EXECUTIVE SUMMARY This study was conducted for the National Highway Traffic Safety Administration (NHTSA). The purpose of this study was to obtain a measure of the current level of misuse of child restraint systems (CRSs) among the general public. The project focused specifically onforms of misuse that can be expected to raise the risk of injury. RESEARCH METHODOLOGY AND DATA COLLECTION Key research task activities included: (1) conducting a workshop with National experts on CRS misuse, injury prevention, and crash data to develop appropriate critical misuse measures; (2) selecting State sites and State site coordinators (SSCs) in six States; (3) holding a train-the-trainer workshop with SSCs and their field site managers (FSMs) to finalize CRS misuse definitions and data collection instruments, and to discuss recruitment and training techniques; (4) conducting field observations; (5) conducting data entry and analysis; and (6) preparing the final report. A workshop was held within five months of the project start date with leading medical, bioengineering, and injury prevention experts in the field. The workshop included discussions on types of child injury severity in crashes; identification of the types of CRS misuse and their relationship to serious injury; and identification of the most important CRS misuse measures to include in data collection. Areas identified for critical CRS misuse measures in the workshop were: age and weight appropriateness of CRS; direction of CRS; placement of CRS in relation to air bags; installation of CRS to the vehicle seat; secureness/tightness of harness straps and crotch strap of the CRS; secureness/tightness of the vehicle safety belt (SB) to the CRS; locking clip use for vehicle SBs; fit of SBs across the children in belt-positioning booster seats; and damaged CRSs. Six States (Arizona, Florida, Mississippi, Missouri, Pennsylvania, and Washington) were selected as data collection sites. States were selected based on their representing diverse geographic regions across the country; and encompassing diversity in socio-demographic and economic characteristics across field sites. The field sites included urban, suburban, and rural areas. It was critical for each State to have a qualified and experienced State Site Coordinator (SSC) who was familiar with managing field observations and could easily gain permission to conduct observations at sites in their area. State regional areas included Tucson (AZ); Boca Raton, Fort Myers, Fort Pierce, and Miami (FL); Jackson (MS); St. Louis (MO); Carlisle, and Harrisburg (PA); and Seattle and Tacoma (WA). Field observation sites were primarily parking areas at community shopping centers, child merchandise department stores, fast food restaurants, health and medical facilities, and community events. A train-the-trainer workshop was held with SSCs and FSMs approximately one month before data collection to finalize CRS misuse definitions and data collection instruments. The workshop also covered recruitment of greeters and field observation staff; training techniques; data collection procedures; and project administration details. Most sites conducted recruitment and training of staff in September 2002. It was required that all field observers be AAA certified CPS technicians. Data were collected from late September 2002 to early January 2003. Most of the States completed their data collection efforts by mid-November 2002. Project staff checked quality of data on a daily basis for the first two weeks and then weekly. Data entry and analysis were then conducted. A summary of the results was presented to NHTSA before the completion of the draft final report. RESULTS A total of 4,126 vehicles and 5,527 children weighing less than the driver-estimated weight of 80 lb were included in the study. There were 511 children weighing less than 20 lb; 2,483 children weighing 20 to 39 lb; and 2,533 children weighing 40 to 79 lb. By age category there were 676 children less than 1 year of age; 2,021 children from the ages of 1 through 3; 2,571 children from ages 4 through 8; and 259 children ages 9 and older. For the total sample of children weighing less than 80 lb, 62.3% were restrained in CRSs, 25.9% were restrained in SBs, and 11.8% were unrestrained. CRS use by weight classification was as follows: 97.1% for children weighing less than 20 lb; 86.4% for children weighing 20 to 39 lb; 41.7% for children weighing 40 to 59 lb; and 10.9% for children weighing 60 to 79 lb. Safety belt (SB) use increased markedly for children 40 lb and over. SB use was 43.1% for children weighing 40 to 59 lb; and 64.9% for children weighing 60 to 79 lb. By age category, CRS use was 97.3% for children younger than 1 year, 90% for children ages 1 through 3, 37.2% for children ages 4 through 8, and 3.1 percent for children ages 9 and older. SB use increased markedly for children 4 through 8 years of age to 45.5%. One or more critical misuses were found in 72.6% of all CRSs observed. Percentages of CRSs with critical misuses by seat type were as follows: infant (83.9%); rear-facing convertible (83.5%); forward-facing convertible (81.9%); forward-facing only (79.3%); belt-positioning booster (39.5%); and shield booster (60.5%). The most common critical misuses were loose harness straps securing the child to the CRS and loose SB securing the CRS to the vehicle. Harness retainer clip misuse was also prevalent, but not deemed as a critical misuse in this study. Other CRS misuse problem areas were also observed. They included visible damage to the CRS (e.g., cracked seat shell, torn harness strap, broken harness parts); percent of CRS base (bottom of seat) contacting the vehicle seat; and presence of aftermarket devices. Forty-two observations of LATCH (Lower Anchors and Tethers for CHildren) were noted. Field observers found three seats with improperly used lower anchors, and three seats with improperly used tethers. Six seats were observed with both a safety belt and the lower anchors in use. Children being transported by drivers who are restrained in SBs are more likely to be restrained than children who are being transported by unrestrained drivers. Of the observed drivers, 77.4% were restrained. When drivers were wearing a SB, 91.7% of children were restrained in either a CRS (44.2%) or SB (47.5%). When drivers were not restrained, only 62.3% of children were restrained in either a CRS (43.2%) or SB (19.1%). Air bag systems were also observed. Driver frontal air bags were noted in 83.1% of the vehicles; passenger frontal air bags were found in 71.8% of the vehicles. Side air bags protecting passengers in the front seat were observed in 4.6% of the vehicles, and side air bags protecting passengers in both the front and rear seats were observed in 1.3% of the vehicles. Only a small percentage of the vehicles (less than 5%) had on/off air bag switches. For 88 vehicles checked for status of this on/off switch, 51 vehicles had the switch "on" and 37 had the switch "off." Sixteen children were in the front passenger seat with an activated ("on") air bag switch; 2 of these children were in a rear-facing CRS, 4 were in a forward-facing seat, 3 were in a SB, and 7 were unrestrained. RECOMMENDATIONS Periodic monitoring of CRS misuse among the general public is recommended because of continuous upgrades and design changes to vehicle occupant protection systems and CRS models (e.g., LATCH systems, side air bags); frequent changes (strengthening) of occupant protection laws in States; and a perpetual supply of new parents responsible for protecting child passengers in vehicles. CRS misuse observations with the general public should be made by qualified and experienced personnel who are trained to gather CRS misuse data. Train-the-trainer workshops for supervisors, field managers and senior field observers prepare staff for field observations. At least two full days should be spent training field observers. Continued enforcement of CPS laws is recommended, and there is general public acceptance for the enforcement of these laws. Programs need to be developed to make law enforcement agencies aware of the importance of correct CRS use, as well as keeping children in CRSs for as long as possible. Enforcement strategies need to be developed to identify booster seat law violations. Law enforcement should also continue to take an active role in communitybased CRS education programs. Education programs should continue to promote proper use of CRSs, enforcement of laws, availability of CRS inspection stations, loaner programs, and local and National CRS hotline information assistance. Programs should also provide information about LATCH systems and correct usage of the LATCH system. 1.0 INTRODUCTION This chapter presents background information on the purpose of this study; characteristics of child restraint systems (CRSs); observational studies of CRS use and misuse; injuries associated with types of CRS misuse; and field observation techniques. 1.1 BACKGROUND The number of young child passengers in vehicles who are either improperly placed in child restraint systems (CRSs) or moved into adult vehicle safety belts (SBs) prematurely is at an alarmingly high rate (Spurlock, Kidd, Mays, McCool, Buckner, Clatos, Rochussen, and Leach, 1998; Taft, Mickalide, and Taft, 1999; Kohn, Chausmer, and Flood 2000; and Morris, Arbogast, Durbin, and Winston 2000). In the mid-1990s, a National CRS misuse study involving random observation checks of the public found 79.5 percent CRS misuse (Decina and Knoebel, 1997). Current CRS fitting station studies, which involve more of a pre-selected sample of volunteer and safety conscious parents/guardians, report over 90 percent misuse (NHTSA, 2001). Current research also focuses on the injury patterns associated with types of CRS misuse, especially the premature graduation of children to either booster seats or SBs. In general, these studies (Kelleher-Walsh, Walsh, and Duffy, 1995; National Transportation Safety Board, 1996; Winston, Durbin, Kallan, and Moll, 2000; and Morris, Arbogast, Durbin, and Winston, 2000) have shown that head and facial injuries are predominant regardless of impact point or seat position. There is also a high risk of abdominal injury associated with improper booster seat use or premature graduation to a SB. For children restrained