NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Williams S, Whitlock E, Smith P, et al. Primary Care Interventions to Prevent Motor Vehicle Occupant Injuries [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2007 Aug. (Evidence Syntheses, No. 51.)

Cover of Primary Care Interventions to Prevent Motor Vehicle Occupant Injuries

Primary Care Interventions to Prevent Motor Vehicle Occupant Injuries [Internet].

Show details

3Results

Key Question 1: Do primary care behavioral counseling interventions for children, adolescents, and adults to increase the correct use of age-and-weight appropriate restraints and reduce driving/riding with drivers under the influence of alcohol reduce morbidity and/or mortality from motor vehicle occupant injuries?

Summary: We identified one large, fair-quality, prospective, non-randomized, group-level controlled trial that reported the impact of behavioral counseling interventions to increase the correct use of child safety seats for infants and children (up to five years of age) on motor vehicle-related occupant injury (MVOI) rates. 53 MVOI counseling and interventions in inpatient and primary care settings were part of a community-wide approach to reducing MVOI and other injuries. During two years of intervention, there was an absolute reduction in the age-adjusted annual MVOI rate (39.2 fewer injuries per 10,000 children) in intervention communities compared to control communities (Table 6). Given the trial's nature, the impact of the clinical counseling components on MVOI reduction cannot be separately determined, although MVOI strategies were not strongly addressed elsewhere in the community-level intervention, and none of the other injury prevention strategies were statistically significant.

Table 6. Summary of studies evaluating counseling to increase use of infant and child safety seats during pregnancy or birth to age 4 years.

Table 6

Summary of studies evaluating counseling to increase use of infant and child safety seats during pregnancy or birth to age 4 years.

We found no other trials relevant to this key question for other patient groups or for behaviors other than infant and child safety restraint use.

Study details: The single study reporting an improvement in MVOI rates was a fair-quality, nonrandomized, controlled, group-level trial that evaluated the effectiveness of the Statewide Childhood Injury Prevention Program (SCIPP) in 14 communities in Massachusetts (N = 286,676). 53 Five different injury prevention projects were implemented targeting multiple injury prevention behaviors, two of which had elements designed to increase use of infant and child safety seats. The first component was a developmentally-oriented, focused counseling system designed for pediatric primary care settings. It targeted reducing six different types of injuries through educating parents of children under age five years. Parents were asked to complete two developmental surveys (The Framingham Safety Surveys (FSS); 20 questions total; approximately 3.5 minutes to complete 54 ) while waiting to see their pediatricians for well-baby visits and these surveys formed the basis of injury prevention counseling during the office visit. Responses that identified a patient as having “high risk” behaviors were identified for the physician who then delivered individually-tailored injury prevention counseling (approximately three minutes) based on guidelines specified in a manual. Pediatric providers were encouraged to administer these surveys at one month, nine months, and 12 months of age. Supplemental written materials were also distributed, including a pamphlet describing vehicular hazards. An earlier study evaluating the acceptability of the FSS found that 93% of parents reported that they felt the survey should be continued, and 82% of pediatricians showed good compliance with the materials.

The second component that focused on increasing use of child safety seats targeted parents of newborns and young children, health providers with access to maternity patients, and hospitals. It included a survey of hospital practices, in-hospital training of maternity and childbirth educators, a resource center for technical assistance, promotion of child safety seat loan programs, outreach to preschools, an observational study of restraint use, and advocacy for a mandatory child restraint law. Exposure to MV-related injury prevention programs was assessed before and after the 2-year intervention period. In the control communities, exposure increased from 14% to 34 %, and from 19% to 55% in the intervention communities, indicating exposure from sources other than the tested interventions throughout the study population. MVOI rates were measured by surveillance through hospitals for injuries requiring medical treatment in an emergency room, hospitalization, or resulting in death and were age-adjusted. The odds ratio for risk of MVOI during the pre-intervention period, compared with the during-intervention period in the intervention communities, was 2.78 times as large as that for the control communities, after adjusting for socioeconomic status. Intervention communities showed a decrease in MVOI rates during the two-year intervention, while control communities showed an increase in MVOI rates for the same time period. Communities in the intervention and control groups were matched for population size and density, age composition, education level, family income, and other pertinent potential confounding characteristics. Baseline information comparing the control and intervention communities on these characteristics, however, were not presented.

