Ongoing Impact of COVID-19 on Breast Radiologists’ Wellness Free

Abstract

Objective

Assess the ongoing impact of COVID-19 on mental well-being of breast radiologists nationwide two years after the start of the pandemic and compared to early in the pandemic.

Methods

A 27-question survey was distributed from December 2021 to January 2022 to physician members of the Society of Breast Imaging. Psychological distress and anxiety scores were calculated, and factors associated with them were identified with a multivariate logistic model.

Results

A total of 550 surveys were completed (23% response rate); the mean respondent age was 50 +/− 10 years. Fifty percent (265/526) of respondents reported two or more psychological distress symptoms, down from 58% in 2020 (P = 0.002), whereas 70% (362/526) of respondents reported increased anxiety, down from 82% in 2020 (P < 0.001). As in 2020, reporting financial strain and childcare adversely affecting job ability were associated with worse psychological distress scores (OR 3.6, 95% CI: 1.6–8.3, P = 0.02 and OR 6.0, 95% CI: 2.5–14.4, P = 0.002, respectively). Less time spent consulting, educating, and discussing results with patients was associated with higher psychological distress (OR 5.3, 95% CI: 2.1–13.2, P = 0.036) and anxiety (OR 6.4, 95% CI: 2.3–17.5, P < 0.001). Diminished research collaboration was associated with higher anxiety (OR 1.8, 95% CI: 1.1–2.9, P = 0.019).

Conclusion

The COVID-19 pandemic continues to cause mental health symptoms in breast radiologists, especially for those with pandemic-specific childcare needs and financial distress. Pandemic-related decreased opportunities to connect with patients and colleagues negatively impacts radiologists’ mental health.

Introduction

Burnout has been a persistent problem that has plagued physicians for decades (1), but the impact has been abruptly exacerbated by the COVID-19 pandemic (2), with evidence from around the world documenting acute stress and mental health challenges in physicians providing care since 2020 (3–7). Burnout arises as a response to chronic prolonged interpersonal job-related stress defined by emotional exhaustion, depersonalization, and reduced personal accomplishment that can occur among individuals who work with people in some capacity (8). Although burnout has been dichotomized for ease of treatment and intervention based on cutoffs on the Maslach Burnout Inventory (MBI) scale—the current standard in measuring the syndrome—Maslach and colleagues recognize that it is, in fact, a spectrum of symptoms ranging from mild to severe (8). The persistence of acute stressors resulting in mild symptoms that may not quite meet the criteria for burnout can lead to increased symptom severity and ultimate burnout (8).

The acute stressors of the initial COVID-19 pandemic wave in 2020 hurt physician mental well-being overall (9), impacting both frontline and non-frontline health care workers (10), with a disproportionate effect on younger and female physicians (11). There is mounting evidence that female physicians carried and continue to carry the brunt of the increased domestic demands during the pandemic (12). The New York Times coined the term “SheCession” to describe how women were more likely to leave the workforce than men. This impact was particularly palpable in the female-dominated subspecialty of breast imaging (13), with female breast radiologists 1.5 times more likely than their male counterparts to take unpaid leave during the pandemic (11).

Although physician wellness, or well-being, is not formally defined, and as such disparate constructs have been evaluated in prior studies (14), the ability to provide optimal patient care has been found to be intricately associated with physician well-being globally (15). Meanwhile, the ability to provide good patient care has separately been exacerbated by the pandemic: In addition to canceled and delayed screening services affecting patient care (16–18), breast radiologists perceived a diminished ability to connect and communicate with patients in the early phases of the pandemic due to barriers from personal protective equipment (PPE), social distancing, and other safety measures. This negative impact on patient care contributed to feelings of low personal accomplishment and dissatisfaction and was associated with worse mental well-being in breast radiologists (19).

