Residents’ Perspective on Need for Point-of-Care Ultrasound Education During Pediatric Residency

The population of interest was 559 pediatric residents enrolled in the 9 pediatric residency programs across the state of Ohio during the study period (July 2019–June 2020). To attain a suitable level of sampling precision (a 95% confidence interval [CI], with no more than ±3% sampling error), we used guidance from Dillman’s Tailored Design Method.³⁹  The resulting estimate suggested that we could achieve adequate survey coverage by sampling 384 of the 559 (68.7%) pediatric residents in Ohio. We contacted each Ohio program director to request permission to survey their trainees and obtain program rosters with contact information. We used program websites to verify roster accuracy and obtain additional information about each resident, such as training level, medical degree, gender, and specific educational program. We assigned each resident a random number using a random number generator and then drew random samples proportional to the number of residents from each program (from small [n = 12] to large [n = 132]) and stratified by level of training (postgraduate years [PGY] 1–3).⁴⁰  Our final adjusted stratified sample was 371 of 559 (66.4%).

Our survey instrument was designed with guidance from a survey methods expert to assess:

1. pediatric resident’s POCUS education and experience at both undergraduate and graduate medical education levels;

2. resident opinions about POCUS education during residency;

3. their future career plans; and

4. perceived barriers to POCUS training during residency. Many survey items were borrowed directly fro=m instruments used in similar studies.^14,41,42  We piloted the instrument with family medicine and pediatric residents from 1 academic pediatric hospital who graduated before full implementation of the study and modified the survey on the basis of resident feedback. The final survey instrument consisted of multiple choice and open-ended items (Supplemental Fig 3).

We used Dillman’s book on survey methods to guide implementation of our electronic survey.³⁹  We sent a brief e-mail to Ohio pediatric residents to describe the study and notify them that, if selected through sampling, they would be receiving a survey invitation. One week later, we emailed the electronic survey to selected participants using Research Electronic Data Capture.⁴³  Participation was voluntary, and responses were tracked to facilitate follow-up. Nonrespondents received up to 3 reminders to complete the survey from mid-August to mid-October 2019. Participants were not compensated. This study was determined to be exempt by our institutional review board.

Before deidentification of survey data, we matched our survey distribution database with our response database using subject emails to generate information about return rates. This allowed for verification of self-reported demographics (gender, program, level of training, and medical degree) with demographics reported on program websites. We also classified the programs into size categories based on number of residents in the program: small ≤49, medium 50 to 99, and large ≥100.

We conducted bias analyses to make judgments about the representativeness of our respondents, using χ² Tests of Association to compare respondents to nonrespondents on 4 key variables: gender, level of training, type of degree, and program size. We used similar tests to compare our sample to the population on these same variables to ensure that our sample represented the population of Ohio pediatric residents. We used simple and bivariate descriptive statistics and subgroup analyses to profile our subjects’ responses to survey items. All analyses were performed using IBM-SPSS for Windows,⁴⁰  or VassarStats.⁴⁴  We also calculated likelihood ratios and previous probabilities to explain the size of significant effects. These were performed using a web-based calculator provided by Schwartz.⁴⁵ 

We distributed 371 surveys across all 9 residency programs in Ohio. Our return rate was 59.3% (220 of 371), representing 39.4% (220 of 559) of the total number of Ohio pediatric residents in 2019. Respondents were representative of the total sample regarding gender and training level training (Table 1). Every program in the state was also well represented; all but 1 program had return rates >40%. χ² test results indicated a slight overrepresentation of large residencies and slight underrepresentation of medium-sized programs (P = .02; +likelihood ratio = 0.77; posterior probability = 0.36 [95% CI: 0.31–0.34]); suggesting that residents from large programs were 36% more likely to respond than those from medium-sized programs.⁴⁵  We also received slightly more surveys than expected from osteopathic physicians (P = .03; +likelihood ratio = 1.24; posterior probability = 0.25 [95% CI: 0.21–0.28]), suggesting that residents with osteopathic degrees were 25% more likely to respond than those with allopathic degrees.⁴⁵ 

Three-quarters of our respondents identified as female (76%; 167 of 221) and three-quarters (75%; 166 of 221) were doctors of medicine. Most of the respondents attended medical school in the Midwest (50%; 109 of 120) or Middle/South Atlantic (22%; 49 of 220) regions of the United States. A small portion, 15% (33 of 220), graduated from medical schools outside the United States. Nearly 61% (134 of 220) of respondents planned to pursue subspecialty training after residency. The top 5 pediatric subspecialties of interest were critical care, emergency medicine, neonatology, hematology/oncology, and cardiology (Table 2).