in CRSs, there is a high frequency of shoulder injuries related to harness misuse. In addition, spinal cord injuries result from infants being placed in the forward direction in a CRS. As expected, there is also a significantly increased likelihood (3 times) of serious injury among unrestrained children who are involved in a crash. Federal, State, and local governments, as well as health care providers and community safety outreach programs (e.g., SafeKids Coalitions) have made a tireless effort to educate the public on proper CRS use. National and State campaigns, as well as local programs incorporating child safety seat inspection stations, are in progress. However, CRSs and vehicle restraint systems can be complicated. Dozens of CRS makes/models exist. A different CRS may be needed for each early life stage. Many CRSs cannot fit securely in certain vehicle seats; nor can certain vehicle safety belts tighten enough to properly secure CRSs to the vehicle seat. In addition, CRS technology continues to evolve (e.g., LATCH system). To complicate this issue even more, there is always a continuous stream of new parents/guardians who need to be educated on each type of CRS (i.e., infant, convertible, and booster); and there are some economic constraints. To address these concerns, it is important for NHTSA to periodically monitor the status of CRS misuse in the Nation. This study sought to focus specifically on CRS misuse measures that have the most practical consequence in terms of the risk of injury to a child when involved in a motor vehicle crash. Characteristics of Child Restraint Systems (CRSs) Child restraint system designs vary according to the size of the child they are designed to restrain, the direction the child should face, the type of internal restraining system, and the method of installation. CRSs are designed for coupling the CRS securely to the vehicle seat using the vehicle safety belt (SB) or LATCH (lower anchors and tethers for children) system if available; and properly securing the child in the CRS with a separate harness and/or other restraining surface. Securing these two links between the vehicle and the child is critical in reducing injuries or death to a child in the event of a vehicle crash (Weber, 2000). There are five basic types of CRSs in current use; infant seats, convertible seats (converts from rear-facing to forward-facing), forward-facing only seats, booster seats, and integrated (built-in) seats. Other less common CRSs are also in use, including the Laptop car seat for children of booster-seat weight; car beds for newborns and other very small infants; harness vests for toddlers and older children; and restraint systems for children with special needs. Characteristics of these CRSs are described below. The infant seat is primarily designed for children birth to 20 lb. Some infant seats haveupper weight limits that range from 17 to 22 lb. Infant seats are recommended until the child is at least 1 year old and at least 20 lb. Infant seats are typically one-piece, protective molded shells. They are designed for a rear-facing installation only. The seat comes equipped with snap-in pads and slots for the vehicle safety belts (lap or lap portion of lap/shoulder belt). The infant is secured in the CRS with a harness, and, in some cases, a harness retainer (chest) clip to hold the shoulder harness together for correct pre-crash positioning. Infant seats include two to three sets of slots in the back of the seat to allow for harness adjustment to accommodate an increase in the infant"s size. Harness slots should be at or below shoulder level. These seats have either a three-point harness that consists of two straps over the shoulder connecting in a "V" shape at the buckle or to a small hip pad that attaches to the buckle, or a five-point harness that also has straps coming around the hip. The angle of these seats should never be more than 45 degrees from the vertical position. The seats can be anchored in place with a vehicle safety belt or LATCH attachments. Rear-facing child restraints are not required to have top tethers (Weber, 2000; NHTSA, 2001; Stewart and Kern, 2003). The convertible seat is designed for children from birth to 40 lb. The seat incorporates features to allow use for infants as well as toddlers. In the rear-facing position, it is used until the child is at least 1 year old and at least 20 lb. Some convertible seats are approved for rearfacing use up to 30 or 35 lb. For infants, the top of the child"s head should be well contained within the seat"s shell (no less than one inch from the top of the shell). The purpose of the harness system in rear-facing seats is to keep the infant"s body, neck, and head contained within the shell. For infants, the harness slots should be at or below shoulder level (Weber, 2000; and NHTSA, 2001). In the forward-facing position, the convertible seat carries the child until 40 lb and approximately 3 to 4 years of age. Convertible seats have either a five-point harness, three-point harness with T-shield combination, or a three-point harness with tray shield combination. The five-point harness system has straps that secure