It has been over two years since the start of the pandemic, and the ongoing impact of COVID-19 on breast radiologists’ mental well-being remains unclear. Physician surveys show that burnout worsened overall in 2022 compared to 2020 (47% vs 42%, respectively), and radiologists remain one of the top 10 most burned-out specialties (49%) (20), with women physicians reporting much higher burnout rates than men (56% vs 41%) (21). Breast radiologists are predominantly female and have greater patient-centric responsibilities than most other radiologists. Even before the pandemic, an alarming 54% of breast radiologists reported burnout in at least two measures (22). It is therefore crucial to understand the persistent impact of COVID-19 on the mental well-being of breast radiologists. Furthermore, it is important to identify which COVID-related stressors continue to persist, two years after the pandemic began, in order to intentionally craft strategies to mitigate worsening burnout in our field.

The purpose of this nationwide survey study was to evaluate the impact of COVID-19 on the mental well-being of breast radiologists two years after the start of the pandemic and compare to results of a similar survey performed 18 months earlier (11). We sought to identify which demographic factors and survey responses were associated with worse mental well-being.

Methods

This study received an institutional review board waiver. A cross-sectional survey assessing the ongoing psychological impact of the COVID-19 pandemic on the breast imaging community was adapted from a survey administered during early in the COVID-19 pandemic (June–September 2020) (11). The survey addressed the same primary end points as the initial survey: mental health, childcare needs, financial strain, and patient care. Several questions were added to the follow-up survey addressing current work practices, volumes, and outlook toward the future. The survey consisted of 27 multiple-choice questions with a limited number allowing multiple responses as well as free-response answer choices (see Supplementary Material S1). The survey was developed by the Patient Care and Delivery Committee of the Society of Breast Imaging (SBI) and piloted at 10 of their respective institutions. The survey was created in SurveyMonkey (SurveyMonkey, San Mateo, CA) and distributed by e-mail to the 2750 active members of the SBI, including U.S. and international practicing physicians and physicians-in-training. The study was subsequently limited to the responses from practicing physicians in the United States and excluded responses from cohorts with small numbers precluding analysis, specifically physicians in training (5) and physicians from Canada (9). The survey was open from December 15, 2021, to January 31, 2022. An initial e-mail and four subsequent e-mail reminders were sent to encourage participation. Completion of the survey was optional, partial responses were permitted, and participants received no compensation.

Responses were compared to the initial 2020 survey administered at the start of the pandemic. The National Consortium of Breast Centers members, non-radiologists, physicians-in-training, and members practicing outside of the United States were excluded from the initial survey analysis to allow for a more direct comparison to this follow-up survey.

Mental Health—Anxiety and Psychological Distress Score

In the absence of uniform definition of physician wellness, here we refer to physician well-being specifically as a mental construct with an absence of mental ill-being or emotional distress. As such, participants were asked to report increases in any of the following symptoms because of COVID-19: anxiety, sadness, depression, anger, withdrawal, sleep problems, guilt, or “other” (see Supplementary Material S1). A Psychological Distress Score was calculated as per the initial survey: one point was assigned for each of the mental health symptoms answered affirmatively by the respondent. The total psychological distress score was computed as the sum of these points, with a higher score indicating more psychological distress (range, 0–7 points) (11). Free-text responses for the “other” answer choice option were not included in the psychological distress score. In addition, as previously, participants received an anxiety score (range, 1–5 points) based on whether their anxiety had increased compared to before the COVID-19 pandemic, with answer choices on a Likert scale (1–strongly disagree to 5–strongly agree) (11).

Categorical and ordinal measures were descriptively summarized using frequency (percentage), whereas continuous measures, such as age, were summarized using means (SD). Associations between specific ordinal/continuous variables versus specific categorical variables were evaluated using the Wilcoxon rank sum test or t-test as appropriate. Associations between two categorical variables were evaluated using the chi-square test. The multivariable ordinal logistic regression model was used to identify factors that were independently associated with psychological distress and anxiety scores adjusting for basic demographic variables as well as other survey questions. The following specific variables were used as candidate predictors in the models: demographic variables, childcare, financial strain, and various aspects of the work environment that might have been impacted by the pandemic. Backwards procedure was used to simplify the models based on the Akaike information criterion (AIC). Statistical significance was determined using a two-sided P-value <5%. All analyses were performed using SAS 9.4 (SAS Institute, Inc, Cary, NC, USA).

