- Email: [email protected]
Distance to care and relative supply among pediatric surgical subspecialties
Journal of Pediatric Surgery (2009) 44, 483–495
www.elsevier.com/locate/jpedsurg
Original articles
Distance to care and relative supply among pediatric surgical subspecialties Michelle L. Mayera,b , Heather A. Beilb,⁎, Daniel von Allmenc a
Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA b Department of Health Policy and Management, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA c Department of Surgery, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA Received 4 April 2008; revised 8 August 2008; accepted 11 August 2008
Key words: Pediatric subspecialty care; Workforce; Geographic access to care
Abstract Background/Purpose: The aim of this study is to describe geographic proximity to and quantify relative supply of 7 pediatric surgical specialties in the United States. Methods: Data from the 2005 American Medical Association Physician Masterfile and the Claritas Pop-Facts Database were used to calculate subspecialty-specific, population-weighted, straight-line distances between each zip code centroid and the nearest provider. These same data sources were used to calculate the percentage of hospital referral regions with a provider, the percentage of the younger than 18 years population living within selected distances of providers, and provider-to-population ratios for each of the pediatric surgical subspecialties. Further, we calculated the correlation between practice locations and children's hospitals offering pediatric surgical services. Results: Across pediatric surgical specialties, average distances to the nearest provider ranged from 27.1 miles for pediatric surgery to 100.9 miles for pediatric cardiothoracic surgery. The average population-weighted distance to a provider was less than 30 miles for pediatric surgery and pediatric ophthalmology only. For 5 of the 7 pediatric surgical specialties studied, approximately one quarter of the younger than 18 years population lives more than 1-hour drive from a provider. Provider–to– younger than 18 years population ratios range across hospital referral region from 0.04 per 100,000 for pediatric cardiothoracic surgery to 0.97 per 100,000 for pediatric surgery. The correlation between pediatric surgeons and children's hospitals offering services was 0.72. Conclusions: Although the practice locations of pediatric surgical subspecialties parallel the geographic distribution of children in the United States, large percentages of the younger than 18 years population must travel long distance to receive care from these providers. Large coefficients of variation reveal substantial maldistribution. These findings lay the groundwork for workforce assessments of the pediatric surgical subspecialties and underscore the need for future studies that assess access barriers for children in need of surgical care. © 2009 Elsevier Inc. All rights reserved.
⁎ Corresponding author. Tel.: +1 919 966 7350; fax: +1 919 966 6961. E-mail address: [email protected] (H.A. Beil). 0022-3468/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2008.08.015
484 The purpose of this is study is to describe geographic proximity to and quantify relative supply of 7 pediatric surgical specialties in the United States. The study findings lay the groundwork for workforce assessments of the pediatric surgical subspecialties and underscore the need for future studies that assess the extent to which the current supply and distribution of these providers act as access barriers for children in need of surgical care.
1. Background/purpose Pediatric surgical subspecialization has been associated with lower complication rates among children with perforated appendixes [1], decreased risk of mucosal perforation post pyloromyotomy, and greater precision in tumor removal [2]. When compared to children treated by general surgeons, studies have demonstrated lower costs and/or shorter lengths of stay among younger children treated by pediatric surgical specialists for appendectomy [3], uncomplicated pyloromyotomy [3], ureteroneocystostomy [4], and appendectomy complicated by perforation [1,3]. Despite the potential
M.L. Mayer et al. benefits of pediatric surgical care, little is known about geographic access to pediatric surgical care. Workforce policies based on information from adultoriented specialties are often applied to pediatric specialties despite the dissimilarities in these fields in terms of disease prevalence, patient demand, and workforce size. Without detailed data on the availability of pediatric surgical specialties, policy makers and medical educators cannot make informed policies about the pediatric surgical workforce. Yet, only a few studies have assessed the pediatric surgical workforce. Every 5 years for the past 30 years, the field of general pediatric surgery has performed an extensive national study to assess the supply of pediatric surgeons in the United States [5]. The 1995 survey found that pediatric surgeons were located in every standard metropolitan statistical area in the United States with a population of 200,000 or more and in several smaller standard metropolitan statistical areas [6]. A follow-up survey in 2000 found that the ratio of pediatric surgeons to the younger than 18 years population ranged from 1:99,000 to less than 1:200,000 across states [5]. Generally, the estimated need for pediatric surgeons based on population growth was consistent with the
Fig. 1 Locations of pediatric surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties number of trainees in these surveys. Recent surveys have noted trends in pediatric surgeons moving into nonchildren's hospitals and gravitating to regions with smaller populations [5,7]. These same surveys have also shown a narrowing of practice scope as pediatric surgeons compete with other pediatric surgical subspecialists. The recommendation for the past decade has been to control the production of pediatric surgeons through training reductions and the elimination of foreign-trained physicians [5]. Pediatric surgical subspecialties have not been studied as systematically as general pediatric surgery. A study of the pediatric otolaryngology workforce compared the supply of these physicians with a national estimate of need and concluded that pediatric otolaryngology programs should curtail the number of trainees [8]. In contrast, a survey of otolaryngology department chairpersons found that pediatric otolaryngology was expected to experience a growth in open positions from 1999 to 2003 [9], and an FOPE II study of pediatric otolaryngologists found that pediatric otolaryngologists were significantly more likely than general otolaryngologists to report increases in the number of referrals [10]. Other specialty-specific studies have found that caseload was
485
described as “too little” by 23% of pediatric cardiac surgeons vs 20% and 29% of adult cardiac and general thoracic surgeons, respectively [11]; a trend toward increased demand for pediatric neurosurgeons based on increases in the number of advertised positions [12]; and conflicting findings about the adequacy of the pediatric urology workforce [13,14]. These studies suggest that there is little consensus over the adequacy of the various pediatric surgical subspecialty workforces. Few of the existing studies include assessments of geographic access to care. Consequently, we know little about how the availability of these providers varies across the country. To address this gap, we quantify and describe geographic access to pediatric surgical care in United States. The findings provide a foundation for future studies that explore the relationship between geographic access to pediatric surgical subspecialty care and child health outcomes such as variations in procedure rates across locales, volume-outcomes studies for common pediatric surgical procedures, and interspecialty differences in surgical outcomes. This study uses physician data from the 2005 American Medical Association (AMA) Physician Masterfile and zip
Fig. 2 Locations of pediatric orthopedic surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
486 code level data from Claritas [15], a proprietary database, to generate specialty-specific estimates for each of the following measures: 1. the population weighted average distance to a provider; 2. the percent of children who live within relatively short distances of a provider (ie, 40 miles); 3. the percentage of hospital referral regions (HRRs) that have providers; and 4. provider-to-pediatric population ratios in HRR. Study data were also used to produce maps that depict the practice locations of pediatric surgical specialists in the United States.
2. Methods 2.1. Data sources We use physician data from 2005 AMA Physician Masterfile. The Masterfile contains current and historical
M.L. Mayer et al. data on more than 800,000 physicians, including both AMA members and nonmembers, residing in the United States who have completed the educational requirements necessary to be recognized as physicians in the United States (http://www.ama-assn.org/ama/pub/category/2673. html). In addition, we used data from the National Association of Children's Hospitals and Related Institutions (www.childrenshospitals.net) to obtain the number of children's hospitals who offer pediatric surgery services. We studied 7 pediatric surgical subspecialties: pediatric surgery, pediatric orthopedic surgery, pediatric ophthalmology, pediatric otolaryngology, pediatric urology, pediatric neurosurgery, and pediatric cardiothoracic surgery. For comparison purposes, we also performed analyses on general surgery, orthopedic surgery, ophthalmology, otolaryngology, neurosurgery, urology, and thoracic surgery. The Institutional Review Board at the University of North Carolina at Chapel Hill approved the study. The analyses included only providers currently active in direct patient care. Trainees were excluded from all analyses. Provider specialty was identified using self-reported primary and secondary specialty from the AMA Masterfile. A provider might be included in more than 1 specialty if they
Fig. 3 Locations of pediatric otolaryngology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties listed any 2 of the surgical specialties of interest as their primary and secondary specialties. Sensitivity analyses using only the primary specialty were also performed and are available upon request from the authors.
