Risk of burnout and stress in physicians working in a COVID team: A longitudinal survey

Abstract

Background and aims: The COVID-19 pandemic represents a source of stress and potential burnout for many physicians. This single-site survey aimed at assessing perceived stress and risk to develop burnout syndrome among physicians operating in COVID wards.

Methods: This longitudinal survey evaluated stress and burnout in 51 physicians operating in the COVID team of Gemelli Hospital, Italy. Participants were asked to complete the Maslach Burnout Inventory (MBI) and the Perceived Stress Questionnaire on a short run (PSQs) (referring to the past 7 days) at baseline (T0) and then for four weeks (T1-T4). Perceived Stress Questionnaire on a long run (PSQl) (referring to the past 2 years) was completed only at T0.

Results: Compared with physicians board-certified in internal medicine, those board-certified in other disciplines showed higher scores for the Emotional Exhaustion (EE) score of the MBI scale (P < .001). Depersonalisation (DP) score showed a reduction over time (P = .002). Attending physicians scored lower than the resident physicians on the DP scale (P = .048) and higher than resident physicians on the Personal Accomplishment (PA) scale (P = .04). PSQl predicted higher scores on the EE scale (P = .003), DP scale (P = .003) and lower scores on the PA scale (P < .001). PSQs showed a reduction over time (P = .03). Attending physicians had a lower PSQs score compared with the resident physicians (P = .04).

Conclusions: Medical specialty and clinical position could represent risk factors for the development of burnout in a COVID team. In these preliminary results, physicians board-certified in internal medicine showed lower risk of developing EE during the entire course of the study.

FIGURE 1

Linear Mixed Model of Maslach Burnout Inventory—Emotional Exhaustion (EE) scale. A, EE score over time by medical specialty (β, 8.28; 95% CI, 4.50 to 12; P < .001). Compared with physicians board‐certified in internal medicine, physicians board‐certified in other disciplines showed a significant reduction of EE score (β, −1.21; 95% CI, −2.36 to −0.06; P = .04). B, scatter plot and regression line indicating the correlation between EE score and PSQ on a long run score (β, 49.99; 95% CI, 37.89 to 62.08; P < .001)

FIGURE 2

Linear Mixed Model of Maslach Burnout Inventory—Depersonalisation (DP) scale. A, DP score over time by clinical position (attending vs. resident). Attending physicians showed lower score on the DP scale over time than resident physicians (β, −3.05; 95% CI, −6.07 to −0.02; P = .048). The latter showed decreasing score on the DP scale over time than attending physicians (β, 0.505; 95% CI, 0.01 to 1.00; P = .047). B, scatter plot and regression line indicating the correlation between DP score and PSQ on a long run score (β, 14.91; 95% CI, 5.11 to 24.72; P = .003)

FIGURE 3

Linear Mixed Model of Maslach Burnout Inventory—Personal Accomplishment (PA) scale. A, PA score over time by clinical position (attending vs. resident) (β, 3.8; 95% CI, 0.14 to 7.45; P = .04). B, Scatter plot and regression line indicating correlation between PA score and PSQ on a long run score (β, −23.34; 95% CI, −35.37 to −11.31; P < .001)