The Impact of Michigan's Partial Repeal of the Universal Motorcycle Helmet Law on Helmet Use, Fatalities, and Head Injuries.

OBJECTIVES: To evaluate the impact of the partial repeal of Michigan's universal motorcycle helmet law on helmet use, fatalities, and head injuries. METHODS: We compared helmet use rates and motorcycle crash fatality risk for the 12 months before and after the April 13, 2012, repeal with a statewide police-reported crash data set. We linked police-reported crashes to injured riders in a statewide trauma registry. We compared head injury before and after the repeal. Regression examined the effect of helmet use on fatality and head injury risk. RESULTS: Helmet use decreased in crash (93.2% vs 70.8%; P < .001) and trauma data (91.1% vs 66.2%; P < .001) after the repeal. Although fatalities did not change overall (3.3% vs 3.2%; P = .87), head injuries (43.4% vs 49.6%; P < .05) and neurosurgical intervention increased (3.7% vs 6.5%; P < .05). Male gender (adjusted odds ratio [AOR] = 1.65), helmet nonuse (AOR = 1.84), alcohol intoxication (AOR = 11.31), intersection crashes (AOR = 1.62), and crashes at higher speed limits (AOR = 1.04) increased fatality risk. Helmet nonuse (AOR = 2.31) and alcohol intoxication (AOR = 2.81) increased odds of head injury. CONCLUSIONS: Michigan's helmet law repeal resulted in a 24% to 27% helmet use decline among riders in crashes and a 14% increase in head injury.

Classification of motor vehicle crash injury severity: A hybrid approach for imbalanced data.

This study aims to classify the injury severity in motor-vehicle crashes with both high accuracy and sensitivity rates. The dataset used in this study contains 297,113 vehicle crashes, obtained from the Michigan Traffic Crash Facts (MTCF) dataset, from 2016-2017. Similar to any other crash dataset, different accident severity classes are not equally represented in MTCF. To account for the imbalanced classes, several techniques have been used, including under-sampling and over-sampling. Using five classification learning models (i.e., Logistic regression, Decision tree, Neural network, Gradient boosting model, and Naïve Bayes classifier), we classify the levels of injury severity and attempt to improve the classification performance by two training-testing methods including Bootstrap aggregation (or bagging) and majority voting. Furthermore, due to the imbalance present in the dataset, we use the geometric mean (G-mean) to evaluate the classification performance. We show that the classification performance is the highest when bagging is used with decision trees, with over-sampling treatment for imbalanced data. The effect of treatments for the imbalanced data is maximized when under-sampling is combined with bagging. In addition to the original five classes of injury severity in the MTCF dataset, we consider two additional classification problems, one with two classes and the other with three classes, to (1) investigate the impact of the number of classes on the performance of classification models, and (2) enable comparing our results with the literature.