Longitudinal Changes in Work-Related Pleasure, Suffering, and Harm Among Healthcare Workers During the COVID-19 Pandemic

Abstract

Healthcare workers experienced substantial psychosocial challenges during the coronavirus disease 2019 (COVID-19) pandemic, yet longitudinal evidence on changes in work-related pleasure, suffering, and harm remains limited. This study investigated longitudinal changes in psychosocial risk indicators among Brazilian healthcare professionals during two phases of the pandemic. A prospective observational study was conducted with 113 healthcare professionals who completed two online assessments in April–June 2020 (T1) and November 2020–January 2021 (T2). Work-related pleasure and suffering were assessed using the Scale of Indicators of Pleasure and Suffering at Work (EIPST), and work-related harm was evaluated using the Work-Related Harm Assessment Scale (EADRT). Paired comparisons were performed using paired t-tests with bootstrap confidence intervals; effect sizes were estimated using Cohen’s d and probability of superiority. After adjustment for multiple comparisons, statistically significant reductions remained for professional burnout (p = 0.001; d = 0.31) and psychological harm (p = 0.006; d = 0.26). Although overall work-related suffering showed a nominally significant reduction (p = 0.034; d = 0.21), this finding did not remain statistically significant after adjustment for multiple comparisons. The corresponding effect sizes were small, and most psychosocial risk indicators remained within critical risk classifications at both assessments. Professional fulfillment changed from a critical to a satisfactory classification, although its longitudinal difference was not statistically significant. These findings suggest modest reductions in psychosocial distress over time while highlighting the persistence of critical psychosocial risks. Organizational strategies aimed at preventing psychosocial hazards and promoting healthy working conditions remain essential to protect healthcare workers during routine practice and future public health emergencies.

1. Introduction

The COVID-19 pandemic, caused by the SARS-CoV-2 virus, was declared in March 2020 and had profound consequences for economic, social, and healthcare systems worldwide [1]. As of July 2024, more than 775 million confirmed cases and seven million deaths had been reported globally [2,3].

In response to the rapid spread of COVID-19, Brazil substantially expanded its healthcare capacity. Between January and June 2020, the number of intensive care unit (ICU) beds increased by 47%, while the capacity of intensive care services within the Brazilian Public Health System (SUS) expanded by 89% [4,5]. Despite these efforts, healthcare services operated under unprecedented pressure, facing shortages of personal protective equipment, workforce constraints, rapidly changing clinical protocols, and increasing demands for specialized care [3].

Healthcare professionals were particularly exposed to psychosocial risks, understood as aspects of work organization, management, and social context that have the potential to adversely affect workers’ psychological and physical health. During the first months of the pandemic, uncertainty regarding disease transmission, limited scientific knowledge, fear of infection, and concerns about transmitting the virus to family members contributed to an environment of intense occupational stress [6]. Although scientific knowledge and clinical protocols evolved over time, healthcare workers continued to experience exceptionally demanding working conditions throughout successive waves of the pandemic.

These circumstances resulted in substantial physical and psychological burden among healthcare professionals, with repercussions extending beyond the workplace. A longitudinal study conducted in Québec, Canada, reported that 51.8% of healthcare workers experienced burnout, while 35.6% presented severe symptoms of at least one mental disorder, including post-traumatic stress disorder or depression [7]. Similarly, a Brazilian study evaluating indicators of pleasure and suffering among frontline healthcare professionals found that 61.6% of participants experienced psychological distress, with work-related suffering reaching critical levels, suggesting a high potential for reduced pleasure, increased suffering, and severe professional exhaustion during the progression of the pandemic [8].

Within the framework of work psychodynamics, occupational experiences emerge from the dynamic interaction between pathogenic and health-promoting aspects of work organization [9]. Work-related suffering is associated with excessive demands, conflict, overload, and situations in which workers are unable to transform occupational challenges into meaningful experiences. Conversely, work-related pleasure reflects positive experiences derived from professional fulfillment, recognition, autonomy, creativity, cooperation, and the perception that work is socially meaningful [10]. These positive and negative dimensions coexist and may fluctuate over time according to changes in working conditions, organizational support, and individual experiences.

