Dataset Information

Forecasting individual risk for long-term Posttraumatic Stress Disorder in emergency medical settings using biomedical data: A machine learning multicenter cohort study.

ABSTRACT: The necessary requirement of a traumatic event preceding the development of Posttraumatic Stress Disorder, theoretically allows for administering preventive and early interventions in the early aftermath of such events. Machine learning models including biomedical data to forecast PTSD outcome after trauma are highly promising for detection of individuals most in need of such interventions. In the current study, machine learning was applied on biomedical data collected within 48 h post-trauma to forecast individual risk for long-term PTSD, using a multinominal approach including the full spectrum of common PTSD symptom courses within one prognostic model for the first time. N = 417 patients (37.2% females; mean age 46.09 ± 15.88) admitted with (suspected) serious injury to two urban Academic Level-1 Trauma Centers were included. Routinely collected biomedical information (endocrine measures, vital signs, pharmacotherapy, demographics, injury and trauma characteristics) upon ED admission and subsequent 48 h was used. Cross-validated multi-nominal classification of longitudinal self-reported symptom severity (IES-R) over 12 months and bimodal classification of clinician-rated PTSD diagnosis (CAPS-IV) at 12 months post-trauma was performed using extreme Gradient Boosting and evaluated on hold-out sets. SHapley Additive exPlanations (SHAP) values were used to explain the derived models in human-interpretable form. Good prediction of longitudinal PTSD symptom trajectories (multiclass AUC = 0.89) and clinician-rated PTSD at 12 months (AUC = 0.89) was achieved. Most relevant prognostic variables to forecast both multinominal and dichotomous PTSD outcomes included acute endocrine and psychophysiological measures and hospital-prescribed pharmacotherapy. Thus, individual risk for long-term PTSD was accurately forecasted from biomedical information routinely collected within 48 h post-trauma. These results facilitate future targeted preventive interventions by enabling future early risk detection and provide further insights into the complex etiology of PTSD.

SUBMITTER: Schultebraucks K

PROVIDER: S-EPMC7843920 | biostudies-literature | 2021 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Forecasting individual risk for long-term Posttraumatic Stress Disorder in emergency medical settings using biomedical data: A machine learning multicenter cohort study.

Schultebraucks Katharina K Sijbrandij Marit M Galatzer-Levy Isaac I Mouthaan Joanne J Olff Miranda M van Zuiden Mirjam M

Neurobiology of stress 20210118

The necessary requirement of a traumatic event preceding the development of Posttraumatic Stress Disorder, theoretically allows for administering preventive and early interventions in the early aftermath of such events. Machine learning models including biomedical data to forecast PTSD outcome after trauma are highly promising for detection of individuals most in need of such interventions. In the current study, machine learning was applied on biomedical data collected within 48 h post-trauma to ...[more]

PMID: 33553513

Similar Datasets

Project description:BackgroundAccording to the rapid response system's team composition, responding teams were named as rapid response team (RRT), medical emergency team (MET), and critical care outreach. A RRT is often a nurse-led team, whereas a MET is a physician-led team that mainly plays the role of an efferent limb. As few multicenter studies have focused on physician-led METs, we comprehensively analyzed cases for which physician-led METs were activated.MethodsWe retrospectively analyzed cases for which METs were activated. The study population consisted of subjects over 18 years of age who were admitted in the general ward from January 2016 to December 2017 in 9 tertiary teaching hospitals in Korea. The data on subjects' characteristics, activation causes, activation methods, performed interventions, in-hospital mortality, and intensive care unit (ICU) transfer after MET activation were collected and analyzed.ResultsIn this study, 12,767 cases were analyzed, excluding those without in-hospital mortality data. The subjects' median age was 67 years, and 70.4% of them were admitted to the medical department. The most common cause of MET activation was respiratory distress (35.1%), followed by shock (11.8%), and the most common underlying disease was solid cancer (39%). In 7,561 subjects (59.2%), the MET was activated using the screening system. The commonly performed procedures were arterial line insertion (17.9%), intubation (13.3%), and portable ultrasonography (13.0%). Subsequently, 29.4% of the subjects were transferred to the ICU, and 27.2% died during hospitalization.ConclusionsThis physician-led MET cohort showed relatively high rates of intervention, including arterial line insertion and portable ultrasonography, and low ICU transfer rates. We presume that MET detects deteriorating patients earlier using a screening system and begins ICU-level management at the patient's bedside without delay, eventually preventing the patient's condition from worsening and transfer to the ICU.

Dataset Information

Forecasting individual risk for long-term Posttraumatic Stress Disorder in emergency medical settings using biomedical data: A machine learning multicenter cohort study.

Publications

Forecasting individual risk for long-term Posttraumatic Stress Disorder in emergency medical settings using biomedical data: A machine learning multicenter cohort study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets