Longitudinal Comparison of Burnout and Anxiety Among Healthcare and Non-Healthcare Workers During COVID-19 in Turkey

Abstract

The COVID-19 pandemic has placed a considerable psychological burden on healthcare workers, potentially leading to increased burnout and anxiety. This study aimed to evaluate burnout and anxiety levels among healthcare workers compared to non-healthcare professionals during the pandemic. We initially recruited 438 adults; 351 (217 HCWs and 134 non-HCWs) provided complete responses across all three survey waves and were analyzed. Burnout was assessed using the Maslach Burnout Inventory, and anxiety with the State–Trait Anxiety Inventory. Data were collected through an online self-administered survey at three different time points during the pandemic, and analyzed with non-parametric tests and effect sizes. Healthcare workers exhibited significantly higher levels of emotional exhaustion, depersonalization, overall burnout, and anxiety compared to non-healthcare workers across all three periods (p < 0.05). Of 438 consented individuals, 351 (80.1%) completed all waves, allowing within-population longitudinal comparisons. Within the healthcare worker group, women, individuals living alone, those working night shifts, and those considering a career change had notably higher burnout and anxiety scores. No significant differences were observed in personal accomplishment scores. Healthcare workers experienced greater psychological distress than non-healthcare workers during the COVID-19 pandemic. Identifying vulnerable subgroups and implementing supportive strategies are essential to protect the mental health and well-being of healthcare professionals during pandemics and similar crises.

1. Introduction

The COVID-19 pandemic, which emerged in late 2019 and rapidly evolved into a global health crisis, has posed unprecedented challenges to healthcare systems and professionals worldwide. Beyond its immediate infectious threat, the pandemic has significantly strained the mental resilience of frontline workers, leading to a surge in psychological stress, anxiety, and occupational burnout among healthcare workers (HCWs) [1,2,3]. Burnout syndrome, defined as a psychological response to chronic occupational stress, encompasses emotional exhaustion, depersonalization, and a diminished sense of personal accomplishment [4,5]. Its prevalence is particularly notable in professions with high emotional demand, especially in the healthcare sector where continuous patient interaction, life-and-death decision-making, and prolonged exposure to suffering are intrinsic to the job [6,7,8].

Burnout in HCWs is multifactorial. Key contributing elements include increased workload, insufficient staffing, ambiguous role expectations, and a lack of institutional support [9,10]. During the COVID-19 pandemic, these issues were amplified by the rapid redeployment of staff to unfamiliar units, inadequate personal protective equipment (PPE), increased patient mortality, and the fear of transmitting the virus to loved ones [11,12,13]. Numerous studies have documented a heightened risk of anxiety and depressive symptoms in HCWs during the pandemic, with younger and less experienced professionals being particularly vulnerable [14,15,16].

In parallel with burnout, anxiety—a common emotional response characterized by feelings of tension, worry, and physiological arousal—has also been reported to be significantly elevated among HCWs during pandemics [17,18]. State anxiety refers to a transient emotional response to a specific stressful situation, while trait anxiety denotes a more enduring predisposition to perceive situations as threatening [19]. Both constructs are relevant when evaluating the mental health impact of COVID-19 on HCWs, given the sustained exposure to high-stakes, rapidly evolving clinical environments.

Turkey, like many other countries, experienced multiple waves of COVID-19 infections, resulting in fluctuating healthcare demands. The initial surge in early 2020 necessitated swift adaptations in hospital workflows, expansion of intensive care capacity, and redeployment of HCWs to COVID-19 units. These dynamic transitions created a fertile ground for psychological distress and mental fatigue. Although considerable attention has been directed toward infection control and vaccination, the psychological toll of the pandemic on healthcare personnel remains underexplored in certain settings.

Maslach Burnout Inventory (MBI) and the State–Trait Anxiety Inventory (STAI) are validated tools frequently employed to assess the levels of burnout and anxiety in medical professionals [20,21,22]. MBI quantifies three domains of burnout—emotional exhaustion, depersonalization, and personal accomplishment—while STAI distinguishes between transient (state) and chronic (trait) anxiety responses. These tools offer a structured approach to evaluating the multifaceted impact of occupational stress, especially in pandemic scenarios.