From this perspective, work-related harm represents the manifestation of adverse physical, psychological, and social consequences resulting from chronic exposure to unfavorable working conditions [11]. Previous research has shown that limited participation in decision-making, reduced autonomy, and restricted opportunities for creativity and recognition may contribute to psychological distress and negatively affect workers’ physical, psychological, and social health [11].

Despite the extensive literature documenting the psychological impact of COVID-19 on healthcare workers, most studies have adopted cross-sectional designs and have primarily focused on adverse outcomes, such as burnout, anxiety, depression, and psychological distress. Comparatively fewer studies have examined positive dimensions of work experience, including professional fulfillment and pleasure at work, and even fewer have explored how these positive and negative experiences evolve simultaneously over time. Longitudinal evidence examining the dynamic coexistence of work-related pleasure, suffering, and harm within the theoretical framework of work psychodynamics therefore remains limited.

Accordingly, the present study aimed to assess longitudinal changes in work-related pleasure, suffering, and harm among frontline healthcare professionals during the COVID-19 pandemic in Brazil.

2. Materials and Methods

2.1. Study Design and Setting

This prospective longitudinal observational study was conducted between April and June 2020 (T1) and between November 2020 and January 2021 (T2) as part of the Brazilian multicenter project entitled Potential for Burnout and Strengthening of Health Workers in the Care Scenarios of Coronavirus Disease 2019 (COVID-19) [Potenciais de Desgaste e Forta lecimento dos Trabalhadores de Saúde Atuantes nos Cenários de Atendimento à Doença por Coronavírus 2019 (COVID19)]. The approximately six-month interval between assessments was established to evaluate changes in psychosocial outcomes during different phases of the COVID-19 pandemic in Brazil, encompassing the initial period of greater uncertainty and a subsequent stage characterized by increased clinical experience and organizational adaptation.

The baseline assessment included 437 healthcare workers, whose characteristics have been described in detail elsewhere [8,12]. All baseline participants were invited to complete the follow-up assessment (T2). Of these, 113 participants completed the second assessment and comprised the longitudinal sample analyzed in the present study, corresponding to a retention rate of 25.9%. The remaining 324 participants did not complete the follow-up assessment and were considered lost to follow-up. This study was designed and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement.

2.2. Participants and Data Collection

Because access to healthcare institutions was severely restricted during the first wave of the pandemic, participants were recruited using a non-probabilistic convenience sampling strategy complemented by snowball sampling. Eligible participants were physicians, nurses, nursing technicians, physiotherapists, psychologists, and other healthcare professionals directly involved in the care of patients with suspected or confirmed COVID-19 at any level of healthcare throughout Brazil.

Recruitment at T1 was conducted through invitations distributed via e-mail and social media using the researchers’ professional networks. Data were collected through an electronic questionnaire hosted on the Google Forms^® platform. Participation was voluntary, and electronic informed consent was required before access to the questionnaire.

The only inclusion criterion was active professional engagement in the care of patients with suspected or confirmed COVID-19. To minimize duplicate responses, participants were asked to provide an e-mail address, and only one response per participant was retained. All 437 participants who completed the baseline assessment were invited by e-mail to participate in the second assessment approximately 180 days later. The present longitudinal analyses included only participants who completed both assessments.

2.3. Study Variables and Measurements

A structured questionnaire developed specifically for this study was used to collect sociodemographic (sex, age, and region of residence/work), occupational (profession, and per capita family income), and clinical (self-reported morbidity) information.

The Scale of Indicators of Pleasure and Suffering at Work (EIPST) and the Work-Related Harm Assessment Scale (EADRT) are complementary scales of the Inventory on Work and Risk of Illness (ITRA), a multidimensional instrument developed within the Psychodynamics of Work framework to assess work context, human cost, experiences of pleasure and suffering, and work-related health harm. The ITRA was originally developed by Mendes and Ferreira and has subsequently undergone psychometric revalidation in different occupational populations [13,14].

