Search Results (47)

Background: Infection prevention and control (IPC) in hospitals relies heavily on infection control nurses (ICNs) who manage complex consultations to prevent and control infections. This study evaluated large language models (LLMs) as artificial intelligence (AI) tools to support ICNs in IPC decision-making processes. Our goal is to enhance the efficiency of IPC practices while maintaining the highest standards of safety and accuracy. Methods: A cross-sectional benchmarking study at Queen Mary Hospital, Hong Kong assessed three LLMs—GPT-4.1, DeepSeek V3, and Gemini 2.5 Pro Exp—using 30 clinical infection control scenarios. Each model generated clarifying questions to understand the scenarios before providing IPC recommendations through two prompting methods: an open-ended inquiry and a structured template. Sixteen experts, including senior and junior ICNs and physicians, rated these responses on coherence, conciseness, usefulness and relevance, evidence quality, and actionability (1–10 scale). Quantitative and qualitative analyses assessed AI performance, reliability, and clinical applicability. Results: GPT-4.1 and DeepSeek V3 scored significantly higher on the composite quality scale, with adjusted means (95% CI) of 36.77 (33.98–39.57) and 36.25 (33.45–39.04), respectively, compared with Gemini 2.5 Pro Exp at 33.19 (30.39–35.99) (p < 0.001). GPT-4.1 led in evidence quality, usefulness, and relevance. Gemini 2.5 Pro Exp failed to generate responses in 50% of scenarios under structured prompt conditions. Structured prompting yielded significant improvements, primarily by enhancing evidence quality (p < 0.001). Evaluator background influenced scoring, with doctors rating outputs higher than nurses (38.83 vs. 32.06, p < 0.001). However, a qualitative review revealed critical deficiencies across all models, for example, tuberculosis treatment solely based on a positive acid-fast bacilli (AFB) smear without considering nontuberculous mycobacteria in DeepSeek V3 and providing an impractical and noncommittal response regarding the de-escalation of precautions for Candida auris in Gemini 2.5 Pro Exp. These errors highlight potential safety risks and limited real-world applicability, despite generally positive scores. Conclusions: While GPT-4.1 and DeepSeek V3 deliver useful IPC advice, they are not yet reliable for autonomous use. Critical errors in clinical judgment and practical applicability highlight that LLMs cannot replace the expertise of ICNs. These technologies should serve as adjunct tools to support, rather than automate, clinical decision-making. Full article

Background/Objectives: We have previously reported the engineering of a point-of-care (POC) system that fully automates the procedures for nucleic acid extraction and multiplexed real-time RT-PCR, with a major advantage of high-level multiplexing. In this study, we applied and validated the system in a respiratory tract infection setting. Methods: An automatic nested real-time RT-PCR assay was developed (POCm). It was a 40-plex assay that simultaneously detected 39 epidemiologically important respiratory pathogens in 1.5 h in the POC system. The analytical and clinical performance was evaluated. Results: The analytical sensitivities of the POCm assay were comparable to those of its single-plex counterparts performed manually on a bench-top. The minimum detectable concentrations ranged from 53 copies/mL to 5.3 × 10³ copies/mL for all pathogen targets except hCoV-NL63 (5.3 × 10⁴ copies/mL). The quantitative performance was demonstrated by the linear correlations between Ct values and input concentrations for all pathogen targets, with 24 of them demonstrating coefficients of correlation (r) greater than 0.9. The POCm assay was subsequently evaluated in 283 clinical samples. A high level of agreement (98.2–100%) was achieved for pathogen detection results between POCm and standard diagnostic methods. The POCm result was also fully concordant with the result of another commercial POC multiplex platform. For positive clinical samples, pairwise Ct values measured by POCm closely correlated with those of the bench-top reference method (r = 0.70). The feasibility of mutation genotyping of the viral subtype was further demonstrated. Conclusions: This study demonstrated the practicality of POCm for routine testing in clinical laboratories. Further clinical trials are being conducted to evaluate the clinical performance of the system. Full article

