Pediatric Faculty and Trainee Attitudes Toward the COVID-19 Pandemic

We distributed an anonymous, Internet-based survey to all pediatric faculty and pediatric resident and fellow trainees at a 315-bed, freestanding, tertiary-care children’s hospital. Residents included categorical pediatrics, medicine-pediatrics, triple board (pediatrics, psychiatry, and child and adolescent psychiatry), and child neurology trainees. Fellows included pediatric medical subspecialty fellows. We did not include any visiting, nonpediatric residents or surgical fellows because their perspectives may be reflective of their primary institutional environment outside of our stand-alone children’s hospital.

We sent the survey link via an e-mail to an internal distribution list, and data were collected during a 7-day period from April 16, 2020, to April 23, 2020. Study data were collected by using Qualtrics (Qualtrics, Provo, UT) hosted at the study institution. This study was deemed exempt by the institutional review board of the study institution (STUDY20040076).

In the survey, we evaluated respondents’ levels of concern related to SARS-CoV-2 infection, levels of trust in multilevel institutional recommendations related to the COVID-19 pandemic, and perspectives related to trainee roles during the COVID-19 pandemic. The survey consisted of a maximum of 21 mixed-format survey questions including Likert-type scales (8), multiple choice (12), and free responses questions (1) (Supplemental Information). We asked all respondents to create a unique 6-digit identifier to connect future survey responses as the COVID-19 pandemic progresses. All survey questions were voluntary.

Before distribution, the survey was assessed for face validity by a multidisciplinary group of experts in pediatrics, pediatric subspecialties, and medical education. The survey was pilot-tested by 2 pediatric subspecialists to assess clarity and readability and determine estimated time burden (∼3–5 minutes).

We calculated descriptive statistics for respondent demographics as well as their attitudes and perspectives toward the COVID-19 pandemic and trainee roles. We compared demographics between roles (faculty, residents, and fellows) using Fisher’s exact tests. We assessed the associations of respondent attitudes and perspectives by role and sex via logistic regression, adjusting age, race, and international versus American medical graduate and role or sex when appropriate. We did not include respondents who did not complete demographic information (n = 17) in these comparisons. We conducted all analyses using R version 3.6.0 statistical computing software (The R Foundation, Vienna, Austria).

We anticipated the total number of respondents to vary by question, given the voluntary nature of each question and the “skip and display” logic of the survey. The range of missing data for each question was n = 0 to 17. Missing data for each demographic question are denoted in Table 1. No sensitivity analyses were conducted regarding missing data, given the exploratory and descriptive nature of this study.

We analyzed comments from the free response questions using thematic analysis methodology to identify themes. A coinvestigator (E.A.S.) developed a codebook based on review of the comments. Two coders (E.A.S. and M.C.-H.) then reviewed each comment to apply the codebook. Using a consensus coding approach, the coders met virtually to review their work and discuss any discrepancies. The principal investigator (T.M.K.) was available to adjudicate any differences in interpretation. We identified central themes and representative quotations.

A total of 701 respondents (491 faculty members, 115 residents, and 95 fellows) received the survey link via e-mail. Forty-six percent (n = 320) completed the survey. Detailed demographic information of survey respondents is displayed in Table 1. Fifty-six percent of respondents (n = 180) were faculty, 16% (n = 51) were fellows, and 28% (n = 89) were residents. Among faculty respondents, 69% (n = 124) were members of the department of pediatrics. All 19 divisions and groups within the department of pediatrics were represented. Sixty-three percent (n = 189) of all respondents were women, and 75% (n = 230) were white. The demographics of survey respondents reflected the demographics of faculty and trainees at our institution.

Two respondents had tested positive for SARS-CoV-2 infection and no respondents were awaiting test results at the time of the survey. Overall, 73% of all respondents (n = 234) were somewhat or very concerned about their personal risk of SARS-CoV-2 infection. Of all respondents, 84% to 89% were somewhat or very concerned about loved ones, faculty, trainees, staff, and patients being infected.