Work-related experiences were assessed using the EIPST, which comprises four dimensions: two related to work-related pleasure (freedom of expression and professional fulfillment) and two related to work-related suffering (professional burnout and lack of recognition). Mean scores for the pleasure dimensions are interpreted as satisfactory (>4.0), critical (2.1–3.9), or severe (≤2.0), whereas mean scores for the suffering dimensions are interpreted as severe (>4.0), critical (2.1–3.9), or satisfactory (≤2.0) according to the original ITRA guidelines [13].

Work-related health harm was assessed using the EADRT, which comprises 29 items distributed across three dimensions: physical, psychological, and social harm. Recent psychometric evidence has further supported the validity of this scale among Brazilian health professionals [15]. Mean scores are interpreted as bearable (<2.0), critical (2.0–3.0), severe (3.1–4.0), or indicative of occupational illness (>4.0) [13,15].

For descriptive purposes, the original EADRT classifications were additionally grouped into three broader categories: mild risk (bearable), moderate risk (critical and severe), and high risk (occupational illness). This grouping was adopted exclusively to facilitate interpretation of the distribution of participants across clinically meaningful risk levels and did not affect the inferential analyses, which were performed using the original continuous scores.

Internal consistency of the EIPST and EADRT was assessed separately at baseline (T1) and follow-up (T2) using Cronbach’s alpha coefficients. The EIPST showed excellent internal consistency across its four dimensions, with alpha coefficients ranging from 0.908 to 0.932 at T1 and from 0.897 to 0.918 at T2. For the EADRT, Cronbach’s alpha coefficients ranged from 0.882 to 0.929 at T1 and from 0.885 to 0.943 at T2, indicating good to excellent internal consistency across all dimensions. Dimension-specific coefficients and corresponding valid sample sizes are presented in Supplementary Table S1.

2.4. Statistical Analysis

Data were entered into Microsoft Excel^® 2016 and analyzed using IBM SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA).

Continuous variables were summarized using means and standard deviations, whereas categorical variables were described using absolute and relative frequencies. The distribution of continuous variables was assessed through visual inspection of Q-Q plots, skewness and kurtosis coefficients, and the Kolmogorov–Smirnov test. Potential outliers were evaluated using the Grubbs test.

No imputation of missing data was performed. Longitudinal comparisons between T1 and T2 were performed using paired t-tests with 1000 bootstrap resamples and bias-corrected and accelerated (BCa) 95% confidence intervals. For descriptive purposes, EIPST and EADRT results are presented as mean item scores to allow interpretation according to the original instrument cutoffs. Inferential analyses were performed using total scores, as recommended for paired comparisons.

Effect sizes were estimated using Cohen’s d and interpreted as small (0.20–0.49), moderate (0.50–0.79), or large (≥0.80). To complement the interpretation of effect magnitude, the Probability of Superiority (PS) was calculated, representing the probability that a randomly selected participant would present a lower score at T2 than at T1. Statistical significance was established at p < 0.05 [16,17].

To account for multiple comparisons, the Benjamini–Hochberg procedure was used to control the false discovery rate (FDR) at 5%, and statistical significance was interpreted based on the adjusted results.

2.5. Ethical Considerations

The study was conducted in accordance with the ethical principles of Resolution 466/2012 of the Brazilian National Health Council and the Declaration of Helsinki. The protocol was registered in Plataforma Brasil (CAAE: 30599420.0.0000.0008) and approved by the National Research Ethics Commission (CONEP; approval number: 3979223). All participants provided informed consent electronically prior to participation.

3. Results

3.1. Sample Characteristics

A total of 113 healthcare professionals completed both assessments and were included in the longitudinal analyses. The sociodemographic, occupational, and clinical characteristics of the analytical sample are presented in

Table 1. Sociodemographic, occupational, and clinical characteristics of healthcare professionals included in the longitudinal analysis, Brazil, 2020–2021 (n = 113).

