Pediatric Hospitalists’ Performance of Recommended Minor Procedures: A Multicenter Study

This survey-based, cross-sectional study included pediatric hospitalists at 20 participating sites who completed an electronic survey between November 2021 and March 2022. The institutional review board approved the study at the primary institution and each site. Eligible respondents practiced PHM as defined by ABP criteria for board certification eligibility. Sites were primarily recruited using the American Academy of Pediatrics Section on Hospital Medicine and Community Pediatric Hospitalist e-mail listservs. We used snowball sampling and referrals from professional networks to recruit from a wide range of geographic areas and settings (Supplemental Table 4). Each site had a “champion” who was responsible for ensuring completion of the survey(s).

The survey included 2 components:

1. an individual survey completed by participants at each site; and

2. a site survey completed by the site champion.

The individual survey assessed frequency, independence, and success in 11 recommended minor procedures: PIV, BMV, LP, peripheral venipuncture (PV), arterial puncture, intraosseous catheter (IO) insertion, gastrostomy tube (GT) change, tracheostomy (trach) tube change, bladder catheterization, nasogastric tube insertion (NG), and incision and drainage (I&D).

Our construct was informed by a literature review on competence in minor procedures among general pediatricians.^6,7,12  A series of group and individual cognitive interviews with pediatric hospitalists ensured appropriate content and response-process validity for both surveys. The survey was pilot-tested by 5 pediatric hospitalists to obtain feedback and subsequently reviewed by site champions to ensure questions were representative and relevant to various practice settings in assessing our construct.

Data were collected using Research Electronic Data Capture hosted at the University of Cincinnati.^13,14  We e-mailed the survey link to site champions who shared it with eligible respondents at respective institutions. Site champions sent 2 reminders at 1-week intervals. We notified sites with <50% response rate to send an additional reminder to their group. Data were collected in 2 waves, each lasting 4 weeks. No financial incentives were offered for individuals or study sites to participate in the survey.

The individual survey collected demographic data including sex, race, and ethnicity.

In addition, details about residency and fellowship, years since completion of training, and percentage of clinical effort (full-time equivalent [FTE]) were collected. The site survey elicited details on hospitalist group size, number of yearly patient encounters, practice locations, types of clinical service provided, and supervisory roles, as well as institutional resources to support procedural competence and processes for maintaining procedural competence.

Participants were asked how frequently they performed minor procedures; response choices were categorized as 0 times per year, <5 times per year, 5 to 10 times per year, and >10 times per year. Given that data collection occurred during the low-census phase of the coronavirus disease 2019 pandemic, we asked respondents to answer on the basis of procedural frequency in a typical, “nonpandemic” year. Participants could list additional procedures currently within their scope of practice that were not currently recommended by the ACGME and ABP.

The primary outcome was SAC, which was calculated using a tool developed specifically for the purpose of this study. Given the subjective nature and inherent variability in interpretation of the word competence, we used responses about procedural independence and success to define procedural competence. Procedural independence was measured by asking participants, “How would you describe your current level of independence for the following procedures?” Responses were captured using a 5-point Likert scale, with responses ranging from “No experience with the procedure (0)” to “Can perform independently and troubleshoot difficult scenarios when others have attempted first (4).” Procedural success was measured by asking respondents, “How would you describe your current success rate for the following procedures?” Response options for procedural success ranged from “No successful attempts (0)” to “Successful >75% of the time (4).”

An SAC score was derived for each respondent by averaging responses to the independence and success items for all 11 minor procedures (a total of 22 responses). If participants indicated they had never attempted a procedure (selected 0 for independence), no score was generated for success, because no evaluation of success can be made if a person has not attempted a given procedure. However, if a person attempted a procedure but had no success, then a 0 would be factored into the SAC score for the success component. The SAC score for each person could range from 0 to 4, with higher values indicating a higher combined average of success and independence across all 11 procedures. Importantly, SAC scores were calculated for respondents and not for individual procedures. As such, the SAC score was meant to capture self-assessed competence, taking all minor procedures into account, and to assess the influence of individual- and site-level characteristics on global procedural competence.

