Probabilistic Matching Approach to Link Deidentified Data from a Trauma Registry and a Traumatic Brain Injury Model System Center.
Ontology highlight
ABSTRACT:
SUBMITTER: Kesinger MR
PROVIDER: S-EPMC5065730 | biostudies-literature | 2017 Jan
REPOSITORIES: biostudies-literature
Publications
Probabilistic Matching Approach to Link Deidentified Data from a Trauma Registry and a Traumatic Brain Injury Model System Center.
Kesinger Matthew Ryan MR Kumar Raj Gopalan RG Ritter Anne Connelly AC Sperry Jason Lee JL Wagner Amy Kathleen AK
American journal of physical medicine & rehabilitation 20170101 1
<h4>Objective</h4>There is no civilian traumatic brain injury database that captures patients in all settings of the care continuum. The linkage of such databases would yield valuable insight into possible care interventions. Thus, the objective of this article is to describe the creation of an algorithm used to link the Traumatic Brain Injury Model System (TBIMS) to trauma data in state and national trauma databases.<h4>Design</h4>The TBIMS data from a single center was randomly divided into tw ...[more]
Similar Datasets
Project description:In a previous study, individuals from a single Traumatic Brain Injury Model Systems and trauma center were matched using a novel probabilistic matching algorithm. The Traumatic Brain Injury Model Systems is a multicenter prospective cohort study containing more than 14,000 participants with traumatic brain injury, following them from inpatient rehabilitation to the community over the remainder of their lifetime. The National Trauma Databank is the largest aggregation of trauma data in the United States, including more than 6 million records. Linking these two databases offers a broad range of opportunities to explore research questions not otherwise possible. Our objective was to refine and validate the previous protocol at another independent center. An algorithm generation and validation data set were created, and potential matches were blocked by age, sex, and year of injury; total probabilistic weight was calculated based on of 12 common data fields. Validity metrics were calculated using a minimum probabilistic weight of 3. The positive predictive value was 98.2% and 97.4% and sensitivity was 74.1% and 76.3%, in the algorithm generation and validation set, respectively. These metrics were similar to the previous study. Future work will apply the refined probabilistic matching algorithm to the Traumatic Brain Injury Model Systems and the National Trauma Databank to generate a merged data set for clinical traumatic brain injury research use.
| S-EPMC5863735 | biostudies-literature
Project description:The aim of this study was to investigate and compare the injury characteristics, severity, and outcome between underweight and normal-weight patients hospitalized for the treatment of all kinds of trauma injury.This study was based on a level I trauma center Taiwan.The detailed data of 640 underweight adult trauma patients with a body mass index (BMI) of <18.5 kg/m and 6497 normal-weight adult patients (25 > BMI ≥ 18.5 kg/m) were retrieved from the Trauma Registry System between January 1, 2009, and December 31, 2014. Pearson's chi-square test, Fisher's exact test, and independent Student's t-test were performed to compare the differences. Propensity score matching with logistic regression was used to evaluate the effect of underweight on mortality.Underweight patients presented a different bodily injury pattern and a significantly higher rate of admittance to the intensive care unit (ICU) than did normal-weight patients; however, no significant differences in the Glasgow Coma Scale (GCS) score, injury severity score (ISS), in-hospital mortality, and hospital length of stay were found between the two groups. However, further analysis of the patients stratified by two major injury mechanisms (motorcycle accident and fall injury) revealed that underweight patients had significantly lower GCS scores (13.8 ± 3.0 vs 14.5 ± 2.0, P = 0.020), but higher ISS (10.1 ± 6.9 vs 8.4 ± 5.9, P = 0.005), in-hospital mortality (odds ratio, 4.4; 95% confidence interval, 1.69-11.35; P = 0.006), and ICU admittance rate (24.1% vs 14.3%, P = 0.007) than normal-weight patients in the fall accident group, but not in the motorcycle accident group. However, after propensity score matching, logistic regression analysis of well-matched pairs of patients with either all trauma, motorcycle accident, or fall injury did not show a significant influence of underweight on mortality.Exploratory data analysis revealed that underweight patients presented a different bodily injury pattern from that of normal-weight patients, specifically a higher incidence of pneumothorax in those with penetrating injuries and of femoral fracture in those with struck on/against injuries; however, the injury severity and outcome of underweight patients varied depending on the injury mechanism.