2.2. Practice location Practice zip codes were used to produce maps for each pediatric surgical subspecialty using MapInfo 7.0. The maps formed the basis for further analyses of distance and markets. In each map, a circle denotes the presence of a pediatric subspecialist, and the size of the circle indicates whether there are less than 5, between 5 and 10, or more than 10 pediatric subspecialists in each zip code. A map was also generated that has a circle for each children's hospital offering pediatric surgery services in each zip code.
2.3. Distance to care Straight-line distance from residence to the nearest provider is a commonly used measure of travel impedance
487
[16] and accounts for nearly all the variation in actual travel time [17]. For each pediatric surgical specialty, the straightline distances between each zip code in the United States and the nearest specialist were calculated using latitude and longitude data for the zip code centroids. Zip codes in which providers are located were assigned a distance estimate of 0. Zip code level estimates of the pediatric population were obtained from Claritas and merged to distance estimates using the provider's zip code. To generate national estimates of the average distance to a subspecialist, the distances between each zip code and the nearest pediatric subspecialist were averaged across all zip codes for each subspecialty, using the younger than 18 years population as a weight. To date, no individual or group has established the desired distance between patients and providers of subspeciality care. In a 1980 report, the Graduate Medical Education National Advisory Committee recommended 95% of the population have a maximum travel time of 90 minutes for surgical care but provided no specific recommendation about travel times to pediatric surgical subspecialty services [18]. In the absence of any universally
Fig. 4 Locations of pediatric ophthalmology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
488 accepted threshold, zip codes were classified as being within 10 miles, 11 to 20 miles, 21 to 40 miles, 41 to 80 miles, or 80 or more miles of a provider. The thresholds were used assuming that 20 miles corresponds to a drive time of 30 minutes [19]. Specialty-specific percentages of the overall younger than 18 years population living within these distances of a provider were then calculated.
2.4. Defining market areas With the exception of neonatology [20], past studies have not generated market areas for pediatric subspecialty physicians. A reasonable conceptualization of market area for these subspecialties is an area in which inflows and outflows of patients are minimized [21]. Hospital referral regions, developed by the Dartmouth Atlas of Health Care, are mutually exclusive market areas based on Medicare data on referrals for specialized care [22]. They have not been validated for use in children but may serve as a reasonable proxy for pediatric subspecialty market areas. For all market-level analyses, therefore, data were aggregated to the HRR (n = 306).
M.L. Mayer et al.
2.5. Supply relative to pediatric population Physician-to-population ratios allow straightforward, gross comparisons of supply across services areas [16]. Specialty-specific estimates of percent of HRR with a provider, the average ratio of pediatric surgical subspecialists to pediatric population, and the coefficients of variation of the ratios were estimated. The coefficient of deviation 100 and is a measure of variation equals standard mean dispersion. A large standard deviation of the physician-topopulation ratio, relative to the mean, indicates that physician-to-population ratios vary considerably across markets. Coefficients of variation can be compared across subspecialties to assess relative dispersion. In addition, the correlation between the number of children's hospitals offering pediatric surgery and the number of pediatric surgeons with an HRR was calculated.
2.6. General surgery and related subspecialties All of the analyses described in the previous paragraphs were also performed analyses on general surgery, orthopedic
Fig. 5 Locations of pediatric cardiothoracic surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties
489
Fig. 6 Locations of pediatric urology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
surgery, ophthalmology, otolaryngology, neurosurgery, urology, and thoracic surgery. These calculations also use the HRR as a market area but use the total population as a measure of demand.
3. Results Maps depicting the provider locations for each of the pediatric surgical specialties of interest are presented in Figs. 1-8. Pediatric surgery and pediatric ophthalmology are the most geographically accessible of the 7 pediatric surgical specialties studied. On average, a child in the United States must travel 27.1 miles to the nearest pediatric surgeon and 28.5 miles to the nearest pediatric ophthalmologist (Table 1). Average distances to the nearest pediatric orthopedic and urologist are 32.6 and 40.0 miles, respectively, whereas average distances to pediatric otolaryngologists, neurosurgeons, and cardiothoracic surgeons are far greater. In contrast, average distance to general surgery and surgical specialties are quite short. The population weighted average
distance to care ranges from 3.5 miles for general surgery to 11.8 miles for neurosurgery. For most pediatric surgical specialties, one half of the pediatric population live within 20 miles of the nearest provider (Table 2). Geographic access is greatest for pediatric surgery for which 93% of the younger than 18 years population live within 80 miles, or an approximately 2 hour drive, of a pediatric surgeon. Likewise, more than 82% of the younger than 18 years population also live within 80 miles of a pediatric ophthalmologist and pediatric orthopedist, respectively. Slightly more than one half of the younger than 18 years population lives within 80 miles of a pediatric cardiothoracic surgeon. Access to general surgery and surgical specialties is far greater than among pediatric surgical specialties with virtually the entire population living within 80 miles of a general surgeon and each of the surgical specialties studied. Viewing access to pediatric surgical care from a market perspective, we note that all HRR have one or more general surgeon and surgical specialists. In contrast, approximately one half of HRR have a pediatric surgeon, pediatric orthopedist, or pediatric ophthalmologist. Far fewer HRR have a pediatric specialist in otolaryngology, cardiothoracic
490
M.L. Mayer et al.
Fig. 7 Locations of pediatric neurosurgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
surgery, urology, or neurosurgery available. Likewise, the relative supply of general surgeons and surgical subspecialists is far greater than among pediatric surgeons and pediatric surgical specialists. Across pediatric subspecialties provider–to–younger than 18 years population ratios range from 0.04 per 100,000 for pediatric cardiothoracic surgery to 0.97 per 100,000 for pediatric surgery. The correlation between the number of children's hospitals offering pediatric surgery services and the number of pediatric surgeons within an HRR was 0.72. Among pediatric surgical specialties, coefficients of variation range from a low of 166 for pediatric urology to a high of 342 for pediatric cardiothoracic surgery. Coefficients of variation are generally smaller for general surgery and surgical specialties than among the pediatric surgical specialties. In those markets that have providers, the average number of children younger than 18 years per provider ranges from slightly more than 100,000 for pediatric surgery to 2.3 million for pediatric cardiothoracic surgery. Population-toprovider ratios for general surgery and surgical specialties, in contrast, range from 8303 persons per general surgeon to 59,480 persons per neurosurgeon (Table 3).
4. Discussion Geographic access is one of many determinants of overall access to care. This study demonstrates that average distances to pediatric surgical care are greater than for the general surgery and surgical specialties. Naturally, we expect greater distances to care for pediatric surgical specialists given the generally low prevalence of conditions treated by many of these providers. Low prevalence rates limit the demand for and the financial viability of their pediatric surgical subspecialty practice in smaller markets where they compete with general surgical specialists. The extent to which these greater distances to care act as an access barrier is not known. Average distances to pediatric surgical specialists are generally greater than for pediatric medical subspecialties [23]. Compared to pediatric medical subspecialty care, long distances to pediatric surgical care may not have as much effect on access unless there is long-term follow-up by the surgeon and his or her team required. In many cases, the condition requiring surgical care may involve visits to the surgical provider for a defined and limited period. Families living distant from pediatric surgical specialists may
Distance to care and relative supply among pediatric surgical subspecialties
491
Fig. 8 Locations of children's hospitals offering pediatric surgery by zip code. Dots are mapped to zip code centroid. Source: The National Association of Children's Hospitals and Related Institutions (NACHRI), www.childrenshospitals.net. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
relocate temporarily for surgical care, living at Ronald McDonald Houses or other residential facilities. For families with limited financial means or limited sick time benefits, long distances to care may create substantial barriers, however. The widespread availability of general surgeons and surgical specialists, whose training and practice includes the care of both adult and pediatric populations, ensures the availability of surgical services to the overwhelming majority of the pediatric population. Indeed, there is evidence that general surgical specialists, like general otolaryngologists, are heavily involved in the care of children [10]. Several studies suggest, however, that pediatric surgical providers may have better outcomes and lower costs than their generalist peers [1-3]. As such, greater distances to pediatric surgical specialists may have quality implications. The few studies that have compared generalists and pediatric providers have several limitations; however, so more research is needed to determine whether there is a quality gap. Furthermore, the aging of the population may lead to greater demand for the services of general surgeon and surgical specialists, potentially decreasing the availability of pediatric surgical care in
general surgical practices. More research into how population dynamics and provider supply influence general surgeons and surgical specialists' willingness to care for children and outcomes of care is needed. More than one half of HRRs lack pediatric surgical specialists. These findings may indicate that HRRs do not represent markets for specialized pediatric care. Given the rarity of many of the pediatric conditions that require surgical intervention, very large population bases may be required to support a pediatric surgical specialist. Thus, the relevant market for a pediatric surgical specialist may encompass an entire state or portions of multiple states. Average provider to population ratios are less then 1 per 100,000 children for all pediatric surgical specialties. The extent to which this represents an adequate supply is unknown. High coefficients of variation suggest, however, that supply ratios vary considerably across markets and likely indicate distributional inequities. Moreover, the supply of pediatric surgeons was highly correlated with the number of children's hospitals offering pediatric surgical services within an HRR. Therefore, as would be expected, pediatric surgeons are located in areas with children's hospitals that offer pediatric surgical services.