While existing literature confirms elevated burnout and anxiety levels in HCWs during the COVID-19 pandemic, few studies have employed longitudinal assessment across different phases of the pandemic. Additionally, comparative analyses between HCWs and non-HCWs within the same sociocultural context are limited. Such comparisons can offer valuable insights into the specific occupational hazards encountered by HCWs and inform targeted interventions to preserve their mental well-being.

In this context, our study aims to evaluate and compare the levels of burnout and anxiety among HCWs and non-HCWs during three distinct periods of the COVID-19 pandemic in Istanbul, Türkiye. We further explore how these psychological parameters vary according to sociodemographic variables (age, gender, marital status, chronic illness), professional characteristics (occupation, experience, work shifts), and attitudinal factors (career satisfaction, willingness to continue in the profession).

We hypothesize that HCWs experience significantly higher levels of burnout and anxiety compared to non-HCWs and that these levels fluctuate with the severity of the pandemic. By identifying the key demographic and occupational risk factors associated with psychological distress, we aim to provide data-driven recommendations to mitigate the mental health impact of future public health emergencies on healthcare providers. However, similar studies have been conducted worldwide [23,24,25,26,27]. Although numerous studies documented psychological burden in HCWs globally, longitudinal, within-country comparisons of HCWs vs. non-HCWs in the same sociocultural context are scarce. To date, no study from Türkiye has provided such longitudinal comparative data, representing a knowledge gap our study addresses. While previous meta-analyses and multicenter studies confirmed the psychological burden of COVID-19 on HCWs globally, our study adds by repeatedly following the same cohort across three pandemic phases, enabling within-population temporal comparisons that meta-analyses could not capture. This design highlights how psychological distress evolved over time in a single sociocultural setting, complementing broader international evidence.

2. Materials and Methods

This single-center, longitudinal survey study was conducted at Prof. Dr. Cemil Taşcıoğlu City Hospital in Istanbul, Turkey, between May and October 2020. Ethical approval was obtained from the hospital’s ethics committee (Approval No. 222, dated 16 June 2020) before the start of the study. All procedures were carried out in accordance with the latest version of the Declaration of Helsinki and the Guideline for Good Clinical Practice. Written informed consent was obtained from all volunteers prior to their participation.

The study population consisted of two groups: healthcare workers and non-healthcare individuals. The healthcare worker group included physicians (faculty members, specialists, and residents across all departments) and other healthcare staff (nurses, technicians, medical technologists, and allied health personnel) who were actively working at Prof. Dr. Cemil Taşcıoğlu City Hospital during the study period. Recruitment of non-HCWs. Non-HCWs were adults residing in İstanbul and not employed in healthcare, recruited using non-probabilistic convenience and snowball sampling through institutional mailing lists and social networks. We did not match non-HCWs to HCWs; therefore representativeness is limited and should be considered when interpreting between-group differences. Non-HCWs may have included participants in public-facing occupations (e.g., retail, security, police), which carry varying exposure risks and represent a potential confounder not controlled in this study.

Exclusion criteria: Participants were excluded if they met any of the following criteria:

• Did not agree to participate in the study (no informed consent);
• Had a diagnosed psychiatric illness prior to the COVID-19 pandemic;
• Were taking psychiatric medications—psychiatric history and medication use were self-reported (no objective verification);
• Had a history of alcohol or substance dependence;
• (For healthcare workers) Were not actively engaged in clinical service during the study period.

Data were collected using a self-administered electronic questionnaire. The survey gathered demographic information (such as age, gender, education level, occupation, years of professional experience, duration of current employment, marital status, household composition, history of alcohol or substance use, presence of chronic medical conditions, and history of COVID-19 infection) and included two standardized assessment tools: the Maslach Burnout Inventory (MBI) and the State–Trait Anxiety Inventory (STAI).

The Maslach Burnout Inventory (MBI) was used to assess burnout and consists of questions evaluating three subdimensions: emotional exhaustion, depersonalization, and personal accomplishment. Each item in the MBI is rated on a 5-point Likert scale, indicating the frequency of the feeling or attitude described (from “never” to “very often”).