Background/Objectives: The potential of generative artificial intelligence (GenAI) to augment clinical consultation services in clinical microbiology and infectious diseases (ID) is being evaluated. Methods: This cross-sectional study evaluated the performance of four GenAI chatbots (GPT-4.0, a Custom Chatbot based on GPT-4.0, Gemini Pro, and Claude 2) by analysing 40 unique clinical scenarios. Six specialists and resident trainees from clinical microbiology or ID units conducted randomised, blinded evaluations across factual consistency, comprehensiveness, coherence, and medical harmfulness. Results: Analysis showed that GPT-4.0 achieved significantly higher composite scores compared to Gemini Pro (p = 0.001) and Claude 2 (p = 0.006). GPT-4.0 outperformed Gemini Pro and Claude 2 in factual consistency (Gemini Pro, p = 0.02; Claude 2, p = 0.02), comprehensiveness (Gemini Pro, p = 0.04; Claude 2, p = 0.03), and the absence of medical harm (Gemini Pro, p = 0.02; Claude 2, p = 0.04). Within-group comparisons showed that specialists consistently awarded higher ratings than resident trainees across all assessed domains (p < 0.001) and overall composite scores (p < 0.001). Specialists were five times more likely to consider responses as “harmless”. Overall, fewer than two-fifths of AI-generated responses were deemed “harmless”. Post hoc analysis revealed that specialists may inadvertently disregard conflicting or inaccurate information in their assessments. Conclusions: Clinical experience and domain expertise of individual clinicians significantly shaped the interpretation of AI-generated responses. In our analysis, we have demonstrated disconcerting human vulnerabilities in safeguarding against potentially harmful outputs, which seemed to be most apparent among experienced specialists. At the current stage, none of the tested AI models should be considered safe for direct clinical deployment in the absence of human supervision. Full article

The emergence of SARS-CoV-2 mutations poses significant challenges to diagnostic tests, as these mutations can reduce the sensitivity of commonly used RT-PCR assays. Therefore, there is a need to design diagnostic assays with multiple targets to enhance sensitivity. In this study, we identified a novel diagnostic target, the nsp10 gene, using nanopore sequencing. Firstly, we determined the analytical sensitivity and specificity of our COVID-19-nsp10 assay. The COVID-19-nsp10 assay had a limit of detection of 74 copies/mL (95% confidence interval: 48–299 copies/mL) and did not show cross-reactivity with other respiratory viruses. Next, we determined the diagnostic performance of the COVID-19-nsp10 assay using 261 respiratory specimens, including 147 SARS-CoV-2-positive specimens belonging to the ancestral strain and Alpha, Beta, Gamma, Delta, Mu, Eta, Kappa, Theta and Omicron lineages. Using a LightMix E-gene RT-PCR assay as the reference method, the diagnostic sensitivity and specificity of the COVID-19-nsp10 assay were found to be 100%. The median Cp values for the LightMix E-gene RT-PCR and our COVID-19-nsp10 RT-PCR were 22.48 (range: 12.95–36.60) and 25.94 (range 16.37–36.87), respectively. The Cp values of the COVID-19-nsp10 RT-PCR assay correlated well with those of the LightMix E-gene RT-PCR assay (Spearman’s ρ = 0.968; p < 0.0001). In conclusion, nsp10 is a suitable target for a SARS-CoV-2 RT-PCR assay. Full article