492
M.L. Mayer et al.
Table 1 Population weighted average distance to the nearest provider by surgical specialty, US zip codes with any children (n = 30,091) No. of Miles to the nearest provider providers Mean 75 95 percentile percentile Surgical subspecialty General surgery 30124 Otolaryngology 9058 Neurosurgery 4500 Ophthalmology 17275 Orthopedic surgery 19606 Thoracic surgery 5106 Urology 9435 Pediatric surgical subspecialty Pediatric surgery 781 Pediatric 266 otolaryngology Pediatric 113 neurosurgery Pediatric 444 ophthalmology Pediatric orthopedic 466 surgery Pediatric 49 cardiothoracic surgery Pediatric urology 276
3.4 6.8 11.8 5.6 5.0 10.6 6.9
4.6 8.2 14.0 7.0 6.2 12.6 8.4
14.7 26.1 43.3 22.9 20.3 38.2 25.4
27.1 48.5
34.3 59.8
95.0 177.4
67.4
79.7
198.8
28.5
37.9
96.3
32.6
40.0
104.6
100.9 130.4
277.4
39.9
47.7
142.7
This further suggests that there may be regionalization of care, and the location of children's hospitals may play a key role in the location of pediatric surgical providers. The current supply of pediatric surgical specialists may not be adequate to allow diffusion into more markets. Proliferation
of positions at community children's hospitals [24] suggests an increased demand for these providers outside of academia, however. There is a tension in determining the ideal supply of pediatric surgical providers. Insufficient supply of providers may limit access, but oversupply may dilute provider experience by leading to a smaller number of procedures performed per provider. Studies have demonstrated a higher mortality rate for children treated by lower volume hospitals and surgeons for pediatric cardiac surgery [25-28]. Among general surgical specialties, there is a well-established relationship between volume and outcomes [29-32]. Therefore, diluted experience may diminish quality of care [5]. There is also concern that the growth in pediatric surgical specialties, like pediatric otolaryngology and pediatric urology, has led to a narrowing of scope of practice among pediatric surgeons because of competition from other pediatric surgical specialists and nonsurgical specialists such as interventional radiology [6,24]. Thus, both oversupply and undersupply of pediatric surgical specialists can have detrimental consequences. The pediatric population size required to support a pediatric surgical specialist may be quite high given the relatively low prevalence of pediatric conditions requiring surgical intervention. Estimates of the pediatric population size required to support a pediatric surgical specialist and recommended provider-to-population ratios would help to inform pediatric surgical workforce policy. Furthermore, expected changes in demand for such providers should be considered in workforce models. Advances in selected fields, such as fetal surgery, as well as the application of existing surgical techniques, such as bariatric surgery, to younger populations may lead to increases in demand. Planning to meet expected increases
Table 2 Percent of younger than 18 years population within selected distances of physicians by surgical specialty, US zip codes with any children (n = 30,091) Distance to the nearest provider
General surgery and surgical specialties General surgery
Otolaryngology
Neurosurgery
Ophthalmology
Orthopedics
Thoracic
Urology
10 or fewer miles 11-20 miles 21-40 miles 41-80 miles More than 80 miles
89.1 8.6 2.0 0.2 0.0
79.3 12.3 6.6 1.5 0.3
67.6 14.9 11.5 4.9 1.1
82.3 11.0 5.4 1.1 0.2
84.6 10.3 4.2 0.8 0.1
69.5 15.5 10.5 3.5 1.0
79.2 12.8 6.2 1.5 0.3
Distance to the nearest provider
Pediatric surgery and pediatric surgical specialties Pediatric surgery
Pediatric otolaryngology
Pediatric neurosurgery
Pediatric ophthalmology
Pediatric orthopedics
Pediatric cardiothoracic
Pediatric urology
10 or fewer miles 11-20 miles 21-40 miles 41-80 miles More than 80 miles
43.6 18.9 16.0 14.3 7.2
29.5 19.8 16.1 16.3 18.4
21.0 17.9 18.9 17.3 24.9
42.1 17.9 16.2 15.3 8.4
39.4 19.2 16.5 16.2 8.8
14.2 12.0 12.5 18.3 43.0
34.0 19.9 16.8 16.5 12.8
Distance to care and relative supply among pediatric surgical subspecialties Table 3
Relative supply of surgical providers, United States HRRs (n = 306)
Surgical subspecialty
General Orthopedic Ophthalmology Otolaryngology Thoracic Urology Neurosurgery
Pediatric Pediatric Pediatric Pediatric Pediatric Pediatric Pediatric a
493
surgery orthopedic ophthalmology otolaryngology cardiothoracic urology neurosurgery
% of HRR with a provider
Provider: population (in 100,000)
Average no. of people per physician a
Mean
SD
COV
100 100 100 100 100 100 100
12.04 7.18 6.25 3.39 1.91 3.67 1.68
33.70 8.17 14.20 7.36 2.61 9.20 1.91
280 114 227 217 137 250 113
% of HRR with a provider
Provider: younger than 18 years population (in 100,000)
56 53 50 34 13 45 25
Mean
SD
Coefficient of Variation
0.97 0.62 0.57 0.27 0.04 0.37 0.12
3.21 1.83 1.82 0.78 0.15 0.62 0.31
333 296 321 291 342 166 248
8303 13,925 15,998 29,517 52,327 27,217 59,480 Average no. of children per physician a
103,604 161,826 176,874 370,883 2,295,999 266,925 801,020
In those HRR that have at least 1 provider.
in demand must be balanced to prevent oversupply that can lead to decreases in per provider procedure volume, which can undermine surgeon's ability to remain skillful in a variety of procedures.
4.1. Limitations The major limitation of our analysis is that it is based on self-reported data from the AMA Masterfile. Because only pediatric surgery currently offers a certification examination, there are few alternatives to the AMA Masterfile that can be used to identify pediatric surgical specialists. In a technical appendix available upon request, we compared aggregate numbers from the AMA Masterfile and the American Board of Medical Specialties for pediatric surgery and found that the number of physicians self-reporting as pediatric surgeons was higher than the number reported as certified by American Board of Medical Specialties but similar to those reported in studies done by the Study on US Surgical Services [5]. Likewise, numbers for general and pediatric otolaryngologists are comparable to past reports [10,33]. Similar discrepancies between the AMA Masterfile and certification data have been previously reported for pediatric medical subspecialties and may reflect delays between initiation of practice and receipt of certification as well as practicing in a field despite lack of board eligibility and certification. In addition, we compared the number of self-designated providers based on AMA with professional society data where it was available. The total number of self-designated
pediatric urologists, pediatric orthopedists, and pediatric neurosurgeons based on the AMA Masterfile were quite comparable to the number of US members reported by their respective professional organizations. Consequently, we expect that our results for these specialties are likely reflective of the actual number of practicing providers for these specialties. The number of members the American Association for Pediatric Ophthalmology and Strabismus exceeded the number of self-designated providers on the AMA file by a large percentage, however. As such, our estimates of the distance to the nearest provider as well as the relative supply of these providers may be inaccurate. We used both primary and secondary specialty to identify providers in each surgical field studied, which may lead to an overestimate of the number of providers. We performed sensitivity analyses to estimate distances and relative supply using only primary specialty to identify providers. We found that distance estimates were virtually unchanged for general surgery and surgical specialties; for pediatric surgery and pediatric surgical specialties, average population weighted distances increased by 1.8 miles for pediatric surgery and 30.4 miles for pediatric cardiothoracic surgery. Average population weighted distance increased dramatically for pediatric neurosurgery from 67.4 to 132.45 miles. Likewise, using the primary specialty only decreased the estimated percentage of the pediatric population living within 10 miles of a provider and increased the percentage of the population living more than 80 miles from a provider across all pediatric surgical specialties with the exception of pediatric surgery. As expected, all measures of relative supply are influenced
494 by whether primary and secondary specialty are used to identify providers vs primary specialty only. Given that nationwide supply estimates for the pediatric subspecialties based or primary and secondary specialty align more closely with memberships in various pediatric surgical specialty association, we expect that estimates based on primary specialty alone understate the actual supply of these providers and that the estimates presented here are a more valid reflection of the actual number of providers. Another limitation is our inability to account for satellite clinics. Some pediatric surgical specialists may travel to other HRR to provide care on a regular basis. Failure to account for these secondary practice locations may bias our estimates of the supply of providers at the HRR level. Given the size of HRR, especially in the western portions of the United States, we expect that the amount of bias introduced is minimal. Our model also assumes that resources at geographic site of pediatric specialty care can accommodate transfers from other locals. Any local capacity constraints may require patients to travel even farther than our estimates suggest.