It is neither possible to separate this intervention's essential elements, nor to be confident of the independent impact of the clinical components separate from the larger community context. The majority of community-level factors might be viewed as part of the current culture. No positive outcomes, other than MVOI reductions, were reported in the intervention community.

We found no studies that directly evaluated health outcomes from counseling parents of children older than five years, older children and adolescents, or adults to use age- and weight-appropriate restraints. Similarly, we located no trials of primary care counseling to reduce alcohol-related driving/riding behavior in any age group. We did not include studies evaluating the effect of screening and counseling risky or harmful alcohol users to reduce alcohol consumption, a body of literature that the USPSTF has evaluated previously. 49, 55

Key Question 2: Do primary care behavioral counseling interventions for children, adolescents, and adults lead to increased correct use of age-and-weight appropriate restraints?

Seventeen trials that met our inclusion and exclusion criteria evaluated behavioral counseling interventions to increase use of age- or weight-appropriate restraints and reported use or correct use of restraints after the intervention. Studies are described in detail in Appendix F.

Antepartum and birth to age four years

The most extensive literature on interventions to increase the correct use of age- and weight-appropriate restraints involves counseling parents of infants or children up to four years old or expecting mothers (Table 6). Tested interventions represent a wide range of educational approaches, including counseling by clinicians, written materials, films on automotive safety, live demonstrations of child safety seat use, and group-level informational sessions. Four studies also included an infant safety seat distribution program and one trial included reinforcement components. Trials have been conducted in primary care clinics, inpatient maternity wards, and educational courses that are feasible for referral from primary care. We discuss the trials from each of these three settings separately.

Primary care Setting

Summary: Two fair-quality controlled clinical trials (CCT) and two fair-to-poor quality CCTs or randomized clinical trials (RCT)s evaluated counseling by pediatricians during well-child clinic (WCC) visits and reported the effect on use of restraints measured one or more times at least two months after the start of the intervention. 53, 56 58 Both CCTs that measured follow-up at two months reported an increase in restraint use, ranging from an absolute difference between the intervention and control group of 13% to 21%. 56, 57 Trials that reported initial or repeated follow-up at later time points 53, 56, 58 did not find significant differences between restraint use in the intervention and control group. One trial that did not find any differences in restraint use did find, however, a greater reduction in MVOI (after adjustment for SES) for intervention compared with control communities (see results for Key Question 1). The timing of injury-reduction measurement, however, preceded the measurement of restraint use. We do not present outcomes for a trial reporting child restraint use at 8–11 months after the intervention, since outcomes were measured for only 57% of the study population. 57

One fair/poor-quality group level RCT evaluated the effect on parents seeking WCC for their children ages 0–4 years of education plus coercion, incentives, and rewards by non-physician primary care clinic staff and health educators. In this study, observed infant and child safety seat use at 12 months was 10% higher (absolute difference) than at baseline, and statistically significant changes were not seen within the control population. 59

One fair/poor quality RCT 60 and one fair-quality RCT 61 evaluated the effect of counseling pregnant women during the last trimester of pregnancy and measured use or correct use at discharge 61 and/or six to eight weeks after discharge. 60, 61 At discharge, a large absolute difference was seen between the intervention and control group (72%) in one of the studies that evaluated education plus infant safety seat distribution through a loan program. 61 Statistically significant differences were not seen at 6–8 weeks after discharge in either study. 60, 61

Individual studies: The best evidence comes from a fair-quality, non-randomized controlled trial (N = 269) that evaluated the effectiveness of pediatrician counseling during the inpatient postpartum hospitalization and well child visits. 56 The intervention included providing the patient with a formal prescription to obtain an infant safety seat and an informational pamphlet, tailored counseling at the one and two month WCC visits, and demonstration of correct use of an infant seat by a pediatrician at the one month WCC visit. Observed correct use was higher in the intervention group at the two month visit (50% vs. 29%), but was similar at other time points. Correct use among subjects in the control group ranged from 31% at the one-month visit to 50% at 15 months.

