Objective To study the imaging findings of the chest disease due to intravenous heroin abuse. Methods Twenty five cases of clinically confirmed chest disease due to intravenous heroin abuse were retrospectively analyzed. 25 cases had conventional X ray film, 6 cases had CT scanning, and 6 cases had echocardiography scanning. Results On X ray and CT, the following signs were found: lung making manifold ( n =5), small patchy shadow ( n =15), pneumatocele ( n =16), small cavity ( n =16), small node ( n =7), pleural effusion ( n =8), pneumothorax ( n =2), hydropneumothorax ( n =6), pulmonary edema ( n =2), megacardia ( n =11), multiple shaped lesion ( n =20). On echocardiography, tricuspid vegetation ( n =4) and tricuspid insufficiency ( n =4) were found. Conclusion The X ray and CT manifestations of chest inflammation due to intravenous heroin abuse are multiple. The multiple small cavities and pneumatoceles sign are of some value in the diagnosis of lung inflammation due to intravenous heroin abuse among young patients.

Objective: To explore the association of the glycosylated hemoglobin (HbA1c) level at admission with 90 days functional outcome in patients with spontaneous intracerebral hemorrhage (ICH). Methods: Patients admitted to the Department of Neurology, Tongji Hospital from January to December 2017 were prospectively and continuously enrolled in this study. Clinical data were collected at admission and functional outcomes 90 days after ICH were assessed by using the modified RANKIN scale. Univariate and multivariate conditional logistic regression models were constructed. Patients were divided into four groups according to the quartile of HbA1c values. The median value of HbA1c in each group was taken as the substitute value and P for trend was calculated. The logistic regression model was fitted by restricted cubic splines to investigate the association between HbA1c level and outcome of ICH. Results: A total of 345 patients with ICH were enrolled, including 214 with favorable outcomes and 131 with poor outcomes (99 severe disability cases and 32 deaths). The risk of poor 90 days outcomes was significantly associated with HbA1c level at admission indicated by multivariate logistic regression analysis, and the P for trend test was <0.001 (middle-level group vs. low-level group: OR=2.33, 95%CI: 1.07-5.07; high-level group vs. low-level group: OR=2.52, 95%CI: 1.12-5.64; extremely high-level group vs. low-level group: OR=6.80, 95%CI: 3.01-15.34). Results from the restricted cubic spline showed that there was a linear correlation between HbA1c level at admission and poor 90 days outcomes of ICH (χ²=14.81, P<0.001; non- linear test: P=0.118). Compared with patients with HbA1c level of 6.5%, the risk of poor outcomes in patients with HbA1c level of <6.5% decreased linearly with the decrease in HbA1c level at admission, and the risk in patients with HbA1c level >6.5% was higher but not significantly. Conclusion There was correlation between high HbA1c level at admission and 90 days poor outcome of ICH. High HbA1c level is an independent prediction indicator for ICH.