Variable	n	%
Sex (n = 112)
Male	29	25.9
Female	83	74.1
Age group (n = 113)
20–29 years	24	21.2
30–39 years	40	35.4
40–49 years	32	28.3
50–59 years	15	13.3
≥60 years	2	1.8
Per capita family income (BMW 1) (n = 97)
≤1.99	22	22.7
2.00–3.99	34	35.0
4.00–9.99	32	33.0
≥10.00	9	9.3
Region (n = 113)
North	16	14.2
Northeast	6	5.3
Central-West	3	2.6
Southeast	85	75.2
South	3	2.6
Profession (n = 113)
Nurse	70	61.9
Nursing technician/assistant	7	6.2
Physician	14	12.4
Physiotherapist	4	3.5
Psychologist	7	6.2
Others	11	9.7
Self-reported morbidity (n = 113)
No	73	64.6
Yes	40	35.4

¹ BMW = Brazilian minimum wages; minimum wage in 2020 = BRL 1095.00.

.

Most participants were female (74.1%), aged 30–39 years (35.4%), and worked or resided in the Southeast region of Brazil (75.2%). Nurses constituted the largest professional group (61.9%). Approximately one-third of the participants reported at least one morbidity (35.4%).

Information on sex was missing for one participant, and per capita family income was available for 97 participants.

Both instruments demonstrated good to excellent internal consistency at baseline and follow-up, supporting the reliability of the scale scores used in the longitudinal analyses.

3.2. Work-Related Pleasure and Suffering

The mean item scores and corresponding EIPST classifications at T1 and T2 are presented in

Table 2. Mean item scores and classifications of work-related pleasure and suffering, Brazil, 2020–2021 (n = 113).

Indicators and Dimensions	T1		T2
	Mean (±SD)	Classification	Mean (±SD)	Classification
Pleasure	3.9 (±1.2)	Critical	3.9 (±1.2)	Critical
Professional fulfillment	3.9 (±1.3)	Critical	4.1 (±1.2)	Satisfactory
Freedom of expression	3.8 (±1.2)	Critical	3.8 (±1.3)	Critical
Suffering	3.3 (±1.5)	Critical	3.0 (±1.4)	Critical
Professional burnout	4.0 (±1.4)	Severe	3.5 (±1.5)	Critical
Lack of recognition	2.8 (±1.7)	Critical	2.6 (±1.6)	Critical

. The overall pleasure indicator remained in the critical range at both assessments, with the same rounded mean item score at T1 and T2 (3.9).

The overall suffering indicator also remained in the critical range, although its mean item score decreased from 3.3 (SD = 1.5) at T1 to 3.0 (SD = 1.4) at T2.

Among the individual EIPST dimensions, professional fulfillment changed from a critical classification at T1 (mean = 3.9; SD = 1.3) to a satisfactory classification at T2 (mean = 4.1; SD = 1.2). Professional burnout decreased from a severe classification at T1 (mean = 4.0; SD = 1.4) to a critical classification at T2 (mean = 3.5; SD = 1.5). Freedom of expression and lack of recognition remained within the critical range at both assessments.

3.3. Work-Related Harm

Mean item scores and corresponding EADRT classifications are presented in

Table 3. Mean item scores and classifications of work-related harm, Brazil, 2020–2021 (n = 113).

Dimension	T1		T2
	Mean (±SD)	Classification	Mean (±SD)	Classification
Physical harm	2.9 (±1.3)	Critical	2.8 (±1.3)	Critical
Psychological harm	2.7 (±1.7)	Critical	2.3 (±1.7)	Critical
Social harm	2.1 (±1.5)	Critical	2.1 (±1.5)	Critical

. Physical, psychological, and social harm remained within the critical range at both assessments.

The mean item score for psychological harm decreased from 2.7 (SD = 1.7) at T1 to 2.3 (SD = 1.7) at T2. Physical harm showed a smaller numerical decrease, whereas the mean score for social harm remained unchanged after rounding.