| S-EPMC5348189 | biostudies-literature
Project description:Objective: To analyze the epidemiological information of patients with traumatic spinal cord injury (SCI) and concomitant traumatic brain injury (TBI) and to suggest points to be aware of during the initial physical examination of patients with SCI.Methods: This study was a retrospective, observational study conducted in a regional trauma center. All the records of patients diagnosed with traumatic SCI between 2016 and 2020 were reviewed. A total of 627 patients with confirmed traumatic SCI were hospitalized. A retrospective study was conducted on 363 individuals.Results: The epidemiological data of 363 individuals were investigated. Changes in American Spinal Injury Association Impairment Scale (AIS) scores in patients with SCI were evaluated. The initial evaluation was performed on average 11 days after the injury, and a follow-up examination was performed 43 days after. Fourteen of the 24 patients identified as having AIS A and SCI with concomitant TBI in the initial evaluation showed neurologic level of injury (NLI) recovery with AIS B or more. The conversion rate in patients with SCI and concomitant TBI exceeded that reported in previous studies in individuals with SCI.Conclusions: Physical, cognitive, and emotional impairments caused by TBI present significant challenges in rehabilitating patients with SCI. In this study, the influence of concomitant TBI lesions could have caused the initial AIS assessment to be incorrect.
| S-EPMC10620491 | biostudies-literature
Project description:BackgroundIntracranial pressure monitor (ICPm) procedure rates are a quality metric for American College of Surgeons trauma center verification. However, ICPm procedure rates may not accurately reflect the quality of care in TBI. We hypothesized that ICPm and craniotomy/craniectomy procedure rates for severe TBI vary across the United States by geography and institution.MethodsWe identified all patients with a severe traumatic brain injury (head Abbreviated Injury Scale, ≥3) from the 2016 Trauma Quality Improvement Program data set. Patients who received surgical decompression or ICPm were identified via International Classification of Diseases codes. Hospital factors included neurosurgeon group size, geographic region, teaching status, and trauma center level. Two multiple logistic regression models were performed identifying factors associated with (1) craniotomy with or without ICPm or (2) ICPm alone. Data are presented as medians (interquartile range) and odds ratios (ORs) (95% confidence interval).ResultsWe identified 75,690 patients (66.4% male; age, 59 [36-77] years) with a median Injury Severity Score of 17 (11-25). Overall, 6.1% had surgical decompression, and 4.8% had ICPm placement. Logistic regression analysis showed that region of the country was significantly associated with procedure type: hospitals in the West were more likely to use ICPm (OR, 1.34 [1.20-1.50]), while Northeastern (OR, 0.80 [0.72-0.89]), Southern (OR, 0.84 [0.78-0.92]), and Western (OR, 0.88 [0.80-0.96]) hospitals were less likely to perform surgical decompression. Hospitals with small neurosurgeon groups (<3) were more likely to perform surgical intervention. Community hospitals are associated with higher odds of surgical decompression but lower odds of ICPm placement.ConclusionBoth geographic differences and hospital characteristics are independent predictors for surgical intervention in severe traumatic brain injury. This suggests that nonpatient factors drive procedural decisions, indicating that ICPm rate is not an ideal quality metric for American College of Surgeons trauma center verification.Level of evidenceEpidemiological, level III; Care management/Therapeutic level III.