Only 15% (33 of 220) of respondents reported having POCUS education during residency. Residents from 8 of the 9 programs in this study were among the 33 respondents who suggested that they had POCUS education; however, only 1 resident reported that they experienced a formal POCUS curriculum during residency. More than twice as many respondents, 36% (79 of 220), reported having POCUS education during medical school. Cumulatively, more than half of pediatric residents (56%; 124 of 220) received no POCUS training in medical school or residency, and only 7% (16 of 220) received POCUS education at both levels. Residents from large programs (≥100 residents) were 23% more likely to have received POCUS education during residency (P ≤.001; +likelihood ratio = 1.73; posterior probability = 0.23 [95% CI: 0.19–0.28]).⁴⁵ 

Among the residents who received some aspect of POCUS training during residency, most (79%; 26 of 33) experienced formal didactics, simulation, or bedside teaching. With regard to hands-on experience, 84% of residents (185 of 220) reported performing 0 ultrasound scans on patients in clinical settings, 16% (34 of 220) performed 1 to 10 ultrasounds, and only 1 respondent performed more than 10. There was a significant statistical relationship between the number of scans performed and year of training, with PGY-3s (31%; [22 of 71]) being 31% more likely to have performed scans than PGY-2s (121%; [9 of 75]) or PGY-1s (5%; [4 of 74]) (P ≤.001; +likelihood ratio = 2.37; posterior probability = 0.31 [95% CI: 0.24–0.39]).⁴⁵ 

Details from the third (n = 79) of respondents who reported having ultrasound education in medical school suggested wide variability in their experiences. Two-thirds of this group (67%; 53 of 79) said that their POCUS education experience was compulsory. Most reported being taught diagnostic POCUS (89%; 70 of 79), whereas more than half said that they learned POCUS for procedural guidance (58%; 46 of 79). More than two-thirds (67%; 53 of 79) said they used ultrasound to visualize anatomic structures or pathology.

Most (90%; 71 of 79) resident participants who reported having ultrasound education during medical school indicated that POCUS was helpful for learning clinical applications, such as diagnosing, evaluating, and treating patients. Three-fourths (75%; 59 of 79) suggested that POCUS helped them to learn how to perform procedures. Most of them (61%; 48 of 79) said ultrasound helped them to learn topics such as anatomy or physiology. About half the respondents (111 of 219; 51%) thought that POCUS education should be required in medical school.

Because of the uneven distribution of where respondents attended medical school, we were unable to draw conclusions about differences in experience across geographic regions. However, 21% or more of respondents who attended medical school in each of the 6 US regions and outside the United States reported having had POCUS education during medical school.

Although only 13% (29 of 220) claimed that POCUS education was a key factor in their residency program selection, most respondents desired POCUS education during training (86%; 189 of 220). Furthermore, two-thirds of the respondents (67%; 147 of 220) thought POCUS education should be required in residency, whereas nearly all (93%; 205 of 220) of resident respondents believed that POCUS would be useful for patient care. Regarding how POCUS education might look during residency, more than three-fourths of respondents (79%; 174 of 220) said they thought that POCUS should become a formal part of the pediatric residency curriculum. Fig 1 summarizes the respondents’ opinions about POCUS education during residency. Overall, 83% of respondents thought POCUS education should be required at some point during their medical education.

Most respondents also recognized the utility of POCUS for patient care. Almost all (95%; 208 of 220) thought that it would be useful for guiding procedures and as an adjunct to medical decision-making (93%; 205 of 220). Nearly three-fourths also agreed with its utility in improving patient safety or visualizing anatomy (74%; 163 of 220). Most importantly, nearly 61% (134 of 220) felt that POCUS education could provide a significant benefit to them in their future practice.

Almost 61% (134 of 220) of respondents said they were planning to pursue pediatric subspecialty fellowship after residency (Table 2). Nearly half (47%; 72 of 155) of subspecialty choices being considered by these 134 residents (some were considering >1), were in fields that have adopted POCUS into fellowship training or their scope of practice (cardiology, critical care, emergency medicine, rheumatology, or sports medicine)^{11–13,24,25,27,36,46–50}  (Table 2). Individuals pursuing any subspecialty fellowship were 69% more likely to perceive POCUS as beneficial to their future practice (P ≤.001; +likelihood ratio = 1.44; posterior probability = 0.69 [95% CI: 0.40–0.56]).⁴⁵  Among those who were not pursuing a fellowship or remained undecided, nearly half (48.2%; 41 of 85) still endorsed POCUS education as beneficial to their future specialty. Residents from all 3 program sizes were equally likely to pursue fellowships in subspecialties (P = .196) and to endorse POCUS education as a benefit to their future specialty (P = .739).

The most significant barrier to POCUS education was the lack of a formal POCUS curriculum (75%; 165 of 220). Many residents also believed that competition for time with other educational activities was a significant barrier (59%; 129 of 220), as was the lack of POCUS-trained faculty (52%; 115 of 220) (Fig 2). Only 36% (80 of 220) of respondents said that there was a shortage of ultrasound machines for teaching POCUS.

Ohio pediatric residents have had limited to no experience with POCUS. Many never received POCUS education during medical school or residency; most had never performed a POCUS scan. Yet, these residents believed that POCUS should become a formal component of their residency curriculum. Senior residents (PGY-3s) and residents from large programs were more likely to have received POCUS education and to have performed scans on patients. Residents planning on subspecialization were more likely to perceive POCUS education as useful to their future practice. Availability of POCUS education did not factor into these residents’ program choice.

The opinions of Ohio residents were consistent with those of pediatric program directors from other studies. Reaume et al and Good et al also found opportunities for POCUS education were limited in pediatric residencies across the United States.^31,34  An Italian study of pediatric residents found they received little to no formal POCUS training during residency, yet they desired it and believed learning POCUS would benefit them in future practice.³²  With evidence that pediatric residents are not getting the POCUS education they desire, what is preventing residency programs from offering POCUS education?

The barriers to implementing POCUS education identified by Ohio residents were comparable to barriers identified by program leaders in other studies.^31,34  Competing educational priorities and lack of adequately trained faculty were considered substantial barriers.^31,34  In contrast, Ohio residents also suggested that lack of a formal curriculum or structured plan for teaching POCUS was a noteworthy barrier. Unlike studies by Reaume and Good, most Ohio residents did not believe that lack of guidelines/standards from governing bodies, nor lack of access to ultrasound technology, were barriers to POCUS education in pediatric residencies.^31,34 

Overcoming these barriers will require deliberate efforts from both program leaders and accreditation bodies that govern pediatric residencies. The POCUS education literature offers potential solutions to the 3 primary barriers we identified in our study: lack of formal curricula, program time, and qualified teachers.

Developing an effective POCUS curriculum for pediatric residency involves 2 critical components:

1. teaching the knowledge and skills required to operate the ultrasound machine to acquire images at the bedside; and

2. teaching the knowledge and skills required to incorporate bedside sonography into current practice.

The second component covers knowledge about the types of scans likely to be useful for pediatricians and skills needed to interpret them and incorporate them into clinical decision-making. An effective pediatric POCUS curriculum could be adapted from existing curricula from other disciplines, and/or existing POCUS curricula for pediatric fellows.^7,10,51  A curriculum designed to teach the core knowledge and skills to operate a bedside ultrasound machine for pediatric emergency medicine fellows could be easily adapted for pediatric residents.⁵²  Others have already delineated the types of ultrasound scans most likely to be useful for pediatricians in practice.^33,53  Adapting POCUS teaching methods from pediatric subspecialties and adult medicine could serve as the foundation of a pediatric residency curriculum until evidence-based research clarifies the content (most important scans that should be understood and/or mastered by the general pediatrician) and best teaching and assessment methods.

