BACKGROUND: In recent years, the focus on patient safety and housestaff supervision has led to a steady increase in institutions providing 24/7 in-hospital (also known as in-house, henceforth referred to as IH) coverage by pediatric intensivists. Effects of this increased attending physician presence on education of pediatric housestaff have not been studied. We hypothesized that IH coverage would decrease perceived autonomy of housestaff and negatively affect their preparedness to be independent attending physicians on completion of training.
METHODS: A secure, anonymous, Web-based survey was sent to pediatric intensivists in the United States and Canada, and pediatric critical care fellows and pediatric residents at academic centers across the United States. Questions focused on perceptions of IH coverage and housestaff educational experience.
RESULTS: We report 1323 responses from 147 institutions (center response rate 74%). Although 96% of respondents stated that the PICU provides “a good educational experience,” only 50% of pediatric intensivists and 67% of housestaff feel that housestaff are prepared for independent practice after training in an IH model. Compared with those training in home-call models, respondents currently working in IH models have more favorable perceptions of the effects of IH coverage on housestaff autonomy (P < .0001), supervision (P < .0001), and preparation for independent practice (P < .0001).
CONCLUSIONS: Pediatric intensivists and housestaff express concern regarding the preparation of housestaff training in a 24/7 IH attending model. An important priority for institutions using or considering a 24/7 IH attending coverage model is the balance between adequate housestaff supervision and autonomy.
- graduate medical education
- professional autonomy
- patient safety
- personnel staffing
- after hours care
- ACGME —
- Accreditation Council for Graduate Medical Education
- HC —
- home coverage
- IH —
- IQR —
- interquartile range
- PALISI —
- Pediatric Acute Lung Injury and Sepsis Investigators
- PGY —
- postgraduate year
What’s Known on This Subject:
Increasing numbers of hospitals are instituting 24/7 in-hospital pediatric intensivist coverage. Data regarding patient outcomes are mixed and the impact on housestaff education remains unknown.
What This Study Adds:
This study quantifies the perceived impact of in-hospital attending coverage on pediatric resident and critical care fellow education and also investigates the growing concern that increasing supervision may contribute to housestaff being less well prepared for independent clinical practice.
In academic medical centers, increasing emphasis is being placed on housestaff supervision and attending physician presence.1–4 Governing bodies such as the Accreditation Council for Graduate Medical Education (ACGME), the Society of Critical Care Medicine, and the American Academy of Pediatrics have recently developed guidelines that more specifically delineate the necessary supervision of housestaff and coverage of ICUs.1,3 These guidelines have been designed to improve both patient safety and graduate medical education in the ICU environment.5,6 The purported benefits of increased attending physician presence include improved patient outcomes, along with enhanced housestaff supervision and education.5,7–10
However, currently available data do not clearly demonstrate that increased attending physician presence positively affects clinical outcomes.11,12 In lieu of a clear benefit on outcomes, a crucial consideration is the impact of increased attending physician presence on housestaff autonomy and education. Autonomy is a key element of housestaff preparation for independent practice, and there are concerns that increased attending physician presence may negatively affect this element of graduate medical education.13–16 Nowhere is the conflict between patient care and housestaff autonomy stronger than in the PICU, where housestaff care for the sickest children in the hospital. We hypothesized that 24/7 coverage by intensivists at academic institutions would decrease perceived autonomy of housestaff and negatively affect their perception of preparation to be independent attending physicians on completion of training.
In conjunction with the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network, we developed an 80-question, logic-imbedded, Web-based survey (the Survey of In-house Coverage by Pediatric Intensivists trial17) and distributed the survey by using the secure RedCAP database.18 After institutional review board exemption, this survey was delivered in August 2012 to representatives at PALISI network sites across the United States and Canada, who were asked to forward the survey on to all pediatric intensivists at their institutions. In January 2013, the survey was also delivered to pediatric critical care fellowship program directors and pediatric residency program directors in the United States, who were asked to forward the survey to pediatric residents and pediatric critical care fellows (jointly referred to as “housestaff”) and, if not a PALISI network site, pediatric intensivists at their institution.
