Length-of-Stay Prediction for Pediatric Patients With Respiratory Diseases Using Decision Tree Methods.

Accurate prediction of a patient's length-of-stay (LOS) in the hospital enables an efficient and effective management of hospital beds. This paper studies LOS prediction for pediatric patients with respiratory diseases using three decision tree methods: Bagging, Adaboost, and Random forest. A data set of 11,206 records retrieved from the hospital information system is used for analysis after preprocessing and transformation through a computation and an expansion method. Two tests, namely bisection test and periodic test, are designed to assess the performance of the prediction methods. Bagging shows the best result on the bisection test (0.296 RMSE, 0.831 R2, and 0.723 Acc ± 1) for the testing set of the whole data test. The performances of the three methods are similar on the periodic test, whereas Adaboost performs slightly better than the other two methods. Results indicate that the three methods are all effective for the LOS prediction. This study also investigates the importance of different data fields to the LOS prediction, and finds that hospital treatment-related data fields contribute more to the LOS prediction than other categories of fields.

Sequence-dependent anesthesia-controlled times: a retrospective study in an ophthalmology department of a single-site hospital.

BACKGROUND: Anesthesia-controlled time (ACT) generally refers to the time durations before and after the period of surgery. The ACT is typically dependent on the sequence of 2 consecutive surgeries and thus adds to the complexity of operating room scheduling. We report a study on sequence-dependent ACTs at the West China Hospital (WCH), focusing on elective surgeries (also referred to as "procedures" below) performed by the ophthalmology department of WCH over a 5-year period, 2007 to 2012. METHODS: ACTs associated with 4 high-volume procedures: phacoemulsification, vitrectomy, strabismus correction, and glaucoma filtration. A total of 29,452 cases were studied, classified into 4 groups according to the sequence of the procedures involved. Specifically, P-P plots were used to determine the distributions of the ACTs, Kruskal-Wallis H test, Nemenyi test, and Student t test were performed to examine the sequence-dependent nature of the ACTs, and the t test was also applied to examine the advantage of sequencing the same procedures consecutively. Permutations were enumerated to identify the best sequence when different procedures were involved. Monte Carlo simulation was used to compute the total completion time, ACTs plus surgical periods, of any given sequence of procedures. RESULTS: We confirm via statistical tests that the ACTs follow lognormal distributions, and identify their corresponding means and variances. Furthermore, we verify that the ACTs are statistically different in means: they are sequence dependent in general. Using statistical tests, we conclude that it is best to sequence identical procedures consecutively, and we also identify the best sequence involving different procedures. Using Monte Carlo simulation, we compared the daily completion times using the best sequences we have identified against actual data from WCH over a 2-year period; the average reduction is 4.7% (with a standard error (SE) of ± ± 0.5%). CONCLUSIONS: ACTs are usually sequence dependent and hence should be considered in operating room scheduling. Although identifying the best sequence in general is a difficult optimization problem, in certain departments (such as the ophthalmology department of WCH) where a set of high-volume small-variety procedures is present, the best sequences can be systematically identified using a combination of statistical tests and Monte Carlo simulation as illustrated in this study.

Electrochemical control of photoluminescence in two-dimensional MoS(2) nanoflakes.

Two-dimensional (2D) transition metal dichalcogenide semiconductors offer unique electronic and optical properties, which are significantly different from their bulk counterparts. It is known that the electronic structure of 2D MoS2, which is the most popular member of the family, depends on the number of layers. Its electronic structure alters dramatically at near atomically thin morphologies, producing strong photoluminescence (PL). Developing processes for controlling the 2D MoS2 PL is essential to efficiently harness many of its optical capabilities. So far, it has been shown that this PL can be electrically or mechanically gated. Here, we introduce an electrochemical approach to actively control the PL of liquid-phase-exfoliated 2D MoS2 nanoflakes by manipulating the amount of intercalated ions including Li(+), Na(+), and K(+) into and out of the 2D crystal structure. These ions are selected as they are crucial components in many bioprocesses. We show that this controlled intercalation allows for large PL modulations. The introduced electrochemically controlled PL will find significant applications in future chemical and bio-optical sensors as well as optical modulators/switches.