5. Conclusion This is the first study to estimate distances to care and relative supply of pediatric surgical specialists in the United States. The study also provides useful comparison data on general surgery and surgical specialties. The findings suggest that a substantial percentage of the pediatric population in the United States must travel one or more hours to reach a pediatric surgical specialist, whereas access to general surgical providers is much more extensive. Furthermore, our findings suggest that pediatric surgical specialties may face many of the distributional challenges plaguing pediatric medical subspecialties. Further research is needed to establish how the current supply and distribution of these providers influence access to and quality of surgical care for children in the United States.
Acknowledgments We would like to thank Bob Schwartz and Ann Howard for performing the distance analyses; Tom Ricketts, PhD, for providing access to the AMA Masterfile and commenting on previous drafts; and George F. Sheldon, MD, for his helpful insights. The research was supported through grant number 1-K02HS013309-01A1 from the Agency for Healthcare Research and Quality and a University Research Council grant from the University of North Carolina at Chapel Hill. The funding agency was not involved in the performance of the study. Doctor Mayer had full access to all of the data in the study and takes responsibility for the integrity of the data and the
M.L. Mayer et al. accuracy of the data analysis. The authors have no conflicts of interest to report.
References [1] Alexander F, Magnuson D, DiFiore J, et al. Specialty versus generalist care of children with appendicitis: an outcome comparison. J Pediatr Surg 2001;36:1510-3. [2] Albright AL, Sposto R, Holmes E, et al. Correlation of neurosurgical subspecialization with outcomes in children with malignant brain tumors. Neurosurgery 2000;47:879-85 [Discussion 885-7]. [3] Kokoska ER, Minkes RK, Silen ML, et al. Effect of pediatric surgical practice on the treatment of children with appendicitis. Pediatrics 2001; 107:1298-301. [4] Snow BW, Catwright PC, Young MD. Does surgical subspecialization in pediatrics provide high-quality, cost-effective patient care? Pediatrics 1996;97:14-7. [5] O'Neill Jr JA, Gautam S, Geiger JD, et al. A longitudinal analysis of the pediatric surgeon workforce. Ann Surg 2000;232:442-53. [6] O'Neill Jr JA, Cnaan A, Altman RP, et al. Update on the analysis of the need for pediatric surgeons in the United States. J Pediatr Surg 1995; 30:204-10 [discussion 211-3]. [7] Parkerton PH, Geiger JD, Mick SS, et al. The market for pediatric surgeons: a survey of recent graduates. J Pediatr Surg 1999;34:931-9. [8] Zalzal GH. Projected societal needs in pediatric otolaryngology. Laryngoscope 1996;106:1176-9. [9] Smith RV. Workforce issues for the academic otolaryngologist. Laryngoscope 2000;110:1602-6. [10] Tunkel DE, Cull WL, Jewett EA, et al. Practice of pediatric otolaryngology: results of the future of pediatric education II project. Arch Otolaryngol Head Neck Surg 2002;128:759-64. [11] Shemin RJ, Dziuban SW, Kaiser LR, et al. Thoracic surgery workforce: snapshot at the end of the twentieth century and implications for the new millennium. Ann Thorac Surg 2002;73:2014-32. [12] Friedlich DL, Feustel PJ, Popp AJ. Workforce demand for neurosurgeons in the United States of America: a 13-year retrospective study. J Neurosurg 1999;90:993-7. [13] Shapiro E, Hatch DA. Pediatric urological manpower report. Pediatric urological manpower committee of the American Association of Pediatric Urology. J Urol 1993;150:675-9. [14] Retik AB, King LR. Pediatric urology manpower report 1995. Ellen Shapiro on behalf of the American Association of Pediatric Urologists. J Urol 1996;156:488-90 discussion 490-1. [15] Claritas Inc. Pop-facts database for ZIP codes. San Diego: Claitas Inc.; 2003. [16] Guagliardo MF. Spatial accessibility of primary care: concepts, methods and challenges. Int J Health Geogr 2004;3:3. [17] Phibbs CS, Luft HS. Correlation of travel time on roads versus straight line distance. Med Care Res Rev 1995;52:532-42. [18] Report of the Graduate Medical Education National Advisory Committee to the Secretary, Department of Health and Human Services. Vol I GMENAC Summary Report. Hyattsville, MD: Health Resources Administration, DHHS Pub. No. (HRA); September 1980. p. 105. [19] Rosenthal MB, Zaslavsky A, Newhouse JP. The geographic distribution of physicians revisited. Health Serv Res 2005;40:1931-52. [20] Goodman DC, Fisher ES, Little GA, et al. Are neonatal intensive care resources located according to need? Regional variation in neonatologists, beds, and low birth weight newborns. Pediatrics 2001;108: 426-31. [21] Morrisey MA, Sloan FA, Valvona J. Defining geographic markets for hospital care. Law Contemp Probl 1988;51:165-94. [22] Guagliardo MF, Jablonski KA, Joseph JG, et al. Do pediatric hospitalizations have a unique geography? BMC Health Serv Res 2004;4:2.
Distance to care and relative supply among pediatric surgical subspecialties [23] Mayer ML. Are we there yet? Distance to care and relative supply among pediatric medical subspecialties. Pediatrics 2006;118:2313-21. [24] Geiger JD, Drongowski RA, Coran AG. The market for pediatric surgeons: an updated survey of recent graduates. Journal of Pediatric Surgery 2003;38:397-405. [25] Jenkins KJ, Newburger JW, Lock JE, et al. In-hospital mortality for surgical repair of congenital heart defects: preliminary observations of variation by hospital caseload. Pediatrics 1995;95:323-30. [26] Hannan EL, Racz M, Kavey RE, et al. Pediatric cardiac surgery: the effect of hospital and surgeon volume on in-hospital mortality. Pediatrics 1998;101:963-9. [27] Sollano JA, Gelijns AC, Moskowitz AJ, et al. Volume-outcome relationships in cardiovascular operations: New york state, 1990-1995. J Thorac Cardiovasc Surg 1999;117:419-28 [discussion 428-30]. [28] Begg CB, Cramer LD, Hoskins WJ, et al. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747-51.
