Acute respiratory infections (ARIs) driven by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and influenza viruses remain significant contributors to global morbidity and mortality. Since late 2019, SARS-CoV-2 has spread globally, resulting in over 776 million reported cases (1), while seasonal influenza causes an estimated 3–5 million cases of severe illness and 290,000–650,000 deaths annually (2).

Following the 2009 H1N1 influenza pandemic, the World Health Organization (WHO) standardized the case definition for Influenza-Like Illness (ILI)—typically fever (≥38 °C) and cough—specifically to capture influenza activity (3). In response to the COVID-19 pandemic, surveillance systems worldwide have integrated SARS-CoV-2 testing into these existing influenza sentinel networks to maximize resource efficiency (4, 5). However, while ILI serves as a robust benchmark for influenza, its effectiveness in capturing current COVID-19 variants appears to be diminishing. This challenge intensified with the emergence of the Omicron subvariant BA.2.86 and its descendants, which arose in late 2023 carrying numerous spike mutations and rapidly achieved global predominance in early 2024 (6). Recent observations indicate that unlike ancestral strains, Omicron variants are increasingly characterized by upper respiratory symptoms (e.g., sore throat) rather than the systemic febrile response (7). Consequently, the sensitivity of standard fever-based definitions is increasingly being challenged when applied to these modern variants (8), raising concerns that relying strictly on ILI definitions may introduce a surveillance bias that obscures the true burden of COVID-19 compared to influenza.

Beyond surveillance, effective public health preparedness demands a precise understanding of healthcare resource utilization, encompassing illness duration and care-seeking behaviors (9). However, current evidence regarding the comparative healthcare burden of COVID-19 and influenza remains inconsistent. While large-scale hospital-based cohorts have demonstrated that COVID-19 generally carries a higher risk of post-acute multi-organ sequelae and mortality, notable distinctions emerge: the same comparative analyses highlight that influenza imposes a significantly higher burden specifically on the pulmonary system (10). Furthermore, in the acute phase, recent studies indicate that influenza generates heavier outpatient demand driven by a significantly higher prevalence of high-grade fever compared to Omicron variants (11). These distinct profiles suggest that the two viruses drive differential healthcare demands across the acute and post-acute phases. Therefore, clarifying these phase-specific utilization patterns is essential for optimizing resource allocation.

To address these knowledge gaps, the primary objective of this study was to directly compare the clinical profiles, surveillance sensitivity, and healthcare burden of COVID-19 and influenza under a unified framework. We analyzed data from a community-based, prospective syndromic surveillance cohort in Shanghai to achieve this. The novel contributions of this work are threefold. First, unlike hospital-based studies, our prospective community-based design minimizes selection bias and captures the full spectrum of disease severity, including mild cases. Second, we explicitly contrast symptom profiles to quantify the specific sensitivity gap of standard ILI definitions for current COVID-19 variants. Third, we extend the comparative analysis into the post-acute phase, providing critical evidence on long-term healthcare utilization differences. The remainder of the paper is structured as follows: we first detail the study design and data collection; then present the comparative results on symptom profiles and healthcare-seeking behaviors; and finally discuss the implications for surveillance policy and resource allocation.

In this community-based prospective cohort study, we identified a distinct divergence in the clinical profiles, surveillance sensitivity, and healthcare-seeking behaviors of COVID-19 and influenza. While influenza maintained a classic febrile presentation driving high medical demand, Omicron-era COVID-19 manifested as a milder, often afebrile upper respiratory illness that frequently evaded detection under standard ILI surveillance criteria.