Descriptive data were summarized by frequencies and percentages for categorical variables, means with standard errors for normally distributed data, and median with interquartile range for data that were not normally distributed. We used the χ² test to determine associations between categorical variables. The associations between SAC scores and predictor variables were determined through analysis of variance for categorical variables and fitted regression models for continuous variables. Given that a substantial proportion of the cohort cared for patients on newborn and NICU services, we conducted secondary analyses to compare the mean SAC scores of those who did and did not provide such services using a 2-group t test. A P value of <.05 was used to determine statistical significance for all analyses. All statistical analyses were conducted using SAS software (Version 9.4, SAS Institute Inc., Cary, NC, USA). Figures were generated using GraphPad Prism Software (Version 9.5.0, CA, USA).

The individual survey response rate was 62% (n = 360 of 578). Respondent characteristics are summarized in Table 1. Most respondents identified as female, white, and non-Hispanic/Latino. The majority were practicing hospitalists (93%), with a small number of current fellows in training (6%). Most completed residency training in categorical pediatrics (85%), whereas the remainder trained in internal medicine–pediatrics. Seventy-eight percent (n = 262 of 338) had not completed a PHM fellowship. Among the 76 participants who had completed fellowship, only 47 had completed training in PHM. Twenty-two participants did not provide a response to this question (6%). Other fellowships included general pediatrics and a variety of subspecialities.

Figure 1 summarizes the procedural frequency, independence, and success for 11 minor procedures. BMV and LP were the most frequently performed procedures. For each of the remaining 9 procedures, over half of the respondents reported not having performed these procedures. Procedural independence and success were also highest for BMV and LP. Additional procedures considered to be within the scope of practice among respondents were laceration repair (24%), foreign-body removal (24%), intubation (21%), umbilical artery/vein catheterization (16%), circumcision (14%), point-of-care ultrasound (14%), fracture splinting (13%), paracentesis (10%), and administration of vaccines (9%).

The associations between reported frequency of individual procedures and SAC among respondents are shown in Table 2. SAC scores were higher with greater reported procedural frequency (P < .001 for each comparison). There was a significant association between SAC score and procedural frequency when procedural frequency was dichotomized (0–4 and ≥5 times per year) for PIV catheter placement, arterial puncture, NG insertion, bladder catheterization, I&D, and LP (P < .05 for each). Figure 2 shows a fitted regression plot examining the association between year of residency graduation and SAC score. SAC scores among respondents decreased with each decade between 1980 and 2020s. Although statistically significant, the effect size of this correlation was low (P < .0001, adjusted R square: 0.05). The relationships between SAC score and clinical FTE, and SAC score and patient encounters per hospitalist FTE, were not statistically significant.

Fourteen participating sites were in urban settings (292 respondents), 8 sites were in suburban locations (123 respondents), and 2 sites were rural community hospitals (14 respondents). Sixty-nine participants (19%) worked in >1 site. Beyond the inpatient wards, respondents reported providing the following types of care: Comanagement of surgical patients, consultation services, complex care, PICU or step-down care, sedation, newborn nursery, level II or III NICU, and urgent and emergency care.

Seventeen sites had access to a simulation laboratory/center. Fifty percent of the sites had a PHM fellowship training program. Nearly a third had dedicated didactics for procedures, provided access to online educational resources, and had resources to provide just-in-time training. Two sites had protected time for hospitalists to practice procedures in the emergency department, operative room, or ICUs (Supplemental Table 4). Four out of 20 sites reported a process for obtaining or maintaining hospital privileges for minor procedures. All sites required privileges to perform LP, and fewer sites required privileges to perform other minor procedures: Peripheral venipuncture and BMV (95%), intraosseous catheter insertion (90%), I&D (85%), bladder catheterization (80%), NG insertion and PIV (75%), gastrostomy tube change (60%), and tracheostomy tube change (35%).

The associations between study site characteristics and SAC scores are shown in Table 3. SAC scores were higher among respondents whose responsibilities involved comanagement of surgical patients, PICU or step-down care, newborn nursery, and urgent or emergency care (P < .05 for each). A large proportion of the study sample reported also providing newborn or NICU care, >40%. In secondary analyses, participants that provided newborn or NICU care had higher SAC scores compared with those that did not provide such care, 2.53 (SD 0.7) vs 2.36 (SD 0.69), P = .032. Respondents from sites that also cared for adult patients in a children’s hospital had lower SAC scores than those from those sites where respondents did not (P = .01). Having a PHM fellowship program, access to a simulation center, and educational resources were not significantly associated with SAC scores. However, reserved time for hospitalists to practice procedures was associated with significantly higher SAC scores relative to those who were not afforded such opportunities (P = .01). Respondents from institutions with requirements for didactics and procedural shifts reported higher SAC (P < .01) in performing minor procedures. The mean SAC score was higher among those practicing in rural settings in comparison with those in urban or suburban settings; however, these results were not statistically significant (P = .24).