at both shoulders, across the upper thighs, and between the legs. It can be adjusted to fit a variety of toddlers snugly and correctly. Harness/Tshield combinations secure two shoulder harness straps to a "T"- shaped shield or to a broader padded tray positioned in front of the torso and hips. The harness/tray shield combination has a crotch strap for the tray shield. It may be separate or be integrated with the shield. Convertible seat harness systems can be adjusted as the child grows. There are harness slots that accommodate changing the harness straps from the lowest slot position for an infant to the upper slots for toddlers. Most of the harness systems require a harness retainer (chest) clip, placed atthe armpit level of the child. The clip is used as a pre-crash positioner holding the harness straps together for correct positioning in the event of a crash (Weber, 2000; and NHTSA, 2001). Convertible seats also have a reclining mechanism, allowing an infant to sleep at an acceptable reclining angle (not more than 45 degrees) while facing the rear, and a toddler to sit more upright while facing forward. This seat is anchored in place with a vehicle safety belt or LATCH attachments (Weber, 2000 and NHTSA, 2001). The forward-facing only seats are used for children weighing between 20 and 40 lb; and usually at least 1 to 4 years of age. Some models can be used for children up to 60 and 70 lb; especially those used as interchangeable booster seats. This would accommodate children through approximately 7 years of age. The harness systems are either five-point harnesses or overhead shield restraints. For these seats, the height of the shoulder strap is usually above the child"s shoulders to effectively limit head excursion, and the height of the seat back should be above the child"s ear to protect against rearward bending. Some models require a harness retainer (chest) clip, placed at the armpit level of the child to hold the straps in place. The seat is anchored in place with a vehicle"s safety belt or LATCH attachments. Models after 1999 are equipped with top tether straps to be anchored rearward from the seat. A combination child seat/booster used with an internal harness is also manufactured. It is used with an internal harness for child securement with children up to 70 lb. Some of these models can have the internal harnesses removed and thus the seat can be transformed into a belt-positioning booster (BPB) for children weighing over 40 lb. Combination forward-facing child restraints and BPBs must have LATCH attachment systems (Weber, 2000; NHTSA, 2001; Stewart and Kern, 2003). Children in a minimum recommended weight range of 30 to 40 lb and a maximum weight range of 60 to 100 lb should be in a booster seat. This includes most 4 to 8 year olds. Booster seats provide the transition from child seats with internal harness to vehicle lap/shoulder belts. These seats are anchored in place with a vehicle"s safety belt system. Booster seats are not required to have LATCH attachment systems. They are not restraint systems by themselves, but rather positioning devices that depend entirely on the vehicle safety belts to hold the child and booster seat in place (Weber, 2000; NHTSA, 2001; Stewart and Kern, 2003). There are three types of booster seats: belt-positioning; high-back belt-positioning; and shield booster. A belt-positioning booster raises the height of the child"s body in a vehicle to allow a more secure safety belt fit across the child"s torso and hips. Some of these seats are combination child seat/booster, while others come without harness and function only as backless beltpositioning boosters. Most of these types of booster seats have small handles or guides under which the lap belt and the lower end of the shoulder belt are routed. Some seats only have depressions or slots for the belt path. The guides function like a crotch strap, holding the lap belt low and flat across the child"s upper thighs, while the inboard guide also pulls the shoulder belt toward the child and makes its angle more vertical, so that the belt crosses the center of the child"s chest (Weber, 2000). High-back belt-positioning boosters can come with a removable 5- point harness system (to be used as a forward-facing child seat up to 40 lb) or come as just a vehicle lap/shoulder belt positioning device. Both provide support for the child"s head and neck and help avoid whiplash injuries. Many high back boosters have a comfort clip or shoulder belt positioning strap on the side of the seat back (NHTSA, 2001) Shield boosters are still in use today, although they are no longer being manufactured. They are designed to be used in seating positions with only a lap belt (pre-1996 vehicles); and allow for the shields to be used when a child weighs only between 30 and 40 lb. The shields can also be removed and the restraint used as a backless belt positioning booster for children who weigh between 30 and 60 lb, if both the lap and shoulder belts are available. Also, most models require the lap belt to be wrapped around the shield. In one case, the lap belt goes through thebase of the seat (NHTSA, 2001). Integrated (built-in) restraints