| S-EPMC8505004 | biostudies-literature
Project description:Traumatic brain injury (TBI) is a huge public health challenge worldwide. Epidemiological monitoring is important to inform healthcare policy. We aimed at determining the prevalence, outcome, and causes of TBI in Cameroon by conducting a 5-year retrospective study in three referral trauma centers. Data on demographics, causes, injury mechanisms, clinical aspects, and discharge status were recorded. Comparisons between two categorical variables were done using Pearson's chi-square test or Fisher's exact test. A total of 6248 cases of TBI were identified of 18,151 trauma cases, yielding a prevalence of 34%. The number of TBI cases increased across the years (915 in 2016, 1406 in 2020). Demographic data and causes of TBI were available for 6248 subjects and detailed data on clinical characteristics on 2178 subjects. Median age was 30.0 (24.0, 41.0) years. Males were more affected (80%). Road traffic incidents (RTIs; 75%) was the main cause of TBI, with professional bike riders being more affected (17%). Computed tomography (CT) imaging was performed in 67.7% of cases. Of the 597 (27.4%) cases who did not undergo neuroimaging, 311 (52.1%) did not have neuroimaging performed because of financial constraints, among which 7% were severe TBI cases. A total of 341 (19.6%) patients were discharged against medical advice, of which 83% had financial limitations. Mortality was 10.3% (225 of 2178) in the overall population, but disproportionately high in patients with severe TBI (55%) compared to those in high-income settings (27%). TBI occurrence is high in Cameroon, and RTIs are the main causes. Disparities in care provision were identified as attributable to financial constraints regarding CT scanning and continuation of care. The data presented can inform preventive interventions to improve care provision and transport policies. Implementation of a universal health insurance may be expected to improve hospital care and reduce the adverse effects of TBI among Cameroonians.
| S-EPMC9879018 | biostudies-literature
Project description:BackgroundHypotension is associated with worse outcome in patients with traumatic brain injury (TBI) and maintaining a systolic blood pressure (SBP) ≥110 mmHg is recommended. The aim of this study was to assess the incidence of TBI in patients suffering multiple trauma in mountain areas; to describe associated factors, treatment and outcome compared to non-hypotensive patients with TBI and patients without TBI; and to evaluate pre-hospital variables to predict admission hypotension.MethodsData from the prospective International Alpine Trauma Registry including mountain multiple trauma patients (ISS ≥ 16) collected between 2010 and 2019 were analysed. Patients were divided into three groups: 1) TBI with hypotension, 2) TBI without hypotension and 3) no TBI. TBI was defined as Abbreviated Injury Scale (AIS) of the head/neck ≥3 and hypotension as SBP < 110 mmHg on hospital arrival.ResultsA total of 287 patients were included. Fifty (17%) had TBI and hypotension, 92 (32%) suffered TBI without hypotension and 145 (51%) patients did not have TBI. Patients in group 1 were more severely injured (mean ISS 43.1 ± 17.4 vs 33.3 ± 15.3 vs 26.2 ± 18.1 for group 1 vs 2 vs 3, respectively, p < 0.001). Mean SBP on hospital arrival was 83.1 ± 12.9 vs 132.5 ± 19.4 vs 119.4 ± 25.8 mmHg (p < 0.001) despite patients in group 1 received more fluids. Patients in group 1 had higher INR, lower haemoglobin and lower base excess (p < 0.001). More than one third of patients in group 1 and 2 were hypothermic (body temperature < 35 °C) on hospital arrival while the rate of admission hypothermia was low in patients without TBI (41% vs 35% vs 21%, for group 1 vs 2 vs 3, p = 0.029). The rate of hypothermia on hospital arrival was different between the groups (p = 0.029). Patients in group 1 had the highest mortality (24% vs 10% vs 1%, p < 0.001).ConclusionMultiple trauma in the mountains goes along with severe TBI in almost 50%. One third of patients with TBI is hypotensive on hospital arrival and this is associated with a worse outcome. No single variable or set of variables easily obtainable at scene was able to predict admission hypotension in TBI patients.
