This study analyzed the epidemiological characteristics of rabies and the causes of post-exposure management failure in Hunan Province from 2019 to 2024, providing evidence for rabies prevention and control strategies in China.

Data on reported human rabies cases, exposures, and post-exposure prophylaxis (PEP) were analyzed using descriptive epidemiological methods.

240 rabies cases were reported in Hunan Province (2019–2024) with an average annual incidence rate of 0.0592 per 100,000 people. A significant decreasing trend was observed (χ²_trend=32.72, P<0.05). Five factors showed statistically significant differences in their effects on the incubation period: site of exposure, wound management, vaccination after exposure, passive immunization preparations, and sources of animals causing exposure (all P<0.05). In the last six years, t here was no increasing trend in the proportion of failed PEP as a percentage of all rabies cases in that year (χ²_trend=1.809, P=0.86). The median incubation period was 16.0 (Interquartile Range, IQR 14.0–22.0) days for failed PEP cases with exposed areas, including to the head and/or face, compared to 31.0 (IQR 24.0–50.0) days for those without such exposure. The difference was statistically significant (U=20.50, P=0.025).

The current situation of rabies prevention and control in Hunan Province remains dire. Therefore, comprehensive measures should be implemented to help reduce the incidence of rabies. These include adopting standardized dog management practices, strengthening control measures in high-risk areas, and improving public awareness of PEP.

The rapid fluorescent focus inhibition test (RFFIT) is a cell-based virus neutralization assay and the gold standard for quantifying rabies virus neutralizing antibodies (RVNA) in serums. It is used to assess the biological efficacy of rabies vaccines and evaluate protective immunity in both humans and animals. Despite its broad application, RFFIT requires thorough validation to ensure reliability.

RFFIT was validated in this study using the third World Health Organization international standard for anti-rabies immunoglobulin (WHO-3 SRIG) and negative human sera. The validation followed the guidelines outlined by the Food and Drug Administration Guidance for Industry and International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH)Q2 (R1) guidelines and included the assessment of intra-assay and intermediate precision, dilutability, linearity, range, accuracy, specificity, robustness, and stability.

The RFFIT method demonstrated good precision, with intra-assay and intermediate-precision geometric coefficient of variation (GCV) <30%. Dilutability was confirmed, with 95% of positive samples showing geometric mean concentration (GMC) differences within ±30% compared to undiluted controls. The standard and detection values were described by y=1.0091x − 0.1128 (R²⁼0.9948); 95.56% of the samples showed 70%–130% recovery. Specificity was verified using homologous and heterologous antigen competition and a matrix with no significant cross-reactivity. The assay was robust to variations in cells, reagents, and time, with titer differences within ±30%. Stability of samples and reagents under freeze–thaw and different short-term storage conditions was confirmed.

The assay was successfully validated for quantifying RVNA content in serum samples.

We explored risk factors for latent tuberculosis infection (LTBI) and developed a risk prediction model using machine learning algorithms.

Patients with active pulmonary TB in months 3 to 6 of anti-TB treatment in Henan Province, China, July–September 2024 were selected as index cases. Close contacts identified through epidemiological investigation underwent tuberculin-purified protein derivative testing to determine LTBI status. Face-to-face questionnaires were conducted to collect epidemiological data. The dataset was divided into training and testing sets (6:4), using a fixed random seed. Five models — logistic regression (LR), decision tree (DT), random forest (RF), support vector machines (SVM), and multilayer perceptron (MLP) — were trained and evaluated using the mean squared error (MSE) and coefficient of determination. The test set was subjected to external validation. Receiver operating characteristic curve analysis, area under the curve (AUC), and F1-scores were used to quantify predictive performance.

Among 795 close contacts, LTBI prevalence was 401 (50.5%). By MSE, models ranked: SVM (0.121), RF (0.165), DT (0.197), LR (0.229), and MLP (0.233). SVM identified five key predictors: contact type of index case, key population classification, residential area, frequency of participation in group activities, and etiological results. Internal validation showed strong performance (AUC=0.921, F1=0.858), whereas external validation showed moderate performance (AUC=0.752, F1=0.694).

The SVM model incorporating contact type of index case, key population classification, residential area, frequency of group activity participation, and etiological results demonstrated robust predictive value for LTBI risk. This model shows promise for the targeted screening and management of high-risk populations.