Residents were significantly less concerned than fellows and faculty about personal risk (odds ratio [OR] = 0.47; 95% confidence interval [CI]: 0.24–0.92 and OR = 0.50; 95% CI: 0.31–0.82, respectively) (Fig 1A). They were also significantly less concerned than fellows and faculty about risk to faculty and trainees (faculty risk: OR = 0.40; 95% CI: 0.20–0.80 and OR = 0.39; 95% CI: 0.24–0.65; trainee risk: OR = 0.44; 95% CI: 0.22–0.87 and OR = 0.51; 95% CI: 0.31–0.84). No significant differences were noted across roles for risk for loved ones or staff. Fellows were more concerned about patient risk than residents and faculty (OR = 4.2; 95% CI: 1.75–10.0 and OR = 3.0; 95% CI: 1.39–6.67, respectively).

Male respondents were significantly less likely than female respondents to have concerns related to SARS-CoV-2 infection for themselves (OR = 0.46; 95% CI: 0.29–0.73), loved ones (OR = 0.38; 95% CI: 0.24–0.61), faculty (OR = 0.46; 95% CI: 0.28–0.73), trainees (OR = 0.44; 95% CI: 0.27–0.71), staff (OR = 0.62; 95% CI: 0.39–0.98), and patients (OR = 0.53; 95% CI: 0.32–0.87) (Fig 1B).

In Figure 2, responses related to the impact on HCWs are summarized. Nearly half of all respondents (48%, n = 152) were somewhat or very concerned about the impact of the COVID-19 pandemic on their role as the major provider for their family. Fellows and faculty were significantly more concerned than residents (OR = 2.62; 95% CI: 1.36–5.06 and OR = 4.03; 95% CI: 2.48–6.55, respectively). Forty-six percent of all respondents (n = 143) were somewhat or very concerned about changes to their salary. Fellows and faculty were significantly more concerned than residents about salary change (OR = 2.75; 95% CI: 1.46–5.15 and OR = 3.72; 95% CI: 2.29–6.05, respectively).

Approximately one-quarter of all respondents (24%, n = 75) were somewhat or very concerned because of personal risk factors, with no differences noted across roles or sex. Fifty-three percent of all respondents (n = 166) were somewhat or very concerned because of family or loved ones’ risk factors, with men having significantly less concern than women (OR = 0.45, 95% CI: 0.29–0.70).

Forty-three percent of all respondents (n = 136) were somewhat or very concerned about redeployment. Men were less concerned than women (OR = 0.43; 95% CI: 0.27–0.67). Fellows were significantly more concerned than residents or faculty (OR = 2.61; 95% CI: 1.38–4.94 and OR = 2.52; 95% CI: 1.40–4.52, respectively). Seventy-nine percent (n = 250) were somewhat or very concerned about lack of PPE, with no differences noted across roles. Men were less concerned than women (OR = 0.50; 95% CI: 0.32–0.78).

Seventy-one percent of all respondents reported that they somewhat or completely trusted the state government, 74% trusted the Centers for Disease Control and Prevention (CDC), 18% trusted the federal government, and 70% trusted the World Health Organization (WHO). Faculty were 3 times more likely to trust state government recommendations than fellows (OR = 3.01; 95% CI: 1.58–5.72) and 2.4 times more likely than residents (OR = 2.35; 95% CI: 1.37–4.03) (Fig 3). Men were significantly more likely to trust federal government recommendations (OR = 1.61; 95% CI: 1.03–2.50).

Nearly three-fourths of all respondents (72%, n = 226) feel trainees are essential personnel (Fig 4A). Residents were 6.6 times more likely than fellows to believe that trainees are essential (86% vs 52%, OR = 6.67; 95% CI: 2.46–17.58). Faculty were 2.2 times more likely than fellows to believe that trainees are essential (72% vs 52%, OR = 2.18; 95% CI: 1.06–4.47). No differences based on sex or between faculty and residents were noted.

Forty-four percent of all respondents (n = 138) felt trainees participating in direct patient care during the COVID-19 pandemic was very or extremely important (Fig 4B). Residents were 5.2 times more likely than fellows and 2.1 times more likely than faculty to believe that direct patient care for trainees was important (OR = 5.16; 95% CI: 2.60–10.24 and OR = 2.11; 95% CI: 1.29–3.43, respectively). Faculty were 2.5 times more likely than fellows to believe that direct patient care was important (OR = 2.45; 95% CI: 1.32–4.55). Men were 2 times more likely than women to view direct patient care as important for trainees (OR = 1.98; 95% CI: 1.27–3.10). Two-thirds of all respondents (67%, n = 209) felt trainees participating in indirect patient care (including rounding remotely and/or telemedicine) was very or extremely important (Fig 4C).