A fair-to-poor quality non-randomized controlled trial evaluated a less intensive intervention involving brief counseling (one to two minutes) by a pediatrician or an RN at a single contact time (four week WCC visit) in combination with a waiting room display and pamphlet on infant safety seat use. 57 Self-reported correct use at eight weeks post-intervention was higher in groups counseled by pediatricians or RNs, compared to the control group who received no information about infant safety seats. Self-reported correct use among the control group was 9%.

A fair-to-poor quality RCT evaluated tailored counseling by a pediatrician in a three-part series delivered during the 6, 9, and 12 month WCC visits and was conducted in a low-income, primarily minority population clinic setting. 58 The counseling addressed multiple injury prevention behaviors. At the six-month follow-up, no difference in self-reported restraint use was seen between the intervention and control group; correct use was not specified. At the same follow-up time point, parents of infants in the intervention group were less likely to report that their infants were seated in the front seat (33% vs. 53%, p < 0.05).

One large, fair-quality, non-randomized group-level controlled trial measured self-reported use of child safety seats two years after intervention. 53 This trial (described earlier for Key Question 1) evaluated five injury prevention projects that were part of the Statewide Childhood Injury Prevention Program in Massachusetts and were implemented in 7/14 communities. Two of the five intervention projects targeted increasing infant safety seat use through pediatrician-delivered behavioral counseling tailored to the parents' responses on a developmentally-based, pre-visit survey during well child visits (at 1, 9, and 12 months), education targeting parents of newborn infants at the time of discharge from the hospital, or through education delivered in daycare settings. Self-reported child safety seat use was measured for approximately 5% of the population using random-digit dialing survey methods. Self-reported child safety seat use was not different between control and intervention communities two years after the interventions were implemented in the communities.

A fair-to-poor quality group-level RCT evaluated the effects of multiple educational components delivered by office staff and health educators to parents of children aged 0–4 years visiting for WCC visits; the intervention included multiple reinforcement components delivered in the parking lots, waiting rooms, and at monthly educational sessions. 59 This trial was conducted in a medically indigent population that was over two thirds of minority race/ethnicity. Nonuse of infant safety seats was observed in random samples of clinic populations 6 months and 12 months after initiating the interventions (observed population represented different individuals at the three time points). A large baseline difference was present among patients sampled from the intervention and control populations. When comparing observed non-use at 12-month follow-up to baseline, restraint use (calculated from nonuse) in both the intervention and control populations increased, but the difference was only statistically significant in the intervention population.

Two trials evaluated behavioral counseling interventions in primary care settings during the last trimester of pregnancy. 60, 61 One very small, fair-quality RCT (n=14) conducted in a low-income, Hispanic population evaluated education by an unspecified prenatal care provider, an infant safety seat loan, and a demonstration of correct use. 61 At the time of discharge from the peripartum hospitalization, a higher proportion of women in the intervention group were observed using infant safety seats correctly, compared to women in the control group. At 6 weeks after discharge, the difference in observed use was not statistically significant between groups. A larger fair-to-poor quality RCT (n=156) evaluated the effect of scheduling a pediatrician visit to discuss multiple anticipatory guidance topics during the last trimester of pregnancy. 60 This study was conducted in a low-income, primarily African American population. At two months after birth of the infant, no statistically significant difference was measured for self-reported use of an infant safety seat during the last ride.

Inpatient maternity ward setting

Summary: One fair quality RCT 62 and three fair or fair/poor quality CCTs 63 65 evaluated the effect of infant safety seat education delivered to a total of 2383 parents in the inpatient setting during the immediate postpartum period by maternity ward or research staff. Two of three trials that included an infant safety seat distribution program measured a large absolute difference between use in the intervention and control group either at discharge or at nine months follow-up (47–67% absolute difference). 62, 63 A third fairly well-conducted CCT evaluating education plus free infant safety seat distribution found a smaller difference in observed correct use at the time of discharge or two to four months after discharge between intervention and control groups (absolute difference 5–7%). 64 One fair-to-poor CCT that did not include a safety seat distribution program did not find significant differences between intervention and control groups at discharge. 65

Individual studies: The evidence from two fair-quality trials provides mixed results. 62, 64 A small RCT (n=30) evaluated the effect of distributing a free loaner infant safety seat immediately prior to discharge, including offering to demonstrate correct use with the infant in the room, carrying the infant in the seat to the car, and correctly restraining it into the family's vehicle. 62 The intervention was estimated to add approximately two minutes to the normal length of time required to discharge the patients. Control subjects received usual discharge care. Observed correct use (by a third party), was higher for mothers and infants in the intervention group, compared to the control group at the time of discharge (absolute difference 67%), but the effect was diminished at four to six weeks after discharge (absolute difference 6%). Use in the control group was 0% at discharge and increased to 23% at four to six weeks.

