Bacterial etiology, antimicrobial susceptibility, and clinical outcomes in cervicofacial space infections before, during, and after the COVID-19 pandemic: a 10-year single-centre study Authors: Shuang He, Zheng Jiang, Yibei Wang, Mailudan Ainiwaer, Xiaoqin Ji, Huiling Zhao Published: 2026 ABSTRACT AbstractObjectiveTo investigate epidemiological trends of cervicofacial space infections in Western China, including pathogen distribution, antimicrobial resistance, and the impact of the COVID-19 pandemic on pathogen susceptibility and patient outcomes.MethodsClinical data from 542 hospitalized patients with severe cervicofacial space infections who underwent open surgery at a tertiary hospital over the past decade were analyzed. Data on bacterial strains, antibiotic susceptibility, drug resistance, and treatment outcomes were recorded. Statistical analyses included one-way ANOVA, chi-square tests, Fisher’s exact tests, logistic regression, and multiple regression.ResultsFrom January 2014 to June 2024, the annual number of hospitalized CFI cases declined initially but later increased. Bacterial growth was observed in 400 cases (73.8%). Based on the frequency of isolation in positive pus cultures, Streptococcus spp. and Klebsiella spp. were the most frequently identified organisms. The five antimicrobial agents with the highest in vitro susceptibility rates were levofloxacin (66.3%), vancomycin (63.8%), cefepime (61.0%), ceftriaxone (59.5%), and cefotaxime (58.5%). Compared to the pre-pandemic period, Patients treated during the pandemic and post-pandemic periods had higher odds of in-hospital complications than those treated during the pre-pandemic period (OR = 7.28, 95% CI:3.67 ~ 14.45; P < 0.001; OR = 4.96, 95% CI:2.16 ~ 11.41; P < 0.001), including pulmonary infections, and coagulopathy etc.ConclusionsBased on the prevalent pathogen profile and local susceptibility patterns observed in this study, cephalosporins are the preferred empirical choice for cervicofacial space infections in our region. Fluoroquinolones may be considered as an alternative in patients with beta-lactam allergy, but should be used judiciously to preserve their activity for resistant organisms. Antibiotic therapy should be tailored based on culture and susceptibility results as soon as they become available. Additionally, patients treated during and after the COVID-19 pandemic had higher odds of in-hospital complications, which may reflect healthcare system disruptions and delayed presentation.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12879-026-14186-6.Keywords: Cervicofacial space infection, Pathogens, Antimicrobial susceptibility, COVID-19 Introduction Cervicofacial space infections (CFIs) are infections that occur within the potential spaces and fascial planes of the head and neck, potentially involving the orbit, zygomatic space, lateral pharyngeal space, or the submandibular and sublingual spaces (oral floor phlegmon) [1, 2]. The clinical presentation of CFIs varies according to the anatomical spaces involved [3]. Streptococcus spp., Staphylococcus spp., and Klebsiella spp. are among the most frequently identified causative pathogens [4, 5]. Common sources of infection include the pharyngeal tonsils, teeth, salivary glands, regional necrotic lymph nodes, and traumatic injuries involving the esophagus or oropharynx [5]. CFIs may progress rapidly and cause life-threatening complications, such as airway obstruction, jugular vein thrombosis, mediastinitis, and sepsis. Their management generally requires airway protection, antimicrobial therapy, and, when indicated, surgical drainage [5]. Because microbiological results are often unavailable when treatment is initiated, appropriate empirical antibiotic selection is particularly important. The increasing use of antimicrobial agents has exacerbated the emergence and spread of antimicrobial-resistant pathogens [6]. In patients with CFIs, this resistance problem is further complicated by regional differences in pathogen distribution and antimicrobial susceptibility patterns [4], making empirical antibiotic selection particularly challenging. Therefore, local epidemiological and susceptibility data are essential for guiding empirical antibiotic selection and optimising CFI management. Several factors, including pathogen virulence, bacterial load, antibiotic selection [7, 8], and patient-related factors such as immune function [9] and healthcare-seeking behaviour, influence the prognosis of CFIs. The COVID-19 pandemic profoundly disrupted healthcare systems worldwide, leading to delayed patient presentation and reduced access