anchor directly to the vehicle seat. Some are used with harnesses for children weighing up to 40 or 60 lb. Some use five-point harnesses; others are used as belt-positioning booster seats; and some can be used in either mode depending upon thesize of the child. Integrated seats may not be used for rear-facing infants (Weber, 2000 and NHTSA, 2001). The Laptop car seat is an energy absorbing child restraint. The device provides an alternative for those children over 40 lb that still can not sit correctly using a belt- positioning booster seat. It can be used for those children who need booster seats but are being transported in vehicles without shoulder belts in the back seat. It looks like a shield without a boosting base, but it is designed to fit snugly on the child"s thighs and abdomen. The Laptop can be used with either a lap belt or a lap/shoulder belt combination. The Laptop is placed over a child already sitting in the vehicle seat. The vehicle safety belt is then threaded through the grooves in front of the laptop. The laptop can be pushed down to get a tighter fit on the child. The seat does not have head support and should be used in a seating position where the top of the child"s ears are below the top of the vehicle seat back (NHTSA, 2001). Car beds are usually for small, premature, or medically fragile infants who should ride prone or supine. The infant lies flat. The vehicle safety belt is used to anchor the car bed perpendicular to the direction of travel. The infant"s head is placed toward the center of the vehicle and not next to the door. An internal harness secures the child in the car bed (NHTSA, 2001). Harness/safety vests, often called travel vests, have a rigid back for attachment of the vehicle belt and use a five-point harness to distribute crash forces across a child"s body. Although these vests differ in appearance and function from most child restraints, they meet federal law requirements. Most travel vests are for children who weigh 25 to 40 lb. They are often used on school buses to restrain children. Some require the use of a tether in conjunction with a vehicle lap belt for securement (NHTSA, 2001). Special-needs restraint systems are used for infants, toddlers, and older children who can not be accommodated by conventional child restraint systems because of respiratory, orthopedic (hip spica and full body casts), neuromuscular, and/or behavioral conditions. These systems are also used for low birth weight and premature infants who weigh less than 5 lb. Car beds and harness vests are included in this category. These systems are secured to a vehicle safety belt system. (Talty, Sheese, Gunn, Stone, Chappelow, Wyatt, Cox, and Bull, 2000 and NHTSA, 2001). Observational Studies of CRS Use and Misuse Extensive observations of CRS use have been documented since the 1970s (Williams, 1976). NHTSA began conducting observational studies in 19 cities as part of periodic observation of child restraint, safety belt, and motorcycle helmet use in the early 1980s; and continued until 1991 (NHTSA, 1991). Other CRS use studies in the 1980s and 1990s were conducted in the following States or provinces: Texas (Hatfield, et al., 1986 and Womack, 1992); Michigan (Streff and Molnar, 1990); Virginia (Stoke, 1992); Ontario, Canada (Canada Market Research Ltd., 1992); and Pennsylvania (Decina, Temple, and Dorer, 1994). CRS use data were collected in four States as part of the NHTSA CRS misuse study in the mid-1990s as well (Decina and Knoebel, 1996). The most recent large-scale CRS use observation studies were conducted in 2000 and 2002 as part of the NHTSA National Occupant Protection Use Survey (NOPUS) (Glassbrenner, 2003). By the mid- to late-1980s CRS studies collected data on the type of CRS (i.e., infant, toddler, and booster) and associated types of misuse errors. These were primarily out-of-vehicle observations (Bulger, 1983; Cynecki and Goryl, 1984; Shelness, 1984; Bull, Stroup, and Gerhart, 1988; and Streff and Molnar, 1990). CRS misuse data were often gathered by having data collectors peer into windows of vehicles stopped at signalized intersections. Other data collection techniques involved walking through shopping centers and peering into unoccupied vehicles and observing the misuse status of empty CRSs. When researchers started to review the findings across studies, it became evident that it was difficult to compare CRS misuse rates to determine "National" rates because researchers had used different data collection techniques, and their own definitions for CRS misuse errors. In the early 1990s, the need to collect more accurate CRS misuse data was realized. Government agencies provided clearance for researchers to use more intrusive techniques to get a closer look at child occupants in vehicles. Field observers were entering vehicles (parked) and making close-up observations of children in CRSs. From these "in-the-vehicle" observations, researchers had the opportunity to identify such CRS characteristics as type/model, looseness of harness and vehicle safety belt systems, and other types of CRS misuse not easily detected from outside the vehicle. These studies provided