Consistent with prior evidence regarding the Omicron variant (7), COVID-19 cases in our cohort presented with less fever and more sore throat. This phenotypic divergence is likely driven by distinct viral tropisms and altered cell entry mechanisms. Unlike influenza viruses, which typically induce robust systemic inflammation, recent virological studies indicate that Omicron variants have evolved a reduced reliance on the cell-surface protease TMPRSS2 for entry, favoring the endosomal pathway instead (15). This evolutionary shift limits viral replication in the lung parenchyma while enhancing efficiency in the ACE2-rich oropharyngeal epithelium, thereby mechanistically explaining the predominance of upper respiratory symptoms such as sore throat over systemic severity (16, 17). This clinical profile stands in stark contrast to the high-grade fever that continues to characterize seasonal influenza A infections (11). Consequently, this poses a fundamental challenge to traditional surveillance. Our findings suggest that strictly fever-based ILI criteria may underestimate COVID-19 activity. To enhance surveillance sensitivity, adopting a broader ARI framework warrants consideration. This approach aligns with the integrated surveillance strategy employed by the US CDC, which utilizes broad syndromic metrics to capture the aggregate burden of co-circulating respiratory viruses (18). Although including afebrile symptoms inevitably lowers clinical specificity (19), ARI serves as a sensitive “screening net” for sentinel surveillance. In this model, high sensitivity ensures the capture of mild COVID-19 cases, while specificity is subsequently confirmed by laboratory testing. Thus, monitoring ARI trends alongside virological positivity rates is essential for accurate epidemic assessment during periods of co-circulation.

Healthcare utilization patterns mirrored symptomatic differences. COVID-19 cases exhibited significantly lower short-term healthcare-seeking rates (13.87% vs. 33.71%) and shorter illness duration (6.66 vs. 8.25 days). While pre-Omicron variants were associated with greater severity (20), our findings align with recent evidence showing shorter hospital stays for Omicron (6.20 days) than influenza (8.30 days) (21). In the post-acute phase, influenza cases retained a higher outpatient burden over 90 days, aligning with findings from Taiwan showing lower post-acute utilization for COVID-19 (22). This sustained demand challenges the perception that influenza recovery is immediate. While Long COVID rightly garners attention for its multi-organ sequelae, recent large-scale cohorts clarify that for the pulmonary system specifically, the long-term risks associated with seasonal influenza can actually exceed those of Omicron (10). Mechanistically, influenza’s higher propensity for secondary bacterial co-infections likely compounds this persistent burden, driving both prolonged symptomatic recovery and a sustained reliance on outpatient care beyond the acute phase (23).

This divergence in clinical severity likely shapes transmission dynamics. In Shanghai, COVID-19 activity persisted across three seasons, contrasting with the sharper, shorter winter peak of influenza. The more frequent mild or afebrile presentation of Omicron-era COVID-19 may reduce the effectiveness of symptom-triggered care seeking and isolation, allowing community activity to persist and remain harder to detect (24, 25). Conversely, influenza more often causes acute febrile illness that physically incapacitates patients (26); however, due to its strong seasonality, it tends to concentrate healthcare demand into a compressed, high-intensity period (27). This creates a distinct public health paradox: COVID-19 requires a wide, sensitive net to detect invisible transmission, whereas influenza demands a robust, resilient buffer to absorb visible clinical surges.

A key strength of this study is its prospective, community-based design, which captures mild, non-attending cases often missed by hospital-based surveillance. Furthermore, linkage with the Shanghai Health Cloud allowed for the integration of longitudinal follow-up with real-world healthcare utilization records. However, limitations exist. First, self-reported symptoms are subject to recall bias, though weekly follow-ups aimed to mitigate this. Second, voluntary testing may introduce selection bias; yet, such bias is likely non-differential and unlikely to alter comparative conclusions. Third, as our findings reflect specific variants (Omicron BA.2.86 and A (H1N1)pdm09), continuous re-evaluation is warranted given the rapid evolution of respiratory viruses.

In conclusion, this prospective community-based study characterized the distinct burden profiles of Omicron-era COVID-19 and seasonal influenza. Our findings demonstrated that while influenza maintained a classic febrile presentation, Omicron variants manifested primarily as mild upper respiratory infections that frequently evaded detection under standard ILI surveillance criteria. Furthermore, longitudinal follow-up revealed that influenza imposed a more significant burden on outpatient resources during the acute and early post-acute phases compared to the milder clinical trajectory of Omicron cases.

Looking forward, effective post-pandemic management requires calibrating public health strategies to these distinct viral profiles. We recommend that surveillance systems transition from fever-based ILI indicators to broader ARI metrics to accurately capture the burden of current and future SARS-CoV-2 variants. Additionally, resource allocation planning must account for the divergent demands of these pathogens: maintaining high-sensitivity monitoring to track community transmission of SARS-CoV-2, while ensuring resilient clinical capacity to manage the intense healthcare surges driven by seasonal influenza.