The State–Trait Anxiety Inventory (STAI) was used to measure anxiety levels. This instrument has two components: the State Anxiety scale, which assesses the individual’s anxiety at a given moment (current state), and the Trait Anxiety scale, which assesses the general tendency to experience anxiety (baseline anxiety). Responses on the STAI are given on a 4-point Likert scale for each item. For interpretability, STAI ≥ 42 was used as a reference threshold for clinically relevant anxiety, per prior literature and Turkish adaptation practice. The Turkish adaptation of STAI [21] underpins cultural validity.

The questionnaire was administered at three distinct time periods corresponding to the progression of the COVID-19 pandemic in Istanbul:

• Period 1 (First Wave): 1–15 May 2020, during the first wave of the pandemic when the number of COVID-19 cases in the city was at its peak and hospital visits and admissions were very high.
• Period 2 (Controlled Phase): 1–15 July 2020, when the pandemic was relatively under control, reflected by a substantial decrease in new cases and a marked reduction in hospital visits and admissions.
• Period 3 (Second Wave): 1–15 October 2020, at the beginning of the second wave, characterized by a renewed rise in COVID-19 cases and a corresponding increase in hospital visits and admissions.

During each of these three periods, the same cohort of participants completed the survey. In total, 351 individuals (217 healthcare workers and 134 non-healthcare participants) were enrolled and all completed responses at each of the three time points; therefore, no attrition occurred. Missing data were minimal and handled by casewise exclusion.

No a priori sample size calculation was performed; the sample size was determined by feasibility. A post hoc power analysis indicated adequate power for the primary between-group comparisons; however, some subgroup analyses may be underpowered, and a type II error cannot be excluded.

3. Statistical Analysis

All data were analyzed using SPSS Statistics version 22.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics were calculated for all variables. Continuous variables were reported as mean ± standard deviation (SD), and categorical variables were summarized as counts and percentages. The Shapiro–Wilk test was applied to determine whether continuous data approximated a normal distribution.

Because the continuous variables were not normally distributed, non-parametric tests were used for group comparisons. The Mann–Whitney U test was employed to compare two groups, and the Kruskal–Wallis test was used for comparisons involving three or more groups. When the Kruskal–Wallis test indicated a significant difference, post hoc pairwise comparisons were performed to identify which specific groups differed. The Chi-square test was used to analyze categorical variables. A p-value < 0.05 was considered statistically significant. In addition to p-values, we report effect sizes: for Mann–Whitney U, r = Z/√N; for Kruskal–Wallis, epsilon-squared (ε²); and for Chi-square, Cramér’s V. No multiplicity adjustment was applied; hence, type I error inflation is possible, and results should be interpreted cautiously. Two-sided α = 0.05 was used.

4. Results

A total of 438 individuals initially consented to participate in the study. Of these, 351 (80.1%) provided complete responses across all three survey waves and were included in the final analysis (217 healthcare workers and 134 non-healthcare workers). Participants with incomplete responses in any wave (n = 87) were excluded to ensure within-subject longitudinal validity. The demographic characteristics of the included participants are summarized in

Table 1. Demographics and COVID-Related Features.

Variable	Healthcare Workers (n = 217)	Non-HCWs (n = 134)	p-Value
Age 20–29	115 (53.0%)	65 (48.5%)	0.176
Age 30–39	61 (28.1%)	31 (23.1%)
Age 40–49	31 (14.3%)	26 (19.4%)
Age 50–59	10 (4.6%)	12 (9.0%)
Female	132 (60.8%)	76 (56.7%)	0.446
Married	113 (52.1%)	62 (46.3%)	0.291
Would choose same profession	108 (49.8%)	86 (64.2%)	0.008
COVID+ or first-degree relative	98 (45.2%)	55 (41.0%)	0.611
Fear of COVID infection	196 (90.3%)	121 (90.3%)	0.999
Physicians	67 (61.5%)	—
Other HCWs	41 (38.5%)	—

Note: Values are presented as n (%). COVID+: self-reported history of COVID-19 infection in the participant or a first-degree relative

. There were no significant differences between the groups in terms of age, gender, marital status, education level, or living arrangements. The proportion of participants who chose their profession willingly was similar in both groups. However, the rate of those who would choose the same profession again was significantly lower among healthcare workers (49.8%) compared to non-healthcare workers (64.2%) (p = 0.008).