We described minor procedure performance among a diverse group of pediatric hospitalists. Participants had varying levels of training and experience and worked in a variety of clinical practice settings. BMV and LP were the most frequently performed procedures. Performing procedures more frequently, and longer duration since training completion, were associated with higher SAC scores. Respondents providing certain types of clinical care (comanagement of surgical patients, PICU or step-down care, newborn nursery, and urgent or emergency care) had significantly higher SAC scores. Sites with training requirements (didactics and procedural shifts) also had significantly higher SAC scores, whereas those practicing at sites with adult patients in children’s hospitals reported lower SAC scores. Presence of a PHM fellowship, access to a simulation center, and educational resources were not significantly associated with SAC. However, higher SAC scores were found among respondents at institutions with reserved time to practice procedures, opportunities for didactics, and procedural shifts.

Our findings indicate that performing minor procedures more frequently has implications for SAC. Thus, exploring barriers to increasing procedural frequency may be important for pediatric hospitalist programs, especially because many hospitalists work in multiple clinical sites. Opportunities to practice procedures may be rare because of practice setting, the presence of learners who need procedural experience, time constraints, competing priorities, or other factors. Reserving time for procedural practice may mitigate some of these factors and lead to higher procedural competence. Physicians in our survey with time dedicated to procedural practice had significantly higher SAC scores. However, only 2 out of 20 study sites reported providing protected time for hospitalists. Although reserving this time may be costly and challenging to implement, resource-rich environments could consider integrating protected procedural time into staffing models.

Having access to a simulation center was not independently associated with a higher SAC score in our study; however, pairing dedicated procedural time with use of a simulation laboratory may increase procedural competence. Previous studies on simulation for procedural competence have been promising. One study by Hale et al evaluating baseline competency in LP and paracentesis found a remarkable improvement in the proportion of adult hospitalists that demonstrated competency (<33%–100%) after a simulation-based training.¹⁵  However, this figure dropped to 40% after a 6-month interval, underscoring the need for both initial competency training and skill maintenance among practicing hospitalists. Another study by Carney et al implemented a multimodal, simulation-based curriculum among pediatric hospitalists with varying levels of experience.¹⁶  The educational sessions included high-fidelity, case-based scenarios with facilitated debriefs, procedural skills station, and brief didactic content geared toward the adult learner. Posttraining scores were higher for the procedural skills evaluated, especially among those who considered themselves not competent at baseline. These studies suggest simulation may enhance baseline competence and skill development. Further study is needed, however, to determine best practices for retaining these gains.

Although our study did not identify practice location as being significantly associated with SAC score among respondents, practitioners in rural locations with limited providers may be asked to cover a wider variety of patients or could be expected to provide care in higher-acuity settings such as the emergency department or ICUs. Hospitalists who practiced in PICU or step-down care, newborn nursery, level II or III NICU, or urgent and emergency care had higher SAC scores relative to those who did not practice in such settings. Further investigations to better understand practice patterns, and procedural competency in these settings specifically, may be beneficial.

Although our site-level data suggest substantial variability in formal processes for obtaining and maintaining privileges, procedural competency is likely to be a factor that remains important for retaining hospital privileges. Thus, hospital-based clinicians may be asked to demonstrate procedural competence, and programs may be asked to support those efforts. We can speculate that there is not a “one size fits all” approach to establishing and maintaining procedural competence. Simulation may help some groups or individuals. Others may prioritize hands-on practice with patients, even if that requires more effort and planning. Other hospitals have used alternate strategies such as development of a core group of hospitalists who lead a “procedure service.”^17,18  This strategy can help ensure that a subset of clinicians retains a high level of procedural proficiency. Quality improvement and patient-safety–focused approaches may also help improve procedural skills, lower complication rates, and reduce cost.¹  Finally, national accrediting bodies may seek to revise which procedures are required for PHM trainees, balancing the low frequency of certain procedures in clinical practice with the importance of maintaining critical skills.

