Impact of COVID-19 on the Sense of Job Security Among Community Pediatric Hospitalists

This study was part of a larger quantitative project investigating community pediatric hospitalists’ career motivators. The survey development began in 2019 and questions regarding the effects of COVID-19 on community pediatric hospitalists were added during the onset of the pandemic. There were 42 survey items, including 15 on career motivators, 11 regarding the impact of COVID-19, and 16 regarding self-reported demographic and workplace characteristics. The survey was drafted through an iterative process by the authors. Contributors to the study design included a community pediatric hospitalist, a community pediatric hospitalist with institutional and national leadership roles, and a university-based hospitalist with institutional and national leadership roles. These investigators had a combined 36 years in PHM at the time of survey development. The order of the questions, format, and wording were reviewed and revised by the institution’s Survey Laboratory staff before cognitive think aloud interviews with 5 additional pediatric community hospitalists. Questions were further refined, and the revisions were finalized in collaboration with the Survey Laboratory staff and investigators. The Survey Laboratory staff programmed the questions into Qualtrics, an online Web survey platform. This study was deemed IRB exempt by the institution’s IRB. The survey questionnaire is available in Supplemental Information.

A convenience sample of community pediatric hospitalists was obtained through direct contact with several community PHM programs known to the investigators. Based on the preferences of the individual sites, a survey link was either emailed directly to participants (147 direct e-mail links) or disseminated by the site lead (122 total individuals) for a total of 269 links. The surveys were identical, though the participants who received their link from the site leads could not re-enter the survey to complete it if they were interrupted.

Initial invitation emails were sent in March 2021. Reminders were sent 2, 3, and 4 weeks after the initial e-mail to the direct e-mail participants. Site leads were sent reminders twice after the initial e-mail. One day before the survey closing, the 4 direct-link cases that were started but incomplete were sent an e-mail reminder to complete the survey. The survey closed in April 2021.

The survey gathered data regarding demographics and workplace characteristics, changes in compensation, furlough, and concern about job security. Geographic region was defined in accordance with the United States Census Bureau and the remaining variables were self-reported. Worry about job security was defined as the respondents’ self-reported degree of worry about their job being permanently terminated utilizing a 5-point Likert scale from 1 = “extremely worried” to 5 = “not worried at all.” Response rate was calculated using the American Association for Public Opinion Research’s standard response rate 6, which includes partial cases in the numerator as well as the denominator and no adjustment for unknown eligibility. Exploratory data analysis was performed to assess the univariate association between worry about job security and demographic, workplace, compensation, and furlough variables examined. The data were then analyzed using the cumulative logit proportional odds models as indicated to assess whether a change in degree of worry about job security was associated with gender, age, years of employment in pediatric hospital medicine, years of employment specifically in community pediatric hospital medicine, holding a leadership role, geographic region of practice, hospital setting (urban, suburban, or rural), being affiliated with a university-based center or free-standing children’s hospital, change in base pay, decrease in benefits, and having experienced furlough.

Gender comparisons were only able to be assessed between males and females as only a single participant reported their gender as “other.” The 5 age groups were consolidated into 3 groups for the analysis: 40 years or younger, 41 to 50 years, and 51 years or older. For analysis of the change in base pay, respondents reported that their pay was “increased, “unchanged,” “initially reduced, and now back to normal,” “reduced,” or “other, please explain.” For the small minority (6%) who reported experiencing “other” impact on pay, the free text responses were reviewed by the research team and recoded as “unchanged” or “reduced” based on consensus for the analysis (Supplemental Table 5). The single respondent who reported increased pay (n = 1) and the single respondent (n = 1) who reported increased benefits were excluded from the statistical analysis. All analyses were performed using SAS version 9.4 (SAS Institute, Inc, Cary, NC) and a 2-sided P value of <.05 was considered statistically significant.

Data were collected from 31 hospitals across the United States with 4 from the Northeast (13%), 11 from the Midwest (36%), 8 from the South (26%), and 8 from the West (26%). There were 126 completed cases, with 71 (56%) from the direct e-mail link group and 55 (44%) from the intermediary group. There were 6 partial cases, 2 from the direct e-mail link group and 4 from the intermediary group. The direct e-mail link cases had a response rate of 50% and the intermediary group had a response rate of 48% for a combined rate of 49%.

