基于Document Triage的TF-IDF算法的改进

针对TF-IDF算法在加权时没有考虑特征词本身在文档中重要度的问题,提出利用用户阅读时的阅读行为来改进TF-IDF。将Document Triage引入到TF-IDF中,利用IPM收集用户阅读中行为的相关信息,计算文档评分。由于用户的标注内容往往是文章的重要内容,或者反映了用户的兴趣。因此,赋予用户标注词项更大的权重,将文档评分和用户的标注信息等作为因子引入到TF-IDF中,设计出改进的加权算法DT-TF-IDF。实验结果表明,相对传统TF-IDF算法,DT-TF-IDF的查全率、查准率,以及查准率和查全率的调和均值都有了一定的提高。DT-TF-IDF算法比传统TF-IDF算法更加有效,提高了文本相似度计算的准确性。     

Injury thresholds after motor vehicle crash—Important factors for patient triage and vehicle design

Introduction

The Committee on Trauma recommends that older motor vehicle crash (MVC) victims or victims of crashes with significant vehicle intrusion of more than 12 in. be transferred to a trauma center since those older than 55 have an increased risk of death after injury. Yet, the precise injury thresholds as they relate to age, gender and velocity remain ill-defined. To maintain a low rate of under triage, reliable methods to identify patients at moderate injury risk are needed. We therefore characterized the likelihood of moderate to severe injury in MVC victims to determine the influence of age, gender and velocity.

Methods

An analysis of drivers from the National Automotive Sampling System (1993-2001) was performed. Weighted logistic regression models were developed to predict the probability of head, leg, and torso injuries as a function of vehicle speed, age, and gender while controlling for confounders. A 10% probability of injury threshold was set and differences in velocity, gender and age were identified in terms of reaching this probability of injury threshold.

Results

The analysis yielded 56,459 drivers which is equivalent to a population of 28,877,696 drivers nationwide. Restraint use, steering away prior to impact, breaking maneuver, gender, delta velocity, driver height and age were independent predictors of injury. Women had a higher velocity injury threshold than men for the 10% probability of injury cut-off to the torso or head which disappeared with increasing age. Conversely, men had a higher velocity injury threshold than women for the 10% probability of injury cut-off to the extremity which persisted even in older victims.

Conclusions

Our data indicate that age and gender must be considered in addition to crash velocity when making triage decisions. Furthermore, Federal Motor Vehicle Safety Standards may need to be modified to address the increased risk of injury among older adults at lower velocities given the increasing number of elderly drivers in the US.     

核应急医学救援现场伤员的分类与转运

核与辐射突发事件现场伤员伤情复杂,可能涉及到外照射损伤、放射性核素内污染、放射性核素体表污染及常规损伤等。核应急医学现场救援要遵循快速有效、先重后轻、先救命后去污的原则。根据放射损伤和常规损伤分类标准,按第一优先处理、第二优先处理、延期处理、最后处理四个等级进行检伤分类和开展现场医学救治。伤员转运时要严密观察伤情变化,避免出现放射性核素污染扩散,做好伤员转送记录、伤员资料和物品的交接工作。     

Scheduling prioritized patients in emergency department laboratories

This research focuses on scheduling patients in emergency department laboratories according to the priority of patients’ treatments, determined by the triage factor. The objective is to minimize the total waiting time of patients in the emergency department laboratories with emphasis on patients with severe conditions. The problem is formulated as a flexible open shop scheduling problem and a mixed integer linear programming model is proposed. A genetic algorithm (GA) is developed for solving the problem. Then, the response surface methodology is applied for tuning the GA parameters. The algorithm is tested on a set of real data from an emergency department. Simulation results show that the proposed algorithm can significantly improve the efficiency of the emergency department by reducing the total waiting time of prioritized patients.     

Predictive validity and internal consistency of the pre-hospital index measured on-site by physicians

Physiological measures of injury are used as triage tools to identify patients that require treatment in trauma centres. The Pre-Hospital Index (PHI) is based on systolic blood pressure, pulse, respiratory rate, (level of) consciousness, and presence of penetrating injury. The present study evaluated the validity and internal consistency of the PHI. The study was based on 628 patients assessed by physicians at the scene. Mean age was 38.7 years (SD = 24.8), and 65% were male. Motor vehicle collisions caused the injury for 45%. The majority had head/neck (56%) and extremity (45%) injuries. Mean PHI was 4.62 (SD = 5.77), 40% had a PHI of zero, 6% between 1 and 3, 32% between 4 and 7, and 21% greater than 7. The associations between PHI and rates of hospital admission, surgery, ICU treatment, mortality, duration of hospitalization, and length of ICU stay were significant (p < 0.001). A total of 260 (41.4%) patients had major trauma requiring treatment at a trauma centre. A PHI> 3 had 83% sensitivity and 67% specificity for identifying these patients. Internal consistency of the PHI variables was above the acceptable limits. This study has shown that the PHI is a valid and reliable physiological measure of injury severity and field triage tool.     

Triage and Injury Severity Scores as predictors of mortality and hospital admission for injuries: A validation study

Background

Many emergency departments use a rating system to establish priority based on urgency: “triage”. The aim of this study was to evaluate the validity of triage in predicting hospitalization and mortality compared to that of the ICD-9-CM based Injury Severity Score (ISS).

