Harvard University. Cambridge (United States of America)
November 16th, 1999
Summary:
This dissertation consists of three articles prepared for publication in the scientific literature; accordingly, it is structured in three chapters. At the time of submission, the first chapter had already been published, the second was under review by another journal, and the third had not yet been submitted. This heterogeneous publication status is characteristic of most dissertations defended and approved at the Harvard School of Public Health.
This summary presents the rationale and general conclusions of the dissertation, along with a brief overview of each of the three chapters.
Introduction
Injuries are a major public health problem worldwide. In the United States, injuries are the leading cause of death among individuals aged 1 to 43 years. Approximately 140,000 people die each year in the country as a result of injury, and an additional 2.3 million sustain injuries severe enough to require hospital admission. An estimated 50 million more individuals suffer injuries requiring outpatient treatment. Injuries represent an important source of disability. The societal costs of such injuries in the United States have been estimated to exceed $158 billion, ranking as the third most important contributor to national health care expenditures.
Motor vehicle crashes constitute the leading cause of injury, accounting for approximately one-third of all trauma victims. About half of those injured in car crashes are drivers, making them the predominant victim group. In 1996, 14,879 passenger-vehicle drivers died in the U.S.; another 611,000 sustained serious but nonfatal injuries, and more than 1,000,000 suffered minor injuries. It is estimated that 7% of nonfatal injuries among front-seat occupants (including drivers) result in permanent functional limitations, potentially leading to long-term disability.
Over recent years, multiple studies have characterized these injuries and examined their incidence and prevalence using descriptive and analytical epidemiological methods. Numerous additional investigations have focused on estimating the economic costs associated with such injuries. Other researchers have identified risk factors, while still others have evaluated the effectiveness of various preventive interventions, including mandatory seat belt laws, speed limits, vehicle manufacturing standards, vehicle inspection laws, drunk-driving regulations, and restrictions on driver licensing. By contrast, relatively few studies to date have assessed the efficiency of these preventive measures.
As a result of past research, driving today is considerably safer than it was in the past. The development and implementation of an impressive number of primary, secondary, and tertiary prevention measures have succeeded in reducing injury rates. These declines have occurred despite substantial increases in both the number of vehicles on the road and the total miles traveled. Nevertheless, the problem remains severe, and many organizations have called for the application of rigorous scientific methods to the evaluation of injury prevention.
One of the most controversial preventive interventions for drivers has been the introduction of airbag systems. In 1984, legislation was passed requiring that all vehicles sold in the United States from September 1, 1997, onward be equipped with a driver-side frontal airbag. The United States is the only country in the world where airbags are mandatory. The adoption of these laws, coupled with growing consumer interest in road safety, led to the exceptionally rapid introduction of airbags into the vehicle fleet, often ahead of schedule. Since 1989, more than 60 million passenger vehicles equipped with driver-side airbags have been sold, and by 1999, it was estimated that 47% of the U.S. fleet had such systems.
Driver-side frontal airbags are designed to protect drivers in frontal (or near-frontal) collisions by deploying an inflatable cushion that covers interior vehicle surfaces (steering wheel and dashboard), thereby reducing the probability of injury during a crash. Legislation requires airbags to pass laboratory crash tests in which a 50th-percentile male dummy, unbelted and positioned in the driver’s seat, must not be subjected to forces above defined thresholds for head and chest when the car collides with a rigid barrier at approximately 20 km/h.
A typical airbag consists of sensors (to detect crash severity), a central control unit (to process sensor information and decide whether deployment is warranted), and the airbag itself. However, there are many design variations across manufacturers, models, and production years, leading to differences in deployment thresholds, inflation dynamics, and speed.
Initial projections, based on experimental data, by the U.S. National Highway Traffic Safety Administration (NHTSA) suggested that airbags would reduce the risk of death by 40% for unbelted drivers and by 10% for belted drivers. Government estimates in 1977–1987 suggested airbags could save between 6,000 and 9,000 lives annually. Notably, these projections did not account for the impact of airbags on nonfatal injuries.
More recent real-world crash data have revealed the overly optimistic nature of these early estimates. Statistical analyses of large crash databases and special crash investigations indicate that airbags offer fewer benefits than initially projected and may sometimes be harmful. By mid-1999, 52 drivers had died in moderate crashes due to airbag deployment. At the same time, official estimates of airbag effectiveness in reducing driver fatality risk suggested only a 10% reduction.
Evaluations of the impact of airbags on nonfatal injuries have yielded inconsistent results. The risk of moderate or severe head trauma appears to be significantly reduced. Conversely, the risk of moderate or severe injuries to other body regions appears to be increased, as documented in government surveillance systems, medical reports, and crash investigations. Approximately 40% of deployed airbags cause some form of driver injury, though most are relatively minor.
Several strategies have been proposed to improve airbag performance. Some involve interventions for existing vehicles, while others pertain to future generations of airbags. For example, NHTSA permits certain drivers who meet specific criteria to apply for authorization to disconnect airbags temporarily or permanently. In terms of future developments, NHTSA has allowed manufacturers to introduce less aggressive airbags by reducing deployment speed and/or bag volume. Other proposed strategies include raising deployment thresholds, altering bag design (fabric type, shape, folding pattern, etc.), or installing sensors to prevent deployment when the driver is positioned too close to the airbag module.
The three chapters of this dissertation use driver airbags as a case study to illustrate three critical areas:
1. The need to anticipate unintended negative consequences of preventive design and implementation.
2. The need for continuous monitoring of the effectiveness of adopted interventions.
3. The need to apply advanced public health methodologies to characterize the effectiveness and efficiency of such interventions.
[...]
Keywords: airbags; road traffic injuries; injury prevention; cost-effectiveness analysis; risk perception; effectiveness of interventions
Citation:
M. Seguí-Gómez, "Evaluating the performance of Injury Prevention Efforts. The Case of Driver-Side Airbags", PhD. dissertation, Harvard University, Cambridge, United States of America, 1999.
