Modeling Probability-Based Injury Severity Scores in Logistic Regression Models: The Logit Transformation Should Be Used

Delays in emergency department intervention for patients with traumatic brain injury in Uganda

Background

In Sub-Saharan African countries, the incidence of traumatic brain injury (TBI) is estimated to be many folds higher than the global average and outcome is hugely impacted by access to healthcare services and quality of care. We conducted an analysis of the TBI registry data to determine the disparities and delays in treatment for patients presenting at a tertiary care hospital in Uganda and to identify factors predictive of delayed treatment initiation.

Methods

The study was conducted at the Mulago National Referral Hospital, Kampala. The study included all patients presenting to the emergency department (ED) with suspected or documented TBI. Early treatment was defined as first intervention within 4 hours of ED presentation-a cut-off determined using sensitivity analysis to injury severity. Descriptive statistics were generated and Pearson's χ² test was used to assess the sample distribution between treatment time categories. Univariable and multivariable logistic regression models with <0.05 level of significance were used to derive the associations between patient characteristics and early intervention for TBI.

Results

Of 3944 patients, only 4.6% (n=182) received an intervention for TBI management within 1 hour of ED presentation, whereas 17.4% of patients (n=708) received some treatment within 4 hours of presentation. 19% of those with one or more serious injuries and 18% of those with moderate to severe head injury received care within 4 hours of arrival. Factors independently associated with early treatment included young age, severe head injury, and no known pre-existing conditions, whereas older or female patients had significantly less odds of receiving early treatment.

Discussion

With the increasing number of patients with TBI, ensuring early and appropriate management must be a priority for Ugandan hospitals. Delay in initiation of treatment may impact survival and functional outcome. Gender-related and age-related disparities in care should receive attention and targeted interventions.

Level of evidence

Prognostic and epidemiological study; level II evidence.

The FASILA Score: A Novel Bio-Clinical Score to Predict Massive Blood Transfusion in Patients with Abdominal Trauma

Background

Early identification of patients who may need massive blood transfusion remains a major challenge in trauma care. This study proposed a novel and easy-to-calculate prediction score using clinical and point of care laboratory findings in patients with abdominal trauma (AT).

Methods

Patients with AT admitted to a trauma center in Qatar between 2014 and 2017 were retrospectively analyzed. The FASILA score was proposed and calculated using focused assessment with sonography in trauma (0 = negative, 1 = positive), Shock Index (SI) (0 = 0.50–0.69, 1 = 0.70–0.79, 2 = 0.80–0.89, and 3 ≥ 0.90), and initial serum lactate (0 ≤ 2.0, 1 = 2.0–4.0, and 2 ≥ 4.0 mmol/l). Outcome variables included mortality, laparotomy, and massive blood transfusion (MT). FASILA was compared to other prediction scores using receiver operating characteristics and areas under the curves. Bootstrap procedure was employed for internal validation.

Results

In 1199 patients with a mean age of 31 ± 13.5 years, MT, MT protocol (MTP) activation, exploratory laparotomy (ExLap), and hospital mortality were related linearly with the FASILA score, Injury Severity Score, and total length of hospital stay. Initial hemoglobin, Revised Trauma Score (RTS), and Trauma Injury Severity Score (TRISS) were inversely proportional. FASILA scores correlated significantly with the Assessment of Blood Consumption (ABC) (r = 0.65), Revised Assessment of Bleeding and Transfusion (RABT) (r = 0.63), SI (r = 0.72), RTS (r = − 0.34), and Glasgow Coma Scale (r = − 0.32) and outperformed other predictive systems (RABT, ABC, and SI) in predicting MT, MTP, ExLap, and mortality.

Conclusions

The novel FASILA score performs well in patients with abdominal trauma and offers advantages over other scores.

Electronic supplementary material

The online version of this article (10.1007/s00268-019-05289-0) contains supplementary material, which is available to authorized users.

Decreased risk adjusted 30-day mortality for hospital admitted injuries: a multi-centre longitudinal study

Background

The interpretation of changes in injury-related mortality over time requires an understanding of changes in the incidence of the various types of injury, and adjustment for their severity. Our aim was to investigate changes over time in incidence of hospital admission for injuries caused by falls, traffic incidents, or assaults, and to assess the risk-adjusted short-term mortality for these patients.

Methods

All patients admitted to hospital with injuries caused by falls, traffic incidents, or assaults during the years 2001–11 in Sweden were identified from the nationwide population-based Patient Registry. The trend in mortality over time for each cause of injury was adjusted for age, sex, comorbidity and severity of injury as classified from the International Classification of diseases, version 10 Injury Severity Score (ICISS).