The rebound characteristics of respiratory infections after lifting pandemic control measures were uncertain. From January to November 2023, patients presenting at a teaching hospital were tested for common respiratory viruses and Mycoplasma pneumoniae using a combination of antigen, nucleic acid amplification, and targeted next-generation sequencing (tNGS) tests. The number and rate of positive tests per month, clinical and microbiological characteristics were analyzed. A rapid rebound of SARS-CoV-2 was followed by a slower rebound of M. pneumoniae, with an interval of 5 months between their peaks. The hospitalization rate was higher, with infections caused by respiratory viruses compared to M. pneumoniae. Though the pediatric hospitalization rate of respiratory viruses (66.1%) was higher than that of M. pneumoniae (34.0%), the 4094 cases of M. pneumoniae within 6 months posed a huge burden on healthcare services. Multivariate analysis revealed that M. pneumoniae-infected adults had more fatigue, comorbidities, and higher serum C-reactive protein, whereas children had a higher incidence of other respiratory pathogens detected by tNGS or pathogen-specific PCR, fever, and were more likely to be female. A total of 85% of M. pneumoniae-positive specimens had mutations detected at the 23rRNA gene, with 99.7% showing A2063G mutation. Days to defervescence were longer in those not treated by effective antibiotics and those requiring a change in antibiotic treatment. A delayed but significant rebound of M. pneumoniae was observed after the complete relaxation of pandemic control measures. No unusual, unexplained, or unresponsive cases of respiratory infections which warrant further investigation were identified. Full article

Developing recombinant proteins as nasal vaccines for inducing systemic and mucosal immunity against respiratory viruses is promising. However, additional adjuvants are required to overcome the low immunogenicity of protein antigens. Here, a self-adjuvanted protein-RNA ribonucleoprotein vaccine was developed and found to be an effective nasal vaccine in mice and the SARS-CoV-2 infection model. The vaccine consisted of spike RBD (as an antigen), nucleoprotein (as an adaptor), and ssRNA (as an adjuvant and RNA scaffold). This combination robustly induced mucosal IgA, neutralizing antibodies and activated multifunctional T-cells, while also providing sterilizing immunity against live virus challenge. In addition, high-resolution scRNA-seq analysis highlighted airway-resident immune cells profile during prime-boost immunization. The vaccine also possesses modularity (antigen/adaptor/RNA scaffold) and can be made to target other viruses. This protein-RNA ribonucleoprotein vaccine is a novel and promising approach for developing safe and potent nasal vaccines to combat respiratory virus infections. Full article

Staphylococcus argenteus is a novel Staphylococcus species derived from Staphylococcus aureus. Information on the prevalence and genetic characteristics of invasive S. argenteus in Asia is limited. In this study, 275 invasive S. aureus complex strains were retrieved from blood culture specimens in Hong Kong and re-analyzed using MALDI-TOF mass spectrometry and an in-house multiplex real-time PCR for S. argenteus. The prevalence of invasive S. argenteus in Hong Kong was found to be 4.0% (11/275). These strains were primarily susceptible to commonly used antibiotics, except penicillin. Whole-genome sequencing revealed the circulation of three S. argenteus genotypes (ST-2250, ST-1223, and ST-2854) in Hong Kong, with ST-2250 and ST-1223 being the predominant genotypes. The local ST-2250 and ST-1223 strains showed close phylogenetic relationships with isolates from mainland China. Antimicrobial-resistant genes (fosB, tet-38, mepA, blaI, blaZ) could be found in nearly all local S. argenteus strains. The ST-1223 and ST-2250 genotypes carried multiple staphylococcal enterotoxin genes that could cause food poisoning and toxic shock syndrome. The CRISPR/Cas locus was observed only in the ST-2250 strains. This study provides the first report on the molecular epidemiology of invasive S. argenteus in Hong Kong, and further analysis is needed to understand its transmission reservoir. Full article