to routine dental and medical services [10, 11]. These disruptions may have influenced the epidemiology of bacterial infections, including CFIs. A study from Western China found that the types of pathogens involved in acute respiratory infections changed significantly following the COVID-19 pandemic [12]. However, whether similar changes occurred in CFIs remains unclear. In particular, long-term data comparing the pathogen distribution, antimicrobial susceptibility patterns, and clinical outcomes of patients with CFIs across the pre-pandemic, pandemic, and post-pandemic periods remain limited. Therefore, this study aimed to characterize the epidemiological and microbiological trends of CFIs in a large tertiary hospital in Western China and to evaluate differences in clinical outcomes across three periods associated with the COVID-19 pandemic. Specifically, we aimed to: (i) describe the pathogen distribution, antimicrobial susceptibility patterns, antimicrobial resistance profiles, and changes in bacterial composition among patients with CFIs from January 2014 to June 2024, thereby providing local evidence to inform empirical antibiotic selection; and (ii) compare in-hospital complications, length of hospital stay, and transfer to the intensive care unit (ICU) across the pre-pandemic, pandemic, and post-pandemic periods. Materials and methods Subjects This retrospective study analyzed medical records of patients with CFIs who underwent surgical treatment at the Department of Otolaryngology-Head and Neck Surgery, West China Hospital of Sichuan University, from January 2014 to June 2024. The study included both paediatric and adult patients, ranging in age from 3 to 96 years. Ethical approval was obtained from the Institutional Review Board (202512). Inclusion criteria were: (i) confirmed CFIs involving one or more anatomical spaces; (ii)open surgical intervention (incision and drainage). Exclusion criteria were: (i) history of head and neck trauma or malignancy near the primary site; (ii) prior chemoradiotherapy; (iii) absence of abscess formation or failure to submit pus for pathogen testing; (iv) incomplete medical records (e.g., missing key clinical variables or outcome data). Of 643 eligible patients, 69 were excluded because pus was not sent for pathogen examination and 32 for incomplete records, resulting in a final cohort of 542 patients (Fig. 1). Patient selection flow diagram Definition of study periods and COVID-19 status We defined the study periods and COVID-19 infection status based on major shifts in China’s national pandemic control strategy: pre-pandemic period (January 2014–December 2019). Pandemic period (January 2020–December 2022): This period commenced when COVID-19 was officially classified as a Category B infectious disease managed with Category A measures, marking the onset of the public health emergency in China. During this phase, China implemented stringent disease control measures. Post-pandemic period (January 2023–June 2024): This phase commenced with the official downgrade of COVID-19 to a Category B infectious disease in China and the subsequent discontinuation of its containment measures, which facilitated the widespread transmission of the SARS-CoV-2 Omicron variant [13]. A patient was classified as"COVID-19-positive” only upon laboratory confirmation, specifically a documented positive RT-PCR test for SARS-CoV-2 nucleic acid in respiratory specimens. It is important to note that our analysis examines calendar period effects (pre-pandemic, during-pandemic, post-pandemic) rather than individual COVID-19 infection status, as only four patients in our cohort had laboratory-confirmed COVID-19. Variable selection Data were collected from the hospital’s electronic medical record system and included: (i) patient demographics at admission (age, gender, discharge diagnosis, operative approach); and (ii) disease characteristics, including comorbidities (hypertension, diabetes, immunological disorders), pathogen distribution, culture results, and antibiotic sensitivity findings. The treating clinical team determined ICU transfer based on the following criteria: (i) airway compromise requiring mechanical ventilation; (ii) haemodynamic instability (systolic BP < 90 mmHg despite adequate fluid resuscitation); (iii) sepsis or septic shock; (iv) need for vasopressor support. Data quality control To ensure data completeness and accuracy, all information was extracted from the hospital’s structured information system and manually verified. A final diagnosis of CFIs required both typical clinical features and either imaging evidence