In a recent paper (I)*, the author and colleagues attempted to bring data from three general source areas to form a unified picture of belt restraint performance. Although the picture is still fuzzy, a prominent pattern can be discerned. In the interest of brevity of the literature, the entire volume of the former paper will not be repeated here;​ rather a brief review of the major points is presented. The abstract of that paper stated:​

"Belt restraint systems have proven effective in the prevention of injuries associated with automobile accidents. Properly worn belts can significantly improve the odds of survival in a crash. Recent accident studies in the USA and Europe, together with biomechanical test results, indicate that conventional lap/shoulder belt systems are most effective in low-to-moderate crash severities, but are somewhat ineffective in frontal collisions more severe than a 30 mph fixed-barrier impact. The primary benefit of belt restraints in side and rollover collision modes is the prevention of ejection. However, other means of preventing ejection (e.g., improved door locks, glazing retention, etc. are also proving to be effective. Extrapolation of available data from the Australian mandatory belt use legislation suggests that U. S. fatalities would be reduced by less than 40 percent, even with 100 percent lap/shoulder belt use."