Background: Olfactory dysfunction (OD) is a common neurosensory manifestation in long COVID. An effective and safe treatment against COVID-19-related OD is needed. Methods: This pilot trial recruited long COVID patients with persistent OD. Participants were randomly assigned to receive short-course (14 days) oral vitamin A (VitA; 25,000 IU per day) and aerosolised diffuser olfactory training (OT) thrice daily (combination), OT alone (standard care), or observation (control) for 4 weeks. The primary outcome was differences in olfactory function by butanol threshold tests (BTT) between baseline and end-of-treatment. Secondary outcomes included smell identification tests (SIT), structural MRI brain, and serial seed-based functional connectivity (FC) analyses in the olfactory cortical network by resting-state functional MRI (rs–fMRI). Results: A total of 24 participants were randomly assigned to receive either combination treatment (n = 10), standard care (n = 9), or control (n = 5). Median OD duration was 157 days (IQR 127–175). Mean baseline BTT score was 2.3 (SD 1.1). At end-of-treatment, mean BTT scores were significantly higher for the combination group than control (p < 0.001, MD = 4.4, 95% CI 1.7 to 7.2) and standard care (p = 0.009) groups. Interval SIT scores increased significantly (p = 0.009) in the combination group. rs–fMRI showed significantly higher FC in the combination group when compared to other groups. At end-of-treatment, positive correlations were found in the increased FC at left inferior frontal gyrus and clinically significant improvements in measured BTT (r = 0.858, p < 0.001) and SIT (r = 0.548, p = 0.042) scores for the combination group. Conclusions: Short-course oral VitA and aerosolised diffuser OT was effective as a combination treatment for persistent OD in long COVID. Full article

An intranasal COVID-19 vaccine, DelNS1-based RBD vaccines composed of H1N1 subtype (DelNS1-nCoV-RBD LAIV) was developed to evaluate the safety and immunogenicity in healthy adults. We conducted a phase 1 randomized, double-blinded, placebo-controlled study on healthy participants, age 18–55 and COVID-19 vaccines naïve, between March and September 2021. Participants were enrolled and randomly assigned (2:2:1) into the low and high dose DelNS1-nCoV-RBD LAIV manufactured in chicken embryonated eggs or placebo groups. The low and high-dose vaccine were composed of 1 × 10⁷ EID₅₀/ dose and 1 × 10^7.7 EID₅₀/ dose in 0.2 mL respectively. The placebo vaccine was composed of inert excipients/dose in 0.2 mL. Recruited participants were administered the vaccine intranasally on day 0 and day 28. The primary end-point was the safety of the vaccine. The secondary endpoints included cellular, humoral, and mucosal immune responses post-vaccination at pre-specified time-points. The cellular response was measured by the T-cell ELISpot assay. The humoral response was measured by the serum anti-RBD IgG and live-virus neutralizing antibody against SARS-CoV-2. The saliva total Ig antibody responses in mucosal secretion against SARS-CoV-2 RBD was also assessed. Twenty-nine healthy Chinese participants were vaccinated (low-dose: 11; high-dose: 12 and placebo: 6). The median age was 26 years. Twenty participants (69%) were male. No participant was discontinued due to an adverse event or COVID-19 infection during the clinical trial. There was no significant difference in the incidence of adverse events (p = 0.620). For the T-cell response elicited after full vaccination, the positive PBMC in the high-dose group increased to 12.5 SFU/10⁶ PMBC (day 42) from 0 (baseline), while it increased to 5 SFU/10⁶ PBMC (day 42) from 2.5 SFU/10⁶ PBMC (baseline) in the placebo group. The high-dose group showed a slightly higher level of mucosal Ig than the control group after receiving two doses of the vaccine (day 31, 0.24 vs. 0.21, p = 0.046; day 56 0.31 vs. 0.15, p = 0.45). There was no difference in the T-cell and saliva Ig response between the low-dose and placebo groups. The serum anti-RBD IgG and live virus neutralizing antibody against SARS-CoV-2 were undetectable in all samples. The high-dose intranasal DelNS1-nCoV-RBD LAIV is safe with moderate mucosal immunogenicity. A phase-2 booster trial with a two-dose regimen of the high-dose intranasal DelNS1-nCoV-RBD LAIV is warranted. Full article