4.1. Burnout Scores Across the Survey Periods

HCWs had higher emotional exhaustion, depersonalization, and overall burnout than non-HCWs across all waves, while personal accomplishment did not differ. Depersonalization rose from May to October, with the third wave higher than the second (

Table 2. Burnout Comparison Across Periods.

Period (Dates)	Measure	HCWs (Mean ± SD)	Non-HCWs (Mean ± SD)	p-Value
1–15 May 2020	Emotional Exhaustion	19.41 ± 6.44	14.39 ± 7.25	<0.001
1–15 May 2020	Depersonalization	6.89 ± 3.69	4.97 ± 3.22	<0.001
1–15 May 2020	Personal Accomplishment	19.84 ± 3.10	20.07 ± 3.08	0.505
1–15 May 2020	Total Burnout	46.15 ± 8.50	39.44 ± 9.45	<0.001
1–15 July 2020	Emotional Exhaustion	19.00 ± 7.82	13.90 ± 7.46	<0.001
1–15 July 2020	Depersonalization	6.49 ± 4.15	5.33 ± 3.83	0.051
1–15 July 2020	Personal Accomplishment	19.29 ± 3.36	19.46 ± 3.51	0.352
1–15 July 2020	Total Burnout	45.40 ± 10.73	38.70 ± 10.31	<0.001
1–15 October 2020	Emotional Exhaustion	20.42 ± 6.50	16.20 ± 4.77	<0.001
1–15 October 2020	Depersonalization	7.55 ± 3.61	6.04 ± 3.14	<0.001
1–15 October 2020	Personal Accomplishment	19.30 ± 3.00	19.57 ± 2.53	0.393
1–15 October 2020	Total Burnout	47.28 ± 8.49	41.79 ± 7.28	<0.001

).

4.2. Anxiety Scores Across the Survey Periods

HCWs consistently showed higher anxiety than non-HCWs, except for state anxiety in the second wave, which was not significant (

Table 3. Comparison of Anxiety Scores Across Periods (STAI-1 and STAI-2).

Period	Measure (STAI)	HCWs (Mean ± SD)	Non-HCWs (Mean ± SD)	p-Value
1–15 May 2020	State Anxiety (STAI-1)	43.21 ± 11.38	38.57 ± 10.78	<0.001
	Trait Anxiety (STAI-2)	44.67 ± 8.58	42.33 ± 8.22	0.013
1–15 July 2020	State Anxiety (STAI-1)	42.65 ± 11.69	39.81 ± 10.39	0.085
	Trait Anxiety (STAI-2)	45.45 ± 8.70	41.85 ± 9.83	0.007
1–15 October 2020	State Anxiety (STAI-1)	44.74 ± 12.21	40.89 ± 9.48	0.002
	Trait Anxiety (STAI-2)	46.68 ± 9.16	44.66 ± 8.19	0.038

Values are presented as mean ± standard deviation (SD). Mann–Whitney U test was used for group comparisons. STAI-1: State Anxiety; STAI-2: Trait Anxiety; HCW: Healthcare worker. Two-sided tests were used.

).

Female HCWs reported higher emotional exhaustion, overall burnout, and anxiety, with no gender differences in depersonalization or personal accomplishment (

Table 4. Gender Differences in Burnout and Anxiety Among Healthcare Workers.

