The aims of this study were: (1) to review the rate of concurrent endometrial cancer in patients with a preoperative diagnosis of atypical endometrial hyperplasia (AEH); and (2) to determine the features of concurrent endometrial carcinoma and their impact on the subsequent management of AEH. We reviewed a retrospective series of 219 AEHs diagnosed locally in routine practice, over 24 years, and followed by a repeat biopsy or hysterectomy. Another series of 65 cases with a malignant diagnosis on preoperative sampling was included as a control group. Clinicopathologic parameters were obtained. In addition, published data on the risk of malignancy and features of malignant tumors after a diagnosis of AEH were collected and analyzed. This study reported on 2571 patients diagnosed in 31 published studies in addition to the current one. This showed a wide variation in the positive predictive value (PPV) of AEH in detecting endometrial cancer (6% to 63%) with an overall PPV of 37%. This variation is not only based on the differences among studies but also on the degree of atypia [mild/moderate (PPV 13%) or severe (PPV 50%)], the type of subsequent intervention (biopsy vs. hysterectomy), and more importantly the time period of diagnosis (around 20% in studies published before 1990s and up to 40% to 48% in recently published cases). Of the benign outcome cases, nearly 40% to 50% showed AEH with a potential risk of progressing to invasive carcinoma in 25% of cases. Malignant tumors after AEH diagnosis are associated with features of good prognosis with endometrioid morphology, lower grade, and early stage. Although the overall PPV of AEH is 37%, a figure of 40% to 48% is expected in the cases currently diagnosed in routine practice. Providing qualifying criteria for AEH will help identify its different associated risks and therefore should be included in routine pathology reports whenever possible. Unless there is a clinical contraindication, hysterectomy should be performed to treat concurrent carcinoma and to reduce the risk of subsequent carcinoma in nonmalignant cases with residual AEH.

Urban road tunnels provide an increasingly cost-effective engineering solution, especially in compact cities like Singapore. For some urban road tunnels, tunnel characteristics such as tunnel configurations, geometries, provisions of tunnel electrical and mechanical systems, traffic volumes, etc. may vary from one section to another. These urban road tunnels that have characterized nonuniform parameters are referred to as nonhomogeneous urban road tunnels. In this study, a novel quantitative risk assessment (QRA) model is proposed for nonhomogeneous urban road tunnels because the existing QRA models for road tunnels are inapplicable to assess the risks in these road tunnels. This model uses a tunnel segmentation principle whereby a nonhomogeneous urban road tunnel is divided into various homogenous sections. Individual risk for road tunnel sections as well as the integrated risk indices for the entire road tunnel is defined. The article then proceeds to develop a new QRA model for each of the homogeneous sections. Compared to the existing QRA models for road tunnels, this section-based model incorporates one additional top event-toxic gases due to traffic congestion-and employs the Poisson regression method to estimate the vehicle accident frequencies of tunnel sections. This article further illustrates an aggregated QRA model for nonhomogeneous urban tunnels by integrating the section-based QRA models. Finally, a case study in Singapore is carried out.