Results

Participant Demographics

A total of 550 participants responded to the survey for an overall response rate of 23% (550/2423 U.S. physician members). Most participants were women (397/550, 76%) with a mean age of 50 ± 10 years and had been practicing for >10 years (371/550, 67%). Although the proportion of female respondents is consistent with a larger proportion of female SBI members (66% female, 32% male, 2% not disclosed), female SBI members were more likely to respond to the survey than male members given a 10% higher proportion of female respondents to both surveys. This is in line with data consistently demonstrating that women are more likely to self-select for survey participation (23). Participants were affiliated with private practice (237/525, 44%), academic (134/525, 27%), or a hybrid community practice affiliated with an academic medical center (125/525, 24%). Demographics of the respondents were similar for the initial and follow-up surveys (Table 1).

Mental Health and Wellness

Two years since the start of the COVID-19 pandemic, 50% (265/526) of breast radiologists continue to report two or more psychological distress symptoms attributed to COVID-19, a decrease from 59% (284/482) in 2020 (P = 0.005). In the follow-up survey, 70% (362/526) of respondents agreed or strongly agreed that their anxiety continued to be elevated compared to before the pandemic, a decrease from 83% (363/482) in 2020 (P < 0.001).

Anxiety was the most reported mental health symptom attributed to COVID-19 in 2022 (322/550, 61%), followed by sadness (163/550, 31%), sleep problems (163/550, 31%), anger (144/550, 27%), depression (105/550, 20%), and withdrawal (78/550, 15%). While reporting of most mental health symptoms decreased in 2022 compared to 2020 (Table 2), statistically significant decreases were seen only for anxiety (328/482, 68% vs 322/550, 61%, respectively) (P = 0.02) and sadness (195/482, 40% vs 163/550, 31%, respectively) (P < 0.002).

In contrast with the initial survey in 2020, younger age and female gender were no longer significantly associated with greater anxiety or psychological distress scores in the multivariable analyses (Tables 3 and 4) in this 2-year follow-up survey.

Workplace wellness efforts specific to COVID-19 declined over the past two years, according to survey responses. There was a decrease in COVID-19-specific wellness efforts from 42% (202/482) in 2020 to 30% (156/521) in 2022 (P < 0.0001). However, the presence or absence of these wellness efforts was not associated with increased psychological distress or anxiety in the multivariable analysis.

Childcare

Nearly half (161/362, 45%) of respondents with children or other dependents at home reported that increased childcare or dependent needs specific to the pandemic adversely affected their ability to do their job within the past year, not significantly changed from 2020 (86/229, 37%) (P = 0.6) (Table 5). Also similar to 2020, those reporting adverse effects of increased childcare/dependent needs on the ability to do their jobs were more likely to be female (177/362, 49% vs 105/362, 29%) (P = 0.01) (2020 results 92/229, 40% vs 66/229, 29%) (P < 0.001) and younger [61% in 31- to 40-year-olds (221/362), 50% in 41- to 50-year-olds (181/362), 27% in 51- to 60-year-olds (98/362), and 37% in > 60-year-olds (134/362); P < 0.001; similar to 62% in 31- to 40-year-olds (142/229), 54% in 41- to 50-year-olds (115/229), 13% in 51- to 60-year-olds (30/229), 10% in 61- to 70-year-olds (23/229), and 39% in ≥70-year-olds (89/229); P < 0.001 in 2020]. Those who reported that pandemic-related childcare and dependent needs adversely affected their jobs were significantly more likely to have elevated psychological distress scores [agree (124/362, 34%): OR 2.5, 95% CI 1.3–4.9, P = 0.005; strongly agree (36/362, 10%): OR 6.0, 95% CI: 2.5–14.4, P = 0.002]. Similarly, respondents who reported negative impact of childcare and dependent needs on their ability to do their jobs were significantly more likely to have elevated anxiety scores [agree (124/362, 34%): OR 3.3, 95% CI: 1.8–6.3, P = 0.002; strongly agree (36/362, 10%): OR 10.8, 95% CI: 4.3–27.5, P < 0.001] compared to those who strongly disagreed on the multivariable analysis (Tables 3 and 4).