A fair-quality CCT evaluated distribution of two educational pamphlets and a free safety seat by research staff, with demonstrations of correct use to women who accepted the seat, during the post-partum hospital stay. 64 Although 94% of women in the intervention group acquired a safety seat in the hospital, observed correct use was not different between the intervention and control groups at the time of discharge or at two to four months post-partum (absolute difference 5–7%). At the time of discharge, observed correct use was extremely low (6–11%) with 1% of women within the intervention and control groups using the infant carrier without fastening it to the vehicle with a safety belt, 1–2% using a device not designed for use in a car, and 87–91% of infants in both groups being transported in someone's arms.

One large fair-to-poor-quality group CCT conducted in Sweden evaluated the effect of free loaner infant safety seat program that included viewing a videotape and a demonstration of correct use. 63 Self-reported use at nine months was higher among subjects in the intervention group compared to a usual care control (absolute difference 46.8%), but there were extremely high levels of self-reported use at (>97%) 15 months, with no between group differences. Any difference in effect between the intervention and control group may be an overestimate because no effort was made to follow-up on 13% of subjects in the intervention group who did not accept a car seat loan for various reasons. A priori exclusion criteria were not stated.

One fair-to-poor quality non-randomized controlled trial set during the peripartum hospitalization evaluated the effectiveness of education, including a film on automotive safety and demonstration of correct infant safety seat use, without any distribution program. 65 The study population was mostly white and nearly 3/4 were college educated. Observed correct use at discharge in the control group in was 63%. Observed correct use at the time of discharge for women in the intervention group was higher (74%), but the results were not statistically significant.

Settings feasible for primary care-referral

Summary: Two fair/poor quality trials evaluated group-level educational programs aimed at parents of infants and toddlers. In a CCT (n=163), 66 self-reported child safety seat use during the last ride at four to six months follow-up was higher in the intervention group (absolute difference 17.8%). In a small RCT (n=79), 67 100% correct use was observed in both the experimental and control groups.

Individual studies: A fair/poor-quality group-level CCT evaluated a hospital-based prenatal class that included an enhanced educational component designed to increase safety seat use compared to usual cursory mention of child passenger safety. 66 The educational component included a 30-minute lecture on infant safety seat use by a social worker, including a film by the Insurance Institute for Highway Safety, a demonstration of correct use, a question and answer session, and a brochure. Self-reported use during the last ride measured at four to six months after birth was higher among participants in the intervention group (96.1% vs. 78.3%). Results were not different at one of the hospitals where both the intervention group and control group reported high use (98–100%).

A fair-to-poor quality RCT evaluated the effect of education on multiple childhood injury prevention behaviors as part of a hospital-affiliated educational course for parents of toddlers. 67 In this study, observed correct use was 100% in both intervention and control groups. All participants were self-referred to the educational course and represented a highly-educated population.

Intervention components

All thirteen trials that targeted increased child safety seat use among infants and children up to four years old included some form of education, the type of which varied widely across interventions (e.g., counseling, films, written materials, waiting room displays, etc.). Eight trials included a demonstration of correct child safety seat use. 56, 59, 61 66 Of these, six reported an increase in child safety seat use. 56, 59, 61 63, 66 Five trials did not include a demonstration of correct use as part of the intervention 53, 57, 58, 60, 67 and only one of these (a fair/poor-quality CCT) 57 reported an increase in use.

Four trials included free or discounted child safety seat distribution as part of the intervention. 61 64 Of these trials, three demonstrated a large absolute increase in use (47–72%). Nine trials did not include a distribution program and these varied widely by other factors such as setting, timing, who delivered the intervention, and length of follow-up. Of the trials testing education-only interventions delivered during well child care visits, 53, 56, 57, 59 findings were consistent with increased use at short-term follow-up and diminished use at later time points, as we have previously described.

