Clostridioides difficile is a Gram-positive spore-forming bacillus that causes pathogenicity by producing toxin A and toxin B. Some strains can also produce a binary toxin, CDT, which significantly enhances virulence (Di Bella et al., 2024). From a global public health perspective, the Centers for Disease Control and Prevention (2019) of the United States has classified it as an “urgent threat pathogen”. Globally, systematic reviews and disease burden studies indicate that the burden of CDI continues to rise in regions with high Socio-Demographic Index and among elderly populations, although significant spatiotemporal distribution variations exist across different areas (Xia et al., 2025).

Over the past two decades in North America and Europe, the hypervirulent strain ST1/RT027, which caused large-scale outbreaks in the mid-2000s, gradually declined, while RT014/020, RT002, and ST11/RT078 increased instead (2024) (European Centre for Disease Prevention and Control, 2024). Over the past decade in China, researchers and clinicians have increasingly focused on CDI, but most of the research has come from Beijing and eastern China, with limited data from other regions. A systematic review and meta-analysis of the literature, covering the sampling years between 2009 and 2020, reported an overall CDI detection rate of 11.4% (2,696/26,852) (Wen et al., 2023). However, only 4.1% of the data came from Southwest China, excluding Sichuan and Henan provinces, which are among the Top 5 provinces by population. To date, a nationwide surveillance network for C. difficile, similar to those in Europe or North America, has not been established.

The COVID-19 pandemic has imposed profound challenges on global health systems, including those related to CDI. During the pandemic, a bundle of robust infection prevention and control measures was implemented, influencing the epidemiology of CDI (Vendrik et al., 2022). Interestingly, the incidence of CDI rose in one study (Karampatakis et al., 2023) but fell in another (Ponce-Alonso et al., 2021). The extensive use of broad-spectrum antibiotics and the escalating issue of AMR severely disrupted gut microbiota and increased the risk of CDI (Markovska et al., 2023). Additionally, reports indicated a rise in severe CDI cases during the mid-to-late stages of the pandemic, potentially linked to delayed medical consultation and diagnosis (Vendrik et al., 2022). These dual pressures may have accelerated adaptive evolution in C. difficile. For instance, studies have identified the emergence of a new epidemic strain, RT176, in central Europe during the pandemic (Krutova et al., 2025). Therefore, the long-term impacts of the pandemic on the dynamics of evolution and AMR of C. difficile require further monitoring and research. In December 2022, in China, the “zero-COVID” policy was relaxed, and both population-wide screening and the digital travel code were canceled. Characteristic changes in CDI cases and C. difficile strains after the pandemic are unclear.

In this study, we analyzed clinical data from CDI patients admitted 6 months after policy relaxation to a 5,000-bed tertiary hospital that also serves as a national center for managing complex cases and critical patients transferred from all hospitals in Southwest China. C. difficile isolates were recovered from fecal samples of these patients and subjected to WGS. By combining clinical data and genomic analyses, this study seeks to clarify the profiles of C. difficile and CDI cases after the pandemic, offering healthcare professionals additional insights to enhance future monitoring and prevention of CDI.

This study systematically delineated the clinical and pathogenic landscape of CDI in Southwest China in the post-COVID-19 era. Clinically, the disease burden remains substantial, driven by host susceptibility and healthcare-associated selection pressure. Genomic analysis revealed diverse sequence types and AMR profiles, confirming persistent regional-dominant lineages, alerting to potential clonal relatedness, and providing evidence for clinical treatment considerations. These findings align with the globally recognized spatiotemporal heterogeneity of C. difficile and uncover previously underdocumented region-specific transmission dynamics in the post-pandemic period. By integrating genomic and clinical data, this study enhances understanding of C. difficile epidemiological characteristics, resistance patterns, and transmission risks in healthcare settings, extending beyond previous studies that primarily relied on pre-2020 isolates and geographically limited cohorts.