Respondents were asked about the balance between increased trainee risk of infection and reduced clinical training at the current time, in 2 months, and in 6 months (Supplemental Fig 5). At the current time, residents and faculty were 2.5 (OR = 2.51; 95% CI: 1.10–5.71) and 2.6 times (OR = 2.60; 95% CI: 1.24–5.45), respectively, more likely than fellows to believe that the balance lies in the direction of increased infection risk and usual patient care responsibilities. When considering the balance in 2 and 6 months’ time, there were no differences across the roles. Men were more likely than women to believe that the balance lies in the direction of increased infection risk and usual patient care responsibilities at all time points (current: OR = 1.89; 95% CI: 1.17–3.04; at 2 months: OR = 1.98; 95% CI: 1.23–3.17; at 6 months: OR = 1.67; 95% CI: 1.05–2.65).

Twenty-two percent (n = 69) of all respondents elected to type a free response comment at the end of the survey. The following 4 major themes emerged: (1) concerns related to institutional messaging and policy; (2) uncertainty during the COVID-19 pandemic; (3) concerns related to trainee education; and (4) concerns related to resource availability and use. Major themes are illustrated by representative quotations in Supplemental Table 2.

In this study, we both assess the concerns of pediatric faculty, fellows, and residents simultaneously and explore their perceptions of the role of trainees as essential personnel during the COVID-19 pandemic. Notably, all respondents endorsed high levels of concern related to SARS-CoV-2 infection and shortages of PPE. However, faculty and trainee responses were not uniform across levels of training. Pediatric fellows and faculty were more concerned than residents about risk of infection and the personal financial impact of the pandemic, which may be influenced by differences in age and life stages. Faculty and fellows were older than resident respondents and, therefore, potentially more likely to be responsible for supporting families of their own and/or caring for elderly relatives. Most pediatric trainees and faculty believe that trainees are “essential personnel” and important in direct patient care, although there were significant differences in this belief across roles. The majority of faculty and trainees are receptive to returning to usual patient care and training over time.

These survey results echo similar investigations from both the United States and abroad, in which trainees express a willingness to serve in direct patient care settings during the pandemic coupled with concerns related to the balance of service with safety.⁶  In a survey of Singaporean resident trainees across a variety of subspecialties, respondents also reflected on the challenging balance of service versus training and reported varying levels of stress during the pandemic.⁷ 

The 2013 Essential Services Act defines the term “essential employee” as a person who “performs work involving the safety of human life or the protection of property, as determined by the head of the agency.”⁸  Our survey found that, although the majority of respondents believed that pediatric trainees are essential, fellows were significantly less likely to endorse this belief compared with residents or faculty. Fellows were also less likely to believe that trainees should participate in direct patient care and favored limiting exposure risk at the expense of training more than residents or faculty. In our institution, residents and faculty are typically mandatory components of patient care, whereas fellows are often an additional layer of care on subspecialty or consulting services, which may send an implicit message of “non-essentialism.” Exploring fellow roles and attitudes across institutions with different clinical care structures and staffing models may help to conceptualize these findings.

The restructuring of medical training programs and clinical care requirements to minimize exposure to SARS-CoV-2 infection and protect trainees has been described by a number of programs as the COVID-19 pandemic progresses in the United States.^7,9–12  The impact of decreased direct patient care and training during the COVID-19 pandemic on residency and fellowship program accreditation remains a matter of active speculation.¹³  The Accreditation Council for Graduate Medical Education released principles to guide actions early in the COVID-19 pandemic and a set of inviolate requirements that should be ensured for trainee safety.¹⁴  Additionally, program directors reiterate the delicate balance between trainee risk, need for continued training, and “duty to serve” as physicians in a public health crisis.^10,13,14 

Our study revealed significant differences in concerns based on sex, with female respondents endorsing higher levels of concern than their male colleagues. In a recent survey of trainees, Kannampallil et al¹⁵  also found that female trainees were significantly more likely to report higher stress levels than their male colleagues. We hypothesize that this difference may be reflective of existing sex inequities in pediatric medicine because women are more likely to be frontline practitioners whereas men are more likely to have administrative and leadership roles.^16–18  Additionally, women are more likely to be responsible for household obligations and caring for children and family members than their male partners,¹⁹  which may make them more aware of the implications of school and child care closures. Because women are more often primary caregivers, female physicians may harbor increased anxiety about becoming infected because it would preclude them from being able to provide this care. Similar fears may make female physicians more concerned than male physicians about the risk of exposing loved ones to the virus.²⁰ 

At the time of this survey, guidelines related to the COVID-19 pandemic were changing constantly, which likely contributed to the varying degrees of trust seen in our results. The high levels of trust in the state government may be reflective of the strong state-level response with frequent, clear communication. Because all respondents were physicians, the high levels of trust in health care organizations (the CDC and the WHO) reported in our survey may be reflective of historical reliance and trust in these institutions. At an institutional level, daily communication and updates were essential to mitigating stress and increasing trust.

