Use of low-dose computed tomography to assess pulmonary tuberculosis among healthcare workers in a tuberculosis hospital.

BACKGROUND: According to the World Health Organization, China is one of 22 countries with serious tuberculosis (TB) infections and one of the 27 countries with serious multidrug-resistant TB strains. Despite the decline of tuberculosis in the overall population, healthcare workers (HCWs) are still at a high risk of infection. Compared with high-income countries, the TB prevalence among HCWs is higher in low- and middle-income countries. Low-dose computed tomography (LDCT) is becoming more popular due to its superior sensitivity and lower radiation dose. However, there have been no reports about active pulmonary tuberculosis (PTB) among HCWs as assessed with LDCT. The purposes of this study were to examine PTB statuses in HCWs in hospitals specializing in TB treatment and explore the significance of the application of LDCT to these workers. METHODS: This study retrospectively analysed the physical examination data of healthcare workers in the Beijing Chest Hospital from September 2012 to December 2015. Low-dose lung CT examinations were performed in all cases. The comparisons between active and inactive PTB according to the CT findings were made using the Pearson chi-square test or the Fisher's exact test. Comparisons between the incidences of active PTB in high-risk areas and non-high-risk areas were performed using the Pearson chi-square test. Analyses of active PTB were performed according to different ages, numbers of years on the job, and the risks of the working areas. Active PTB as diagnosed by the LDCT examinations alone was compared with the final comprehensive diagnoses, and the sensitivity and positive predictive value were calculated. RESULTS: A total of 1 012 participants were included in this study. During the 4-year period of medical examinations, active PTB was found in 19 cases, and inactive PTB was found in 109 cases. The prevalence of active PTB in the participants was 1.24%, 0.67%, 0.81%, and 0.53% for years 2012 to 2015. The corresponding incidences of active PTB among the tuberculosis hospital participants were 0.86%, 0.41%, 0.54%, and 0.26%. Most HCWs with active TB (78.9%, 15/19) worked in the high-risk areas of the hospital. There was a significant difference in the incidences of active PTB between the HCWs who worked in the high-risk and non-high-risk areas (odds ratio [OR], 14.415; 95% confidence interval (CI): 4.733 - 43.896). Comparisons of the CT signs between the active and inactive groups via chi-square tests revealed that the tree-in-bud, cavity, fibrous shadow, and calcification signs exhibited significant differences (P = 0.000, 0.021, 0.001, and 0.024, respectively). Tree-in-bud and cavity opacities suggest active pulmonary tuberculosis, whereas fibrous shadow and calcification opacities are the main features of inactive pulmonary tuberculosis. Comparison with the final comprehensive diagnoses revealed that the sensitivity and positive predictive value of the diagnoses of active PTB based on LDCT alone were 100% and 86.4%, respectively. CONCLUSIONS: Healthcare workers in tuberculosis hospitals are a high-risk group for active PTB. Yearly LDCT examinations of such high-risk groups are feasible and necessary.

Prediction models for solitary pulmonary nodules based on curvelet textural features and clinical parameters.

Lung cancer, one of the leading causes of cancer-related deaths, usually appears as solitary pulmonary nodules (SPNs) which are hard to diagnose using the naked eye. In this paper, curvelet-based textural features and clinical parameters are used with three prediction models [a multilevel model, a least absolute shrinkage and selection operator (LASSO) regression method, and a support vector machine (SVM)] to improve the diagnosis of benign and malignant SPNs. Dimensionality reduction of the original curvelet-based textural features was achieved using principal component analysis. In addition, non-conditional logistical regression was used to find clinical predictors among demographic parameters and morphological features. The results showed that, combined with 11 clinical predictors, the accuracy rates using 12 principal components were higher than those using the original curvelet-based textural features. To evaluate the models, 10-fold cross validation and back substitution were applied. The results obtained, respectively, were 0.8549 and 0.9221 for the LASSO method, 0.9443 and 0.9831 for SVM, and 0.8722 and 0.9722 for the multilevel model. All in all, it was found that using curvelet-based textural features after dimensionality reduction and using clinical predictors, the highest accuracy rate was achieved with SVM. The method may be used as an auxiliary tool to differentiate between benign and malignant SPNs in CT images.