495
[29] Hannan EL, Wu C, Walford G, et al. Volume-Outcome relationships for percutaneous coronary interventions in the stent era. Circulation 2005;112:1171-9. [30] Norton EC, Garfinkel SA, McQuay LJ, et al. The effect of hospital volume on the in-hospital complication rate in knee replacement patients. Health Serv Res 1998;33:1191-210. [31] Thiemann DR, Coresh J, Oetgen WJ, et al. The association between hospital volume and survival after acute myocardial infarction in elderly patients. N Engl J Med 1999;340:1640-8. [32] Edwards EB, Roberts JP, McBride MA, et al. The effect of the volume of procedures at transplantation centers on mortality after liver transplantation. N Engl J Med 1999;341:2049-53. [33] Pillsbury III HC, Cannon CR, Sedory Holzer SE, et al. The workforce in otolaryngology-head and neck surgery: moving into the next millennium. Otolaryngol Head Neck Surg 2000;123: 341-56.
www.elsevier.com/locate/jpedsurg
Original articles
Distance to care and relative supply among pediatric surgical subspecialties Michelle L. Mayera,b , Heather A. Beilb,⁎, Daniel von Allmenc a
Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA b Department of Health Policy and Management, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA c Department of Surgery, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA Received 4 April 2008; revised 8 August 2008; accepted 11 August 2008
Key words: Pediatric subspecialty care; Workforce; Geographic access to care
Abstract Background/Purpose: The aim of this study is to describe geographic proximity to and quantify relative supply of 7 pediatric surgical specialties in the United States. Methods: Data from the 2005 American Medical Association Physician Masterfile and the Claritas Pop-Facts Database were used to calculate subspecialty-specific, population-weighted, straight-line distances between each zip code centroid and the nearest provider. These same data sources were used to calculate the percentage of hospital referral regions with a provider, the percentage of the younger than 18 years population living within selected distances of providers, and provider-to-population ratios for each of the pediatric surgical subspecialties. Further, we calculated the correlation between practice locations and children's hospitals offering pediatric surgical services. Results: Across pediatric surgical specialties, average distances to the nearest provider ranged from 27.1 miles for pediatric surgery to 100.9 miles for pediatric cardiothoracic surgery. The average population-weighted distance to a provider was less than 30 miles for pediatric surgery and pediatric ophthalmology only. For 5 of the 7 pediatric surgical specialties studied, approximately one quarter of the younger than 18 years population lives more than 1-hour drive from a provider. Provider–to– younger than 18 years population ratios range across hospital referral region from 0.04 per 100,000 for pediatric cardiothoracic surgery to 0.97 per 100,000 for pediatric surgery. The correlation between pediatric surgeons and children's hospitals offering services was 0.72. Conclusions: Although the practice locations of pediatric surgical subspecialties parallel the geographic distribution of children in the United States, large percentages of the younger than 18 years population must travel long distance to receive care from these providers. Large coefficients of variation reveal substantial maldistribution. These findings lay the groundwork for workforce assessments of the pediatric surgical subspecialties and underscore the need for future studies that assess access barriers for children in need of surgical care. © 2009 Elsevier Inc. All rights reserved.
⁎ Corresponding author. Tel.: +1 919 966 7350; fax: +1 919 966 6961. E-mail address: [email protected] (H.A. Beil). 0022-3468/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2008.08.015
484 The purpose of this is study is to describe geographic proximity to and quantify relative supply of 7 pediatric surgical specialties in the United States. The study findings lay the groundwork for workforce assessments of the pediatric surgical subspecialties and underscore the need for future studies that assess the extent to which the current supply and distribution of these providers act as access barriers for children in need of surgical care.
1. Background/purpose Pediatric surgical subspecialization has been associated with lower complication rates among children with perforated appendixes [1], decreased risk of mucosal perforation post pyloromyotomy, and greater precision in tumor removal [2]. When compared to children treated by general surgeons, studies have demonstrated lower costs and/or shorter lengths of stay among younger children treated by pediatric surgical specialists for appendectomy [3], uncomplicated pyloromyotomy [3], ureteroneocystostomy [4], and appendectomy complicated by perforation [1,3]. Despite the potential
M.L. Mayer et al. benefits of pediatric surgical care, little is known about geographic access to pediatric surgical care. Workforce policies based on information from adultoriented specialties are often applied to pediatric specialties despite the dissimilarities in these fields in terms of disease prevalence, patient demand, and workforce size. Without detailed data on the availability of pediatric surgical specialties, policy makers and medical educators cannot make informed policies about the pediatric surgical workforce. Yet, only a few studies have assessed the pediatric surgical workforce. Every 5 years for the past 30 years, the field of general pediatric surgery has performed an extensive national study to assess the supply of pediatric surgeons in the United States [5]. The 1995 survey found that pediatric surgeons were located in every standard metropolitan statistical area in the United States with a population of 200,000 or more and in several smaller standard metropolitan statistical areas [6]. A follow-up survey in 2000 found that the ratio of pediatric surgeons to the younger than 18 years population ranged from 1:99,000 to less than 1:200,000 across states [5]. Generally, the estimated need for pediatric surgeons based on population growth was consistent with the
Fig. 1 Locations of pediatric surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties number of trainees in these surveys. Recent surveys have noted trends in pediatric surgeons moving into nonchildren's hospitals and gravitating to regions with smaller populations [5,7]. These same surveys have also shown a narrowing of practice scope as pediatric surgeons compete with other pediatric surgical subspecialists. The recommendation for the past decade has been to control the production of pediatric surgeons through training reductions and the elimination of foreign-trained physicians [5]. Pediatric surgical subspecialties have not been studied as systematically as general pediatric surgery. A study of the pediatric otolaryngology workforce compared the supply of these physicians with a national estimate of need and concluded that pediatric otolaryngology programs should curtail the number of trainees [8]. In contrast, a survey of otolaryngology department chairpersons found that pediatric otolaryngology was expected to experience a growth in open positions from 1999 to 2003 [9], and an FOPE II study of pediatric otolaryngologists found that pediatric otolaryngologists were significantly more likely than general otolaryngologists to report increases in the number of referrals [10]. Other specialty-specific studies have found that caseload was
485
described as “too little” by 23% of pediatric cardiac surgeons vs 20% and 29% of adult cardiac and general thoracic surgeons, respectively [11]; a trend toward increased demand for pediatric neurosurgeons based on increases in the number of advertised positions [12]; and conflicting findings about the adequacy of the pediatric urology workforce [13,14]. These studies suggest that there is little consensus over the adequacy of the various pediatric surgical subspecialty workforces. Few of the existing studies include assessments of geographic access to care. Consequently, we know little about how the availability of these providers varies across the country. To address this gap, we quantify and describe geographic access to pediatric surgical care in United States. The findings provide a foundation for future studies that explore the relationship between geographic access to pediatric surgical subspecialty care and child health outcomes such as variations in procedure rates across locales, volume-outcomes studies for common pediatric surgical procedures, and interspecialty differences in surgical outcomes. This study uses physician data from the 2005 American Medical Association (AMA) Physician Masterfile and zip
Fig. 2 Locations of pediatric orthopedic surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
486 code level data from Claritas [15], a proprietary database, to generate specialty-specific estimates for each of the following measures: 1. the population weighted average distance to a provider; 2. the percent of children who live within relatively short distances of a provider (ie, 40 miles); 3. the percentage of hospital referral regions (HRRs) that have providers; and 4. provider-to-pediatric population ratios in HRR. Study data were also used to produce maps that depict the practice locations of pediatric surgical specialists in the United States.
2. Methods 2.1. Data sources We use physician data from 2005 AMA Physician Masterfile. The Masterfile contains current and historical
M.L. Mayer et al. data on more than 800,000 physicians, including both AMA members and nonmembers, residing in the United States who have completed the educational requirements necessary to be recognized as physicians in the United States (http://www.ama-assn.org/ama/pub/category/2673. html). In addition, we used data from the National Association of Children's Hospitals and Related Institutions (www.childrenshospitals.net) to obtain the number of children's hospitals who offer pediatric surgery services. We studied 7 pediatric surgical subspecialties: pediatric surgery, pediatric orthopedic surgery, pediatric ophthalmology, pediatric otolaryngology, pediatric urology, pediatric neurosurgery, and pediatric cardiothoracic surgery. For comparison purposes, we also performed analyses on general surgery, orthopedic surgery, ophthalmology, otolaryngology, neurosurgery, urology, and thoracic surgery. The Institutional Review Board at the University of North Carolina at Chapel Hill approved the study. The analyses included only providers currently active in direct patient care. Trainees were excluded from all analyses. Provider specialty was identified using self-reported primary and secondary specialty from the AMA Masterfile. A provider might be included in more than 1 specialty if they
Fig. 3 Locations of pediatric otolaryngology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties listed any 2 of the surgical specialties of interest as their primary and secondary specialties. Sensitivity analyses using only the primary specialty were also performed and are available upon request from the authors.
2.2. Practice location Practice zip codes were used to produce maps for each pediatric surgical subspecialty using MapInfo 7.0. The maps formed the basis for further analyses of distance and markets. In each map, a circle denotes the presence of a pediatric subspecialist, and the size of the circle indicates whether there are less than 5, between 5 and 10, or more than 10 pediatric subspecialists in each zip code. A map was also generated that has a circle for each children's hospital offering pediatric surgery services in each zip code.