3.4. Longitudinal Changes in Pleasure, Suffering, and Work-Related Harm

The longitudinal comparisons based on total scale scores are presented in

Table 4. Longitudinal comparisons of EIPST and EADRT total scores, Brazil, 2020–2021 (n = 113).

Variable	T1 Mean (95% CI)	T1 SD	T2 Mean (95% CI)	T2 SD	Cohen’s d	PS (%)	p-Value
EIPST
Pleasure	65.67 (61.8–69.4)	20.88	66.93 (63.5–70.4)	20.16	−0.07	52.0%	0.488
Professional fulfillment	35.19 (32.9–37.4)	12.11	36.91 (34.8–39.1)	11.32	−0.16	54.5%	0.104
Freedom of expression	30.49 (28.5–32.3)	10.14	30.02 (28.2–31.8)	10.01	0.05	51.4%	0.636
Suffering	49.94 (45.8–53.9)	22.24	45.44 (41.6–49.2)	21.73	0.21	55.9%	0.034
Professional burnout	27.85 (25.8–29.8)	10.07	24.68 (22.7–26.7)	10.26	0.31	58.7%	0.001
Lack of recognition	22.08 (19.6–24.6)	13.64	20.76 (18.3–23.1)	12.94	0.10	52.8%	0.302
EADRT
Physical harm	35.07 (32.1–37.8)	15.87	34.04 (31.1–37.1)	16.05	0.08	52.3%	0.396
Psychological harm	27.31 (24.2–30.4)	16.75	23.18 (20.0–26.5)	17.63	0.26	57.3%	0.006
Social harm	14.68 (12.8–16.4)	10.66	14.49 (12.6–16.3)	10.48	0.02	50.6%	0.828

Notes: T1 = April–June 2020; T2 = November 2020–January 2021; paired t-test with bootstrapping (1000 resamples; 95% BCa CI); EIPST = Scale of Indicators of Pleasure and Suffering at Work; EADRT = Work-Related Harm Assessment Scale; SD = standard deviation; CI = confidence interval; PS = probability of superiority. Statistical significance was additionally evaluated using the Benjamini–Hochberg procedure to control the false discovery rate (FDR = 5%).

. The overall suffering score decreased from 49.94 (SD = 22.24) at T1 to 45.44 (SD = 21.73) at T2. Although the unadjusted comparison reached statistical significance (p = 0.034), this finding did not remain significant after adjustment for multiple comparisons.

Professional burnout also decreased significantly, from 27.85 (SD = 10.07) at T1 to 24.68 (SD = 10.26) at T2, with a small effect size (d = 0.31; p = 0.001; PS = 58.7%).

No statistically significant longitudinal differences were observed in overall pleasure, professional fulfillment, freedom of expression, or lack of recognition.

Among the EADRT dimensions, psychological harm was the only dimension showing a statistically significant change. Its mean total score decreased from 27.31 (SD = 16.75) at T1 to 23.18 (SD = 17.63) at T2, with a small effect size (d = 0.26; p = 0.006; PS = 57.3%). No statistically significant differences were identified for physical or social harm.

Thus, statistically significant reductions were identified in professional burnout and psychological harm. Although overall work-related suffering showed a nominally significant reduction (p = 0.034), this finding did not remain statistically significant after controlling for multiple comparisons using the Benjamini–Hochberg procedure. All corresponding effect sizes were small, and most categorical risk classifications remained unchanged between assessments.

4. Discussion

This longitudinal study identified statistically significant reductions in professional burnout and psychological harm among healthcare professionals between the two assessments conducted during the COVID-19 pandemic. Although overall work-related suffering showed a nominally significant reduction, this finding did not remain statistically significant after adjustment for multiple comparisons. Nevertheless, the magnitude of all observed changes was small, and most indicators remained within critical risk classifications at T2.