Our study is limited by the design, which relied on retrospective recall data and self-assessment; hence, causality cannot be inferred from any observed relationships. It is possible that reported self-assessed competency may not be reflective of actual, observed competency. Compounding matters further, our study was conducted during the coronavirus disease 2019 pandemic and potentially influenced participants’ recall of their procedural frequency and SAC. Although our sample of hospitalists was diverse, we had a small number of respondents practicing at rural sites, and this subgroup should be studied further with larger samples. This study was a pilot effort and used a survey that had not been previously validated; the clinical relevance of the numerical SAC scale has not been determined. Survey items may require revision for future studies to ensure clarity. For example, the item that asks respondents to describe their current success rate should be clarified further to indicate exactly what time frame current refers to. Lastly, we did not use statistical weighting procedures, and therefore responses to independence and success questions for each of the procedures contributed equally to the SAC score. Future studies may consider weighing potentially life-saving procedures higher in measures of procedural competence.

Procedural competency is essential for pediatric hospitalists, and a variety of strategies, alone or in combination, may help establish and maintain procedural skills. Defining the scope of clinical practice is critical for this newly established specialty, and this study provides insight into current practice, which may inform future curricular and faculty development.

Minor Procedure Study Group: Laura Goldstein, Section of Hospital Medicine, Children’s Hospital of Philadelphia, Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; Joanne Mendoza, Division of Pediatric Hospital Medicine, Children’s Specialty Group, Children’s Hospital of the King’s Daughters and Eastern Virginia Medical School, Norfolk, Virginia; Meghan E. Hofto, Division of Hospital Medicine, Department of Pediatrics, University of Alabama at Birmingham and Children’s of Alabama, Birmingham, Alabama; Sean O’Brien, Division of Hospital Medicine, Department of Pediatrics, University of Minnesota Masonic Children’s Hospital, Minneapolis, Minnesota; Melissa Cameron, Division of Pediatric Hospital Medicine, Rady Children’s Hospital-UC San Diego, San Diego, California; Walter Wickremasinghe, Department of Medicine, Boston Children’s Hospital, Boston, Massachusetts; Austin Ostermeier, Division of Hospital Medicine, St Louis Children’s Hospital, Department of Pediatrics, Washington University in St Louis, St Louis, Missouri; Bo Stapler, Department of Pediatrics, St Vincent Hospital, Billings, Montana; Kruti Bhavsar, Pediatric Hospital Medicine, Spence and Becky Wilson Baptist Children’s Hospital, Memphis, Tennessee; Joseph Fakhoury, Pediatric Hospital Medicine, Bronson Children’s Hospital, Department of Pediatric and Adolescent Medicine, Homer Stryker MD School of Medicine, Western Michigan University, Kalamazoo, Michigan; Ann Allen, Division of Hospital Medicine, Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin; Suresh Nagappan, Cone Health, Department of Pediatrics, University of North Carolina, Greensboro, North Carolina; Jeff Lancaster, Division of Pediatric Hospital Medicine, Department of Pediatrics, West Virginia University/WVU Medicine Children’s Hospital, Morgantown, West Virginia; Heather Dyer, Division of Hospital Medicine, Dayton Children’s Hospital, Department of Pediatrics, Boonshoft School of Medicine, Wright State University, Dayton, Ohio; Constanza Burciaga Calderoni, Department of Medicine, Loma Linda University Hospital. Department of Pediatric Hospitalists, Loma Linda University Children’s Hospital, Loma Linda, California; J.B. Lykes, Wakemed Children’s Hospital, Raleigh, North Carolina; Marie Dawlett, Division of Pediatric Hospital Medicine, Department of Pediatrics, University of Texas Medical Branch, Galveston, Texas; Melissa Drewry, Division of Pediatric Hospital Medicine, Corewell Health East William Beaumont University Hospital, Royal Oak, Michigan; Emilee Colella Lewis, Division of Hospital Pediatrics, University of North Carolina Chapel Hill, N.C. Children’s Hospital, Chapel Hill, North Carolina; Lauren Marlowe, Pediatric Hospital Medicine, Children’s Hospital of Philadelphia Care Network at Virtua Health System, Voorhees, New Jersey; Anh Vu Le, Division of Pediatric Hospital Medicine, Department of Pediatrics, Loma Linda University Children’s Hospital, Loma Linda, California.