2.3. Distance to care Straight-line distance from residence to the nearest provider is a commonly used measure of travel impedance
487
[16] and accounts for nearly all the variation in actual travel time [17]. For each pediatric surgical specialty, the straightline distances between each zip code in the United States and the nearest specialist were calculated using latitude and longitude data for the zip code centroids. Zip codes in which providers are located were assigned a distance estimate of 0. Zip code level estimates of the pediatric population were obtained from Claritas and merged to distance estimates using the provider's zip code. To generate national estimates of the average distance to a subspecialist, the distances between each zip code and the nearest pediatric subspecialist were averaged across all zip codes for each subspecialty, using the younger than 18 years population as a weight. To date, no individual or group has established the desired distance between patients and providers of subspeciality care. In a 1980 report, the Graduate Medical Education National Advisory Committee recommended 95% of the population have a maximum travel time of 90 minutes for surgical care but provided no specific recommendation about travel times to pediatric surgical subspecialty services [18]. In the absence of any universally
Fig. 4 Locations of pediatric ophthalmology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
488 accepted threshold, zip codes were classified as being within 10 miles, 11 to 20 miles, 21 to 40 miles, 41 to 80 miles, or 80 or more miles of a provider. The thresholds were used assuming that 20 miles corresponds to a drive time of 30 minutes [19]. Specialty-specific percentages of the overall younger than 18 years population living within these distances of a provider were then calculated.
2.4. Defining market areas With the exception of neonatology [20], past studies have not generated market areas for pediatric subspecialty physicians. A reasonable conceptualization of market area for these subspecialties is an area in which inflows and outflows of patients are minimized [21]. Hospital referral regions, developed by the Dartmouth Atlas of Health Care, are mutually exclusive market areas based on Medicare data on referrals for specialized care [22]. They have not been validated for use in children but may serve as a reasonable proxy for pediatric subspecialty market areas. For all market-level analyses, therefore, data were aggregated to the HRR (n = 306).
M.L. Mayer et al.
2.5. Supply relative to pediatric population Physician-to-population ratios allow straightforward, gross comparisons of supply across services areas [16]. Specialty-specific estimates of percent of HRR with a provider, the average ratio of pediatric surgical subspecialists to pediatric population, and the coefficients of variation of the ratios were estimated. The coefficient of deviation 100 and is a measure of variation equals standard mean dispersion. A large standard deviation of the physician-topopulation ratio, relative to the mean, indicates that physician-to-population ratios vary considerably across markets. Coefficients of variation can be compared across subspecialties to assess relative dispersion. In addition, the correlation between the number of children's hospitals offering pediatric surgery and the number of pediatric surgeons with an HRR was calculated.
2.6. General surgery and related subspecialties All of the analyses described in the previous paragraphs were also performed analyses on general surgery, orthopedic
Fig. 5 Locations of pediatric cardiothoracic surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Distance to care and relative supply among pediatric surgical subspecialties
489
Fig. 6 Locations of pediatric urology surgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
surgery, ophthalmology, otolaryngology, neurosurgery, urology, and thoracic surgery. These calculations also use the HRR as a market area but use the total population as a measure of demand.
3. Results Maps depicting the provider locations for each of the pediatric surgical specialties of interest are presented in Figs. 1-8. Pediatric surgery and pediatric ophthalmology are the most geographically accessible of the 7 pediatric surgical specialties studied. On average, a child in the United States must travel 27.1 miles to the nearest pediatric surgeon and 28.5 miles to the nearest pediatric ophthalmologist (Table 1). Average distances to the nearest pediatric orthopedic and urologist are 32.6 and 40.0 miles, respectively, whereas average distances to pediatric otolaryngologists, neurosurgeons, and cardiothoracic surgeons are far greater. In contrast, average distance to general surgery and surgical specialties are quite short. The population weighted average
distance to care ranges from 3.5 miles for general surgery to 11.8 miles for neurosurgery. For most pediatric surgical specialties, one half of the pediatric population live within 20 miles of the nearest provider (Table 2). Geographic access is greatest for pediatric surgery for which 93% of the younger than 18 years population live within 80 miles, or an approximately 2 hour drive, of a pediatric surgeon. Likewise, more than 82% of the younger than 18 years population also live within 80 miles of a pediatric ophthalmologist and pediatric orthopedist, respectively. Slightly more than one half of the younger than 18 years population lives within 80 miles of a pediatric cardiothoracic surgeon. Access to general surgery and surgical specialties is far greater than among pediatric surgical specialties with virtually the entire population living within 80 miles of a general surgeon and each of the surgical specialties studied. Viewing access to pediatric surgical care from a market perspective, we note that all HRR have one or more general surgeon and surgical specialists. In contrast, approximately one half of HRR have a pediatric surgeon, pediatric orthopedist, or pediatric ophthalmologist. Far fewer HRR have a pediatric specialist in otolaryngology, cardiothoracic
490
M.L. Mayer et al.
Fig. 7 Locations of pediatric neurosurgeons by zip code. Dots are mapped to zip code centroid. Source: American Medical Association Physician Masterfile, October 31, 2005. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
surgery, urology, or neurosurgery available. Likewise, the relative supply of general surgeons and surgical subspecialists is far greater than among pediatric surgeons and pediatric surgical specialists. Across pediatric subspecialties provider–to–younger than 18 years population ratios range from 0.04 per 100,000 for pediatric cardiothoracic surgery to 0.97 per 100,000 for pediatric surgery. The correlation between the number of children's hospitals offering pediatric surgery services and the number of pediatric surgeons within an HRR was 0.72. Among pediatric surgical specialties, coefficients of variation range from a low of 166 for pediatric urology to a high of 342 for pediatric cardiothoracic surgery. Coefficients of variation are generally smaller for general surgery and surgical specialties than among the pediatric surgical specialties. In those markets that have providers, the average number of children younger than 18 years per provider ranges from slightly more than 100,000 for pediatric surgery to 2.3 million for pediatric cardiothoracic surgery. Population-toprovider ratios for general surgery and surgical specialties, in contrast, range from 8303 persons per general surgeon to 59,480 persons per neurosurgeon (Table 3).
4. Discussion Geographic access is one of many determinants of overall access to care. This study demonstrates that average distances to pediatric surgical care are greater than for the general surgery and surgical specialties. Naturally, we expect greater distances to care for pediatric surgical specialists given the generally low prevalence of conditions treated by many of these providers. Low prevalence rates limit the demand for and the financial viability of their pediatric surgical subspecialty practice in smaller markets where they compete with general surgical specialists. The extent to which these greater distances to care act as an access barrier is not known. Average distances to pediatric surgical specialists are generally greater than for pediatric medical subspecialties [23]. Compared to pediatric medical subspecialty care, long distances to pediatric surgical care may not have as much effect on access unless there is long-term follow-up by the surgeon and his or her team required. In many cases, the condition requiring surgical care may involve visits to the surgical provider for a defined and limited period. Families living distant from pediatric surgical specialists may
Distance to care and relative supply among pediatric surgical subspecialties
491
Fig. 8 Locations of children's hospitals offering pediatric surgery by zip code. Dots are mapped to zip code centroid. Source: The National Association of Children's Hospitals and Related Institutions (NACHRI), www.childrenshospitals.net. Produced by: North Carolina Rural Health Research and Policy Analysis Center, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
relocate temporarily for surgical care, living at Ronald McDonald Houses or other residential facilities. For families with limited financial means or limited sick time benefits, long distances to care may create substantial barriers, however. The widespread availability of general surgeons and surgical specialists, whose training and practice includes the care of both adult and pediatric populations, ensures the availability of surgical services to the overwhelming majority of the pediatric population. Indeed, there is evidence that general surgical specialists, like general otolaryngologists, are heavily involved in the care of children [10]. Several studies suggest, however, that pediatric surgical providers may have better outcomes and lower costs than their generalist peers [1-3]. As such, greater distances to pediatric surgical specialists may have quality implications. The few studies that have compared generalists and pediatric providers have several limitations; however, so more research is needed to determine whether there is a quality gap. Furthermore, the aging of the population may lead to greater demand for the services of general surgeon and surgical specialists, potentially decreasing the availability of pediatric surgical care in
general surgical practices. More research into how population dynamics and provider supply influence general surgeons and surgical specialists' willingness to care for children and outcomes of care is needed. More than one half of HRRs lack pediatric surgical specialists. These findings may indicate that HRRs do not represent markets for specialized pediatric care. Given the rarity of many of the pediatric conditions that require surgical intervention, very large population bases may be required to support a pediatric surgical specialist. Thus, the relevant market for a pediatric surgical specialist may encompass an entire state or portions of multiple states. Average provider to population ratios are less then 1 per 100,000 children for all pediatric surgical specialties. The extent to which this represents an adequate supply is unknown. High coefficients of variation suggest, however, that supply ratios vary considerably across markets and likely indicate distributional inequities. Moreover, the supply of pediatric surgeons was highly correlated with the number of children's hospitals offering pediatric surgical services within an HRR. Therefore, as would be expected, pediatric surgeons are located in areas with children's hospitals that offer pediatric surgical services.
