To our knowledge, this longitudinal study is the first in our country to systematically compare the six consecutive pandemic waves, with a specific focus on in-hospital fatality rates, providing a comprehensive understanding within the Tunisian context. The study included all COVID-19-infected patients admitted to both public and private health facilities in the Monastir governorate from March 2020 to March 2022, covering six infection waves. The in-hospital case fatality rate fluctuated across the six waves, aligning with global trends and highlighting the dynamic nature of the pandemic. The initial wave, marked by a prompt response and effective preventive measures, successfully contained the virus. However, subsequent waves, influenced by factors such as border reopening and the relaxation of preventive measures, presented new challenges. The fourth wave, driven by the Alpha variant, saw the highest in-hospital CFR, peaking at 27.5%, due to the variant's characteristics, strain on the healthcare system, and pandemic fatigue. The fifth wave, driven by the Delta variant, experienced the highest numbers of infections, hospitalizations, and in-hospital deaths, reflecting the complex interaction of variant aggressiveness, delayed vaccination, and relaxed measures. The final wave demonstrated lower hospitalization and mortality rates, likely influenced by the less virulent nature of the Omicron variant, along with increased vaccination coverage and community immunity.

Our study revealed a male predominance among hospitalizations and in-hospital deaths. These observations align with previous findings in both national and international literature, suggesting that men are more prone to developing severe forms of the disease and require hospitalization due to COVID-19 [18,19,20,21,22,23,24,25,26]. A national longitudinal study covering all COVID-19-related deaths in Tunisia between March 2020 and February 2021, totaling 8051 deaths, reported a sex ratio (M/F) of 1.8 [27]. Besides, a metanalysis (MA) incorporated data from 43 studies spanning 12 countries indicated a male predominance among hospitalized patients in all countries studied, with Europe reporting the highest percentage of males (69.6%), followed by North America (58.2%), and Asia (54.5%) [28]. These disparities based on gender could be attributed to biological sex distinctions, including sex-based variations in immune responses, hormonal influences and genetic polymorphism. It is crucial to recognize that gender differences in health behaviors, such as smoking, and the presence of pre-existing conditions, such as diabetes and hypertension, can also significantly contribute to these disparities [20, 29, 30].

In terms of age, the elderly demonstrated the highest rates of hospitalization and in hospital death. The thorough national study on all Covid-19-related fatalities in Tunisia from March 2020 to November 2022 disclosed a median age of 72 years, ranging from 0 days (newborns) to 112 years [31]. This difference underscores the variation in population distribution among the country's governorates. Besides, an umbrella review comprising 120 systematic reviews revealed a consistent, linear relationship between age and the risk of disease severity [32]. Thus, it is well established in the literature that the elderly are more vulnerable to the COVID-19 pandemic [33, 34]. Concerning the pediatric population (18 and under), our study revealed significantly lower hospitalization and in-hospital case fatality rates compared to other age groups at 1.3% and 8.3% respectively. Our CFR among children is lower than the rates reported in the United States [35, 36], Mexico [37], India [38], Indonesia [39], and Iran (13%) [40]. These differences may be explained by variables such as predominant virus driving the outbreak at the time of reporting, social restriction measures enforced, prevalence of comorbidities, socioeconomic issues, health infrastructure, or even ethnicity as multisystem inflammatory syndrome has been more frequently reported in Hispanic children [41].

Additionally, in our study, we observed an increase in the number of infections and hospitalizations within this age cohort over the course of the pandemic, rising from 0% in W1 to 18.6% in W6 for infections, and from 0% in W1 to 6.6% in W6 for hospitalizations. The observed evolution among children during the pandemic aligns with findings in the literature and could be attributed to changes in prevention policies [42]. Initially, schools were closed, reducing contact among adolescents and youth and consequently decreasing the number of cases. However, with the reopening of schools, the relaxation of preventive measures and the implementation of a selective vaccination approach deliberately excluding children, the situation changed. Such strategies, adopted in various countries due to concerns over vaccine safety, resulted in limited vaccine coverage among children [42, 43].

Concerning the distribution of comorbidities, the three most prevalent comorbidities in our study among hospitalizations and deaths were hypertension (25.9% vs. 28.2%), diabetes (23.8% vs. 24.0%), and cardiovascular diseases (11.3% vs. 13.2%), observed both overall and across different waves. This aligns with findings from national and international studies [21, 28, 44,45,46,47,48,49,50]. Studies extensively demonstrated a robust correlation between the patient's medical condition and the severity of the infection [51].