Financial

A decrease in financial strain due to the pandemic was observed in this follow-up survey. Twenty-two percent (116/526) of respondents were under more financial strain now compared to before the pandemic, down from 37% (86/229) reported on the initial survey at the start of the pandemic (P < 0.0001). However, those who were under greater financial strain due to the pandemic were significantly more likely to have higher psychological distress scores [agree (91/526, 17%): OR 3.7, 95% CI: 1.6–8.3, P = 0.002; strongly agree (25/526, 5%): OR 4.5, 95% CI: 1.5–13.4, P = 0.006] and anxiety [agree (91/526, 17%): OR 8.5, 95% CI: 3.9–18.3, P < 0.001, strongly agree (25/526, 5%): OR 17.5, 95% CI: 5.8–53.0, P < 0.0001] compared to those who strongly disagreed on the multivariable analysis (Tables 3 and 4).

Patient Care

There was no significant difference between 2020 and 2022 in the proportion of respondents who reported that the overall level of patient care was decreased because of the pandemic (162/439, 37% in 2020 vs 184/526, 35% in 2022) (P = 0.5). Fewer respondents reported a diminished ability to fulfill the psychological needs of their patients because of COVID-19 on the follow-up survey (202/438, 46% in 2020 vs 192/522, 37% in 2022) (P < 0.001). Fewer respondents reported that PPE made it harder to communicate with patients (385/438, 88% in 2020 vs 414/524, 79% in 2022) (P < 0.001).

Perceived limitations in patient interactions and referring physician collaborations were significantly associated with greater psychological distress and anxiety. Diminished reported time spent consulting, educating, and discussing results with patients was associated with higher psychological distress score (OR 5.3, 95% CI: 2.1–13.2, P = 0.036) and increased anxiety (OR 6.4, 95% CI: 2.3–17.5, P < 0.001). Similarly, diminished research collaboration with referring physicians was associated with higher anxiety (OR 1.8, 95% CI: 1.1–2.9, P = 0.019). A minority of respondents (89/550, 15%) noted that patient interactions are more likely to be negative or hostile since COVID-19; however, the nature of patient interactions was not significantly associated with either greater psychological distress or anxiety scores on multivariate analysis (P = 0.580).

Despite nearly half of the respondents (239/550, 44%) reporting that overall breast imaging volume has increased to more than 100% of pre-pandemic volumes, this factor was not significantly associated with either greater psychological distress or anxiety scores on multivariate analysis (P = 0.668). Similarly, changes in workflow compared to before the pandemic were not significantly associated with either psychological distress or anxiety scores on multivariate analysis. Workflow changes included: patient wait times during imaging appointments (190/515, 37% increased; 291/515, 57% unchanged; 34/515, 6% decreased) (P = 0.471), patient wait time to schedule screening appointments (290/516, 56% increased; 201/516, 39% unchanged; 25/516, 5% decreased) (P = 0.393), practice of same-day add-on biopsies (55/514, 11% increased; 316/514, 61% unchanged; 143/514, 28% decreased) (P = 0.171), and same-day screening interpretation (40/513, 8% increased; 355/513, 69% unchanged; 118/513, 23% decreased) (P = 0.609).

Notwithstanding, 66% (363/550) of breast radiologists reported a positive or extremely positive outlook toward the future of our field, whereas only a minority (52/550, 9%) reported a negative or extremely negative outlook.