The experience of pediatric emergency medicine fellowship programs in New York City at the height of the pandemic supports the importance of open communication. The pediatric emergency medicine programs’ approach involved, “constant communication, with most programs communicating with individual fellows daily, instituting weekly or biweekly huddles, and engaging in frequent group chats or text messages.”²¹  The differences in concerns and beliefs across training levels highlighted in our survey results underscore the importance of transparent, timely, and bidirectional communications with frequent solicitation of feedback from HCWs at all levels of employment and training. Although many people face emotional stressors, uncertain financial futures, and social isolation during the pandemic, HCWs make additional unique sacrifices with additional unique stressors.^22–25  Additionally, trainees exposed to COVID-19 patients have reported higher levels of stress and burnout.¹⁵  Thus, the baseline high levels of concern noted in our study despite low local COVID-19 prevalence are notable. We believe these results highlight the need for additional support for trainees and faculty alike during this time and during future disasters. Financial support and creative child care solutions should be strongly considered for trainees and faculty with child care responsibilities in the setting of school and day care closures. Special consideration should be given to career development and scholarly productivity because these efforts are likely to be compromised during this time.

Respondents endorsed anxieties related to access to PPE, exposing loved ones to SARS-CoV-2 infection, inconsistent messaging, and overall uncertainty. These concerns were similar to those recorded by Sanghavi et al,²⁶  who found 50% of pediatric residents had an increased level of anxiety, 50% were concerned about spreading the virus to loved ones, and 60% were concerned about limited PPE. Notably, the aforementioned study was performed at the same time as ours but in a different region of the country. The similarities in resident concerns across geographical distance support the generalizability of our results. Moreover, these concerns were also raised by health care providers during the 2003 severe acute respiratory syndrome epidemic and 2009 influenza A/H1N1 pandemic.^27–29  Acknowledging these shared concerns can increase social connectedness in this time of distance and isolation.

Our study has several limitations. This study was conducted at a single US tertiary-care pediatric hospital, which may limit generalizability to other practice settings. Our response rate of 46% (77% for pediatric residents, 54% for fellows, and 37% for faculty) is consistent with existing literature regarding response rates of health care professionals.^30–32  In addition, our study relied on respondent self-report of their attitudes, which may be subject to social desirability bias. Furthermore, respondents most interested in the topic may have completed the survey, leading to response bias. We also did not assess the perspectives of community providers, further limiting generalizability. Finally, although this survey took place at a set point in time, the United States has struggled with containment, and the first wave of the pandemic has continued. In this study, we capture the responses of pediatric residents, fellows, and faculty during a period of exponential pandemic growth worldwide, with relatively low rates of local infection with adequate availability of PPE. Thus, we believe our results offer a unique perspective of the concerns of pediatric trainees and faculty in a low prevalence area. Because the incidence of COVID-19 in children has remained low, our experience is likely to reflect that of most pediatric training programs across the country.

The differences noted across training level in this survey suggest that a simultaneous national survey of pediatric residents, fellows, and program directors would provide valuable insight. As the pandemic progresses, continued inquiry into evolving concerns and perspectives of pediatric faculty and trainees will be vital, and a follow-up survey at our institution to assess changes over time and the impact of reopening is ongoing. Future work should include perspectives of outpatient pediatricians and providers as well as other pediatric staff.

Because SARS-CoV-2 is likely to remain a prominent aspect of the health care environment for the foreseeable future, increased trainee exposure is seemingly unavoidable. Most pediatric trainees and faculty are receptive to the increased exposure risk that accompanies the return to routine schedules. However, the high levels of concern endorsed by all providers underscore the importance of adequate PPE, consistent institutional messaging, and respect for the personal and professional sacrifices incurred as a result of the COVID-19 pandemic.