Survey questions focused on perceptions of education related to in-hospital (IH) coverage. IH coverage was defined as the presence of a pediatric intensivist in the hospital 24 hours per day, 7 days a week, whereas home coverage (HC) was defined as pediatric intensivists taking calls from outside the hospital during nontraditional hours (nights and weekends). Mixed coverage models involved fewer than 7 nights per week covered by an intensivist in the hospital.
A 4-point Likert scale (strongly agree, agree, disagree, or strongly disagree) was used for subjective questions. Data were analyzed by using JMP version 10 (SAS Institute, Inc, Cary, NC). Continuous variables were compared by using the Wilcoxon rank sum test; dichotomous variables were compared by using the Fisher exact test. P values <.05 were considered significant. Unless otherwise stated, providers in mixed models were excluded from analyses comparing IH with HC. Postgraduate year (PGY)-1 residents were excluded from analysis of questions regarding procedures, as many had not yet rotated in the PICU at the time of survey distribution. To account for sampling bias from clustering (ie, multiple responses from a single institution), we completed a secondary analysis excluding all data from centers that individually generated >1% of the responses.
We received 1323 responses from pediatric intensivists, pediatric critical care fellows (subsequently referred to as “fellows”), and pediatric and medicine-pediatric residents (subsequently referred to as “residents”). Response rate by center was 147/200 (74%), including 100/200 (50%) for intensivists, 50/59 (85%) for pediatric critical care fellowship programs, and 94/192 (49%) for pediatric and medicine-pediatric residency programs. Four residency program directors declined to forward the survey; 3 cited “survey fatigue” and 1 cited institutional review board restrictions. Characteristics of respondents are summarized in Table 1. Sixty-seven percent of intensivists trained in an HC model, compared with only 23% of current housestaff (P < .0001).
Respondents from institutions with IH coverage were more likely to have a fellowship program (66% vs 44%, P < .0001) and coverage by midlevel providers (physician assistants and/or nurse practitioners) (69% vs 31%, P < .0001). Respondents from centers with mixed-coverage models were less likely to report 24/7 IH coverage by fellows or midlevel providers (58%), compared with HC (75%) or IH (70%) models (P = .007). Centers with IH coverage also had more pediatric critical care beds (median 26 vs 18, P < .0001).
Seventy percent of housestaff stated that they would prefer to work in an institution with IH coverage, but this preference was much greater for those housestaff currently working in a center with IH coverage (85% vs 40%, P < .0001). Seventy-seven percent of housestaff stated that they would prefer to work in an institution featuring the same coverage model in which they are currently training.
Overall Educational Experience
Ninety-six percent of all respondents (1038/1079) felt that working in the PICU was a good educational experience; 68% of intensivists and 68% of fellows stated that IH models are good for education, in comparison with 88% of residents (P < .0001). Intensivists, fellows, and residents working in an IH model each were more likely to state that IH coverage was good for education (all P values <.002). Table 2 further compares perceptions of the PICU educational experience among providers in IH and HC models.
Autonomy and Supervision
Intensivists (72%) and fellows (68%) were more likely than residents (50%) to state that IH models limit autonomy (P < .0001 for each). Across all roles, providers currently working in IH models were less likely to state that IH coverage limits autonomy (all P values <.03). Similarly, intensivists and fellows currently working in IH models were more likely to state that IH coverage improves supervision of housestaff (P < .0001); however, this perception was not uniformly stated by residents (P = .07). Fellows (95%) were more likely than residents (72%) to state that an attending physician would be at the bedside when assistance was needed (P < .0001). However, residents in IH coverage models were more likely to state that an attending would be at the bedside when assistance was needed, in comparison with those in HC models (84% vs 61%, P < .0001). Ninety-six percent of fellows stated that they had appropriate clinical responsibility in the PICU; however, this was less for fellows currently working in IH coverage models (94% vs 100%, P = .03). The frequency of residents stating that they had appropriate clinical responsibility in the PICU (83%) did not differ by coverage model.