Discussion

Despite some areas of improvement since 2020, our study demonstrates that there are persistent challenges affecting the mental health of breast radiologists due to COVID-19. Although we noted that psychological distress improved since 2020, levels continue to overall be high in 2022 with 50% of breast radiologists reporting two or more psychological distress symptoms and 70% reporting elevated anxiety attributed to the pandemic. The sustained mental health decline of breast radiologists may exacerbate physician burnout and adversely affect patient care in an already stressed health care environment (24).

Despite high levels of anxiety and psychological distress, according to our survey institutional physician wellness efforts appear to be decreasing, with fewer than a third (30%) of respondents reporting any pandemic-related wellness initiatives at their institution, down from 42% during the early phase of the pandemic. Our findings are in accord with the 2022 national Physician Burnout & Depression Report 2022, which demonstrated that physician mental health needs related to the pandemic have not resolved (20) and that there is an ongoing need for physician support. Although the decline in institutional wellness efforts noted in our study is concerning, absence of any discernable impact this decline had on psychological distress or anxiety scores between the initial and follow-up surveys may point to lack of effectiveness of those efforts present. Evidence demonstrates that wellness interventions addressing individual factors such as mental health care, counseling, and physical activity resources have a modest and short-lasting effect on wellness compared to organization-directed workplace interventions (25). Both recent randomized controlled trials and meta-analyses demonstrate that, in fact, organizational interventions work and are critical to creating a culture and practice environment that cultivates professional fulfillment, which mitigates burnout (25–27). Ample resources are available as a roadmap to institutions (28,29), including those from the American College of Radiology (ACR) (30) and the recently released National Academy of Medicine (NAM) consensus report: Taking Action Against Clinician Burnout: A Systems Approach to Professional Well-being (31). Recognition of the critical nature of the problem and identifying the principal drivers of stress within each institution to initiate timely system-level interventions are vital to the health of the field. Contributing negative factors to breast radiologists’ mental health found in our study can serve as a framework for initial discussions and initial institutional assessment of the problem.

Our study highlights the critical impact of limited patient interactions on increased anxiety and psychological distress among breast radiologists, as was seen in 2020. Even though the patient–physician interaction challenges related to PPE have marginally improved since 2020, breast radiologists continue to report that PPE negatively affects their ability to communicate with patients and fulfill their patients’ psychological needs. This has the potential to worsen patient health care outcomes as patient–physician communication is critical to mitigating health disparities, improving adherence to follow-up recommendations, reducing procedural pain, and adjusting to a new cancer diagnosis (32–35). Furthermore, our results demonstrate that the impact of the patient–physician relationship is not one-sided and has a significant effect on the physician’s well-being, stress levels, anxiety, and ultimately burnout (36–38). Of particular significance is the finding in our study that it is, in fact, the shortened duration of the patient–physician interaction rather than the nature of that interaction that impacts radiologists’ wellbeing. As breast imaging workflows continue to be revised in the face of changing pandemic-related operational demands (16,39,40), it is crucial to not obviate patient interactions in favor of increased patient throughput and rather be intentional in building in adequate time for patient–breast radiologist interaction. Incorporating adequate patient–physician interaction time will also diminish the potential negative impact of PPE on communication by allowing time to establish rapport and rely on body language and other social cues, thus allowing for a greater balance between patient communication and infection risk prevention. The negative impact of limitation in patient–physician communication seen in our study further suggests that refocusing on patient-centered communication training in the context of PPE and social distancing limitations may provide for increased adaptability of the next generation of breast radiologists (41).

Breast imaging is an important component of the multidisciplinary breast cancer care and research environment. Our survey demonstrated that a diminished ability to interact with nonradiology colleagues was associated with greater psychological distress. Although many health systems have transitioned to virtual meetings for multidisciplinary conferences, such as tumor board or journal club, these venues lack the same opportunities for interpersonal connections. Furthermore, increasing patient volumes, as noted in this study and prior reports, coupled with reduced staffing (17,18) place additional time constraints on radiologists. New strategies are needed to nurture physician interactions and build relationships in a post-pandemic world—for example, monthly multidisciplinary research meetings may increase opportunities for cross-departmental research collaborations; cross-departmental mentorship lunches may allow for organic relationship building without impinging on time after work; and volunteer endeavors may support group socialization and allow physicians to find meaning and connect with colleagues by giving back to others. The challenge is to seamlessly weave these strategies into a physician’s day without adding a time burden that could backlash and exacerbate burnout.