| S-EPMC8086074 | biostudies-literature
Project description:BackgroundProviding optimal care for trauma, the leading cause of death for young adults, remains a challenge e.g., due to field triage limitations in assessing a patient's condition and deciding on transport destination. Data-driven On Scene Injury Severity Prediction (OSISP) models for motor vehicle crashes have shown potential for providing real-time decision support. The objective of this study is therefore to evaluate if an Artificial Intelligence (AI) based clinical decision support system can identify severely injured trauma patients in the prehospital setting.MethodsThe Swedish Trauma Registry was used to train and validate five models - Logistic Regression, Random Forest, XGBoost, Support Vector Machine and Artificial Neural Network - in a stratified 10-fold cross validation setting and hold-out analysis. The models performed binary classification of the New Injury Severity Score and were evaluated using accuracy metrics, area under the receiver operating characteristic curve (AUC) and Precision-Recall curve (AUCPR), and under- and overtriage rates.ResultsThere were 75,602 registrations between 2013-2020 and 47,357 (62.6%) remained after eligibility criteria were applied. Models were based on 21 predictors, including injury location. From the clinical outcome, about 40% of patients were undertriaged and 46% were overtriaged. Models demonstrated potential for improved triaging and yielded AUC between 0.80-0.89 and AUCPR between 0.43-0.62.ConclusionsAI based OSISP models have potential to provide support during assessment of injury severity. The findings may be used for developing tools to complement field triage protocols, with potential to improve prehospital trauma care and thereby reduce morbidity and mortality for a large patient population.
| S-EPMC10561449 | biostudies-literature
Project description:Bladder rupture occurs in only 1.6% of blunt abdominopelvic trauma cases. Although rare, bladder rupture can result in significant morbidity if undiagnosed or inappropriately managed. AUA Urotrauma Guidelines suggest that urethral catheter drainage is a standard of care for both extraperitoneal and intraperitoneal bladder rupture regardless of the need for surgical repair. However, no specific guidance is given regarding the length of catheterization. The present study seeks to summarize contemporary management of bladder trauma at our tertiary care center, assess the impact of length of catheterization on bladder injuries and complications, and develop a protocol for management of bladder injuries from time of injury to catheter removal. A retrospective review was performed on 34,413 blunt trauma cases to identify traumatic bladder ruptures over the past 10 years (January 2008-January 2018) at our tertiary care facility. Patient data were collected including age, gender, BMI, mechanism of injury, and type of injury. The primary treatment modality (surgical repair vs. catheter drainage only), length of catheterization, and post-injury complications were also assessed. Review of our institutional trauma database identified 44 patients with bladder trauma. Mean age was 41 years, mean BMI was 24.8 kg/m2, 95% were Caucasian, and 55% were female. Motor vehicle collision (MVC) was the most common mechanism, representing 45% of total injuries. Other mechanisms included falls (20%) and all-terrain vehicle (ATV) accidents (13.6%). 31 patients had extraperitoneal injury, and 13 were intraperitoneal. Pelvic fractures were present in 93%, and 39% had additional solid organ injuries. Formal cystogram was performed in 59% on presentation, and mean time to cystogram was 4 hours. Gross hematuria was noted in 95% of cases. Operative management was performed for all intraperitoneal injuries and 35.5% of extraperitoneal cases. Bladder closure in operative cases was typically performed in 2 layers with absorbable suture in a running fashion. The intraperitoneal and extraperitoneal injuries managed operatively were compared, and length of catheterization (28 d vs. 22 d, p=0.46), time from injury to normal fluorocystogram (19.8 d vs. 20.7 d, p=0.80), and time from injury to repair (4.3 vs. 60.5 h, p=0.23) were not statistically different between cohorts. Patients whose catheter remained in place for greater than 14 days had prolonged time to initial cystogram (26.6 d vs. 11.5 d) compared with those whose foley catheter was removed within 14 days. The complication rate was 21% for catheters left more than 14 days while patients whose catheter remained less than 14 days experienced no complications. The present study provides a 10-year retrospective review characterizing the presentation, management, and follow-up of bladder trauma patients at our level 1 trauma center. Based on our findings, we have developed an institutional protocol which now includes recommendations regarding length of catheterization after traumatic bladder rupture. By providing specific guidelines for initial follow-up cystogram and foley removal, we hope to decrease patient morbidity from prolonged catheterization. Further study will seek to allow multidisciplinary trauma teams to standardize management, streamline care, and minimize complications for patients presenting with traumatic bladder injuries.