Preparation for Independent Practice
Ninety-eight percent of fellows and 69% of residents stated that they were comfortable responding to pediatric emergencies, and this did not differ by coverage model. Only 50% of intensivists stated that housestaff will be prepared to be independent attending physicians after training in an IH model, in comparison with 67% of housestaff (P < .0001). In addition, fellows (59%) were less likely than residents (70%) to state that housestaff would be prepared after training in an IH model (P = .008). Although intensivists working in an IH model were significantly more likely to state that IH models provide adequate training, almost half (46%) of these intensivists still disagreed that IH coverage prepared housestaff for independent practice. Furthermore, 11% of intensivists and 10% fellows strongly disagreed that housestaff would be prepared after training in an IH coverage model. In contrast, most (81%) respondents stated that housestaff would be prepared after training in an HC model, with providers currently in HC models more likely to state this perception (89% vs 75%, P < .0001). This perception did not vary by respondent’s role (intensivist versus housestaff).
Fellowship Versus Nonfellowship
Thirty-nine percent of residents stated that their institution has a pediatric critical care fellowship program. Table 3 summarizes comparisons of resident perceptions from institutions with and without fellowship programs. Fewer residents at institutions with fellowship programs perceived that they had appropriate responsibility for patient care during their PICU rotation, but this finding did not reach statistical significance (P = .07). In addition, there was no difference in overall educational experience, respect as a team member, or comfort with pediatric emergencies across the 2 models. Generally, residents’ perception of IH coverage effects on their education did not differ based on presence of a fellowship program.
Resident Procedural Experience
Only 41% of residents reported that they had adequate procedural opportunity while on their PICU rotation. Residents were less likely to state that they had adequate procedural opportunity if their institution had IH coverage (36% vs 51%, P = .008) or a fellowship program (25% vs 48%, P < .0001). Reported procedural experience for residents (PGY-2 and above) is summarized in Table 4. Of note, fewer than half of residents attempted any of the surveyed procedures during their PICU rotations, with the exception of bag mask ventilation.
We identified 18 institutions, which individually accounted for >1% of the total response; the largest single-center response was 2.6% of the total (35 responses). Generally, these institutions were large hospitals, with respondents reporting a median of 32 pediatric critical care beds (interquartile range [IQR] 20–44). A fellowship program was also reported by 92% of respondents from these institutions. Cumulatively, these institutions accounted for 352/1323 (27%) of the total responses, 164/457 (36%) of intensivists, 100/174 (58%) of fellows, and 102/619 (16%) of residents. When the responses from these 18 centers were removed from data analysis, the previously reported statistical analysis remained largely unchanged.
Notably, for the data presented in Table 2, changes on exclusion of these high-response centers included loss of statistical significance for comparisons of fellow perception of preparation after training in an HC model and resident perception of the effects of IH coverage on housestaff autonomy. In contrast, clustering analysis led to the following comparisons becoming statistically significant: intensivist perception of limited housestaff autonomy when the attending was IH (IH 67% versus HC 81%, P = .02), likelihood for residents to again choose this training program (IH 83% versus HC 93%, P = .02), and resident perception of supervision when attending was IH (IH 92% versus HC 83%, P = .02). For the data in Table 3, clustering analysis negated the statistical difference between resident perceptions of supervision related to IH coverage based on presence of a fellowship.
Concerns for patient care and safety have appropriately been the impetus for increased IH coverage by intensivists, but the impact on housestaff autonomy and education are important additional considerations that have not been adequately studied. This investigation demonstrates that both intensivists and housestaff express significant concerns regarding the preparedness of housestaff after training at institutions with IH coverage. The increased presence of attending physicians likely leads to more direct intensivist involvement in decision-making, potentially adversely affecting the development of housestaff independence. Similar concerns regarding critical care fellow autonomy were evident in a recent survey of primarily adult critical care fellows and program directors.19 In addition to preparation for independent practice, concerns have been raised regarding pediatric resident preparation for fellowship, as evidenced by the formation of a “Fellowship Development” action team by the Council of Pediatric Subspecialties.20
In general, intensivists in this survey had more negative perceptions of the effects of IH coverage on education than housestaff. This difference may simply represent a bias toward the type of coverage model in which most of these intensivists trained. However, most pediatric critical fellows are currently training in either an IH or a mixed coverage model, and their views on the educational effects of IH coverage often aligned more with intensivists than with residents. This alignment may be related to the advanced level of training among fellows, making them more equipped to make independent decisions and, therefore, placing a higher priority on autonomy. Similarly, as residents only briefly rotate through the PICU, they may be more appreciative of supervision and less focused on gaining independence in the management of critically ill infants and children.
