To ensure access to pediatric subspecialty care, recent national efforts to inform workforce planning include the American Medical School Pediatric Department Chairs Pediatrics 2025 Workforce Initiative1 and a National Academies consensus study.2 To supplement available data, the American Board of Pediatrics (ABP) added questions regarding perceived job availability (PJA) to its annual surveys with subspecialty fellows in 2019 and 2021, with plans to continue in odd-numbered years moving forward, permitting cross-sectional analysis of trends over time. We review results from 2019 and 2021 in this brief.
Methods: Procedures
Subspecialty fellows who took the ABP annual Subspecialty In-Training Examination (SITE) in February of 2019 or 2021 were asked to complete a brief, 4-question survey that assessed respondents’ perceptions of (1) the availability of jobs in academic centers nationwide, (2) the availability of jobs in nonacademic centers or private practices, (3) the availability of jobs where I want to live, and (4) confidence in finding a job to practice in my subspecialty (see Table 1). Responses (“Strongly Disagree”, “Disagree”, “Agree”, “Strongly Agree”, “Not Applicable/Don’t Know”) were coded ordinally to permit the calculation of means (Strongly Disagree = 1; Strongly Agree = 4); Not Applicable/Don’t Know responses were excluded.
Adjusted Expected Probabilities of Selecting Strongly Agree/Agree for Graduating Subspecialty Fellows Across Years by Subspecialty
Subspecialtya,b . | Adjusted Estimated Probabilities . | P . | |
---|---|---|---|
2019, % . | 2021, % . | ||
Q1. “There are available jobs in academic centers nationwide.” | |||
Cardiology | 61 | 50 | .08 |
Critical care medicine | 56 | 26 | <.0001c |
Emergency medicine | 94 | 25 | <.0001c |
Endocrinology | 87 | 55 | .0002c |
Gastroenterology | 87 | 53 | <.0001c |
Hematology-oncology | 37 | 24 | .04 |
Infectious diseases | 60 | 73 | .23 |
Neonatal-perinatal medicine | 77 | 60 | .0005c |
Pulmonology | 93 | 80 | .09 |
Other | 88 | 84 | .38 |
Q2. “There are available jobs in non-academic centers or private practices.” | |||
Cardiology | 71 | 65 | .34 |
Critical care medicine | 87 | 64 | <.0001c |
Emergency medicine | 97 | 58 | <.0001c |
Endocrinology | 80 | 82 | .75 |
Gastroenterology | 91 | 76 | .01 |
Hematology-oncology | 46 | 51 | .39 |
Infectious diseases | 49 | 44 | .69 |
Neonatal-perinatal medicine | 97 | 88 | .003c |
Pulmonology | 79 | 89 | .21 |
Other | 57 | 61 | .50 |
Q3. “There are available jobs where I want to live.” | |||
Cardiology | 43 | 43 | .95 |
Critical care medicine | 44 | 29 | .01 |
Emergency medicine | 88 | 39 | <.0001c |
Endocrinology | 55 | 50 | .59 |
Gastroenterology | 61 | 52 | .25 |
Hematology-oncology | 26 | 22 | .47 |
Infectious diseases | 42 | 47 | .64 |
Neonatal-perinatal medicine | 73 | 62 | .02 |
Pulmonology | 84 | 56 | .005c |
Other | 76 | 71 | .44 |
Q4. “I am confident that I will find a job to practice in my subspecialty.” | |||
Cardiology | 76 | 65 | .07 |
Critical care medicine | 79 | 51 | <.0001c |
Emergency medicine | 99 | 67 | <.0001c |
Endocrinology | 94 | 82 | .04 |
Gastroenterology | 93 | 72 | .001c |
Hematology-oncology | 51 | 49 | .72 |
Infectious diseases | 73 | 79 | .60 |
Neonatal-perinatal medicine | 95 | 93 | .25 |
Pulmonology | 91 | 96 | .35 |
Other | 94 | 92 | .61 |
Composite mean for all 4 questions | |||
Cardiology | 73 | 62 | .08 |
Critical care medicine | 77 | 48 | <.0001c |
Emergency medicine | 97 | 51 | <.0001c |
Endocrinology | 88 | 75 | .07 |
Gastroenterology | 92 | 73 | .003c |
Hematology-oncology | 42 | 36 | .26 |
Infectious diseases | 59 | 67 | .48 |
Neonatal-perinatal medicine | 94 | 83 | .002c |
Pulmonology | 93 | 90 | .61 |
Other | 86 | 88 | .60 |
Subspecialtya,b . | Adjusted Estimated Probabilities . | P . | |
---|---|---|---|
2019, % . | 2021, % . | ||
Q1. “There are available jobs in academic centers nationwide.” | |||
Cardiology | 61 | 50 | .08 |
Critical care medicine | 56 | 26 | <.0001c |
Emergency medicine | 94 | 25 | <.0001c |