The emergence of new immune-evasive severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants and subvariants outpaces the development of vaccines specific against the dominant circulating strains. In terms of the only accepted immune correlate of protection, the inactivated whole-virion vaccine using wild-type SARS-CoV-2 spike induces a much lower serum neutralizing antibody titre against the Omicron subvariants. Since the inactivated vaccine given intramuscularly is one of the most commonly used coronavirus disease 2019 (COVID-19) vaccines in developing regions, we tested the hypothesis that intranasal boosting after intramuscular priming would provide a broader level of protection. Here, we showed that one or two intranasal boosts with the Fc-linked trimeric spike receptor-binding domain from wild-type SARS-CoV-2 can induce significantly higher serum neutralizing antibodies against wild-type SARS-CoV-2 and the Omicron subvariants, including BA.5.2 and XBB.1, with a lower titre in the bronchoalveolar lavage of vaccinated Balb/c mice than vaccination with four intramuscular doses of inactivated whole virion vaccine. The intranasally vaccinated K18-hACE2-transgenic mice also had a significantly lower nasal turbinate viral load, suggesting a better protection of the upper airway, which is the predilected site of infection by Omicron subvariants. This intramuscular priming and intranasal boosting approach that achieves broader cross-protection against Omicron variants and subvariants may lengthen the interval required for changing the vaccine immunogen from months to years. Full article

Hong Kong SAR has adopted universal masking, social distancing, testing of all symptomatic and high-risk groups for isolation of confirmed cases in healthcare facilities, and quarantine of contacts as epidemiological control measures without city lockdown or border closure. These measures successfully suppressed the community transmission of pre-Omicron SARS-CoV-2 variants or lineages during the first to the fourth wave. No nosocomial SARS-CoV-2 infection was documented among healthcare workers in the first 300 days. The strategy of COVID-19 containment was adopted to provide additional time to achieve population immunity by vaccination. The near-zero COVID-19 situation for about 8 months in 2021 did not enable adequate immunization of the eligible population. A combination of factors was identified, especially population complacency associated with the low local COVID-19 activity, together with vaccine hesitancy. The importation of the highly transmissible Omicron variant kickstarted the fifth wave of COVID-19, which could no longer be controlled by our initial measures. The explosive fifth wave, which was partially contributed by vertical airborne transmission in high-rise residential buildings, resulted in over one million cases of infection. In this review, we summarize the epidemiology of COVID-19 and the infection control and public health measures against the importation and dissemination of SARS-CoV-2 until day 1000. Full article

The epidemiology of patients with gastrointestinal colonization of carbapenem-resistant Acinetobacter baumannii (CRAB) has not been systematically analyzed. We aimed to analyze the incidence, risk factors, and clinical outcomes of patients with newly identified gastrointestinal colonization of CRAB in a healthcare region in Hong Kong, where a multi-pronged screening strategy for gastrointestinal colonization of CRAB, together with other multidrug-resistant organisms (MDROs), was conducted by collecting fecal specimens (rectal swab or stool) upon admission and during hospitalization. From 1 October 2015 to 31 December 2019, a total of 161,339 fecal specimens from 63,588 patients, 61,856 (97.3%) of whom were hospitalized patients, and 54,525 (88.1%) were screened upon admission, with 1309 positive for CRAB (2.4% prevalence). Among patients positive for CRAB in fecal specimens, 698 (53.3%) had newly detected gastrointestinal colonization of CRAB, giving an incidence of 10.03 per 10,000 patient admissions and constituting 2646 CRAB colonization days in the general wards. Excluding the 164 patients with co-colonization of other MDROs, 534 patients had gastrointestinal colonization with only CRAB, and 12.5% (67/534) developed symptomatic CRAB infections at a median of 61 days (range: 2 to 671 days), during prospective follow-up for 2 years. Compared with age- and sex-matched controls, patients being referred from residential care homes for the elderly, the presence of indwelling devices, use of beta-lactam/beta-lactamase inhibitors, carbapenems, and proton pump inhibitors in the preceding 6 months, and history of hospitalization in the past 6 months were significantly associated with gastrointestinal colonization with CRAB, as shown by multivariable analysis. Log-rank test showed that cases had significantly shorter survival duration than controls (p < 0.001). The adjusted hazard ratio of gastrointestinal colonization of CRAB was 1.8 (95% CI: 1.5–2.2; p < 0.001), as shown by Cox regression analysis. Whole-genome sequencing of eight patients with CRAB isolates in their blood cultures and rectal swabs during the same episode of hospitalization revealed ST-195 as the predominant type, as shown by multilocus sequencing type. Gastrointestinal colonization of CRAB poses a considerable challenge for infection prevention and control. Full article