Respondents currently at institutions with IH coverage were uniformly more positive of IH coverage models than those currently working in HC models. This difference could be explained by familiarity with 1 model over another, self-selection into 1 style of coverage, or by IH centers that may have found an appropriate balance between supervision and autonomy. In addition, this finding may also indicate that perceptions of deleterious effects of IH coverage are worse than the reality of IH coverage. One of the challenges in investigating the impact of IH coverage models in actual practice is that different programs use a wide variety of staffing models to achieve “off-hour” attending coverage, which may also explain the conflicting data on the patient effects of IH coverage.9,11,21–23 Each individual model may place greater emphasis on either supervision or autonomy, and just as increased supervision may not always lead to improved outcomes,11,23 increased attending physician presence may not always lessen housestaff autonomy.24
The new ACGME standards establishing the various levels of supervision for housestaff appropriately address the importance of autonomy as an element of the progression toward independent practice.1 Many of the clinical decisions made during times traditionally covered by attending physicians from home involve substantial independent thought and decision-making on the part of housestaff. Although there may be less autonomy when an attending physician is present, a potential benefit of increased attending physician presence during off hours is additional individualized educational time for housestaff. Bedside teaching during patient management is an important element of housestaff education, and given the increasing prevalence of night shifts in the era of duty-hour reform, structured nighttime curriculums may be added to enhance the housestaff educational experience.25
Bedside teaching in the PICU is also an important expectation of critical care fellows. In programs with fellowship training programs, a fellow is often omnipresent for resident education, assistance, and supervision. However, as fellows are providing care in the busy PICU environment and also in training themselves, resident education may not always receive a high priority. In this context, it is possible that intensivist educational efforts may be directed primarily toward the fellows instead of the residents. Given these factors, there is the potential that the impact of critical care fellowship training on resident education in the PICU could be either positive or negative. In this investigation, we demonstrated that although overall responsibility (autonomy) and procedural experience may have been slightly lower, the overall perceived educational experience did not suffer due to the presence of fellows.
Resident procedural experience is a small but potentially important aspect of pediatric critical care education. Opportunities for procedures, such as central line insertion and intubation, vary among training programs,26 and there remains controversy regarding the safety of learners performing high-risk procedures.27 This survey included 3 procedures (peripheral intravenous line insertion, neonatal intubation, and bag-mask ventilation) for which the ACGME requires competence for general pediatric housestaff.28 However, our data demonstrate that residents consistently have minimal exposure to these required procedures, regardless of intensivist coverage model. To achieve basic procedural competency among general pediatricians, training programs must investigate opportunities for residents to safely learn and practice these procedures, which may or may not be as a part of PICU rotations. Thus, as this survey asks only about the PICU experience, it is also important for training programs to explore other venues to teach procedures, including high-fidelity simulation.29
Although these data are interesting and revealing, all surveys are limited by self-reporting, which may not reflect the reality of the studied actions or behaviors. Reporting bias in this study is mitigated by our large response rate. Another limitation of this study is reliance on institutional representatives for distribution of the survey. Due to this methodology, we do not have an individual response rate; however, this survey represents a large sample size, encompassing at least 147 academic centers, more than 20% of the US pediatric intensivists (likely a higher percentage of academic intensivists),30 more than 60% of the US pediatric critical care fellows,31 and a large number of pediatric residents (albeit a much smaller percentage [∼8%] of the US total).32 It must also be noted that this survey is focused on perception of IH coverage models rather than objective measures of IH coverage effects.