Methods

Sources: The Emergency Information System 2000, the Hospital Information System 2000–2001 and the Mortality Register 2000–2001, of the Lazio Region.Case selection: Emergency department visits for traumas that occurred on the road or at home.Outcomes: Hospitalization and 30-day mortality.For each case, trauma diagnoses from the ICD-9-CM were given a corresponding ISS value. We performed logistic models, including age, sex and, alternatively, triage or ISS. We compared discrimination measures and calibration of the models.

Results

Out of 264,709 emergency department visits, 22,249 (8.4%) were followed by a hospitalization and 655 (0.2%) died within 30 days. ISS scores were calculated for 72,179 (27%) cases. Of the most urgent triage (840 patients), 78.3% (658) were hospitalized and 9% (76) died, while among patients with ISS ≥ 16 value (1276) 36.4% (464) of were hospitalized and 1.8% (23) died.Measures of discrimination and calibration showed similar results. The triage model had a better fitness in predicting hospitalization probability for home accidents (Hosmer–Lemeshow statistic: χ²_triage = 5.5 vs χ²_ISS = 34.3) and had a better performance for road accidents (ROC_triage = 0.71 vs ROC_ISS = 0.66). There were no differences between the models in predicting the probability of death.

Conclusions

The agreement between the two scales confirms the validity of triage as a clinical management tool in the emergency department, and as a proxy of trauma severity.     

On scene injury severity prediction (OSISP) algorithm for car occupants

Many victims in traffic accidents do not receive optimal care due to the fact that the severity of their injuries is not realized early on. Triage protocols are based on physiological and anatomical criteria and subsequently on mechanisms of injury in order to reduce undertriage. In this study the value of accident characteristics for field triage is evaluated by developing an on scene injury severity prediction (OSISP) algorithm using only accident characteristics that are feasible to assess at the scene of accident. A multivariate logistic regression model is constructed to assess the probability of a car occupant being severely injured following a crash, based on the Swedish Traffic Accident Data Acquisition (STRADA) database. Accidents involving adult occupants for calendar years 2003–2013 included in both police and hospital records, with no missing data for any of the model variables, were included. The total number of subjects was 29 128, who were involved in 22 607 accidents. Partition between severe and non-severe injury was done using the Injury Severity Score (ISS) with two thresholds: ISS > 8 and ISS > 15. The model variables are: belt use, airbag deployment, posted speed limit, type of accident, location of accident, elderly occupant (>55 years old), sex and occupant seat position. The area under the receiver operator characteristic curve (AUC) is 0.78 and 0.83 for ISS > 8 and ISS > 15, respectively, as estimated by 10-fold cross-validation. Belt use is the strongest predictor followed by type of accident. Posted speed limit, age and accident location contribute substantially to increase model accuracy, whereas sex and airbag deployment contribute to a smaller extent and seat position is of limited value. These findings can be used to refine triage protocols used in Sweden and possibly other countries with similar traffic environments.     

Providing early resource allocation during emergencies: The mobile triage tag

Quick response is critical during an emergency situation. This paper describes a system based on mobile electronic triage tags that makes victim information available at the base of operations as soon as possible, thus allowing an early medical resource allocation and immediate action. The cornerstone of the system is mobile agent technology, which allows information to be transported asynchronously and reliably from terminal to terminal and not requiring any network infrastructure at all. This novel approach is ready to be used in the worst case scenario, where only small handheld devices carried by the emergency personnel are available, but also integrates well when synchronous connections are possible, for instance when a mesh network can be created. The system has been successfully implemented, showing the feasibility of the proposal. By using this low-budget system, the number of casualties during the triage stage of an emergency is expected to drop off.     

Ambulance transport rates after motor vehicle collision for older vs. younger adults: A population-based study

Older adults are at greater risk than younger adults for life-threatening injury after motor vehicle collision (MVC). Among those with life-threatening injury, older adults are also at greater risk of not being transported by emergency medical services (EMS) to an emergency department. Despite the greater risk of serious injury and non-transportation among older adults, little is known about the relationship between patient age and EMS transportation rates for individuals experiencing MVC. We describe transport rates across the age-span for adults seen by EMS after experiencing MVC using data reported to the North Carolina Department of Motor Vehicles between 2008 and 2011. Of all adults aged 18 years and older experiencing MVC and seen by EMS (n = 484,310), 36.3% (n = 175,768) were transported to an emergency department. Rates of transport for individuals seen by EMS after MVC increased only a small amount with increasing patient age. After adjusting for potential confounders of the relationship between patient age and the decision to transport (patient gender, patient race, air bag deployment, patient trapped or ejected, and injury severity), transport rates were: age 18–64 = 36.0% (95% confidence interval [CI], 35.9–36.2%); age 65–74 = 36.6% (95% CI, 36.0–37.1%); age 75–84 = 37.3% (95% CI, 36.5–38.1%), and age 85–94 = 38.2% (95% CI, 36.7–39.8%). In North Carolina between 2008 and 2011, the transportation rate was only slightly higher for older adults than for younger adults, and most older adults experiencing MVC and seen by EMS were not transported to the emergency department. These findings have implications for efforts to improve the sensitivity of criteria used by EMS to determine the need for transport for older adults experiencing MVC.