Ages four to eight

Summary: One fair/poor-quality RCT evaluated booster seat education with and without distribution of a free booster seat among families visiting the emergency department for any chief complaint who reported not using booster seats at baseline 68 (Table 7). One month after the intervention, self-reported use by families in the control group and the education-only groups were low and not statistically different from each other. In contrast, 98% of families in the group that received both education and a free booster reported using the seat at one month.

Table 7. Summary of Studies Evaluating Counseling to Increase Use of Booster Seats in Children 4–8 Years of Age.

Table 7

Summary of Studies Evaluating Counseling to Increase Use of Booster Seats in Children 4–8 Years of Age.

Individual study: One fair/poor-quality RCT evaluated the effect of brief (5- minute) educational counseling about the importance of booster seat use delivered by a certified car seat technician. It was conducted through an emergency department in a low-income, predominantly African American population among families with children ages four to seven years who reported not using booster seats at baseline. 225 families were randomized to three groups: 1) a control group that received standard discharge instructions, 2) an education-only group that received 5 minutes of booster seat training, or 3) an education + distribution group that received the education plus a free booster seat properly installed in their car when they left the ED. Based on self-reported data at 1 month after the intervention was delivered, 98% of families in the education + distribution group reported using a booster seat compared to 5.5% of families in the other two groups (control or education-only) combined (p < 0.001).

The trial had several methodological problems that could have introduced bias including high overall attrition (35%), differential attrition across treatment groups (40%/39%/25%), self-reported outcomes, analyzing the completers only, and not reporting process measures. The trial also has several limitations with respect to generalizing the findings to the primary care setting. The intervention was delivered by certified car seat technicians rather than clinicians who typically deliver care in primary care settings. In addition, the results are only among families that did not use a booster seat when presenting the to ED. Therefore, the magnitude of benefit from the education + distribution programs in a general primary care population cannot be directly determined from these findings. Outcomes were measured at 1 month post-intervention, so it is unknown whether or not differences were sustained in the group that received the free booster seat.

Ages nine to nineteen years

Summary: One fair-quality CCT (Table 8) reported short-term improvement in observed seat belt use among children ages 5–19 years old immediately after the intervention, but analyzed only the children who were not wearing seat belts when arriving to the visit. 69 A fair-quality RCT reported no difference between intervention and control groups in seat belt use by fifth and sixth graders at 12–36 months. 70

Table 8. Summary of studies evaluating counseling to increase child safety seats or seatbelt use in children 9–19 years.

Table 8

Summary of studies evaluating counseling to increase child safety seats or seatbelt use in children 9–19 years.

Individual studies: The Dartmouth Prevention Project 70 was a large fair-quality, cluster-randomized RCT evaluating an office-based structured prevention intervention delivered to 3145 fifth and sixth graders receiving care in 12 matched pediatric primary care clinics in rural and urban communities. It was designed to test the effectiveness of clinician-delivered advice to promote family communication and prevent adolescent high-risk behaviors. Half of participating practices delivered an injury prevention message about gun safety, seatbelt use, and bicycle helmet use. The other half delivered an alcohol- and tobacco use- prevention message. The authors do not report baseline injury prevention counseling practices among providers delivering care at the control sites. At baseline, 72–74% of the children in both intervention and control groups reported always wearing seatbelts during the previous 30 days. Counseling by a pediatrician or nurse practitioner during well child visits was supplemented by a contract for a family policy, reinforcement of the message at subsequent office visits over 36 months, and written materials mailed to the home and phone calls alternately targeting the parent and child. Process measures suggest a high fidelity for initial intervention delivery, with reinforcement at subsequent office visits for at least half of the children seen. Outcomes were measured using child and parent responses on self-administered mailed surveys. No differences were found in the proportion of children who reported always wearing seat belts during the last month at 12, 24, or 36 months follow-up time point.