richer data on the types of misuse errors by the types of CRS components (e.g., safety belt connection, locking clip, harness/shield, chest clip, tether strap, etc.) (Margolis, Wagenaar, and Molnar, 1992; Womack, 1992; Canada Market Research Ltd., 1992; Decina, Temple, Dorer, 1994 a,b; Frank and Ascheim, 1996; Taft, Mickalide, and Taft, 1999). By the mid 1990s, observational studies also included more interaction with targeted drivers (Decina and Knoebel, 1996; Eby and Kostyniuk, 1999). These Federal and State-funded studies focused not only on collecting extensive CRS misuse information (i.e., CRS make/model, misuse errors by each CRS component) and vehicle characteristics (i.e., vehicle make, model, restraint systems), but also interviewing parents and other drivers to determine knowledge, attitudes and beliefs regarding child passenger protection, acquisition of CRSs (e.g., purchase, gift, loan), frequency of installing and moving CRSs, and other factors that could possibly influence proper use of CRSs. Results of NHTSA"s most comprehensive CRS misuse observation effort in the mid 1990s showed that there was very high CRS misuse overall (80%); and that certain types of misuse errors (e.g., locking clips, harness retainer clip, harness strap) were more common than others (Decina and Knoebel, 1996, 1997). These findings confirmed patterns of misuse that had been reported in earlier studies (Bull, et al., 1988 and Margolis, et al., 1992). By the end of the 1990s, States and their local communities began CRS inspection/fitting stations and clinics. These programs provided individuals with the opportunity to go to a location (e.g., hospital, State police barracks, car dealership, etc.) to learn whether they were putting their children in CRSs correctly and properly installing CRSs in their vehicles. Some CRS educators and researchers used these events as good opportunities to collect CRS misuse data in a setting that would provide ample time to talk with parents/drivers, and the ability to thoroughly check all potential misuse errors (SafetyBeltSafe USA, 1994; Spurlock, Kidd, Mays, McCool, Buckner, Clatos, Rochussen, and Leach, 1998; Kohn, Chausmer, and Flood, 2000; and Morris, Arbogast, Durbin, and Winston, 2000). Results of these studies and other checkup events held across the Nation have shown higher CRS misuse rates (90% to 98%) than traditional observation studies, which involve randomly selected drivers with young passengers who have no prior knowledge that they will be stopped and asked to participate in a safety check of their CRS usage (NHTSA, 2001). Injuries Associated with Types of CRS Misuse By the mid 1990s, CRS research not only focused on determining the reasons why young children were not being restrained properly in CRSs (Decina and Knoebel, 1996,1997); and the reasons why children were prematurely moving into safety belts or riding unrestrained; but also on assessing the performance of restraint systems and identifying injury patterns associated with young children involved in motor vehicle crashes (Kelleher-Walsh, Walsh, and Duffy, 1995; National Transportation Safety Board, 1996; Winston, Durbin, Kallan, and Moll, 2000; Winston, Arbogast, Lee, and Menon, 2000). Kelleher-Walsh, et al. (1995) focused on injuries observed in young children (ages 5 and younger) who had been involved in automobile crashes while restrained in CRSs. They created a child injury database containing information on 371 children who sustained a total of 601 injuries. The study was a retrospective case review of medical and police records on child injuries caused in motor vehicle crashes from 1986 through 1990 in western counties of New York. The database provided an increased understanding of child impact injury mechanisms and an increased knowledge of the characteristics of injuries (i.e., body regions, type of injury, severity levels) that have been suffered by children improperly restrained in CRSs. The study revealed that head and facial injuries were predominant, regardless of impact direction and seating position. A comparison of toddler abdominal injury and CRS design (forward-facing harness or booster/shield) indicated that abdominal injury might be associated to a higher degree with booster/shield type restraints than to harness-type restraints. Comparisons of toddler shoulder injuries and CRS design showed that shoulder injuries may be associated to a higher degree with toddler harness type restraints than to booster/shield-type restraints. Injuries to children restrained by rear-facing infant seats included a high percentage of facial injuries in frontal impacts, even though the infant"s back is toward the principal direction of force in this crash mode. No injuries to the neck, abdomen, or thorax of infants in rear-facing infant restraints were found regardless of impact direction. In the mid-1990s, the National Transportation Safety Board (NTSB) conducted a study to examine the performance and use of occupant protection systems (i.e., CRSs, vehicle SBs, andair bags) for children. They investigated 133 crash incidents in which the vehicle was transporting a child passenger younger than age 11 and