Results

Both the incidence of fall (689 to 636/100000 inhabitants: p = 0.047, coefficient − 4.71) and traffic related injuries (169 to 123/100000 inhabitants: p < 0.0001, coefficient − 5.37) decreased over time while incidence of assault related injuries remained essentially unchanged during the study period. There was an overall decrease in risk-adjusted 30-day mortality in all three groups (OR 1.00; CI95% 0.99–1.00). Decreases in traffic (OR 0.95; 95% CI 0.93 to 0.97) and assault (OR 0.93; 95% CI 0.87 to 0.99) related injuries was significant whereas falls were not during this 11-year period.

Discussion

Risk-adjustment is a good way to use big materials to find epidemiological changes. However after adjusting for age, year, sex and risk we find that a possible factor is left in the pre- and/or in-hospital care.

Conclusions

The decrease in risk-adjusted mortality may suggest changes over time in pre- and/or in-hospital care. A non-significantdecrease in risk-adjusted mortality was registered for falls, which may indicate that low-energy trauma has not benefited for the increased survivability as much as high-energy trauma, ie traffic- and assault related injuries.

Prediction of mortality risk in victims of violent crimes

BACKGROUND

To predict mortality risk in victims of violent crimes based on individual injury diagnoses and other information available in health care registries.

METHODS

Data from the Swedish hospital discharge registry and the cause of death registry were combined to identify 15,000 hospitalisations or prehospital deaths related to violent crimes. The ability of patient characteristics, injury type and severity, and cause of injury to predict death was modelled using conventional, Lasso, or Bayesian logistic regression in a development dataset and evaluated in a validation dataset.

RESULTS

Of 14,470 injury events severe enough to cause death or hospitalization 3.7% (556) died before hospital admission and 0.5% (71) during the hospital stay. The majority (76%) of hospital survivors had minor injury severity and most (67%) were discharged from hospital within 1day. A multivariable model with age, sex, the ICD-10 based injury severity score (ICISS), cause of injury, and major injury region provided predictions with very good discrimination (C-index=0.99) and calibration. Adding information on major injury interactions further improved model performance. Modeling individual injury diagnoses did not improve predictions over the combined ICISS score.

CONCLUSIONS

Mortality risk after violent crimes can be accurately estimated using administrative data. The use of Bayesian regression models provides meaningful risk assessment with more straightforward interpretation of uncertainty of the prediction, potentially also on the individual level. This can aid estimation of incidence trends over time and comparisons of outcome of violent crimes for injury surveillance and in forensic medicine.

Mortality prediction models in the general trauma population: A systematic review

BACKGROUND

Trauma is the leading cause of death in individuals younger than 40 years. There are many different models for predicting patient outcome following trauma. To our knowledge, no comprehensive review has been performed on prognostic models for the general trauma population. Therefore, this review aimed to describe (1) existing mortality prediction models for the general trauma population, (2) the methodological quality and (3) which variables are most relevant for the model prediction of mortality in the general trauma population.

METHODS

An online search was conducted in June 2015 using Embase, Medline, Web of Science, Cinahl, Cochrane, Google Scholar and PubMed. Relevant English peer-reviewed articles that developed, validated or updated mortality prediction models in a general trauma population were included.

RESULTS

A total of 90 articles were included. The cohort sizes ranged from 100 to 1,115,389 patients, with overall mortality rates that ranged from 0.6% to 35%. The Trauma and Injury Severity Score (TRISS) was the most commonly used model. A total of 258 models were described in the articles, of which only 103 models (40%) were externally validated. Cases with missing values were often excluded and discrimination of the different prediction models ranged widely (AUROC between 0.59 and 0.98). The predictors were often included as dichotomized or categorical variables, while continuous variables showed better performance.

CONCLUSION

Researchers are still searching for a better mortality prediction model in the general trauma population. Models should 1) be developed and/or validated using an adequate sample size with sufficient events per predictor variable, 2) use multiple imputation models to address missing values, 3) use the continuous variant of the predictor if available and 4) incorporate all different types of readily available predictors (i.e., physiological variables, anatomical variables, injury cause/mechanism, and demographic variables). Furthermore, while mortality rates are decreasing, it is important to develop models that predict physical, cognitive status, or quality of life to measure quality of care.

Internationally comparable diagnosis-specific survival probabilities for calculation of the ICD-10-based Injury Severity Score

BACKGROUND

The International Statistical Classification of Diseases, 10th Revision (ICD-10)-based Injury Severity Score (ICISS) performs well but requires diagnosis-specific survival probabilities (DSPs), which are empirically derived, for its calculation. The objective was to examine if DSPs based on data pooled from several countries could increase accuracy, precision, utility, and international comparability of DSPs and ICISS.

METHODS

Australia, Argentina, Austria, Canada, Denmark, New Zealand, and Sweden provided ICD-10-coded injury hospital discharge data, including in-hospital mortality status. Data from the seven countries were pooled using four different methods to create an international collaborative effort ICISS (ICE-ICISS). The ability of the ICISS to predict mortality using the country-specific DSPs and the pooled DSPs was estimated and compared.

RESULTS

The pooled DSPs were based on a total of 3,966,550 observations of injury diagnoses from the seven countries. The proportion of injury diagnoses having at least 100 discharges to calculate the DSP varied from 12% to 48% in the country-specific data set and was 66% in the pooled data set. When compared with using a country's own DSPs for ICISS calculation, the pooled DSPs resulted in somewhat reduced discrimination in predicting mortality (difference in c statistic varied from 0.006 to 0.04). Calibration was generally good when the predicted mortality risk was less than 20%. When Danish and Swedish data were used, ICISS was combined with age and sex in a logistic regression model to predict in-hospital mortality. Including age and sex improved both discrimination and calibration substantially, and the differences from using country-specific or pooled DSPs were minor.

CONCLUSION

Pooling data from seven countries generated empirically derived DSPs. These pooled DSPs facilitate international comparisons and enables the use of ICISS in all settings where ICD-10 hospital discharge diagnoses are available. The modest reduction in performance of the ICE-ICISS compared with the country-specific scores is unlikely to outweigh the benefit of internationally comparable Injury Severity Scores possible with pooled data.

LEVEL OF EVIDENCE

Prognostic and epidemiological study, level III.

Prehospital injury deaths--strengthening the case for prevention: nationwide cohort study

BACKGROUND

To determine the frequency and characteristics of prehospital deaths compared with hospital deaths in different subpopulations with severe injuries.

METHODS

Population-based cohort study using person-based linkage of the Swedish nationwide hospital discharge register with death certificate data. In all, 28,715 injury deaths were identified among 419,137 cases of severe injury during 1998 to 2004. Prehospital deaths were defined as autopsied out-of-hospital deaths with injury as the underlying cause. Their impact on mortality prediction was assessed using the International Classification of Disease Injury Severity Score with the C statistic as a measure of discrimination.

RESULTS

The majority of all injury deaths occurred either at the scene or before hospitalization. Among persons younger than 65 years, for each hospital death there were nine prehospital deaths. A high proportion of deaths from drowning, suffocation, and firearm injuries were prehospital (85, 82, and 67% of all cases, respectively). More than 90% of hospital deaths resulted from unintentional injuries, while only 43% of prehospital deaths were unintentional. The largest increase in a cause-specific case fatality risk estimate was seen for poisoning, where inclusion of prehospital deaths increased the risk estimate from 1.6% to 22.8%. Injury mortality prediction based on International Classification of Disease Injury Severity Score improved when prehospital deaths were added to hospital data (C statistic increased from 0.86 to 0.93).

CONCLUSIONS

Prehospital deaths constitute the majority of trauma deaths and differ in major characteristics from hospital deaths. The high proportion of prehospital deaths among young and middle aged people highlights the potential impact of preventive efforts.

LEVEL OF EVIDENCE

III.

Prehospital Trauma Life Support (PHTLS) training of ambulance caregivers and impact on survival of trauma victims

BACKGROUND

The Prehospital Trauma Life Support (PHTLS) course has been widely implemented and approximately half a million prehospital caregivers in over 50 countries have taken this course. Still, the effect on injury outcome remains to be established. The objective of this study was to investigate the association between PHTLS training of ambulance crew members and the mortality in trauma patients.

METHODS

A population-based observational study of 2830 injured patients, who either died or were hospitalized for more than 24 h, was performed during gradual implementation of PHTLS in Uppsala County in Sweden between 1998 and 2004. Prehospital patient records were linked to hospital-discharge records, cause-of-death records, and information on PHTLS training and the educational level of ambulance crews. The main outcome measure was death, on scene or in hospital.

RESULTS

Adjusting for multiple potential confounders, PHTLS training appeared to be associated with a reduction in mortality, but the precision of this estimate was poor (odds ratio, 0.71; 95% confidence interval, 0.42-1.19). The mortality risk was 4.7% (36/763) without PHTLS training and 4.5% (94/2067) with PHTLS training. The predicted absolute risk reduction is estimated to correspond to 0.5 lives saved annually per 100,000 population with PHTLS fully implemented.

CONCLUSIONS

PHTLS training of ambulance crew members may be associated with reduced mortality in trauma patients, but the precision in this estimate was low due to the overall low mortality. While there may be a relative risk reduction, the predicted absolute risk reduction in this population was low.

Battlefield resuscitation

PURPOSE OF REVIEW

To bring together in one review article, the most current and relevant evidence relating to military trauma resuscitation.

RECENT FINDINGS

The main themes highlighted by this review are coagulopathy of trauma shock (CoTS), damage control resuscitation, haemostatic resuscitation, the management of massive transfusion, use of adjuvant drugs for haemostasis and use of an empiric massive transfusion protocol.

SUMMARY

The review aims to educate the readership in recent advances in trauma practice, culminating in a novel empiric massive transfusion algorithm seamlessly guiding the clinician through the initial resuscitation stage resulting in reduced mortality, morbidity, coagulopathy and decreased overall blood product usage.