| S-EPMC6930775 | biostudies-literature
Project description:Vestibular dysfunction, causing dizziness and imbalance, is a common yet poorly understood feature in patients with TBI. Damage to the inner ear, nerve, brainstem, cerebellum and cerebral hemispheres may all affect vestibular functioning, hence, a multi-level assessment-from reflex to perception-is required. In a previous report, postural instability was the commonest neurological feature in ambulating acute patients with TBI. During ward assessment, we also frequently observe a loss of vertigo sensation in patients with acute TBI, common inner ear conditions and a related vigorous vestibular-ocular reflex nystagmus, suggesting a 'vestibular agnosia'. Patients with vestibular agnosia were also more unbalanced; however, the link between vestibular agnosia and imbalance was confounded by the presence of inner ear conditions. We investigated the brain mechanisms of imbalance in acute TBI, its link with vestibular agnosia, and potential clinical impact, by prospective laboratory assessment of vestibular function, from reflex to perception, in patients with preserved peripheral vestibular function. Assessment included: vestibular reflex function, vestibular perception by participants' report of their passive yaw rotations in the dark, objective balance via posturography, subjective symptoms via questionnaires, and structural neuroimaging. We prospectively screened 918 acute admissions, assessed 146 and recruited 37. Compared to 37 matched controls, patients showed elevated vestibular-perceptual thresholds (patients 12.92°/s versus 3.87°/s) but normal vestibular-ocular reflex thresholds (patients 2.52°/s versus 1.78°/s). Patients with elevated vestibular-perceptual thresholds [3 standard deviations (SD) above controls' average], were designated as having vestibular agnosia, and displayed worse posturography than non-vestibular-agnosia patients, despite no difference in vestibular symptom scores. Only in patients with impaired postural control (3 SD above controls' mean), whole brain diffusion tensor voxel-wise analysis showed elevated mean diffusivity (and trend lower fractional anisotropy) in the inferior longitudinal fasciculus in the right temporal lobe that correlated with vestibular agnosia severity. Thus, impaired balance and vestibular agnosia are co-localized to the inferior longitudinal fasciculus in the right temporal lobe. Finally, a clinical audit showed a sevenfold reduction in clinician recognition of a common peripheral vestibular condition (benign paroxysmal positional vertigo) in acute patients with clinically apparent vestibular agnosia. That vestibular agnosia patients show worse balance, but without increased dizziness symptoms, explains why clinicians may miss treatable vestibular diagnoses in these patients. In conclusion, vestibular agnosia mediates imbalance in traumatic brain injury both directly via white matter tract damage in the right temporal lobe, and indirectly via reduced clinical recognition of common, treatable vestibular diagnoses.
| S-EPMC7880674 | biostudies-literature
Project description:BackgroundThe Enhanced Matching System (EMS) is a probabilistic record linkage program developed by the tuberculosis section at Public Health England to match data for individuals across two datasets. This paper outlines how EMS works and investigates its accuracy for linkage across public health datasets.MethodsEMS is a configurable Microsoft SQL Server database program. To examine the accuracy of EMS, two public health databases were matched using National Health Service (NHS) numbers as a gold standard unique identifier. Probabilistic linkage was then performed on the same two datasets without inclusion of NHS number. Sensitivity analyses were carried out to examine the effect of varying matching process parameters.ResultsExact matching using NHS number between two datasets (containing 5931 and 1759 records) identified 1071 matched pairs. EMS probabilistic linkage identified 1068 record pairs. The sensitivity of probabilistic linkage was calculated as 99.5% (95%CI: 98.9, 99.8), specificity 100.0% (95%CI: 99.9, 100.0), positive predictive value 99.8% (95%CI: 99.3, 100.0), and negative predictive value 99.9% (95%CI: 99.8, 100.0). Probabilistic matching was most accurate when including address variables and using the automatically generated threshold for determining links with manual review.ConclusionWith the establishment of national electronic datasets across health and social care, EMS enables previously unanswerable research questions to be tackled with confidence in the accuracy of the linkage process. In scenarios where a small sample is being matched into a very large database (such as national records of hospital attendance) then, compared to results presented in this analysis, the positive predictive value or sensitivity may drop according to the prevalence of matches between databases. Despite this possible limitation, probabilistic linkage has great potential to be used where exact matching using a common identifier is not possible, including in low-income settings, and for vulnerable groups such as homeless populations, where the absence of unique identifiers and lower data quality has historically hindered the ability to identify individuals across datasets.
| S-EPMC4547731 | biostudies-literature