Not surprisingly, our detection rate of CDI among suspected diarrhea inpatients was 5.4%. Over half of our patients were over 65 years old with typical high-risk medication exposures (such as PPIs and broad-spectrum antibiotics), consistent with two pre- and intra-pandemic studies from Chongqing (Dai et al., 2020; Cui et al., 2023) (Table 3) and previous global literature (Di Bella et al., 2024). In contrast to international studies reporting a higher proportion of females in CDI (Di Bella et al., 2024), males comprised the majority of CDI patients in this research, which aligns with the two previous studies from southwestern China (Dai et al., 2020; Cui et al., 2023) (Table 3). More importantly, the clinical burden of CDI in this region should not be underestimated. Lacking hypervirulent strains (no ST1 and only 1 ST11 were found), 20.4% of patients still developed severe infections, including two fulminant episodes. The causative strains from these patients were distributed across 8 distinct STs: ST3, ST2, ST54, ST37, ST81, ST42, ST17, and ST5 (Figure 2). The proportion of severe cases was comparable to the 21.5% reported in Zhejiang Province for the sampling years 2009–2016 (Xu et al., 2017) and lower than 41.4% in the United States (Drwiega et al., 2024), highlighting the combined regional differences in host susceptibility, strain genotype and toxicity, clinical management, and infection control strategies. Based on the above, there is a clear need to optimize antimicrobial and PPIs prescribing practices, refine risk stratification, and monitor elderly or immunocompromised patients more closely.

In this study, the predominant lineage was ST3, which consistently produces toxin A and B. This aligns with previous systematic reviews involving Chinese research before the pandemic (Wen et al., 2023), particularly with studies from other regions in Southwest China (Liao et al., 2018; Cui et al., 2023; Gu et al., 2025) (Table 3), further supporting that ST3 remains dominant in China in the post-pandemic era. Still, isolates that produce binary toxin and carry cdtA/B were rare, and no hypervirulent ST1/RT027 strains were identified. Instead, ST5, as a hypervirulent phenotype that appeared sporadically and was rarely reported in China (Shaw et al., 2020), was the main clone producing binary toxin prevalent in this region. ST11, occasionally identified in elderly patients in Beijing (Suo et al., 2017), was also isolated from one patient in this region. Since it was previously identified in animals in east China and farms were regarded as the reservoirs, we agree that both zoonotic and patient infections warrant ongoing surveillance (Zhang et al., 2019).

In this study, a notable shift in the dynamics of tcdA − tcdB + cdtA/B− strains was observed. ST37 has historically dominated in China and across Asia (Collins and Riley, 2018). But now, unlike other regions in Southwestern China (Dai et al., 2020; Cui et al., 2023) (Table 3), ST81 has replaced ST37 as the second-most prevalent clone in this region, following ST3. All our ST81 strains carried gyrA Thr82Ile and gyrB Ser366Ala/Asp426Val mutations (Figure 2), conferring universal resistance to moxifloxacin. A Japanese review (Senoh and Kato, 2022) and a study from Beijing (Cheng et al., 2020), both of which analyzed samples collected before the COVID-19 pandemic, reported an increased prevalence of ST81 with high fluoroquinolone resistance. Moreover, in the present study, approximately 20% of ST81 isolates also exhibited metronidazole resistance. Although they differ by only 1 allele, genomic and phenotypic comparisons suggested that ST81 had advantages over ST37 in sporulation, environmental adaptation, and AMR (Su et al., 2021). The persistent rapid expansion of ST81 with enhanced fitness and high AMR underscores the need for strengthened clinical surveillance to alert to a potential outbreak.

Despite strict prescription restrictions on fluoroquinolones at our hospital since 2017, resistance to moxifloxacin remained remarkably high at 56.0%, comparable to the intra-pandemic fluoroquinolone resistance levels of Chongqing (62.2%) (Cui et al., 2023), and significantly exceeding the pre-pandemic levels in southwestern China (13.0–14.5%) (Dai et al., 2020; Gu et al., 2025) (Table 3) and levels outside the Southwest China (16–43%) (Yang et al., 2020; Zhang et al., 2024). This aligns with the high exposure rate to fluoroquinolone antibiotics (second only to β-lactams) among CDI patients before the onset of illness, indicating that a history of fluoroquinolone usage remains a significant selective pressure. When integrated with molecular typing results (Figure 4), it becomes evident that both the dominant ST3 and clade 4, particularly ST81, contribute substantially to this trend, reinforcing the view that fluoroquinolone restriction stewardship should remain our critical intervention measure. Regarding metronidazole, the resistance rate in this study was 3.7%, higher than that reported in Shanghai (1.07%) but lower than some earlier data (15.7%) (Jin et al., 2017; Yang et al., 2020). In contrast, metronidazole resistance was not detected in other studies from southwestern China (Dai et al., 2020; Cui et al., 2023; Gu et al., 2025) (Table 3), collectively highlighting significant regional and temporal variations. No pCD-METRO plasmid was detected in this cohort, consistent with its reported global low prevalence (<0.2%) (Olaitan et al., 2023). Among the 14 ST81 isolates carrying PnimB^G, only 4 demonstrated metronidazole resistance, consistent with a previous report of the genotype-phenotype discrepancy (Kolte and Nubel, 2024). Therefore, PnimB^G alone cannot determine the resistance phenotype, and its contribution may depend on other genetic and metabolic backgrounds. Notably, as shown in Figure 2, PnimB^G was observed exclusively among ST81 strains with the Thr82Ile substitution in gyrA, supporting the recent finding that strains carrying both mutations spread more rapidly (Olaitan et al., 2023). So, ST81 should be closely followed to prevent further dissemination, and genomic surveillance of gyrA/gyrB and nimB promoter variants is warranted.