492
M.L. Mayer et al.
Table 1 Population weighted average distance to the nearest provider by surgical specialty, US zip codes with any children (n = 30,091) No. of Miles to the nearest provider providers Mean 75 95 percentile percentile Surgical subspecialty General surgery 30124 Otolaryngology 9058 Neurosurgery 4500 Ophthalmology 17275 Orthopedic surgery 19606 Thoracic surgery 5106 Urology 9435 Pediatric surgical subspecialty Pediatric surgery 781 Pediatric 266 otolaryngology Pediatric 113 neurosurgery Pediatric 444 ophthalmology Pediatric orthopedic 466 surgery Pediatric 49 cardiothoracic surgery Pediatric urology 276
3.4 6.8 11.8 5.6 5.0 10.6 6.9
4.6 8.2 14.0 7.0 6.2 12.6 8.4
14.7 26.1 43.3 22.9 20.3 38.2 25.4
27.1 48.5
34.3 59.8
95.0 177.4
67.4
79.7
198.8
28.5
37.9
96.3
32.6
40.0
104.6
100.9 130.4
277.4
39.9
47.7
142.7
This further suggests that there may be regionalization of care, and the location of children's hospitals may play a key role in the location of pediatric surgical providers. The current supply of pediatric surgical specialists may not be adequate to allow diffusion into more markets. Proliferation
of positions at community children's hospitals [24] suggests an increased demand for these providers outside of academia, however. There is a tension in determining the ideal supply of pediatric surgical providers. Insufficient supply of providers may limit access, but oversupply may dilute provider experience by leading to a smaller number of procedures performed per provider. Studies have demonstrated a higher mortality rate for children treated by lower volume hospitals and surgeons for pediatric cardiac surgery [25-28]. Among general surgical specialties, there is a well-established relationship between volume and outcomes [29-32]. Therefore, diluted experience may diminish quality of care [5]. There is also concern that the growth in pediatric surgical specialties, like pediatric otolaryngology and pediatric urology, has led to a narrowing of scope of practice among pediatric surgeons because of competition from other pediatric surgical specialists and nonsurgical specialists such as interventional radiology [6,24]. Thus, both oversupply and undersupply of pediatric surgical specialists can have detrimental consequences. The pediatric population size required to support a pediatric surgical specialist may be quite high given the relatively low prevalence of pediatric conditions requiring surgical intervention. Estimates of the pediatric population size required to support a pediatric surgical specialist and recommended provider-to-population ratios would help to inform pediatric surgical workforce policy. Furthermore, expected changes in demand for such providers should be considered in workforce models. Advances in selected fields, such as fetal surgery, as well as the application of existing surgical techniques, such as bariatric surgery, to younger populations may lead to increases in demand. Planning to meet expected increases
Table 2 Percent of younger than 18 years population within selected distances of physicians by surgical specialty, US zip codes with any children (n = 30,091) Distance to the nearest provider
General surgery and surgical specialties General surgery
Otolaryngology
Neurosurgery
Ophthalmology
Orthopedics
Thoracic
Urology
10 or fewer miles 11-20 miles 21-40 miles 41-80 miles More than 80 miles
89.1 8.6 2.0 0.2 0.0
79.3 12.3 6.6 1.5 0.3
67.6 14.9 11.5 4.9 1.1
82.3 11.0 5.4 1.1 0.2
84.6 10.3 4.2 0.8 0.1
69.5 15.5 10.5 3.5 1.0
79.2 12.8 6.2 1.5 0.3
Distance to the nearest provider
Pediatric surgery and pediatric surgical specialties Pediatric surgery
Pediatric otolaryngology
Pediatric neurosurgery
Pediatric ophthalmology
Pediatric orthopedics
Pediatric cardiothoracic
Pediatric urology
10 or fewer miles 11-20 miles 21-40 miles 41-80 miles More than 80 miles
43.6 18.9 16.0 14.3 7.2
29.5 19.8 16.1 16.3 18.4
21.0 17.9 18.9 17.3 24.9
42.1 17.9 16.2 15.3 8.4
39.4 19.2 16.5 16.2 8.8
14.2 12.0 12.5 18.3 43.0
34.0 19.9 16.8 16.5 12.8
Distance to care and relative supply among pediatric surgical subspecialties Table 3
Relative supply of surgical providers, United States HRRs (n = 306)
Surgical subspecialty
General Orthopedic Ophthalmology Otolaryngology Thoracic Urology Neurosurgery
Pediatric Pediatric Pediatric Pediatric Pediatric Pediatric Pediatric a
493
surgery orthopedic ophthalmology otolaryngology cardiothoracic urology neurosurgery
% of HRR with a provider
Provider: population (in 100,000)
Average no. of people per physician a
Mean
SD
COV
100 100 100 100 100 100 100
12.04 7.18 6.25 3.39 1.91 3.67 1.68
33.70 8.17 14.20 7.36 2.61 9.20 1.91
280 114 227 217 137 250 113
% of HRR with a provider
Provider: younger than 18 years population (in 100,000)
56 53 50 34 13 45 25
Mean
SD
Coefficient of Variation
0.97 0.62 0.57 0.27 0.04 0.37 0.12
3.21 1.83 1.82 0.78 0.15 0.62 0.31
333 296 321 291 342 166 248
8303 13,925 15,998 29,517 52,327 27,217 59,480 Average no. of children per physician a
103,604 161,826 176,874 370,883 2,295,999 266,925 801,020
In those HRR that have at least 1 provider.
in demand must be balanced to prevent oversupply that can lead to decreases in per provider procedure volume, which can undermine surgeon's ability to remain skillful in a variety of procedures.
4.1. Limitations The major limitation of our analysis is that it is based on self-reported data from the AMA Masterfile. Because only pediatric surgery currently offers a certification examination, there are few alternatives to the AMA Masterfile that can be used to identify pediatric surgical specialists. In a technical appendix available upon request, we compared aggregate numbers from the AMA Masterfile and the American Board of Medical Specialties for pediatric surgery and found that the number of physicians self-reporting as pediatric surgeons was higher than the number reported as certified by American Board of Medical Specialties but similar to those reported in studies done by the Study on US Surgical Services [5]. Likewise, numbers for general and pediatric otolaryngologists are comparable to past reports [10,33]. Similar discrepancies between the AMA Masterfile and certification data have been previously reported for pediatric medical subspecialties and may reflect delays between initiation of practice and receipt of certification as well as practicing in a field despite lack of board eligibility and certification. In addition, we compared the number of self-designated providers based on AMA with professional society data where it was available. The total number of self-designated
pediatric urologists, pediatric orthopedists, and pediatric neurosurgeons based on the AMA Masterfile were quite comparable to the number of US members reported by their respective professional organizations. Consequently, we expect that our results for these specialties are likely reflective of the actual number of practicing providers for these specialties. The number of members the American Association for Pediatric Ophthalmology and Strabismus exceeded the number of self-designated providers on the AMA file by a large percentage, however. As such, our estimates of the distance to the nearest provider as well as the relative supply of these providers may be inaccurate. We used both primary and secondary specialty to identify providers in each surgical field studied, which may lead to an overestimate of the number of providers. We performed sensitivity analyses to estimate distances and relative supply using only primary specialty to identify providers. We found that distance estimates were virtually unchanged for general surgery and surgical specialties; for pediatric surgery and pediatric surgical specialties, average population weighted distances increased by 1.8 miles for pediatric surgery and 30.4 miles for pediatric cardiothoracic surgery. Average population weighted distance increased dramatically for pediatric neurosurgery from 67.4 to 132.45 miles. Likewise, using the primary specialty only decreased the estimated percentage of the pediatric population living within 10 miles of a provider and increased the percentage of the population living more than 80 miles from a provider across all pediatric surgical specialties with the exception of pediatric surgery. As expected, all measures of relative supply are influenced
494 by whether primary and secondary specialty are used to identify providers vs primary specialty only. Given that nationwide supply estimates for the pediatric subspecialties based or primary and secondary specialty align more closely with memberships in various pediatric surgical specialty association, we expect that estimates based on primary specialty alone understate the actual supply of these providers and that the estimates presented here are a more valid reflection of the actual number of providers. Another limitation is our inability to account for satellite clinics. Some pediatric surgical specialists may travel to other HRR to provide care on a regular basis. Failure to account for these secondary practice locations may bias our estimates of the supply of providers at the HRR level. Given the size of HRR, especially in the western portions of the United States, we expect that the amount of bias introduced is minimal. Our model also assumes that resources at geographic site of pediatric specialty care can accommodate transfers from other locals. Any local capacity constraints may require patients to travel even farther than our estimates suggest.