Variable	Female HCWs (n = 132)	Male HCWs (n = 85)	p-Value
Emotional Exhaustion	20.50 ± 6.81	18.51 ± 6.78	<0.001
Depersonalization	7.08 ± 3.75	7.02 ± 3.90	0.903
Personal Accomplishment	19.65 ± 3.05	19.67 ± 3.30	0.573
Total Burnout	47.22 ± 8.96	45.17 ± 9.29	0.004
State Anxiety	44.87 ± 11.82	41.88 ± 11.60	0.004
Trait Anxiety	46.40 ± 9.11	44.56 ± 8.40	0.012

STAI reference values: State Anxiety (20–80), Trait Anxiety (20–80). Burnout reference values: Emotional Exhaustion (0–54), Depersonalization (0–30), Personal Accomplishment (0–48). Values are presented as mean ± standard deviation (SD). Effect sizes are reported as r (Mann–Whitney U), ε² (Kruskal–Wallis), and Cramér’s V (Chi-square).

).

Further subgroup analyses stratified by profession (physicians vs. other healthcare workers) and gender are presented in Supplementary Table S1.

4.3. Burnout and Anxiety According to Willingness to Change Profession

Participants who reported considering a change in profession had significantly higher emotional exhaustion, depersonalization, overall burnout, and both state and trait anxiety scores, regardless of whether they were healthcare workers or not. Among healthcare workers, the personal accomplishment score was significantly lower in those considering a career change, whereas this difference was not significant in non-healthcare workers (see Supplementary Table S2).

Among physicians, assistant doctors had the highest burnout and anxiety, followed by specialists and faculty (

Table 5. Burnout and Anxiety Scores among Physicians by Career Stage. Mean (± SD) scores of Maslach Burnout Inventory (MBI) subscales and State–Trait Anxiety Inventory (STAI) among physicians, stratified by career stage (assistant doctors, specialists, and faculty).

Career Stage	Emotional Exhaustion Mean ± SD	Depersonalization Mean ± SD	Personal Accomplishment Mean ± SD	Overall Burnout Mean ± SD	STAI-State Mean ± SD	STAI-Trait Mean ± SD
Assistant Doctors	20.91 ± 6.65	8.81 ± 3.46	19.09 ± 3.04	48.82 ± 9.06	44.29 ± 11.96	46.84 ± 9.52
Specialists	19.84 ± 6.70	6.83 ± 3.67	19.32 ± 2.94	46.00 ± 8.54	43.71 ± 11.76	45.91 ± 8.82
Faculty	16.52 ± 7.65	5.19 ± 3.79	21.38 ± 2.22	43.10 ± 9.81	42.48 ± 12.36	41.19 ± 8.62

Abbreviations: MBI, Maslach Burnout Inventory; STAI, State–Trait Anxiety Inventory; SD, standard deviation. p-values: Emotional exhaustion p = 0.019; Depersonalization p < 0.001; Personal accomplishment p = 0.002; Overall burnout p < 0.001; STAI-State p = 0.727; STAI-Trait p = 0.036.

).

Living arrangement was associated with burnout among healthcare workers, with significantly higher emotional exhaustion, depersonalization, and overall burnout scores in those living alone compared with those living with family, while no such differences were observed in non-healthcare workers (Supplementary Table S3).

5. Discussion

The COVID-19 pandemic, which began in Wuhan, China, in December 2019, has affected all segments of society to varying degrees. Among the most impacted were healthcare workers, who stood on the frontlines of the pandemic response. According to a six-month evaluation report by the Turkish Medical Association, healthcare workers constituted 11.5% of all COVID-19 cases in Türkiye at that time, and 72 healthcare workers had died from the disease. Burnout is especially common in professions involving constant interpersonal communication, and healthcare workers, by nature of their duties, are at heightened risk.

Our findings confirm higher burnout and anxiety in HCWs compared with non-HCWs, consistent with international meta-analyses [14]. However, unlike some longitudinal studies that observed progressive increases across waves, our data suggest relative stability, possibly reflecting adaptation or coping strategies in Turkish HCWs. Country-specific factors such as family support or institutional resilience measures may have buffered anxiety trajectories despite persistent workload.

Meta-analyses and studies conducted in various countries, including Germany, China, and Hong Kong, have consistently shown elevated levels of depression, anxiety, and insomnia among healthcare providers. Lai et al. reported that over 50% of healthcare professionals in Wuhan exhibited depressive symptoms, with anxiety and insomnia being similarly prevalent. In Türkiye, studies predating the pandemic already showed moderate levels of emotional exhaustion and depersonalization among healthcare workers, with some reports indicating even higher levels among ICU nurses.