Regarding symptoms, consistent with the literature, the majority (63.8%) of individuals experienced mild to moderate symptoms [52], a trend observed across different waves. Additionally, our findings align with previous studies, highlighting cardiorespiratory signs (69.1%) and fever (39.4%) as the most prevalent clinical signs [21, 53,54,55,56]. In an international study which included 60,109 symptomatic COVID-19 inpatients, the most frequently reported symptoms were fever, cough, and dyspnea [55].

In our study, we discussed symptoms in broad categories based on systems. During W6, we noticed a decline in the occurrence of respiratory symptoms, fever, weakness, neurological, and rheumatological symptoms. However, the signs related to ear, nose, and throat (ENT) stayed relatively unchanged. Similarly, various studies in the literature revealed a decrease in symptoms like cough, fever, shortness of breath, taste/smell loss, muscle pain, fatigue, and headache following the prevalence of the Omicron variant. However, a notable increase in the incidence of sore throat was specifically observed among the symptoms [57]. Studies showed that ageusia, anosmia and severe hypoxemia were less prevalent for Omicron than for the other variants, and cold-like symptoms were more indicative of Omicron compared to the earlier variants [58] [59].

According to our findings, the median LoS was 6 days [2,3,4,5,6,7,8,9,10,11]. The LoS in the ICU was twice the hospitalization duration at 12 days [6,7,8,9,10,11,12,13,14,15,16,17,18]. A systematic review of 52 studies, with 46 originating from China, revealed variations in the median LoS ranging from 4 to 53 days within China and 4 to 21 days outside of China. The decrease in the LoS over the course of the pandemic could be attributed to an enhanced understanding of the disease's pathophysiology, improved secondary care due to scientific advancements, and the exchange of information among different teams worldwide. Furthermore, the introduction of vaccination may have played a role in this improvement. Indeed, scientific literature reports that vaccinated individuals have a significantly shorter hospital stay (both with and without ICU admission) compared to those who are unvaccinated [60].

Concerning the evolution of in-hospital fatality across waves, the pandemic unfolded in six distinct waves in the region of Monastir during the study period, mirroring the overall pattern of the pandemic in Tunisia [61]. Despite fluctuations across waves, the in-hospital CFR consistently remained among the elevated figures globally [4].

The first wave had the lowest number of confirmed cases. Since January 22, 2020, Tunisia has implemented a series of stringent and proactive measures to curb the spread of the virus [62, 63]. As the virus began circulating within the country in early March 2020, authorities took swift and diverse actions to reduce COVID-19 transmission. Thanks to this wise and prompt strategy coupled with the Tunisian population's rigorous adherence to hygiene measures, the country was able to maintain an exceptionally low infection rate [61, 62]. As such, Tunisia's response to the COVID-19 pandemic during its first wave can offer valuable guidance for similar epidemic situations [63].

The second wave in our study saw a higher number of infections than the first wave (6,450 individuals), with an in-hospital CFR of 14.4%. According to the report by the National Observatory of New and Emerging Diseases (NONED), during this period, the Monastir governorate was classified as having a very high-risk level, with an incidence rate equal to or greater than 100 cases per 100,000 inhabitants, which explains the high hospitalization rate of 16.7% compared to the national Fig. (4.8%) [64].

The third wave exhibited a higher incidence of infections, hospitalizations, and in-hospital deaths, with a significant increase in the in-hospital CFR to 23.8%. This rise can be attributed to the occupancy of ICU beds and the emergence of the Alpha variant. The decrease in hospitalization rates during the third wave compared to the second in the Monastir region, coupled with the increase in the CFR, can be primarily attributed to the improvement in care provided by primary care physicians, thus contributing to the reduction of excessive hospitalizations. Additionally, subsequent adjustments to admission and discharge criteria were made to prioritize services for those who needed them the most [25].

The fourth wave exhibited a lower total number of infections compared to W2 and W3, with a recorded count of 3,835 cases. However, there was a significant increase in the hospitalization rate, in-hospital CFR, and the percentage of severe cases in comparison to previous waves. Notably, there was a significant decrease in the percentages of hospitalized individuals and deaths within the older age groups compared to the third wave. On the other hand, the number of deaths increased among young people, particularly those aged between 19 and 45. The fourth wave coincided with the onset of the first phase of the national COVID vaccination program, which was launched in March 2021, which focused on healthcare workers and the elderly due to limited vaccine availability. This prioritization of vaccines could explain the reduction in hospitalization and lethality percentages among the elderly and its increase among younger individuals, a consequence of the vaccination's selection strategy. However, simultaneously, this wave was characterized by the emergence of the Alpha variant in Tunisia, which quickly became the most dominant variant [12], replacing the original Wuhan strain.