The majority of respondents (79%) identified as female. Approximately half (52%) of respondents were between 31 and 40 years old. Participants had a wide range of years of experience with the largest proportion of respondents working for ≤5 years in both pediatric hospital medicine in general (41%) and community pediatric hospital medicine specifically (47%). More than a third of respondents (41%) held a leadership role in their current job. Additional demographic data are in Table 1.

Each region of the United States was represented with nearly half of the respondents from the Midwest (49%). About two-thirds of the respondents worked primarily in a suburban setting (68%). About two-thirds (68%) of the respondents reported that their hospital was affiliated with a university-based center or a free-standing children’s hospital (Table 2).

Community pediatric hospitalists experienced tangible consequences as a result of the response to the COVID-19 pandemic, including changes in base pay, changes in benefits, and furlough. Although about two-thirds of respondents (67%) reported unchanged base pay at the time of survey administration, the remainder reported reduced pay (10%), initial reduction in pay that had been reversed (16%), or other change in pay (6%). Only 1 respondent reported increased pay. For nearly all respondents who experienced pay reductions, the pay reduction was involuntary (97%). Additionally, 41% reported reduced benefits. Finally, although most respondents did not experience mandatory (92%) nor voluntary furlough (97%) at the time of the survey, the small number of respondents who reported experiencing mandatory (8%) or voluntary (3%) furlough subsequently resumed their positions (Table 3).

Nearly two-thirds of respondents (64%) reported some degree of worry about job security (Fig 1). The univariate associations between the respondents’ degree of self-reported worry about job security and the variables are in Table 4. Demographic characteristics were assessed and odds of greater worry about job security did not vary with gender (P = .2), age (P = .4), years of employment in PHM (P = .8), years of employment in community PHM (P = .3), or holding a leadership role (P = .5).

Workplace characteristics were evaluated with respect to job security. Degree of worry about job security did not vary with geographic region of practice (P = .6). However, degree of worry about job security was significantly associated with primary hospital setting (P = .01). Hospitalists in suburban settings had greater odds of reporting worry about job security than those in rural settings (odds ratio [OR] = 4.23, 95% confidence interval [CI] = 1.48–12.04). There was no difference in degree of worry between hospitalists practicing in suburban areas compared with urban (OR = 1.97, CI = 0.85–4.58) nor in urban compared with rural (OR = 2.14, CI = 0.63–7.22). Notably, those affiliated with a university-based center or free-standing children’s hospital had greater odds of reporting worry about job security than those who were not (OR = 3.17, CI = 1.55–6.51, P = .002).

Compensation in relation to concerns about job security was also examined. Worry about job security varied with changes in base pay (P = .03). Hospitalists who had “initially reduced, but now back to normal” pay had greater odds of reporting worry about job security compared with those with “unchanged” pay (OR = 3.26, CI = 1.33–8.00). Worry about job security did not vary with reduction in benefits compared with unchanged benefits (P = .2) nor with mandatory furlough compared with no furlough (P = .2).

The COVID-19 pandemic was a significant stressor on the health care system.^7–12  This study is the first to examine the impact of such a stressor on compensation, furlough, and sense of job security among community pediatric hospitalists. About one-third of the hospitalists who participated experienced negative changes in pay, and nearly half experienced negative changes in benefits. Few participants experienced furlough. Additionally, nearly two-thirds of the community pediatric hospitalists surveyed reported some degree of worry about job security. None of the demographic factors examined were associated with increased odds of worry about job security, suggesting that workplace factors may have played a larger role.

Community pediatric hospitalists who experienced “initially reduced but now back to normal” pay demonstrated greater worry about job security. This initial reduction in pay may have been perceived as a sign of financial instability in the PHM program or the hospital itself, resulting in worry about job security. Although the sample may have been too small to capture all of the differences in sense of job security across the groups, this piece of information supports a growing collection of research and opinion statements that suggest that financial transparency can improve workplace culture.¹⁷  Thus, community hospitals may be able to support the well-being of its workforce and alleviate anxieties about job loss by sharing information about the financial health and fiscal plans of the hospital in times of normalcy and in times of stress.