5. Conclusion This is the first study to estimate distances to care and relative supply of pediatric surgical specialists in the United States. The study also provides useful comparison data on general surgery and surgical specialties. The findings suggest that a substantial percentage of the pediatric population in the United States must travel one or more hours to reach a pediatric surgical specialist, whereas access to general surgical providers is much more extensive. Furthermore, our findings suggest that pediatric surgical specialties may face many of the distributional challenges plaguing pediatric medical subspecialties. Further research is needed to establish how the current supply and distribution of these providers influence access to and quality of surgical care for children in the United States.
Acknowledgments We would like to thank Bob Schwartz and Ann Howard for performing the distance analyses; Tom Ricketts, PhD, for providing access to the AMA Masterfile and commenting on previous drafts; and George F. Sheldon, MD, for his helpful insights. The research was supported through grant number 1-K02HS013309-01A1 from the Agency for Healthcare Research and Quality and a University Research Council grant from the University of North Carolina at Chapel Hill. The funding agency was not involved in the performance of the study. Doctor Mayer had full access to all of the data in the study and takes responsibility for the integrity of the data and the
M.L. Mayer et al. accuracy of the data analysis. The authors have no conflicts of interest to report.
References [1] Alexander F, Magnuson D, DiFiore J, et al. Specialty versus generalist care of children with appendicitis: an outcome comparison. J Pediatr Surg 2001;36:1510-3. [2] Albright AL, Sposto R, Holmes E, et al. Correlation of neurosurgical subspecialization with outcomes in children with malignant brain tumors. Neurosurgery 2000;47:879-85 [Discussion 885-7]. [3] Kokoska ER, Minkes RK, Silen ML, et al. Effect of pediatric surgical practice on the treatment of children with appendicitis. Pediatrics 2001; 107:1298-301. [4] Snow BW, Catwright PC, Young MD. Does surgical subspecialization in pediatrics provide high-quality, cost-effective patient care? Pediatrics 1996;97:14-7. [5] O'Neill Jr JA, Gautam S, Geiger JD, et al. A longitudinal analysis of the pediatric surgeon workforce. Ann Surg 2000;232:442-53. [6] O'Neill Jr JA, Cnaan A, Altman RP, et al. Update on the analysis of the need for pediatric surgeons in the United States. J Pediatr Surg 1995; 30:204-10 [discussion 211-3]. [7] Parkerton PH, Geiger JD, Mick SS, et al. The market for pediatric surgeons: a survey of recent graduates. J Pediatr Surg 1999;34:931-9. [8] Zalzal GH. Projected societal needs in pediatric otolaryngology. Laryngoscope 1996;106:1176-9. [9] Smith RV. Workforce issues for the academic otolaryngologist. Laryngoscope 2000;110:1602-6. [10] Tunkel DE, Cull WL, Jewett EA, et al. Practice of pediatric otolaryngology: results of the future of pediatric education II project. Arch Otolaryngol Head Neck Surg 2002;128:759-64. [11] Shemin RJ, Dziuban SW, Kaiser LR, et al. Thoracic surgery workforce: snapshot at the end of the twentieth century and implications for the new millennium. Ann Thorac Surg 2002;73:2014-32. [12] Friedlich DL, Feustel PJ, Popp AJ. Workforce demand for neurosurgeons in the United States of America: a 13-year retrospective study. J Neurosurg 1999;90:993-7. [13] Shapiro E, Hatch DA. Pediatric urological manpower report. Pediatric urological manpower committee of the American Association of Pediatric Urology. J Urol 1993;150:675-9. [14] Retik AB, King LR. Pediatric urology manpower report 1995. Ellen Shapiro on behalf of the American Association of Pediatric Urologists. J Urol 1996;156:488-90 discussion 490-1. [15] Claritas Inc. Pop-facts database for ZIP codes. San Diego: Claitas Inc.; 2003. [16] Guagliardo MF. Spatial accessibility of primary care: concepts, methods and challenges. Int J Health Geogr 2004;3:3. [17] Phibbs CS, Luft HS. Correlation of travel time on roads versus straight line distance. Med Care Res Rev 1995;52:532-42. [18] Report of the Graduate Medical Education National Advisory Committee to the Secretary, Department of Health and Human Services. Vol I GMENAC Summary Report. Hyattsville, MD: Health Resources Administration, DHHS Pub. No. (HRA); September 1980. p. 105. [19] Rosenthal MB, Zaslavsky A, Newhouse JP. The geographic distribution of physicians revisited. Health Serv Res 2005;40:1931-52. [20] Goodman DC, Fisher ES, Little GA, et al. Are neonatal intensive care resources located according to need? Regional variation in neonatologists, beds, and low birth weight newborns. Pediatrics 2001;108: 426-31. [21] Morrisey MA, Sloan FA, Valvona J. Defining geographic markets for hospital care. Law Contemp Probl 1988;51:165-94. [22] Guagliardo MF, Jablonski KA, Joseph JG, et al. Do pediatric hospitalizations have a unique geography? BMC Health Serv Res 2004;4:2.
Distance to care and relative supply among pediatric surgical subspecialties [23] Mayer ML. Are we there yet? Distance to care and relative supply among pediatric medical subspecialties. Pediatrics 2006;118:2313-21. [24] Geiger JD, Drongowski RA, Coran AG. The market for pediatric surgeons: an updated survey of recent graduates. Journal of Pediatric Surgery 2003;38:397-405. [25] Jenkins KJ, Newburger JW, Lock JE, et al. In-hospital mortality for surgical repair of congenital heart defects: preliminary observations of variation by hospital caseload. Pediatrics 1995;95:323-30. [26] Hannan EL, Racz M, Kavey RE, et al. Pediatric cardiac surgery: the effect of hospital and surgeon volume on in-hospital mortality. Pediatrics 1998;101:963-9. [27] Sollano JA, Gelijns AC, Moskowitz AJ, et al. Volume-outcome relationships in cardiovascular operations: New york state, 1990-1995. J Thorac Cardiovasc Surg 1999;117:419-28 [discussion 428-30]. [28] Begg CB, Cramer LD, Hoskins WJ, et al. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747-51.
495
[29] Hannan EL, Wu C, Walford G, et al. Volume-Outcome relationships for percutaneous coronary interventions in the stent era. Circulation 2005;112:1171-9. [30] Norton EC, Garfinkel SA, McQuay LJ, et al. The effect of hospital volume on the in-hospital complication rate in knee replacement patients. Health Serv Res 1998;33:1191-210. [31] Thiemann DR, Coresh J, Oetgen WJ, et al. The association between hospital volume and survival after acute myocardial infarction in elderly patients. N Engl J Med 1999;340:1640-8. [32] Edwards EB, Roberts JP, McBride MA, et al. The effect of the volume of procedures at transplantation centers on mortality after liver transplantation. N Engl J Med 1999;341:2049-53. [33] Pillsbury III HC, Cannon CR, Sedory Holzer SE, et al. The workforce in otolaryngology-head and neck surgery: moving into the next millennium. Otolaryngol Head Neck Surg 2000;123: 341-56.