or purulent secretion found during surgery. Microbiological and antimicrobial susceptibility testing All microbiological analyses were performed on intraoperatively collected pus specimens under strict aseptic conditions, following hospital SOPs and the CLSI M40-A2 guideline [14]. Cultures were incubated under both aerobic (35–37℃, 5% CO₂, up to 72 h) and anaerobic (35℃, anaerobic atmosphere, 48 h to 5 days) conditions using appropriate agar plates. Bacterial identification was carried out using conventional biochemical methods and, after 2016, supplemented or replaced by MALDI-TOF mass spectrometry(Bruker Daltonics MALDI Biotyper LT/SH system, Bruker Corporation, Billerica, MA, USA). Antimicrobial susceptibility testing was performed in accordance with CLSI M100 guidelines using the VITEK 2 XL system [15], E-test strips, or disk diffusion as appropriate. All procedures employed standard quality control strains, and culture media were obtained from specified suppliers (e.g., Zhengzhou Autobio). A bacterial isolate was classified as an MDRO if it demonstrated non-susceptibility to at least one agent in three or more distinct antimicrobial categories, as determined by antimicrobial susceptibility testing results. Data analysis Continuous variables that were normally distributed were presented as means with standard deviations and were compared using one-way analysis of variance (ANOVA). Categorical variables were presented as frequencies or percentages and were compared using chi-square tests or Fisher’s exact tests. To control for confounding, we first performed univariate analysis to assess the association between patient characteristics (age, gender, MDRO status, time period, etc.) and complication status, length of hospital stay, and ICU transfer status. Variables with p < 0.05 in the univariate analysis were subsequently entered into logistic regression to identify independent predictors of complication occurrence, with results reported as odds ratios (OR) and 95% confidence intervals. Data were analysed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). A p < 0.05 was considered statistically significant for all analyses. Multiple comparisons among the three time periods were adjusted using the Bonferroni correction. Results Patient demographics and clinical characteristics A total of 542 CFI patients were enrolled in this study, including 359 males (66.2%) and 183 females (33.8%), with a mean age of 53.6 ± 20.2 years(range: 3–96 years). Patients aged ≤ 18 years accounted for 45 cases (8.3%), and patients aged ≥ 65 years accounted for 171 cases (31.5%). The average length of hospital stay was 13.0 ± 8.0 days, and 207 patients (38.2%) had a hospital stay longer than 2 weeks. Regarding comorbidities, 118 (21.8%) had diabetes and 54 (10.0%) had hypertension. Additionally, 12 patients had immunodeficiency diseases, including 1 patient (0.3%) before the COVID-19 pandemic, 6 patients (3.7%) during the pandemic, and 5 patients (5.4%) after the pandemic. Diagnoses included rheumatoid arthritis, immune thrombocytopenic purpura, herpes zoster, decompensated autoimmune cirrhosis, and immune dysfunction. In-hospital complications were defined as any adverse events occurring during the hospital stay, including pulmonary infections, gastrointestinal bleeding, sepsis, coagulopathy, and thrombocytopenia. During hospitalization, 65 patients (12.0%) developed in-hospital complications. Stratified by period, complications occurred in 13 patients (4.5%) before the pandemic, 39 (24.2%) during the pandemic, and 13 (14.1%) after the pandemic. This comparison showed a statistically significant difference among the three groups (χ² = 39.47, P < 0.001). The proportion of patients transferred to the ICU was 59 (20.4%) before the pandemic, 77 (47.8%) during the pandemic, and 28 (30.4%) after the pandemic, showing a statistically significant difference among the three groups (χ² = 36.82, P < 0.001). (Table 1). Table 1Clinical characteristics, ICU transfer, and in-hospital complications among patients with CFIs across the three study periods (n = 542)CharacteristicsBefore the COVID-19 Pandemic(n = 289)Duringthe COVID-19 Pandemic(n = 161)After the COVID-19 Pandemic(n = 92)Test statistic () P Age(Mean ± SD)54.4 ± 20.854.4 ± 19.749.36 ± 18.902.330.099Multidrug-resistant Bacterial infection (n%)54.4 ± 20.8 (95% CI: 52.0-56.8)54.4 ± 19.7 (95% CI: 51.3–57.5)49.36 ± 18.90 (95% CI: 45.44–53.28)5.350.069Length of Hospital Stay(Mean ± SD)40 (13.8%, 95% CI: 10.3–18.3%)11 (6.8%, 95% CI:3.8–11.9%)9 (9.8%, 95% CI: 5.2–17.2%)1.010.365Transferred to the ICU (n%)12.9 ± 8.3 (95% CI: 