Burnout and anxiety levels among HCWs remained relatively stable across the three pandemic phases in Türkiye. This finding contrasts with international studies reporting progressive increases over time [13,14], suggesting that country-specific organizational measures may have influenced the trajectory. The significant rise in depersonalization scores between the second and third wave indicates cumulative stress effects consistent with the Job Demands–Resources model [9] and Maslach’s burnout framework [4].

Although HCWs reported higher exhaustion and depersonalization, personal accomplishment did not differ between groups. This may reflect resilience and a sustained sense of purpose among HCWs, as highlighted in qualitative studies [2]. Alternatively, the Maslach PA subscale may lack sensitivity to capture changes during acute crises, and thus findings should be interpreted cautiously.

State and trait anxiety scores in our cohort exceeded the clinical threshold of 42, reflecting clinically significant distress. Compared to pre-pandemic Turkish studies, levels were moderately higher, in line with findings from China and Italy during the early pandemic [1,28]. This suggests that heightened risk perception and uncertainty contributed substantially to psychological burden.

Female healthcare workers consistently reported higher burnout and anxiety levels, consistent with prior meta-analyses showing greater vulnerability among women during the pandemic [14]. Older healthcare professionals and those with longer work experience reported lower levels of burnout, likely reflecting improved coping mechanisms and lower unmet expectations. These results are in line with other studies showing that younger and less experienced professionals are more vulnerable to emotional fatigue.

Among healthcare professionals, physicians demonstrated significantly higher levels of burnout than nurses and technicians, especially in the dimensions of depersonalization and overall burnout. Assistant doctors were particularly affected, showing the highest burnout levels among physician subgroups (Table 5), possibly reflecting demanding working conditions during residency. Assistant doctors exhibited the highest burnout, consistent with systematic reviews identifying residents as a high-risk group [29]. No significant differences were observed among nurses, technicians, or allied health staff, suggesting that the risk of burnout was particularly pronounced among physicians.

Those considering a career change reported significantly higher burnout and anxiety, in line with studies showing that intention to leave the profession is strongly linked with emotional exhaustion [2]. This trend has been observed in other studies as well, indicating a strong link between job dissatisfaction and emotional exhaustion.

Living arrangements also played a role; those living alone reported higher burnout levels than those living with family, consistent with evidence linking social isolation to increased psychological distress [18,23]. On the other hand, fear of transmitting the virus to family members remained a major concern for many healthcare professionals, often exacerbating stress.

Overall, our findings underscore that during the COVID-19 pandemic, healthcare workers were more vulnerable to burnout and anxiety than the general population. Specific subgroups, including women, younger professionals, and those with fewer years of experience, were particularly at risk.

Limitations

This study has several limitations. First, it was single-center with non-probabilistic recruitment, limiting external validity. Second, outcomes were based on self-reported measures, subject to recall and social desirability bias. Third, although all participants completed three waves, response fatigue may have influenced answers. Fourth, we did not stratify non-HCWs by public-facing roles, which may confound comparisons. Fifth, no multivariable modeling was performed to adjust for potential confounders. Sixth, the absence of pre-pandemic baseline data prevents determining whether observed levels were pandemic-specific. Finally, the time lag between data collection (2020) and submission (2025) may reduce immediacy.

6. Conclusions

This study demonstrates that healthcare workers experienced significantly higher levels of burnout and anxiety during the COVID-19 pandemic compared to the general population. Younger, less experienced, and female healthcare workers were disproportionately affected. These findings highlight the need to address modifiable workplace factors—such as excessive workload, inadequate staffing, and limited institutional support—through targeted interventions. Interventions should particularly focus on vulnerable subgroups such as women, younger professionals, and residents, who consistently demonstrated higher distress. Strengthening organizational support and resilience-building programs may mitigate long-term consequences. However, as this was an observational survey study without pre-pandemic baseline data, causal inference cannot be established; our findings should be interpreted as associations rather than causal effects.