Traumatic transfers: calibration is adversely affected when prediction models are transferred between trauma care contexts in India and the United States

Pediatric Thoracolumbar Spinal Injuries in United States Trauma Centers

OBJECTIVES

Injuries are the leading cause of morbidity and mortality in children ages 1 to 18 years. There are limited studies about pediatric thoracolumbar (TL) spinal injuries; the purpose of this study was to characterize TL spinal injuries among pediatric patients evaluated in US trauma centers.

METHODS

This was a retrospective cohort study of the National Trauma Data Bank. Patients aged 1 to 18 years with a thoracic or lumbar spinal injury sustained by blunt trauma during calendar years 2011 through 2016 were included. Cervical spinal injuries, death before arrival, or penetrating trauma were excluded. The data was abstracted, and missing data was addressed by imputations. Data was analyzed using descriptive statistics and multinomial logistic regressions.

RESULTS

A total of 20,062 patients were included in the study. Thoracolumbar spinal injuries were more commonly sustained by 16- to 17-year-olds (45.7%), boys (56.6%), and White (74.8%). The injuries were often from a motor vehicle collision (MVC) (55.2%) and resulted in a bone injury (82.3%). Mechanism of injury and age were significant in predicting injury type. A fall was more likely than MVC to result in disc injury (odds ratio [OR], 1.70; 95% confidence interval [CI], 1.24-2.33), strain injury (OR, 1.18; 95% CI, 1.05-1.34), or cord injury (OR, 1.27; 95% CI, 1.12-1.45). Younger children were more likely than adolescents to present with disc injury (OR, 2.79; 95% CI, 1.75-4.45), cord injury (OR, 1.46; 95% CI, 1.18-1.81), or strain injury (OR, 1.37; 95% CI, 1.09-1.72).

CONCLUSIONS

To our knowledge, this is the largest pediatric TL spinal study. Clinicians should consider TL spinal injuries when adolescents present after an MVC, and specifically, TL spinal cord injuries when young children present after a fall. Additionally, pediatric TL spinal injury prevention should highlight motor vehicle and fall safety.

The advanced machine learner XGBoost did not reduce prehospital trauma mistriage compared with logistic regression: a simulation study

Background

Accurate prehospital trauma triage is crucial for identifying critically injured patients and determining the level of care. In the prehospital setting, time and data are often scarce, limiting the complexity of triage models. The aim of this study was to assess whether, compared with logistic regression, the advanced machine learner XGBoost (eXtreme Gradient Boosting) is associated with reduced prehospital trauma mistriage.

Methods

We conducted a simulation study based on data from the US National Trauma Data Bank (NTDB) and the Swedish Trauma Registry (SweTrau). We used categorized systolic blood pressure, respiratory rate, Glasgow Coma Scale and age as our predictors. The outcome was the difference in under- and overtriage rates between the models for different training dataset sizes.

Results

We used data from 813,567 patients in the NTDB and 30,577 patients in SweTrau. In SweTrau, the smallest training set of 10 events per free parameter was sufficient for model development. XGBoost achieved undertriage rates in the range of 0.314–0.324 with corresponding overtriage rates of 0.319–0.322. Logistic regression achieved undertriage rates ranging from 0.312 to 0.321 with associated overtriage rates ranging from 0.321 to 0.323. In NTDB, XGBoost required the largest training set size of 1000 events per free parameter to achieve robust results, whereas logistic regression achieved stable performance from a training set size of 25 events per free parameter. For the training set size of 1000 events per free parameter, XGBoost obtained an undertriage rate of 0.406 with an overtriage of 0.463. For logistic regression, the corresponding undertriage was 0.395 with an overtriage of 0.468.

Conclusion

The under- and overtriage rates associated with the advanced machine learner XGBoost were similar to the rates associated with logistic regression regardless of sample size, but XGBoost required larger training sets to obtain robust results. We do not recommend using XGBoost over logistic regression in this context when predictors are few and categorical.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-021-01558-y.

The transfer of clinical prediction models for early trauma care had uncertain effects on mistriage

OBJECTIVES

This study aimed to assess how transfers of clinical prediction models for early trauma care between different care contexts within a single health system affected mistriage rates.

STUDY DESIGN AND SETTING

Patients aged 15 years or older, registered between 2011 and 2016 in the Swedish national trauma registry, SweTrau, were included. Three data set groups were created: high- and low-volume centers, metropolitan and nonmetropolitan centers, and multicenters and single centers. Clinical prediction models were developed using logistic regression in each data set group and transferred between data sets within groups. Model performance was evaluated using mistriage rate, undertriage rate, and overtriage rate. Multiple imputation using chained equations was used to handle missing data. Model performance was reported as medians with 95% confidence intervals (CIs).

RESULTS

A total of 26,965 patients were included. Changes in mistriage rates after transfer ranged from -0.25 (95% CI -0.21 to 0.04) to 0.29 (95% CI 0.13-0.39). Both overtriage and undertriage rates were affected.

CONCLUSIONS

Transferring clinical prediction models for early trauma care is associated with substantial uncertainty in regards to the effect on model performance. Depending on the care context, model transfer led to either increased or decreased mistriage. Overtriage was more affected by model transfer than undertriage.

The clinical usefulness of prognostic prediction models in critical illness

Critical illness is any immediately life-threatening disease or trauma and results in several million deaths globally every year. Responsive hospital systems for managing critical illness include quick and accurate identification of the critically ill patients. Prognostic prediction models are widely used for this aim. To be clinically useful, a model should have good predictive performance, often measured using discrimination and calibration. This is not sufficient though: a model also needs to be tested in the setting where it will be used, it should be user-friendly and should guide decision making and actions. The clinical usefulness and impact on patient outcomes of prediction models has not been greatly studied. The focus of research should shift from attempts to optimise the precision of models to real-world intervention studies to compare the performance of models and their impacts on outcomes.