A survey of nearly 59 000 health care workers during the height of the pandemic (April–December of 2020) found that those with childcare stress had greater odds (OR 2.15 95% CI: 2.04–2.26; P < 0.001) of experiencing anxiety and depression, as well as burnout (OR, 1.80; 95% CI: 1.70–1.90; P < 0.001) (42). Our findings suggest that the stress related to dependent care persists well into the second year of the pandemic. We saw no significant change in proportion of respondents reporting that childcare needs are adversely affecting their ability to do their job compared to the initial survey, which may be related to unpredictable school closures and periodic quarantines. Younger and female physicians were more likely to report increased childcare needs. It is important to recognize the vulnerability and needs of this population given both the starkly elevated odds ratio of those with childcare needs reporting anxiety (OR 10.6) or psychological distress (OR 6.0) found in our study, as well as earlier findings demonstrating increased rate in reduction of hours worked and intent to leave health care for those with childcare stress (42). Institutions should consider operational changes—greater flexibility and virtual options, for example—to address childcare stress and dependent needs and to help mitigate the pandemic’s impact on breast radiologists’ mental wellbeing and ultimate burnout.

Similar to the 2020 survey, our 2022 follow-up survey demonstrated that women and younger physicians are significantly more likely to report that the pandemic has negatively impacted their ability to do their job. In contrast to the initial survey, however, neither female gender nor younger age were associated with worse anxiety or psychological distress scores. Perhaps women have become accustomed to a “new normal” pandemic life and developed coping strategies, or perhaps male and older physicians—groups traditionally less likely to report mental health symptoms (43)—have become increasingly comfortable acknowledging psychological distress in the face of a more universal perception of the widespread negative impact of the pandemic on mental well-being (44). Nonetheless, our results support the notion of a potential “SheCession” due to a disproportionate impact of the pandemic on women in the workforce compared to men, a problem with great potential consequences for the female-dominated specialty of breast radiology, in which women represent 79% of breast radiologists compared to 34% of the field of radiology (45).

This study has several limitations, including volunteer selection bias and other biases inherent to a survey study design. Specifically, radiologists experiencing distress may be more likely to participate in the survey, artificially elevating the assessed rate of anxiety and psychological distress. Although the same SBI physician membership cohort was invited to participate in the follow-up study, the cohorts of respondents between the initial and follow-up studies are unlikely to be identical. Since the survey was anonymous, it was not possible to determine whether the same members responded to both surveys. Nonetheless, the demographics of the respondents to both surveys were very comparable in terms of gender, age, years out of practice, and practice type, supporting validity of the comparison. Similarly, although the anxiety and psychological distress measures utilized are nonvalidated tools based on self-reporting of mental health symptoms, utilization of the same survey questions for the initial previously published and follow-up survey allows for a valid comparison.

Conclusion

Despite some improvement in the self-reported mental health of breast radiologists, this follow-up survey study demonstrates the continued impact of the COVID-19 pandemic on breast radiologists’ mental health, with the majority still reporting increased anxiety and other mental health symptoms. Greater pandemic-specific childcare needs and financial distress continue to be significant contributors to breast radiologists’ psychological distress and anxiety. This study further underscores the negative psychological impact of diminished opportunities to connect with patients and colleagues—now a “new normal” likely stemming from operational changes aimed to increase patient throughput and decrease disease spread. As the pandemic continues indefinitely, an increased awareness and focus on providing adequate patient–physician communication interactions and multidisciplinary collaborations are needed to ensure the best value-based care for patients and to mitigate continued pandemic-specific burnout in physicians.

Funding

None declared.

Conflict of Interest Statement

None declared.

References