11.94–13.86)13.3 ± 7.8 (95% CI: 12.08–14.52)11.9 ± 5.5 (95% CI: 10.76–13.04)36.82< 0.001Complications(n%)59 (20.4%, 95% CI: 16.2–25.2%)77 (47.8%, 95% CI: 40.2–55.5%)28 (30.4%, 95% CI: 21.9–40.4%)39.47< 0.001Pulmonary infections (n%)11 (3.8%)22 (13.7%)7 (7.6%)//Gastrointestinal bleeding (n%)2 (0.7%)4 (2.5%)1 (1.1%)//Sepsis (n%)02 (1.24%)0//Coagulopathy(n%)05(3.1%)4 (4.4%)//Thrombocytopenia (n%)06 (3.7%)1 (1.1%)//Note: Based on the Chi-square test, multiple comparisons were conducted using the Bonferroni method. Each subscript letter indicates a subset of the “group” categories whose column proportions do not differ significantly from each other at the 0.05 level. Groups that share the same letter (a or b) are not statistically different, while those labeled with different letters are statistically different. Owing to the limited number of complication events, statistical comparisons between groups were not conducted Clinical characteristics, ICU transfer, and in-hospital complications among patients with CFIs across the three study periods (n = 542) Complications (n%) Note: Based on the Chi-square test, multiple comparisons were conducted using the Bonferroni method. Each subscript letter indicates a subset of the “group” categories whose column proportions do not differ significantly from each other at the 0.05 level. Groups that share the same letter (a or b) are not statistically different, while those labeled with different letters are statistically different. Owing to the limited number of complication events, statistical comparisons between groups were not conducted Microbiological findings Among the 542 patients with CFIs, bacterial growth was observed in 400 cases (73.8%), while 142 (26.2%) showed no bacterial growth. Among the 400 positive cultures, 34 (8.5%) yielded only normal skin, mucosal, or oral flora, which were considered non-pathogenic contaminants and not reported in the pathogen distribution. The most frequently identified pathogenic organisms were: Streptococcus spp. 272 (68.0%), Klebsiella spp. 53 (13.3%), Staphylococcus spp. 15 (3.8%), Actinomyces spp.9 (2.3%), unidentified Gram-negative bacilli 15 (3.8%), and unidentified Gram-positive cocci 2 (0.5%). Among the Streptococcus species, Streptococcus constellatus was the most prevalent, accounting for 84 (21.0%). Among the Klebsiella genus, Klebsiella pneumoniae had the highest proportion, at 37 (9.2%). The period-specific distribution (Table 2) showed that streptococcus spp. remained the predominant pathogen across all periods, with an increasing trend, while Klebsiella spp. decreased during the pandemic. Table 2Distribution of organisms identified in positive pus cultures across the three study periods (n = 400)PathogenBefore the COVID-19 Pandemic(n = 209)Duringthe COVID-19 Pandemic(n = 118)After the COVID-19 Pandemic(n = 73)Streptococcus spp.128 (61.3%)90 (76.3%)54 (74.0%)Klebsiella spp.36 (17.2%)9 (7.6%)8 (11.0%)Staphylococcus spp.7 (3.3%)6 (5.1%)2 (2.7%)Actinomyces spp.07 (6.8%)2 (2.7%)Unidentified bacteria8 (3.8%)4 (3.4%)5 (6.9%)Normal flora30 (14.4%)2 (1.7%)2 (2.7%) Distribution of organisms identified in positive pus cultures across the three study periods (n = 400) Obligate anaerobic organisms were isolated in 9 patients (2.3% of positive cultures), including Actinomyces spp. 7 (1.8%) and Prevotella spp.2 (0.5%). Polymicrobial infections involving both aerobic and anaerobic organisms were observed in 6 patients (1.5%). Combinations included Streptococcus spp. with Actinomyces spp. (strict anaerobe), Streptococcus spp. with Staphylococcus aureus, and Klebsiella pneumoniae with Streptococcus spp. Strict anaerobes (Actinomyces spp.) were identified in mixed cultures in 2 cases(0.5%). From January 2014 to June 2024, 542 patients with CFIs were hospitalized, including 60 patients (11.1%) with MDRO infections. The annual number of patients with MDRO infections ranged from 3 in 2020 and 2022 to 13 in 2015. In contrast, the annual proportion of patients with MDRO infections ranged from 5.5% in 2020 to 21.0% in 2015 (Fig. 2). Across the three study periods, MDRO infections occurred in 40 patients (13.8%) during the pre-pandemic period, 11 patients (6.8%) during the pandemic period, and 9 patients (9.8%) during the post-pandemic period. Annual proportion of MDRO infections among patients with CFIs, 2014–2024 The distribution of positive pus-culture findings was broadly similar across the three study periods. Streptococcus spp. predominated throughout, followed by Klebsiella spp. Staphylococcus spp. and normal flora were less frequently identified, although their relative rankings varied