The delicate balance between supervision and autonomy depends on a number of variables, including the personality and style of the attending and trainee, previous experience and comfort level of the attending/trainee, and complexity of the clinical decision.16,33,34 Despite patient safety concerns of housestaff independence, there are data to suggest that autonomous decisions made by housestaff during the learning process do not have a negative impact on patient care.23,35,36 Moving forward, it is imperative for attending physicians to promote autonomous decisions by housestaff, with proper support and supervision. Although focus must remain on the safest possible care of our patients, we must also ensure that we are adequately preparing future pediatricians and pediatric intensivists to optimally and independently care for patients.
Although the overall perceived educational experience in the PICU is positive, pediatric intensivists, pediatric critical care fellows, and pediatric residents express concern regarding the preparation of housestaff training under 24/7 IH attending models. Balancing autonomy with adequate supervision should be an important priority for institutions using or considering 24/7 IH intensivist coverage.
- Accepted September 19, 2013.
- Address correspondence to Kyle J. Rehder, MD, Division of Pediatric Critical Care Medicine, DUMC Box 3046, Durham, NC 27710. E-mail:
Dr Rehder conceptualized and designed the study, supervised data collection, completed data analysis, and drafted the initial manuscript; Dr Cheifetz conceptualized and designed the study and critically reviewed the manuscript; Dr Willson assisted with study design and critically reviewed the manuscript; Dr Turner assisted with conceptualization and study design, participated in data analysis, and critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
COMPANION PAPER: A companion to this article can be found on page 131, and online at www.pediatrics.org/cgi/doi/10.1542/peds.2013-3493.
- ↵Duty hour standards. 2011. Available at: www.acgme.org/acgmeweb/tabid/271/GraduateMedicalEducation/DutyHours.aspx. Accessed October 25, 2013
- Angus DC,
- Shorr AF,
- White A,
- Dremsizov TT,
- Schmitz RJ,
- Kelley MA,
- Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS)
- Haupt MT,
- Bekes CE,
- Brilli RJ,
- et al.,
- Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine
- Gajic O,
- Afessa B,
- Hanson AC,
- et al
- Leung WC
- Rehder KJ,
- Cheifetz IM,
- Markovitz BP,
- Turner DA
- ↵Council of Pediatric Subspecialties. Action teams. Available at: www.pedsubs.org/taskforce/index.cfm. Accessed April 18, 2013
- Reriani M,
- Biehl M,
- Sloan JA,
- Malinchoc M,
- Gajic O
- ↵Haber LA, Lau CY, Sharpe BA, Arora VM, Farnan JM, Ranji SR. Effects of increased overnight supervision on resident education, decision-making, and autonomy. J Hosp Med. 2012;7(8):606–610
- Sanders RC Jr,
- Giuliano JS Jr,
- Sullivan JE,
- et al.,
- National Emergency Airway Registry for Children Investigators and Pediatric Acute Lung Injury and Sepsis Investigators Network
- ↵Accreditation Council for Graduate Medical Education. ACGME program requirements for graduate medical education in pediatrics. 2012. Available at: www.acgme.org/acgmeweb/Portals/0/PFAssets/2013-PR-FAQ-PIF/320_pediatrics_07012013.pdf. Accessed September 4, 2013
- ↵American Board of Pediatrics. Number of diplomate certificates granted. 2012. Available at: https://www.abp.org/ABPWebStatic/?anticache=0.20577792472794532#murl%3D%2FABPWebStatic%2FaboutPed.html%26surl%3D%2Fabpwebsite%2Fstats%2Fnumdips.htm. Accessed April 9, 2013
- ↵National Resident Matching Program. Pediatric specialties fall match. 2011. Available at: www.nrmp.org/match-data/fellowship-match-data/. Accessed April 18, 2013
- ↵National Resident Matching Program. 2013 Main residency match. Available at: www.nrmp.org/match-data/main-residency-match-data/. Accessed April 18, 2013
- Johnson T
- Copyright © 2014 by the American Academy of Pediatrics