Endocrinology | 87 | 55 | .0002c |
Gastroenterology | 87 | 53 | <.0001c |
Hematology-oncology | 37 | 24 | .04 |
Infectious diseases | 60 | 73 | .23 |
Neonatal-perinatal medicine | 77 | 60 | .0005c |
Pulmonology | 93 | 80 | .09 |
Other | 88 | 84 | .38 |
Q2. “There are available jobs in non-academic centers or private practices.” | |||
Cardiology | 71 | 65 | .34 |
Critical care medicine | 87 | 64 | <.0001c |
Emergency medicine | 97 | 58 | <.0001c |
Endocrinology | 80 | 82 | .75 |
Gastroenterology | 91 | 76 | .01 |
Hematology-oncology | 46 | 51 | .39 |
Infectious diseases | 49 | 44 | .69 |
Neonatal-perinatal medicine | 97 | 88 | .003c |
Pulmonology | 79 | 89 | .21 |
Other | 57 | 61 | .50 |
Q3. “There are available jobs where I want to live.” | |||
Cardiology | 43 | 43 | .95 |
Critical care medicine | 44 | 29 | .01 |
Emergency medicine | 88 | 39 | <.0001c |
Endocrinology | 55 | 50 | .59 |
Gastroenterology | 61 | 52 | .25 |
Hematology-oncology | 26 | 22 | .47 |
Infectious diseases | 42 | 47 | .64 |
Neonatal-perinatal medicine | 73 | 62 | .02 |
Pulmonology | 84 | 56 | .005c |
Other | 76 | 71 | .44 |
Q4. “I am confident that I will find a job to practice in my subspecialty.” | |||
Cardiology | 76 | 65 | .07 |
Critical care medicine | 79 | 51 | <.0001c |
Emergency medicine | 99 | 67 | <.0001c |
Endocrinology | 94 | 82 | .04 |
Gastroenterology | 93 | 72 | .001c |
Hematology-oncology | 51 | 49 | .72 |
Infectious diseases | 73 | 79 | .60 |
Neonatal-perinatal medicine | 95 | 93 | .25 |
Pulmonology | 91 | 96 | .35 |
Other | 94 | 92 | .61 |
Composite mean for all 4 questions | |||
Cardiology | 73 | 62 | .08 |
Critical care medicine | 77 | 48 | <.0001c |
Emergency medicine | 97 | 51 | <.0001c |
Endocrinology | 88 | 75 | .07 |
Gastroenterology | 92 | 73 | .003c |
Hematology-oncology | 42 | 36 | .26 |
Infectious diseases | 59 | 67 | .48 |
Neonatal-perinatal medicine | 94 | 83 | .002c |
Pulmonology | 93 | 90 | .61 |
Other | 86 | 88 | .60 |
The table displays the estimated probabilities of endorsing Strongly Agree/Agree on 4 survey questions and the composite question with the overall mean, adjusting for age, sex, medical degree type, and program region.
Subspecialty counts for 2019 and 2021 are as follows (n = # in 2019; # in 2021): Total (n = 1207; 1098); cardiology (n = 137; 135); critical care medicine (n = 154; 148); emergency medicine (n = 157; 146); endocrinology (n = 74; 59); gastroenterology (n = 95; 86); hematology-oncology (n = 154; 121); infectious diseases (n = 41; 43); neonatal-perinatal medicine (n = 240; 201); pulmonology (n = 51; 46); other (n = 114; 118).
P values <.01.
The 2 cohorts were further limited to those who had completed a United States-based Accreditation Council for Graduate Medical Education or Royal College of Physicians and Surgeons of Canada accredited residency program and were in their final year of a US fellowship during survey administration (hereafter, “graduating”) to best capture the responses of fellows who were actively job-seeking.
Survey data were linked with ABP administrative data, including age, sex, medical degree type, and program region. Because of the small sample sizes, 5 subspecialties (Adolescent Medicine, Child Abuse Pediatrics, Rheumatology, Developmental-Behavioral Pediatrics, and Nephrology) were grouped as “Other”.
Least squares means between 2019 and 2021 and for each subspecialty were calculated by using separate logistic regression models to obtain the expected probabilities of selecting either Strongly Agree/Agree or Strongly Disagree/Disagree for each of the 4 questions and a composite of those questions holding age, sex, medical degree type, and program region constant and accounting for differing respondent numbers in each subspecialty. An interaction term with survey year and subspecialty was included in all models. SAS Enterprise Guide 8.2 (SAS Institute, Cary NC) was used for data manipulation and modeling.
This study was deemed exempt by the ABP Institutional Review Board of record.