slightly between periods (Fig. 3A). Microbiological findings and antimicrobial susceptibility among patients with CFIs. (A) Distribution of the four most frequently identified organism groups in pus cultures across the pre-pandemic, during-pandemic, and post-pandemic periods. (B) Overall antimicrobial susceptibility of the tested bacterial isolates. (C) Changes in the susceptibility of Klebsiella pneumoniae to quinolones and cephalosporins. (D) Changes in the antimicrobial susceptibility of Streptococcus spp. to quinolones and cephalosporins For the analysis of MDRO phenotypes, one clinically relevant MDRO isolate was included per patient. Among the 60 MDRO isolates, Klebsiella pneumoniae was the predominant organism, accounting for 43 isolates (71.7%). Among these 43 Klebsiella pneumoniae isolates, 42 (97.7%) were extended-spectrum β-lactamase (ESBL)-producing isolates and 1 (2.3%) was carbapenem-resistant Klebsiella pneumoniae (CRKP). Overall, ESBL-producing Klebsiella pneumoniae and CRKP accounted for 42 (70.0%) and 1 (1.7%) of all MDRO isolates, respectively. The remaining MDRO isolates included 9 (15.0%) methicillin-resistant Staphylococcus aureus (MRSA), 5 multidrug-resistant Pseudomonas aeruginosa (8.3%), and 3 multidrug-resistant Acinetobacter baumannii (5.0%). Antibiotic susceptibility testing Antimicrobial susceptibility testing was performed on isolates recovered from pus specimens obtained from 400 patients. Because the number of isolates tested varied among antimicrobial agents, susceptibility rates were calculated using the number of isolates tested for each agent as the denominator. Seven agents showed overall susceptibility rates of at least 50%. The five agents with the highest susceptibility rates were levofloxacin (66.3%), vancomycin (63.8%), cefepime (61.0%), ceftriaxone (59.5%), and cefotaxime (58.5%) (Fig. 3B; Table 3). Susceptibility rates for all tested agents are provided in Supplementary Table S1. Table 3Percentage susceptibility of antimicrobial agentsAntibiotic ClassAntibiotic Agent N PercentageCephalosporinsCeftriaxone23859.5%CephalosporinsCefotaxime23458.5%CephalosporinsCefepime24461.0%QuinolonesLevofloxacin26566.3%GlycopeptidesVancomycin25563.8%AmphenicolsChloramphenicol20350.8%OxazolidinonesLinezolid20250.5%Note: The percentage susceptibility for each antimicrobial agent was calculated as the number of susceptible isolates divided by the total number of isolates tested Percentage susceptibility of antimicrobial agents Note: The percentage susceptibility for each antimicrobial agent was calculated as the number of susceptible isolates divided by the total number of isolates tested We further evaluated the annual susceptibility patterns of Streptococcus spp. and Klebsiella spp. to cephalosporins and quinolones. For each pathogen–antibiotic combination, the annual susceptibility rate was calculated as the proportion of susceptible isolates among all corresponding isolates tested in that year. Susceptibility of Streptococcus spp. to cephalosporins initially decreased and subsequently increased, whereas susceptibility to quinolones fluctuated but showed an overall declining pattern. In contrast, susceptibility of Klebsiella spp. to cephalosporins fluctuated considerably but generally increased over time, while susceptibility to quinolones remained relatively stable (Fig. 3C, D). Outcomes by pandemic period To control for potential confounding, univariable analyses were first performed to evaluate the associations of patient characteristics, including age, sex, MDRO infection, time period, and diabetes mellitus, with in-hospital complications, ICU transfer, and length of hospital stay. Variables with P < 0.05 in the corresponding univariable analyses were subsequently included in the respective multivariable models. In the multivariable logistic regression analysis of in-hospital complications, the during-pandemic and post-pandemic periods were independently associated with higher odds of complications compared with the pre-pandemic period, with adjusted ORs of 7.28 (95% CI, 3.67–14.45; P < 0.001) and 4.96 (95% CI, 2.16–11.41; P < 0.001), respectively. In a separate multivariable model, the during-pandemic period was independently associated with ICU transfer (OR, 2.38; 95% CI, 1.53–3.70; P < 0.001), whereas no significant associations were observed for the post-pandemic period. No significant association was observed between the study period and length of hospital stay. Detailed results of the complication model are presented in Table 4, while the ICU-transfer and length-of-stay analyses are provided in Supplementary