetter, MD, MPH.