Psychological impact of the COVID-19 pandemic on healthcare worker parents

During the early COVID-19 pandemic, healthcare workers (HCW) faced extreme stress, reporting crisis-level rates of depression, anxiety, and acute stress disorder [1]. This burden was felt disproportionately among groups of marginalized HCW, including women, people of color, and HCW parents who faced compounded minority stressors and family demands [2]. During initial surges, HCW in the Bronx, New York City's poorest borough, were impacted by disparities in exposure and death. To respond, hospital systems rolled out support services with yet-unmeasured rates of utilization. We conducted a mixed-methods survey study in May–June 2020 to measure these concerns. The sample was 443 HCW parents at Montefiore Medical Center (Montefiore), the largest hospital system in the Bronx who participated in a consented survey study. Data were collected on Qualtrics.com and analyzed using STATA.

We conducted two unadjusted logistic regression models, the first estimating differential odds for stress, and the second for support. Exogenous factors included demographics (e.g, age, gender, race/ethnicity, income, number of children, prior psychiatric diagnosis, and prior health condition); and COVID-19 stressors (e.g., hours per week of direct COVID-19 patient exposure, redeployment, and location of work). The first outcome, stress, was operationalized as whether HCW parents screened positive (i.e., met scale cutoff) for at least one of the following mental health concerns: psychological distress (6-item Kessler, K6; score ≥ 13), psychological impact of the COVID-19 surge (adapted Impact of Event Scale, IES; score ≥ 33), posttraumatic stress (adapted 5-item Primary Care Post-Traumatic Stress Disorder Screen, PC-PTSD; score ≥ 3), depression (Patient Health Questionnaire-9, PHQ-9; score ≥ 10), anxiety (Generalized Anxiety Disorder-7 Scale, GAD 7; score ≥ 7), insomnia (Insomnia Severity Index, ISI; score ≥ 15), and alcohol misuse (Alcohol Use Disorders Identification Test-Concise (AUDIT-C; score ≥ 3). The second outcome, support, was operationalized as whether HCW parents utilized at least one support service offered by Montefiore: phone support (a 7 day/week volunteer phone support line), in-person support (individual short-term mental health services, Caregiver Support Centers), and other support (e.g., webinars on stress, ad hoc support groups). To contextualize results, we conducted thematic analyses to interpret qualitative responses to the free-text survey question, “What do you wish people knew about your experience during the COVID-19 crisis?” (n = 230). Thematic analysis was used to capture themes in qualitative data from 230 HCW. TR and AZ independently coded data to identify, analyze and interpret open-text responses, then met to resolve discrepancies. Two sets of themes were identified, (1) general and (2) parent-specific, with demonstrative quotes below.

See Table 1 for sample characteristics and associations between demographics, COVID-19 stressors, and each outcome. Among 314 participants who completed all mental health screens, 244 (78%) had ≥1 positive screen. Prevalence of symptom rates across outcomes were high: psychological distress (10%); psychological impact of the COVID-19 surge (36%); posttraumatic stress (33%); depression (20%); anxiety (34%); insomnia (14%); and alcohol misuse (36%). Demographic factors significantly associated with higher odds for ≥1 positive screen included gender (women compared to men, OR = 2.2, p = .02), prior psychiatric diagnosis (OR = 1.9, p < .05); and prior health condition associated with COVID-19 risk (OR = 2.1, p = .02). Unexpectedly, having ≥3 children was associated with lower odds when compared to having one (OR = 0.5, p = .04). COVID-19 stressors with significantly higher odds included hours worked per week with COVID-19 patient exposure (OR = 1.03, p < .001); and redeployment (OR = 3.0, p < .001). Every additional hour per week working with COVID-19 patients was associated with a 3% increase in odds for a positive screen. Working in research conferred lower odds, compared with ambulatory medicine (OR = 0.4, p = .02).

Table 1.

Sample characteristic and associations with ≥1 positive mental health screen and social support utilization among HCW parents surveyed from Montefiore/Einstein in the Bronx, NY (n = 443).

