key: cord-0964835-m9ofc1wk authors: Sonis, Jeffrey; Pathman, Donald E.; Read, Susan; Gaynes, Bradley N.; Canter, Courtney; Curran, Patrick; Miller, Thomas title: Generalized anxiety, depression and posttraumatic stress disorder in a national sample of U.S. internal medicine physicians during the COVID-19 pandemic date: 2021-05-24 journal: Gen Hosp Psychiatry DOI: 10.1016/j.genhosppsych.2021.05.004 sha: 8a3d96a193c041cc25169a565cde8c5b3266a884 doc_id: 964835 cord_uid: m9ofc1wk nan many of those used convenience sampling [2] , focused exclusively on workers at one or a few institutions [3] , or had response rates of 20% or less [4] . The goal of this study was to identify the prevalence of generalized anxiety, depression, and PTSD during the COVID-19 pandemic in a national sample of internal medicine physicians (internists) and assess associations between self-rated risk of COVID-19 and mental health outcomes. We conducted a national survey in 2020 of internists who are members of an online panel Outcomes were measured with short screening scales. Positive generalized anxiety disorder screening was defined as a score of 3 or greater on the GAD-2 (range 0 -6) [5] . Positive depression screening was defined as a score of 3 on the PHQ-2 (range, 0-6) [6] . Positive PTSD screening was defined as a score of 6 or greater on a four-item scale based on the PCL-5 (range, 0 -16) [7] . Among the 2,145 eligible ACP Panel members, 810 responded to the survey, a 37.8% response rate. We used post-stratification weighting and raking [8] to make the age-gender and age-race/ethnicity composition of the sample comparable to that of the ACP Panel (Table S1 ). All analyses were weighted. J o u r n a l P r e -p r o o f Journal Pre-proof Prevalence estimates are reported for the three outcomes. Associations between COVID-19 risks and outcomes were assessed using logistic regression in MPLUS 8.5. Less than 15% of respondents had missing data for self-rated risk of developing or dying from COVID-19. Missing data were addressed using full information maximum likelihood in the logistic regressions using Mplus 8.5. Sample characteristics are shown in Table S2 . The prevalence of positive screening for generalized anxiety was 14.9% (95% CI, 12.4%-17.4%); for depression, 11.7% (95% CI, 9.4%-13.9%); and for PTSD, 12.8% (95% CI, 10.5%-15.1%). The prevalence of previous COVID-19 infection was 13.4% (95% CI, 11.0%-15.8%). Among respondents who had not previously had COVID-19, 28.3% (95% CI, 25.0%-31.8%) rated themselves at high or very high risk of becoming infected with COVID-19 and 8.0% (95% CI, 6.0%-10.0%) rated themselves at high or very high risk of death if infected with COVID-19. In multivariable logistic regression models, respondents who rated their risk of developing COVID-19 as high or very high had moderately elevated odds screening positive for depression and PTSD but not anxiety (Table 1) . Respondents who rated their risk of death, if they developed COVID-19, as high or very high had markedly higher odds of all three adverse outcomes (Table 1) . Those who had previously had COVID-19 infection had moderately lower odds of screening positive for depression but not generalized anxiety or PTSD. In this national study, the estimated prevalences of screening positive for generalized anxiety, depression and PTSD were between 10% and 15%. Our estimates are lower than most [1] prevalence estimates from meta-analyses of the psychosocial impact of the COVID-19 J o u r n a l P r e -p r o o f Journal Pre-proof pandemic on health care workers. The use of convenience samples in most previous studies of the impact on health care workers of the COVID-19 pandemic, with likely greater participation from those more adversely affected, may explain the higher rates they reported [9] . The difference in prevalence of mental health outcomes between our study and most other studies is not due to differences in measurement because those studies also used short screening instruments. Physicians who rated themselves at high risk of developing COVID-19 at work and those who rated themselves at high risk of dying if infected had substantially higher prevalence of mental health outcomes and may merit special attention. Respondents who had had COVID-19 infection prior to the survey were less likely to screen positive for depression, perhaps because surviving COVID-19 lowered concerns about the unknown [10] . This study has limitations. The response rate was 37.9%. We used post-stratification weights for age, gender and race/ethnicity but other differences between the sample and the panel could have affected the findings. Also, prevalence estimates are based on short screening tests, which do not perfectly predict rates of mental disorders. Finally, the prevalence estimates are based on a survey from autumn 2020. In summary, we found that the prevalence of screening positive for anxiety, depression, and PTSD in this national sample of internists in the U.S. during the COVID-19 pandemic was lower than in many other studies, suggesting overall resilience. However, the odds of adverse mental health outcomes were much higher among those who rated themselves at high risk of developing COVID-19 or dying, if infected. Journal Pre-proof Tables Table 1. Associations between self-rated COVID-19 risks or infection and mental Ref. Ref. Ref. a aOR denotes odds ratio adjusted for age category, gender, race-ethnicity, work setting (outpatient only, outpatient and inpatient, inpatient only), clinical subspecialty at high risk of exposure to COVID-19 (Hospital Medicine, Infectious Disease, Pulmonary Medicine, Critical Care Medicine, Emergency Medicine) versus all other specialties), total clinical hours in previous week, number of patients with suspected or confirmed COVID-19 seen in previous two weeks, type of health care setting (hospital, outpatient practice, setting devoted to care for the underserved, and other), respondent's assessment of rate of COVID-19 infection in their local community in previous two weeks (higher than U.S. as a whole, the same as the U.S. as a whole, lower than the U.S. as a whole) All odd ratios are based on a logistic regression model with post-stratification weights as a weight variable, with estimation using full information maximum likelihood. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f The mental health impact of the covid-19 pandemic on healthcare workers, and interventions to help them: A rapid systematic review Protecting the front line: a cross-sectional survey analysis of the occupational factors contributing to healthcare workers' infection and psychological distress during the COVID-19 pandemic in the USA Academic Emergency Medicine physicians' anxiety levels, stressors, and potential stress mitigation measures during the acceleration phase of the COVID-19 pandemic Well-being and education of urology residents during the COVID-19 pandemic: results of an American national survey Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection The Patient Health Questionnaire-2: validity of a two-item depression screener Developing an optimal short-form of the PTSD Checklist for DSM-5 (PCL-5) Poststratification and weighting adjustments Association of torture and other potentially traumatic events with mental health outcomes among populations exposed to mass conflict and displacement: a systematic review and meta-analysis The psychological impact of COVID-19 and other viral epidemics on frontline healthcare workers and ways to address it: A rapid systematic review