Community pediatric hospitalists in suburban areas had increased odds of reporting greater worry about job security than those in rural areas. In other words, working at a hospital in a rural area may be protective against worry about job security. This is interesting given that before the pandemic pediatric inpatient capacity was decreasing with the greatest decline in bed availability in rural areas.¹⁸  Because beds in rural areas are limited, these programs may serve a critical regional need for pediatric care, resulting in greater job security for the hospitalists at the remaining rural PHM programs. Further studies are needed to elaborate on these associations.

Interestingly, hospitalists who worked at sites affiliated with university-based centers or free-standing children’s hospitals were more likely to report greater worry about job security than those who did not. This may be because of the “regionalization of pediatric care,” which is a practice that involves transferring all pediatric patients who present to a primarily adult hospital to pediatric-specific sites to create beds for the influx of adult patients. Some argue that this practice is essential in times of low pediatric and high adult patient volumes, as was seen early in the COVID-19 pandemic.¹⁹  Hospitals affiliated with university-based centers or free-standing children’s hospitals are more likely to have an infrastructure in place that allows for easy communication and transfer of children to the pediatric center. This elimination of general pediatric beds in the community is likely to cause concern over job security for the pediatric hospitalists working at that site. To protect against this anxiety, hospitals that have regionalization of pediatric care as a contingency plan to combat stress on the hospital system should also include a clear employment strategy for their workforce. This could include assuming patient care responsibilities at the main pediatric center, focusing on nonpatient care activities, seeing pediatric patients in the emergency department, or taking care of uncomplicated adult patients.²⁰ 

This study has some limitations. Bias may have been introduced because of sampling strategy, variability in program size, higher response rates from hospitals in the Midwest compared with other geographic regions, the larger proportion of the sample being affiliated with a university or free-standing children’s hospital, and the overlap between those who worked in the suburbs and those who were affiliated with a university or free-standing children’s hospital. Future studies should further explore the connections between geographic region of practice, hospital setting, and affiliation with a university or free-standing children’s hospital in relation to job security and aim to collect a larger sample. Furthermore, because the respondents from the 8 groups contacted by the site leads could not re-enter the survey to complete it if they were interrupted, it was more likely that this group would include some partial cases compared with the group that received the survey through a direct link. In the absence of a comprehensive database of community pediatric hospitalists, it is difficult to evaluate the representativeness of this sample. However, the gender distribution with the majority of respondents (79%) self-identifying as female is consistent with the gender distribution currently described in pediatric hospital medicine.²¹  Similarly, the absence of data regarding feelings of job security before the pandemic limits the ability to attribute concerns about job security to the pandemic, rather than pre-existing stressors. Finally, the sample may have been too small to capture all the nuances in the factors associated with sense of job security, and the study may have been underpowered to detect all differences across groups.

There are several factors that were not examined in the survey that may influence the feeling of job security among community pediatric hospitalists. These include the specific roles of the community pediatric hospitalists, including newborn nursery coverage as birth admissions did not decline during the pandemic,¹²  and the presumed value or reputation of the pediatric hospital medicine program. Such unexamined factors are important variables that may influence the sense of job security and merit further attention in future studies. Despite these limitations, this preliminary study provides valuable insight into the way community pediatric hospital medicine programs respond to a stressor such as the COVID-19 pandemic and highlights some areas that could be strengthened to prepare for the next hardship for this vulnerable yet important system in pediatric care.

This study characterizes the impact of the COVID-19 pandemic on community pediatric hospitalists’ compensation, furlough, and concerns about job security. More broadly, it demonstrates the significant effect that an external stressor can have on an already fragile system. Future studies should aim to identify protective factors that contribute to job security of community pediatric hospitalists.

We thank the site leads for the distribution of the survey and The University of Chicago Survey Laboratory for guidance on the wording, ordering, and formatting of survey questions, as well as support for data collection operations.