A fair-quality controlled clinical trial (n=242) evaluated the effect of brief counseling (< 3 minutes) of children ages 5 – 19 years by a pediatrician during routine well child care visits, plus a signed contract and dashboard sticker. 69 Car safety was not mentioned to the control group. The intervention was delivered during alternate weeks over a five- week period. Data on adherence, contamination, or cross-over were not reported. Blinded parking lot observers assessed seat belt use before and after the visit. At one year post-visit, a self-administered questionnaire that could not be easily linked to the office visit was mailed to participants. No seat belt law was in effect in the state at the time the study was conducted. At the post-visit time point, only children who were not observed to be wearing seat belts upon arriving to the clinic were analyzed. During the intervention weeks, 29/77 children who were not observed to be wearing seat belts upon arrival were wearing them when leaving the clinic (p < 0.001). During control weeks, 4/73 children not observed to be wearing seat belts upon arrival were wearing them when leaving the clinic (n.s.). The between group comparison was statistically significant (p < 0.001). Sibling use was also higher at the post-visit observation during intervention weeks. Forty-two to forty-six percent of siblings who did not wear seatbelts prior to the visit wore them at the post-visit observation compared to 0–9% of siblings during control weeks. Only 65% of participants returned the questionnaire mailed one year after the intervention. Response rates were similar between control and intervention group, but large overall loss of follow-up makes results suspect. Reported seat belt use in intervention group (52/84 (62%)) and control groups (47/70 (67%)) were not different.

Adults

One fair-to-poor-quality RCT (Table 9) set in a rural primary care clinic found no difference in self-reported seat belt use among adults six months after viewing a six-minute film explaining the rationale for wearing a seat belt compared to those who watched a film of comparable length that did not mention seat belts. 71 Data on adherence, crossover, and contamination were not reported, and thus we do not know whether the intervention was delivered as intended. Self-reported seat belt use increased within each group from 20–22% at baseline, to 34–37% at six months post-intervention, indicating that changes in behavior occurred due to factors unrelated to the intervention such as the regulatory environment. Also, the outcome of interest was not well masked on the survey, and thus social desirability bias may have caused over-reporting of seat belt use within each group.

Table 9. Summary of studies evaluating counseling to increase use of seat belts by adults.

Table 9

Summary of studies evaluating counseling to increase use of seat belts by adults.

For all age groups, the volume and quality of research were inadequate to quantitatively address questions about essential elements of efficacious interventions, other positive outcomes from BC interventions addressing seat belt usage, or the maintenance of MVOI-safety behaviors after BC interventions.

Key Question 3: Do primary care behavioral counseling interventions for children, adolescents, and adults reduce driving/riding with drivers under the influence of alcohol?

Our searches found no studies of primary care interventions evaluating behavioral counseling in general populations to reduce driving, or riding with drivers, under the influence of alcohol.

Key Question 4: What are the adverse effects of counseling children, adolescents, and adults to correctly use age-and- weight appropriate restraints and reduce driving/riding with drivers under the influence of alcohol?

Our searches found no studies of adverse effects of counseling to use age- and weight-appropriate restraints or reduced driving/riding with drivers under the influence of alcohol.

Summary of Evidence Quality

Table 10 summarizes the overall quality of evidence according to USPSTF criteria 44 for each of the key questions addressed in this review. The overall quality of evidence is fair for the direct effect (KQ1) of interventions to increase the use of restraints for infants and children up to age four years and is poor for the direct effects of counseling older children, adolescents, or adults to use age- and weight- appropriate restraints or counseling to reduce drinking and driving/riding behavior in any age group due to a lack of studies. The quality of evidence is fair for the effect of behavioral counseling interventions to increase use of safety seats for infants and children up to age four in primary care or peripartum hospital admission settings (KQ2) and fair-to-poor for studies in primary care referable settings. The quality of evidence is fair-to-poor for counseling to increase use of booster seats for children ages four to eight years, fair for increasing use of seat belts in older children and adolescents, and fair-to-poor for increasing seat belt use in adults. The evidence is poor for the linkage between counseling and a reduced alcohol-related driving or riding behavior due to a lack of studies (KQ3), and poor for evidence of adverse effects from counseling interventions (KQ4) due to lack of studies.

Table 10. Summary of evidence quality by key questions.

Table 10

Summary of evidence quality by key questions.

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (2.0M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...