The mean professional fulfillment score increased and changed from a critical to a satisfactory classification, but the longitudinal difference was not statistically significant. Therefore, this finding should be interpreted as a descriptive change in classification rather than evidence of a measurable improvement over time. Similarly, overall pleasure, freedom of expression, lack of recognition, physical harm, and social harm did not change significantly between assessments.

Taken together, the findings indicate that positive and negative experiences continued to coexist in healthcare work. Although suffering, burnout, and psychological harm decreased, the persistence of critical classifications suggests that the psychosocial burden remained substantial. This pattern is consistent with the work psychodynamics perspective, according to which pleasure and suffering are not mutually exclusive but may occur simultaneously within the same work environment.

The sociodemographic profile of the participants was broadly consistent with that reported in studies of healthcare workers during the COVID-19 pandemic [18,19]. Women constituted most of the sample, reflecting the marked feminization of the healthcare workforce, particularly in nursing [18,19]. Most participants were aged between 30 and 39 years, and approximately two-thirds reported no morbidity.

Nursing was the most represented professional group. Although this is consistent with the large contribution of nursing professionals to the healthcare workforce [18,20], nurses were considerably more represented than nursing technicians and assistants in the present sample. This distribution differs from the composition of the Brazilian nursing workforce, in which technicians and assistants constitute the largest occupational group [21]. This discrepancy likely reflects the non-probabilistic recruitment strategy and limits the representativeness of the sample.

Most participants with available income information reported a per capita family income between 2.00 and 3.99 Brazilian minimum wages. Low or insufficient remuneration has been associated with multiple job holding, extended working hours, and increased occupational demands in healthcare [19,22]. However, these relationships were not directly tested in the present study. Income should therefore be interpreted as a descriptive characteristic rather than as an explanatory factor for the observed psychosocial outcomes.

Within the EIPST framework, professional fulfillment refers to experiences of gratification, pride, identification with work, and perceived usefulness, whereas professional burnout encompasses frustration, insecurity, exhaustion, and emotional strain. Lack of recognition reflects experiences of injustice, devaluation, or insufficient acknowledgment of professional effort [13].

Professional fulfillment changed from a critical to a satisfactory classification, but its mean longitudinal difference was not statistically significant. This result may indicate that the sense of meaning and value attributed to healthcare work was maintained or modestly strengthened during the study period. Nevertheless, because the study did not directly evaluate meaning of work, social recognition, patient recovery, or perceived usefulness, these mechanisms cannot be confirmed.

Healthcare work during the pandemic was frequently associated with social purpose, responsibility, and commitment to patient care [23]. Evidence from France suggests that healthcare professionals’ perceptions of the value of their work were related to patient recovery, quality of care, and recognition from patients’ relatives [24]. Such factors may provide a possible context for understanding the maintenance of work-related pleasure and the descriptive increase in professional fulfillment. However, this interpretation should remain cautious given the absence of direct measures of these experiences in the present study.

The significant reduction in professional burnout was the most pronounced longitudinal change identified, although its effect size remained small. The baseline assessment occurred during the initial phase of the pandemic, when healthcare professionals faced considerable uncertainty regarding transmission, prevention, diagnosis, and treatment, as well as high service demands and reported shortages of protective equipment [4,5,25]. By T2, healthcare teams may have accumulated greater clinical experience and become more familiar with protocols and organizational routines. These contextual changes could have contributed to the reduction in burnout, but they were not measured directly and should not be interpreted as demonstrated causal mechanisms.

Previous studies have documented substantial physical and emotional strain among healthcare professionals during the pandemic, particularly among nursing workers. Frequently reported stressors included fear of infection, concern about transmitting the virus to family members, work overload, team conflicts, rapidly changing protocols, and repeated exposure to severe illness and death [26,27]. These conditions may help contextualize the severe burnout classification observed at baseline.

Coping strategies and psychological adaptation may also have changed over time [28,29]. Nevertheless, the present study did not measure resilience, coping, or individual adaptation. Accordingly, these constructs should be considered possible explanatory pathways rather than conclusions derived from the data. Moreover, reductions in distress should not be interpreted as evidence that working conditions had become adequate. Individual adaptation may coexist with persistent organizational deficiencies and should not replace institutional responsibility for protecting workers’ health [28].