in which at least one occupant was admitted to the hospital. The study examined the severity of injuries to children restrained in CRSs and vehicle SBs, including the effects of CRS misuse, crash severity and seating position. Analysis was performed on 207 child passengers, of whom 52 were restrained in CRSs, 80 were restrained in SBs, and 75 were unrestrained. Thirty percent of the children sustained moderate to- severe injuries, or were fatally injured. Many children sustained injuries to the head and face. NTSB found a higher likelihood of severe injury for those children in the front seat, regardless of the presence of a passenger frontal air bag. The analysis also examined the injury severity level by type of CRS misuse error, and there was a greater likelihood of more severe injuries and fatalities when young children were not in the appropriate type of CRS for their age and weight (NTSB, 1996). In the late 1990s, Partners for Child Passenger Safety (PCPS) developed a child-focused crash surveillance system based on a representative sample of children from birth to 15 years of age who were involved in crashes that were reported to State Farm Insurance Companies in 15 States and the District of Columbia. Analyses of the PCPS data showed that many young children were inappropriately graduating from their CRS to a safety belt, putting these children at higher risk for intestinal, liver, spleen and spinal cord injury. In addition, a large number of infants were incorrectly turned to face forward before one year of age, increasing the risk of spinal injury. It was also reported that many children less than age 12 still continue to ride in the vehicle front seat, increasing their risk of injury caused by air bag deployment (Winston, Durbin, Kallan, and Moll, 2000). Other PCPS results showed that there is still a high level of CRS misuse (82%). The common mistakes include failing to attach the seat tightly to the vehicle, failing to fasten the harness tightly around the child, and using the chest clip incorrectly. The PCPS team also reported that unrestrained children were three times more likely to sustain severe injury in a crash when compared with children who were restrained; and reported that 64% of severe injuries sustained by children in a crash were to the head (Winston, 2000). Field Observation Techniques The Patterns of Misuse of Child Safety Seats study conducted for NHTSA provided recommendations for performing efficient and successful field observations (Decina and Knoebel, 1996). In brief, the following attributes are critical to a successful field observation effort: (1) knowledge and community support at test sites; (2) a high volume of target group vehicles traveling in test sites; (3) permission from shopping center proprietors and local police departments; (4) optimal test site characteristics (i.e., limited entrance lanes, ample openness of parking areas, safe designated areas to pull over target vehicles, etc.); (5) a data collection team comprised of personable, well-trained individuals who are knowledgeable about CRS misuse issues and who are familiar with community and test site locations; (6) a comprehensive training program, which includes a training manual and supplemental material, classroom workshops and hands-on CRS demonstration sessions, in-field training exercises, and closely supervised "live" data collection; (7) a state-of-the-art training manual, incorporating child development, CRS characteristics, CRS misuse errors, instructions for interviewing and following data collection protocols and methodology, and copies of all necessary field activity forms; (8) customized data-collection forms for observations and interview questions; and (9) proper safety attire (i.e., orange vests, photo identification) and maintenance of a professional appearance. 1.2 PROJECT OBJECTIVES AND SCOPE OF WORK The main objective of this project was to obtain a measure of the current level of misuse of child restraint systems (CRS) among the general public. The project focused specifically on forms of misuse that can be expected to raise the risk of injury. To reach the objectives of this project, the following task activities were performed: Held an initial meeting with the Contracting Officer"s Technical Representative (COTR) and other NHTSA staff to discuss study objectives and activities. Finalized a work plan based on discussions from the initial meeting. Conducted a workshop with National experts on injury and CRS misuse issues to develop appropriate measurement criteria. Selected observation sites and State site coordinators in six States. Prepared data collection instruments. Held "train-the-trainer" session with State site coordinators and their field managers (finalized CRS misuse measures and data collection methodology). Convened an institutional review board (IRB) panel to review and approve the study plan and methodology. Coordinated staff and site logistics (e.g., gaining permission at sites to collect data). Conducted data collection of CRS misuse among the public, with a goal of 4,000 target vehicles, in six States across the country. Prepared data summary and analysis.