In Tunisia, the maximum peak was reached during W5, which began in July 2021 [25]. In May 2021, the Delta variant was initially identified in the country, leading to an uncontrolled spread across almost all regions throughout the summer of 2021 [7, 65]. The Delta variant proved to be the most lethal in terms of cumulative deaths in Tunisia, leading to an increase in hospitalizations and the demand for oxygen. Consequently, in July, Tunisia experienced the highest mortality rate in the Eastern Mediterranean region and on the African continent, recording 42.3 new deaths per 100,000 inhabitants due to Covid-19 [66]. On a global scale, despite vaccination programs and extensive vaccine coverage, Delta variant has rapidly spread, affecting the majority of territories and causing an increase in the number of cases, hospitalizations, ICU admissions and overall mortality, especially among the younger population [4]. This wave was a consequence of the relaxation of public health measures and inadequate control of preventive measures. On an individual scale, people experienced fatigue and a desire to return to their pre-pandemic habits, contributing to the resurgence of cases.

The sixth wave was characterized by the lowest hospitalization rate at 2.8% and a reduction in the number of severe cases despite the significant number of infected cases (17.855). Many studies corroborated our findings that confirmed Omicron cases had a lower hospitalization rate compared to other waves [18, 67,68,69,70]. In addition, the in-hospital CFR in the present study decreased compared to W3, W4, and W5, reaching 19.4%. This improvement aligns with a lower incidence of respiratory symptoms, fewer severe cases, and shorter hospitalization durations. Studies have consistently reported the milder nature of Omicron symptoms, including reduced rates of dyspnea, lower severity, fewer hospital admissions, and quicker recovery when compared to earlier variants, particularly the Delta variant [67, 71,72,73].

This improvement was likely influenced by factors such as increased vaccination coverage, and the less aggressive nature of the Omicron variant​.

The global transformation in the SARS-CoV-2 infection profile and the attenuation of severity during W6 likely resulted from several factors. Firstly, the highly mutated Omicron variant, despite its increased transmissibility compared to previous variants, potentially exhibited reduced pathogenicity [74, 75]. The Omicron variant tended to replicate more in the upper respiratory tract than in the lungs, which may have reduced the risk of mortality [76, 77]. Secondly, the extensive vaccination coverage resulted in a high level of acquired immunity [78], due to the campaigns that began in March 2021, initially targeting high-risk populations, followed by mass vaccination that started in July 2021 [8].

Thirdly, there was a higher level of natural immunity due to previous infections. Additionally, a better understanding of the disease's pathophysiology, lower epidemic pressure, improvement in secondary care due to scientific advancements, and the exchange of information among different teams worldwide also played a crucial role in reducing the severity level [26].

This study was the first to comprehensively investigate hospital outcomes related to COVID-19, analyzing fatality rates across the six waves of the pandemic over a period of two years. The analysis involved a comparison of data over time (epidemic curve) and among the population based on socio-demographic and clinical characteristics. Secondly, it involved an extensive dataset of high-quality information gathered prospectively and subsequently verified through medical records. Thirdly, we closely followed all patients in the cohort until death or discharge. Ultimately, the exclusion of patients incidentally infected with SARS-CoV-2 (asymptomatic or mildly symptomatic and without diagnostic confirmation) was implemented to maximize the certainty that mortality was directly attributed to COVID-19.

While our discussion highlighted significant strengths, it is important to acknowledge certain limitations. In our study, the absence of detailed data on the direct cause of death may have led to a slight overestimation of in-hospital fatality rates. Establishing a direct causal link between death and COVID-19 was challenging, particularly in a hospital setting, due to the high prevalence of pre-existing health conditions. Additionally, our study focused on hospital deaths, excluding cases where patients were discharged but later passed away at home. This exclusion is consistent with the definition of in-hospital mortality. Furthermore, we did not include data on the vaccination status or previous COVID-19 infections of patients—key factors that likely influenced variations in mortality rates across different phases of the pandemic. To address this, we relied on national data regarding vaccination coverage in Tunisia. While some patient data was missing, the large sample size played a critical role in minimizing the risk of systematic bias when comparing different periods.