BACKGROUND: Trauma-induced coagulopathy (TIC) is a disorder that occurs in one-third of severely injured trauma patients, manifesting as increased bleeding and a 4X risk of mortality. Understanding the mechanisms driving TIC, clinical risk factors are essential to mitigating this coagulopathic bleeding and is therefore essential for saving lives. In this retrospective, single hospital study of 891 trauma patients, we investigate and quantify how two prominently described phenotypes of TIC, consumptive coagulopathy and hyperfibrinolysis, affect survival odds in the first 25 h, when deaths from TIC are most prevalent. METHODS: We employ a joint survival model to estimate the longitudinal trajectories of the protein Factor II (% activity) and the log of the protein fragment D-Dimer ([Formula: see text]g/ml), representative biomarkers of consumptive coagulopathy and hyperfibrinolysis respectively, and tie them together with patient outcomes. Joint models have recently gained popularity in medical studies due to the necessity to simultaneously track continuously measured biomarkers as a disease evolves, as well as to associate them with patient outcomes. In this work, we estimate and analyze our joint model using Bayesian methods to obtain uncertainties and distributions over associations and trajectories. RESULTS: We find that a unit increase in log D-Dimer increases the risk of mortality by 2.22 [1.57, 3.28] fold while a unit increase in Factor II only marginally decreases the risk of mortality by 0.94 [0.91,0.96] fold. This suggests that, while managing consumptive coagulopathy and hyperfibrinolysis both seem to affect survival odds, the effect of hyperfibrinolysis is much greater and more sensitive. Furthermore, we find that the longitudinal trajectories, controlling for many fixed covariates, trend differently for different patients. Thus, a more personalized approach is necessary when considering treatment and risk prediction under these phenotypes. CONCLUSION: This study reinforces the finding that hyperfibrinolysis is linked with poor patient outcomes regardless of factor consumption levels. Furthermore, it quantifies the degree to which measured D-Dimer levels correlate with increased risk. The single hospital, retrospective nature can be understood to specify the results to this particular hospital's patients and protocol in treating trauma patients. Expanding to a multi-hospital setting would result in better estimates about the underlying nature of consumptive coagulopathy and hyperfibrinolysis with survival, regardless of protocol. Individual trajectories obtained with these estimates can be used to provide personalized dynamic risk prediction when making decisions regarding management of blood factors.

Trauma is one of the leading causes of death in the U.S. and is ranked third among death causes across all age groups. This paper presents a novel fuzzy rule-based classification approach based on the concept of General Type-2 Fuzzy sets to predict mortality for trauma patients. In this approach each rule in the rule-base has an IF and a THEN part and parameters of the IF part (antecedents) are automatically extracted using powerful general type-2 fuzzy clustering algorithms which enables the model to deal with noisy and/or missing data. To verify efficacy of the proposed model, it has been implemented on several publicly available datasets. Finally, it is used to predict mortality among patients having traumatic injuries based on a large clinical dataset. Accuracy results demonstrate superior capabilities of the proposed approach compared to crisp and fuzzy classification methods in the literature.

The Cox model has been widely used in time-to-outcome predictions, particularly in studies of medical patients, where prediction of the time of death is desired. In addition, the cure model has been proposed to model times of death for discharged patients. However, neither the Cox model nor the cure model allow explicit cure information and prediction of patient cure times (discharge times). In this paper we propose a new model, the "cure time model", which models the static data for dying patients, surviving patients, and their death/cure times jointly. It models (1) mortality via logistic regression and (2) death and discharge times via Cox models. We extend the cure time model to situations with censored data, where neither time of death nor discharge time are known, as well as to multiple (>2) outcomes. In addition, we propose a joint log-odds ratio which can predict the mortality of patients using the information from both the logistic regression and Cox models. We compare our model with the Cox and cure models on a trauma patient dataset from UCSF/San Francisco General Hospital. Our results show that the cure time model more accurately predicts both mortality and time-to-mortality for patients from these datasets.

BACKGROUND: Physicians use clinical and physiological data to treat patients every day, and it is essential for treating a patient appropriately. However, medical sources of clinical physiological data are only now starting to find use in bioinformatics research. RESULTS: We collected 29 types of physiological and clinical data on a minute-by-minute basis from trauma patients in the intensive care unit along with whether they contracted an infection during their stay. Dividing the patients into two groups based on this criterion, we determined that the correlational network amongst pairs of physiological variables changes based on whether the patient contracted an infection. CONCLUSION: Examining the variable pairs with the largest change in correlation across groups reveals potential changes in the way our treatments affect the patient's physiology and in how our bodies react to physiological insults. These findings highlight the usefulness of physiological informatics and suggest new relationships to study while also validating previously reported relationships.

We describe an application of probabilistic modeling and inference technology to the problem of analyzing sensor data in the setting of an intensive care unit (ICU). In particular, we consider the arterial-line blood pressure sensor, which is subject to frequent data artifacts that cause false alarms in the ICU and make the raw data almost useless for automated decision making. The problem is complicated by the fact that the sensor data are acquired at fixed intervals whereas the events causing data artifacts may occur at any time and have durations that may be significantly shorter than the data collection interval. We show that careful modeling of the sensor, combined with a general technique for detecting sub-interval events and estimating their duration, enables effective detection of artifacts and accurate estimation of the underlying blood pressure values.