Tables S2 and S3. Table 4Univariable and multivariable logistic regression analyses of in-hospital complications among patients with CFIs. (n = 542)VariablesUnivariate analysisLogistic Regression AnalysisβS.EZ P OR (95%CI)βS.EZ P OR (95%CI) Age 0.030.013.85 < 0.001 1.03 (1.02 ~ 1.05)0.040.013.97 < 0.001 1.04 (1.02 ~ 1.06) Sex Male1.00 (Reference) Female0.000.280.010.9881.00 (0.58 ~ 1.74) Whether MDRO No1.00 (Reference) Yes-0.040.43-0.080.9340.97 (0.42 ~ 2.22) Time Period Pre-COVID-191.00 (Reference)1.00 (Reference)During COVID-191.880.345.55 < 0.001 6.56 (3.37–12.75)1.990.355.68 < 0.001 7.28 (3.67–14.45)Post-COVID-191.340.413.30 < 0.001 3.81 (1.72–8.44)1.600.423.77 < 0.001 4.96 (2.16–11.41) Diabetes mellitus No1.00 (Reference)1.00 (Reference) Yes0.630.292.17 0.030 1.87 (1.06–3.29)0.150.310.490.6271.16 (0.63–2.15)OR: Odds Ratio, CI: Confidence Interval Univariable and multivariable logistic regression analyses of in-hospital complications among patients with CFIs. (n = 542) OR: Odds Ratio, CI: Confidence Interval Discussion We retrospectively analyzed clinical data from 542 CFI patients who underwent surgery. The mean age was 53.6 years (range: 3–96), with 66.2% male, consistent with previous studies in Western China [1, 16]. Diabetes mellitus was the most common comorbidity, aligning with prior findings [17]. Diabetes impairs host defenses, increasing the risk of CFIs [18, 19]. The proportion of diabetic patients (22.1%) was lower than in other studies (34.6%), possibly due to inclusion of pediatric cases. The bacterial culture positivity rate (73.8%) was similar to previously reported rates (55.2%–74.5%) [4, 20] for deep neck infections. This relatively high rate may reflect prompt specimen collection during emergency procedures and a notable proportion of MDROs among pathogens. In pus cultures from 542 patients, Streptococcus constellatus and Klebsiella pneumoniae were the most frequently isolated organisms, differing from earlier studies in the region [20, 21]. The longer study period may explain the emergence of Klebsiella pneumoniae as the second most frequent isolate (9.6%). Klebsiella pneumoniae was the predominant Klebsiella species identified. National surveillance data show its proportion increased from 9.91% in 2014 to 18.9% in 2021, highlighting its growing clinical significance. The adoption of MALDI-TOF MS in 2016 may have improved species-level identification and may partly account for the observed temporal variation in the detection of K. pneumoniae. This factor should be considered when interpreting microbiological trends. The lower proportion of patients with MDRO infections during the pandemic period should be interpreted cautiously. Systematic reviews show the impact of COVID-19 on antimicrobial resistance varies by region; for example, MDRO prevalence increased in Asia but decreased in North America [22]. The reduction in MDRO rates at this center may be due to enhanced infection control measures and changes in patient admissions. These factors likely contributed to the decline, independent of antibiotic susceptibility patterns. Antibiotic exposure and prescribing patterns were not systematically recorded; therefore, the potential contribution of changes in antimicrobial use to the observed variation in MDRO infection rates could not be determined. In vitro susceptibility to selected cephalosporins and fluoroquinolones was relatively high among some predominant isolates, although susceptibility varied by organism and antimicrobial agent. Susceptibility rates among Streptococcus spp. appeared to increase after 2016. Klebsiella species had variable sensitivity to cephalosporins and a temporary decline in quinolone sensitivity from 2019 to 2022, followed by improvement. These trends align with other regional studies [23]. Given the high susceptibility of Streptococcus spp. and Klebsiella spp. to cephalosporins, these remain the preferred empirical therapy. Levofloxacin showed good in vitro activity, but fluoroquinolones should be reserved for patients with beta-lactam allergies or suspected ESBL-producing organisms, in line with antimicrobial stewardship. Use quinolones cautiously in elderly patients and those on antiarrhythmic medications [24]. To control the spread of COVID-19, the Chinese government implemented strict measures, including travel restrictions, stay-at-home orders, and a reallocation of healthcare resources. While these interventions effectively curbed virus transmission, they also had unintended consequences for patients with chronic health issues. Many patients were hesitant to seek timely medical care due to fear