Strains in our study harbored ARGs that confer resistance to multiple classes of antimicrobial agents. Of particular concern was the first detection of cfr(B) in one ST54 isolate in China. The cfr gene family confers cross-resistance to oxazolidinones (linezolid), phenicols, lincosamides, pleuromutilins, and streptogramin A, and has been demonstrated to be carried on mobile genetic elements with horizontal transfer potential. Previously, a Chinese study reported the presence of cfr(B) in ST37 and ST3 strains (Fu et al., 2025). Recent reports from animal sources have identified the cfr(B) gene in C. difficile (Brenciani et al., 2025). These findings indicate that the dissemination of cfr(B) is not restricted to a specific ST of C. difficile, and that C. difficile acts as a potential reservoir and vector for this resistance gene, underscoring its public health threat. It is therefore crucial to incorporate cfr(B) into ongoing integrated “One Health” surveillance efforts targeting human-animal-environment transmission pathways.

Integrating WGS with traditional epidemiological investigations facilitates the early identification of transmission. Incorporating genomic and temporal thresholds (e.g., ≤2 cgSNPs within 124 days) into the surveillance framework may provide an actionable early-warning tool for timely interventions. In this study, although genetically closely related strains were identified by cgSNP comparison and certain spatiotemporal overlaps were revealed by epidemiological investigations, no confirmed nosocomial transmission or distinct clonal dissemination due to close contact was found. However, the potential clonal dissemination should be on high alert through cryptic pathways from environmental reservoirs or community sources. Our findings support the importance of ongoing genomic surveillance and strengthening infection control, including enhanced cleaning and disinfection protocols, strict closed-loop management of patient transfers, and targeted screening of asymptomatic carriers or patients readmitted soon after discharge.

These post-pandemic dataset from Southwest China effectively addresses a gap in the existing literature, which has primarily focused on pre-2020 isolates or geographically restricted cohorts. Genomic findings integrated with detailed clinical data enable a more comprehensive interpretation of strain diversity, resistance profiles, and transmission risks in our hospital setting. However, several limitations should be acknowledged. First, this study only included CDI cases by the strict definition of combined toxin A/B assay and positive isolation of C. difficile strains, which offers high specificity but limited sensitivity, thereby underestimating the actual prevalence of C. difficile in this region. Second, the mechanisms underlying metronidazole resistance in ST81 in this study remain unclear and warrant further investigation. Third, the retrospective, single-center data could not fully capture the relationship between clinical outcomes and bacteriological features, necessitating validation in large, multicenter prospective cohorts. We call for establishing coordinated national-level multicenter surveillance networks to provide a more accurate assessment of the true burden of CDI in China and to harmonize surveillance strategies for the whole country.

Our findings indicate that in the post-COVID-19 era in Southwest China, although no global hypervirulent strain is prevalent, the clinical burden of CDI cannot be overlooked, due to its disease severity. All C. difficile isolates are still susceptible to fidaxomicin and vancomycin. ST81 has significantly increased in prevalence as the second-most dominant strain type after ST3, generally exhibiting high-level fluoroquinolone resistance and partial metronidazole resistance, and warrants more attention. Core genome analysis could identify clone groups, prompting clinicians to assess whether cryptic clonal dissemination is present. Our sustained clinical case monitoring, genomic surveillance of isolates, and AMR pattern surveys integrate data on clinical cases and their pathogenic strains, providing further evidence to enhance infection control measures to prevent future clonal dissemination and supporting clinical management and public health decisions.