Results
Of all graduating fellows in 2019 and 2021, 1207 (92.2%) and 1098 (96.3%), respectively, took the SITE examination and participated in the survey. Demographic statistically significant differences between respondents/nonrespondents were noted but were unlikely to be meaningful because of the high response rate. Most were between ages 30 and 39 (92%), female (67%), American medical graduates (74%), and doctors of medicine (80%) or osteopathy (9%). Most were in the South (34%) and the Midwest (25%). Demographic representations were relatively consistent across 2019 and 2021.
Table 1 and Figure 1 display the adjusted probabilities of selecting Strongly Agree/Agree for each subspecialty, controlling for other covariates. When modeling the mean overall response, probabilities in 2019 ranged from 97% (Emergency Medicine) to 42% (Hematology-Oncology). Changes from 2019 to 2021 were unique to each discipline, varying in both direction and slope. From 2019 to 2021, Emergency Medicine showed the greatest decrease in the probability of selecting Strongly Agree/Agree across all questions; contrastingly, Infectious Diseases saw the greatest increase, except for job availability in nonacademic settings. Hematology-Oncology was consistently the lowest-rated subspecialty for both 2019 and 2021. For overall mean response, Critical Care Medicine, Emergency Medicine, Gastroenterology, and Neonatal-Perinatal Medicine showed statistically significant decreases in the probability of selecting Strongly Agree/Agree over time.
Linear changes over time for adjusted expected probabilities of selecting Strongly Agree/Agree by subspecialty. Linear changes over time for adjusted expected probabilities of selecting Strongly Agree/Agree by subspecialty. Figure displays linear changes over time in estimated probabilities of endorsing Strongly Agree/Agree on 4 survey questions and the composite question with the overall mean; probabilities are adjusted for age, sex, medical degree type, and program region. The reference line at 0.5 permits identification of subspecialties more likely to select Strongly Agree/Agree.
Linear changes over time for adjusted expected probabilities of selecting Strongly Agree/Agree by subspecialty. Linear changes over time for adjusted expected probabilities of selecting Strongly Agree/Agree by subspecialty. Figure displays linear changes over time in estimated probabilities of endorsing Strongly Agree/Agree on 4 survey questions and the composite question with the overall mean; probabilities are adjusted for age, sex, medical degree type, and program region. The reference line at 0.5 permits identification of subspecialties more likely to select Strongly Agree/Agree.
Discussion
This national analysis examined the PJA of almost all US graduating pediatric subspecialty fellows in February 2019 and 2021. Baseline responses and changes over time were strikingly different. For Hematology-Oncology, PJA was low in both years, paralleling reports of decreasing fellowship applications during this period.3 In contrast, PJA for other subspecialties may have been impacted by the coronavirus disease 2019 pandemic; Infectious Disease showed marked stability and Emergency Medicine’s PJA precipitously declined, concurrent with a marked reduction in visits due to the pandemic.4
A limitation of this study’s cross-sectional survey design is the lack of data on factors elucidating why fellows’ perceptions changed over this time and why these changes were more pronounced for certain subspecialties. Changes over time may have reflected ongoing trends in job availability regionally or nationally, decreasing emergency department visits, hospital volumes, hiring freezes, and heightened personal and familial stressors due to the pandemic, or a combination of factors. Recent publications examining the pandemic’s impact have focused on fellowship recruitment, educational and clinical experience, or burnout; we identified only 1 survey with pediatric gastroenterology fellows in 2020 linking the pandemic to concerns about jobs.5 In addition, for those who trained internationally, we were unable to determine if they were considering working in the United States or abroad after graduation because the ABP does not collect information on citizenship or visas. Future surveys will allow for cross-sectional tracking of responses over time.
These findings call for deliberate monitoring of the interplay between fellowship positions, numbers of graduating fellows, available jobs over time, and patient outcomes. Additional research will inform national efforts to avoid workforce undersupply, oversupply, or maldistribution that negatively impact child health outcomes.
Acknowledgments
We thank the pediatric subspecialty fellows who complete the SITE annual survey yearly and help to inform pediatric workforce initiatives. We also thank the ABP Research Advisory Committee for its review of this manuscript and Ms Ashley Tucker for her administrative support.
Mr Turner, Ms Gregg, and Drs Barnard and Leslie, conceptualized, designed, and implemented the data analyses and linked dashboard, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Schaechter drafted the initial manuscript and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
The content is solely the responsibility of the authors and does not represent the official views of the American Board of Pediatrics or the American Board of Pediatrics Foundation.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: Drs Leslie and Schaechter, Mr Turner, and Ms Gregg are employees of the American Board of Pediatrics. Dr Barnard is on the American Board of Pediatrics’ Board of Directors.
Comments