Psychological harm was the only EADRT dimension that decreased significantly. Despite this reduction, its mean classification remained critical, as did those of physical and social harm. Therefore, the results suggest attenuation rather than resolution of work-related psychological distress. The small effect size further indicates that the average change was limited, even though it reached statistical significance.

The findings should also be interpreted within the broader structural conditions of healthcare work. Before the pandemic, many healthcare systems already faced staffing shortages, precarious employment, insufficient remuneration, high workloads, and limited organizational support [22,30]. The pandemic likely intensified these pre-existing weaknesses rather than creating them. This context may help explain why some indicators decreased over time while overall psychosocial risk remained elevated.

The second assessment was conducted between November 2020 and January 2021, when the pandemic remained severe in Brazil, and vaccination of the general population had not yet begun. Healthcare professionals continued to face substantial service demand and uncertainty. Therefore, the reductions observed in suffering, burnout, and psychological harm occurred within a continuing crisis rather than after the end of the pandemic. This may partly explain why the improvements in continuous scores were modest and why most categorical classifications remained critical.

From an occupational health perspective, the persistence of critical psychosocial risk classifications indicates that reductions in individual distress indicators should not be interpreted as sufficient recovery. Organizational interventions remain necessary to address the structural and psychosocial conditions of healthcare work.

Potential measures include adequate staffing, workload and shift management, clear communication, access to confidential mental health support, opportunities for worker participation in organizational decisions, and formal recognition of professional contributions. Interventions should prioritize changes in work organization rather than relying predominantly on individual resilience or coping strategies.

The findings also support the inclusion of psychosocial risk assessment and management in occupational health policies and emergency preparedness plans. Future crisis-response strategies should incorporate worker protection, staffing capacity, communication systems, psychological support, and mechanisms for monitoring work-related suffering throughout the emergency and recovery periods.

The longitudinal design is a major strength of this study because it enabled the assessment of within-person changes during two distinct phases of the COVID-19 pandemic. Furthermore, the use of validated instruments addressing both positive and negative work experiences provided a comprehensive assessment of psychosocial risks.

Nevertheless, several limitations should be considered. First, the convenience and snowball sampling strategies, together with the relatively small analytical sample and loss to follow-up, may have introduced selection, attrition, and survivorship bias, limiting the generalizability of the findings. Participants who completed both assessments may have differed from those who were lost to follow-up. Second, all measures were self-reported, making the results susceptible to reporting and common-method bias, while objective indicators of working conditions and health outcomes were not available. Third, potentially relevant contextual factors, such as workload, organizational support, availability of personal protective equipment, and workplace changes, were not measured; therefore, the observed longitudinal differences cannot be attributed to specific individual or organizational mechanisms. Finally, the analyses were based on only two assessment points and involved multiple comparisons, requiring the findings—particularly those with small effect sizes—to be interpreted with caution. Future studies with larger and more diverse samples, repeated follow-up assessments, and broader organizational measures are warranted.

5. Conclusions

This longitudinal study found small but statistically significant reductions in professional burnout and psychological harm among healthcare professionals during the COVID-19 pandemic. Although work-related suffering also decreased numerically, this finding did not remain statistically significant after adjustment for multiple comparisons. Most psychosocial risk indicators remained within critical risk classifications, indicating that the overall psychosocial burden persisted throughout the study period. Although professional fulfillment shifted from a critical to a satisfactory classification, its longitudinal change was not statistically significant.

These findings suggest that some improvement in psychosocial distress may occur over time even during prolonged public health crises; however, such changes are insufficient to eliminate critical occupational risks. From an occupational health perspective, the results reinforce the need for sustained organizational strategies aimed at preventing psychosocial risks through adequate staffing, workload management, supportive work environments, and access to mental health resources. Such measures are essential not only to protect healthcare workers during routine practice but also to strengthen the resilience of healthcare systems in future public health emergencies.