of infection, which resulted in delayed hospital visits and more advanced disease at the time of admission [10, 11]. In our cohort, the adjusted odds of in-hospital complications were higher during both the pandemic and post-pandemic periods than during the pre-pandemic period. Notably, only four patients tested positive for COVID-19, so the higher complication rates may have been related to changes in healthcare access, referral patterns, or timing of presentation. However, because these factors were not directly measured, causal attribution to healthcare-system disruption or SARS-CoV-2 infection is not possible. Limitations This study has several limitations. First, because it is a retrospective, single-center study conducted at a tertiary referral hospital in Southwest China, our findings may be affected by regional bias. They may not be generalizable to other locations. Second, the exclusion of mild cases treated in the emergency department may have introduced selection bias. Third, the small number of patients with confirmed COVID-19 (n = 4) limits our ability to make direct comparisons based on infection status. Therefore, our analysis should be seen as examining calendar period effects rather than individual infection status. Fourth, the three study periods had unequal durations, and we did not have a population denominator for accurate incidence estimation. Fifth, the wide age range of participants (3–96 years), which includes both pediatric and elderly patients, may influence the distribution of pathogens and complication rates; thus, comparisons with adult-only studies should be approached with caution. Sixth, several clinically significant variables (including inflammatory markers, time from symptom onset, surgical extent, and airway compromise) were not included in the regression models due to incomplete documentation, which may have resulted in residual confounding. Additionally, data on infectious diseases consultations were not systematically recorded, limiting our ability to assess their impact on outcomes. Furthermore, data on pre-admission antibiotic use, postoperative antibiotic adjustments, changes after culture results, and antibiotic duration were not systematically recorded in this retrospective dataset, limiting our ability to assess their impact on outcomes. Finally, changes in laboratory methods over time (such as the introduction of MALDI-TOF in 2016) may have improved species identification, potentially causing observed shifts in pathogen distribution; this is an inherent limitation of the retrospective design. Conclusion In this single-centre cohort of surgically managed patients with CFIs, Streptococcus constellatus and Klebsiella pneumoniae were the most frequently isolated organisms among culture-positive specimens. Higher odds of in-hospital complications were observed during and after the COVID-19 pandemic; these findings should be interpreted as calendar-period associations rather than evidence of a direct effect of SARS-CoV-2 infection. Empirical antimicrobial therapy should be guided by local susceptibility data, infection severity, patient-specific factors, and the possibility of polymicrobial and anaerobic infection, with subsequent adjustment according to culture and susceptibility results. Prospective multicentre studies are needed to determine whether these temporal differences persist after accounting for changes in patient characteristics, healthcare access, treatment practices, and laboratory methods. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Author contributions Shuang He, Xiaoqin Ji, Yibei Wang, Mailudan Ainiwaer, Huiling Zhao and Zheng Jiang performed material preparation, data collection and analysis. Huiling Zhao and Shuang He wrote the manuscript. All authors reviewed the manuscript. Funding None. Data availability The data are available from the corresponding author upon reasonable request via email at 15710087@163.com. Declarations Ethics approval and consent to participate The study was approved by the Institutional Review Board (IRB) of West China Hospital, Sichuan University, under the reference number 202512. This study was conducted in compliance with the ethical principles that have their origin in the Declaration of Helsinki, as well as with all currently applicable laws and regulations of the country where the study was conducted. A written informed consent document was obtained from all subjects during the screening period. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data are available from the corresponding author upon reasonable request via email at 15710087@163.com.