Antimicrobial stewardship and infection control measures are equally important in the control of antimicrobial-resistant organisms. We conducted a retrospective analysis of the incidence rate of hospital-onset carbapenem-resistant Acinetobacter baumannii (CRAB) infection (per 1000 patient days) in the Queen Mary Hospital, a 1700-bed, university-affiliated teaching hospital, from period 1 (1 January 2007 to 31 December 2013) to period 2 (1 January 2014 to 31 December 2019), where enhanced infection control measures, including directly observed hand hygiene before meal and medication rounds to conscious patients, and the priority use of single room isolation, were implemented during period 2. This study aimed to investigate the association between enhanced infection control measures and changes in the trend in the incidence rate of hospital-onset CRAB infection. Antimicrobial consumption (defined daily dose per 1000 patient days) was monitored. Interrupted time series, in particular segmented Poisson regression, was used. The hospital-onset CRAB infection increased by 21.3% per year [relative risk (RR): 1.213, 95% confidence interval (CI): 1.162–1.266, p < 0.001], whereas the consumption of the extended spectrum betalactam-betalactamase inhibitor (BLBI) combination and cephalosporins increased by 11.2% per year (RR: 1.112, 95% CI: 1.102–1.122, p < 0.001) and 4.2% per year (RR: 1.042, 95% CI: 1.028–1.056, p < 0.001), respectively, in period 1. With enhanced infection control measures, the hospital-onset CRAB infection decreased by 9.8% per year (RR: 0.902, 95% CI: 0.854–0.953, p < 0.001), whereas the consumption of the extended spectrum BLBI combination and cephalosporins increased by 3.8% per year (RR: 1.038, 95% CI: 1.033–1.044, p < 0.001) and 7.6% per year (RR: 1.076, 95% CI: 1.056–1.097, p < 0.001), respectively, in period 2. The consumption of carbapenems increased by 8.4% per year (RR: 1.84, 95% CI: 1.073–1.094, p < 0.001) in both period 1 and period 2. The control of healthcare-associated CRAB could be achieved by infection control measures with an emphasis on directly observed hand hygiene, despite an increasing trend of antimicrobial consumption. Full article

Formulating termination of isolation (de-isolation) policies requires up-to-date knowledge about viral shedding dynamics. However, current de-isolation policies are largely based on viral load data obtained before the emergence of Omicron variant. In this retrospective cohort study involving adult patients hospitalised for COVID-19 between January and February 2022, we sought to determine SARS-CoV-2 viral shedding kinetics and to investigate the risk factors associated with slow viral decline during the 2022 Omicron wave. A total of 104 patients were included. The viral load was highest (Ct value was lowest) on days 1 post-symptom-onset (PSO) and gradually declined. Older age, hypertension, hyperlipidaemia and chronic kidney disease were associated with slow viral decline in the univariate analysis on both day 7 and day 10 PSO, while incomplete or no vaccination was associated with slow viral decline on day 7 PSO only. However, older age was the only risk factor that remained statistically significant in the multivariate analysis. In conclusion, older age is an independent risk factor associated with slow viral decline in this study conducted during the Omicron-dominant 2022 COVID-19 wave. Transmission-based precaution guidelines should take age into consideration when determining the timing of de-isolation. Full article