Because the pediatric residency curriculum is already densely packed with learning objectives, adding content material will require some finesse. One of the most direct methods of delivering POCUS education to residents would be to develop customized POCUS education on the basis of residents’ interests and future career plans. The Accreditation Council for Graduate Medical Education explicitly requires individualized curricula,⁵⁴  so integrating POCUS education to those residents who desire it might be accomplished through carefully crafted asynchronous learning modules, simulation workshops, and supplemental electives.^55–57  Good et al provide an effective example through implementation of an effective basic POCUS curriculum to residents with limited POCUS experience during a pediatric intensive care rotation.⁵⁸  A longitudinal curriculum thread such as one successfully demonstrated by Brant et al, also shows how customized POCUS training can be offered in smaller bits throughout residency.³³ 

To overcome the shortage of POCUS trained faculty, institutions can combine POCUS development courses for current faculty with recruitment of POCUS-trained faculty to champion POCUS in their programs. Faculty development will require an initial investment of both money and time but will eventually benefit the institution. As we discovered, POCUS education does not currently factor into a resident’s program choice, but it may in the future. Resident learners like having program options and providing a robust POCUS education program may eventually enhance recruitment. Additionally, if reimbursement for POCUS studies is pursued, the cost of training and equipment could be reduced by billing for POCUS services. Finally, institutional efficiencies may be increased by moving ultrasonography from the radiology suite to the bedside. The integration of POCUS into the pediatric residency will take time and institutional support, so until pediatric faculty become effective POCUS teachers, POCUS educators from other disciplines or specialties can be recruited to teach.^59,60 

For reasons related to cost in time and resources, we limited our survey to pediatric residency programs in Ohio. Results, therefore, may not be generalizable to other regions of the United States or other countries. We believe, however, that, because these programs were of variable sizes and contained residents from diverse medical schools, the voice of these residents is likely to be representative of pediatric residents throughout the United States. Our results should also be weighed in the context of our respondents. Although the associated posterior probabilities were small, we found a slight bias in our respondent’s characteristics toward the larger residency programs and graduates of osteopathic medical schools. The keen observer might also have noticed that resident respondents in this study were more likely than their US counterparts to have expressed vocational interests in subspecialties that have already adopted POCUS. Although this may have been true in the past, recent changes in the growth of the medical subspecialty workforce suggest that our results were similar to national trends.⁶¹  Finally, given the lack of direct and purposeful experience Ohio pediatric residents have had with POCUS, we should recognize the speculative nature of their about both the accessibility of ultrasound machines and the utility of POCUS in clinical care and consider these when interpreting our findings.

Pediatric residents in Ohio believe that the absence of POCUS education represents an unmet educational need and desire training in this imaging modality. Trainees are increasingly receiving POCUS education in medical school, and there is evidence of its growing use in pediatric subspecialties. To sustain this educational momentum, enhanced POCUS education should be provided during pediatric residency. Introducing POCUS education into the general residency curriculum, or subspecialty tracks and elective rotations, models successful integration strategies of other medical specialties and subspecialties. Furthermore, developing faculty to become competent POCUS educators can also serve to expand the impact of POCUS into the broader pediatric scope of practice.

We thank and acknowledge the Ohio Pediatric Program directors for their gracious support of this project: Ann E. Burke, MD, Wright State University Program; Philip Fragassi, MD, MetroHealth System/Case Western Reserve University program; Charles Kwon, MD, Cleveland Clinic Foundation Program; Keith Ponitz, MD, Case Western Reserve University/University Hospitals/Cleveland Metro Center/Rainbow Babies and Children’s Hospital Program; Sue E. Poynter, MD, MEd, Cincinnati Children’s Hospital Medical Center Program; Maria Ramundo, MD, MS, Children’s Hospital Medical Center of Akron/NEOMED Program; Sherri A. Thomas, MD, Mercy St. Vincent Medical Center Program; Rebecca Wallihan, MD, Nationwide Children’s Hospital/Ohio State University program; and Mary E. Wroblewski, MD, University of Toledo program. We also thank the 2019–20 Ohio pediatric residents for their contribution to this project.