Sample characteristics			Stress: ≥1 positive mental health screen		Support: utilization of ≥1 HCW support service
Demographic factors	n	%	OR	95% CI	OR	95% CI
Age (M = 37, IQR = 28–46)
18–29 years	129	30	Reference
30–39 years	118	27	1.0	0.5–2.1	1.4	0.8–2.4
40–49 years	127	29	0.6	0.3–1.2	1.1	0.7–2.0
≥50 years	69	14	0.6	0.3–1.5	0.6	0.3–1.3
Gender
Men	72	22	Reference
Women	222	66	2.2*	1.1–4.3	2.4**	1.3–4.6
Nonbinary/Other	40	12	2.1	0.8–5.7	2.5*	1.1–5.9
Race/Ethnicity
Non-Hispanic White	189	46	Reference
Non-Hispanic Black	61	15	1.5	0.6–3.7	1.9*	1.1–3.5
Hispanic	64	16	1.5	0.7–3.3	1.4	0.8–2.4
Asian	69	17	0.9	0.4–01.9	0.6	0.3–1.1
Other	25	6	0.7	0.2–1.9	1.3	0.6–3.1
Income
< $60,001	54	13	1.4	0.6–3.4	0.9	0.5–1.7
$60,001-80,000	51	13	1.0	0.4–2.2	1.8	1.0–3.3
$80,001-100,000	53	13	1.7	0.7–4.1	2.2**	1.2–4.1
>$100,000	246	61	Reference
Number of children
1	129	35	Reference
2	154	42	1.0	0.5–2.1	0.8	0.5–1.3
≥3	87	24	0.5*	0.2–0.9	0.8	0.5–1.4
Prior psychiatric diagnosis	98	24	1.9**	1.0–3.8	3.6***	2.3–5.8
Prior health condition	141	35	2.1*	1.1–3.9	1.3	0.9–2.0
COVID-19 stressors
Hours per week of COVID-19 patient exposure	M = 30	IQR = 2–40	1.03***	1.02–1.05	1.0	0.99–1.02
COVID-19 redeployment	291.0	29.0	3.0**	1.5–6.0	1.8*	1.1–2.7
Location of work
Emergency & ICU	76.0	19.0	1.8	0.7–4.7	1.4	0.8–2.5
Inpatient	113.0	28.0	0.6	0.3–1.4	0.9	0.5–1.6
Outpatient ambulatory medicine	105.0	26.0	Reference
Research	73.0	18.0	0.4*	0.2–0.8	0.3***	0.1–0.6
Other	42.0	10.0	0.7	0.3–2.0	1.0	0.5–2.1

≥1 positive mental health screen indicates at or above cutoff (i.e., positive) on ≥1 of the following: K6 (6-item Kessler), IES (Impact of Event Scale), PC-PTSD (Primary Care PTSD Screen for COVID), PHQ9 (Patient Health Questionnaire 9), GAD7 (7-item Generalized Anxiety Scale), ISI (Insomnia Severity Index), AUDIT-C (Alcohol Use Disorder Identification Test-Concise). Bold denotes significance <0.05. *, <0.05; **, < 0.01; ***, < 0.001. M, median; IQR, interquartile range; OR, odds ratio; 95% CI, 95% confidence interval; HCW, healthcare worker; ICU, intensive care unit.

A third of HCW parents utilized ≥1 HCW support service, with 6% utilizing phone support and 5% utilizing in-person support. The remainder reported utilizing other support, most commonly psychotherapy (12%) and formal peer-to-peer support (6%). Demographic factors significantly associated with higher odds for utilizing ≥1 HCW support service included gender (women compared to men, OR = 2.4, p = .005; and non-binary people compared to men OR = 2.5, p = .03), non-Hispanic Black race compared to white race (OR = 1.9, p = .03), income ($80,001–$100,000 compared to >$100,000, OR = 2.2, p = .009), and prior psychiatric diagnosis (OR = 3.6, p < .001). Redeployment was associated with significantly higher odds (OR = 1.8, p = .01), while odds were significantly lower for those working in research (compared with ambulatory medicine, OR = 0.3, p = .001).

In terms of qualitative data, the following general themes emerged: 1) hardship, n=72, e.g., “It was brutal and very real”; 2) lack of support, n=43, e.g., “zero training and minimal physician support”; 3) fear, n=29, e.g., “I had never felt the level of fear that would question my vocation”; 4) grief, n=25, e.g., “Getting asked to do a brain death exam on a young guy with a devastating stroke”; 5) pride, n=16, e.g., “I chose this profession for a reason and felt proud to be here”; 6) helplessness, n=9, e.g., "It is soul crushing”; 7) uncertainty, n=5, e.g., “There was so much unknown.” Three parent-specific themes emerged: 1) fear of infecting family, n=27, e.g., “Four weeks spent away from family to keep them safe still makes me tearful”; 2) difficulty with family responsibilities/lapses in childcare, n=11, e.g., “Caring for a toddler while attempting to remain productive at work is causing significant burn-out and is not sustainable”; and 3) lack of support for parenting, n=6, e.g., “I am a new mom and my worksite has not been flexible.”

These mixed-methods results, despite limitations due to social desirability bias and cross-sectional design, suggest a high rate of significant stress and low utilization of support services. Findings are pronounced among women and those with prior psychiatric and health conditions, as well as HCW parents with more hours of direct patient exposure and those who were redeployed. Qualitative data were profound, and further reflect the severity of additional stress experienced during this time among HCW parents. Our findings support the demand for psychological support for HCW parents and amplify longstanding calls to better support and retain HCW parents, particularly women, by providing improved childcare [3].

Funding

Dr. Rubinstein's contributions to the research reported in this publication were supported in part by the National Institute of Mental Health of the National Institutes of Health, award R01MH126821 and by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health, award K23AR080803. Dr. Breslow's contributions were supported in part by the National Institute of Mental Health of the National Institutes of Health, award K23MH128582. Dr. Zayde's contributions were supported in part by the National Institutes of Mental Health of the National Institutes of Health, award R01MH126821, and the FAR fund. Dr. Gabbay's contributions were supported in part by the National Institute of Mental Health of the National Institutes of Health, awards R01MH120601, R21MH121920, R21MH126501, R01MH126821, R01DA054885, and RM1DA055437, R01MH128878, R01MH131207. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funding source played no role in study design; collection, analysis, and interpretation of data; writing of the report; or the decision to submit the article for publication.

Declaration of Competing Interest

None.