The National Highway Traffic Safety Administration has reported that consistent safety belt use (SBU) reduces crash-related fatalities by 45%.1 Unfortunately, while SBU nationwide was over 80% in 2004, SBU among Massachusetts residents was only 63%.2 Some suggest that the emergency department (ED) is an ideal setting for providing harm reduction counseling during a “teachable moment.”3 However, given the time and financial pressures of addressing the core function of emergency medicine today, others have speculated that expanding the mission of emergency medicine to include public health interventions should be done with caution.4 Kelen warns that within an already overstretched ED “safety net,” limited resources should incorporate only those interventions with lasting efficacy.4 In a prior work, we found that a brief intervention improved self-reported SBU among high-risk ED patients at 3 months.5 In hopes of demonstrating lasting public health benefits, we sought to determine if such effects were sustained at 6 months postenrollment.
Objectives: Safety belt use (SBU) reduces motor vehicle deaths by 45%. We previously reported that a brief intervention improved self-reported SBU among emergency department (ED) patients at 3 months. We sought to determine if these effects were sustained at 6 months postenrollment.
Methods: This was a prospective, randomized controlled trial of adult patients (age ≥ 21 years) at an academic medical center ED from February 2006 to May 2006. Patients were systematically sampled for self-reported SBU. Those with self-reported SBU less than “always” were asked to participate. Subjects were surveyed at baseline with a nine-item series of situational SBU questions scored on a five-point Likert scale (e.g., 5 = always, 1 = never). This nine-item average comprised the mean SBU score. Subjects were randomized to a control group (CG) and an intervention group (IG). The CG received an injury prevention brochure; the IG received a brief motivation interview by a trained interventionist and the brochure. Subjects were phoned at 3 and 6 months to determine interval change in SBU scores via a standard script. Repeated-measures analysis of covariance and t-tests were used to analyze trends in mean SBU scores between groups, as well as to test mean changes in SBU scores from the 3- to 6-month intervals.
Results: Of 432 eligible patients, 292 enrolled (mean age = 35 years, SD ± 11.2 years; 61% male). At baseline, there were no significant demographic differences; the IG (n = 147) and CG (n = 145) had similar mean SBU scores (2.8 vs. 2.6, p = 0.31), and 66% (n = 96 in each) completed both 3- and 6-month follow-up. The mean SBU score at 6 months in the IG was greater than in the CG group (3.6 vs. 2.9, p < 0.001), as were the mean SBU score differences from baseline (IG = 0.84 vs. CG = 0.29, p < 0.001). These differences were sustained from the 3-month interval (IG = −0.02 vs. CG = −0.06, p > 0.05).
Conclusions: The previously reported finding that ED patients who received a brief motivation interview reported higher SBU scores at 3 months compared to a CG was sustained at 6-month follow-up. Although limited by self-report, a brief intervention may enhance lasting SBU behavior among high-risk ED patients.
This is a follow-up report to a previously published randomized trial of a brief intervention to increase lasting self-reported SBU among ED patients.5 The institutional review board at our institution approved the study.
Study Setting and Population
The study was conducted at a Level 1 trauma center in Boston with over 90,000 annual ED visits. Adult ED patients (age ≥ 21 years) were eligible provided that they spoke English, answered other than “always” wear safety belts on the screening questionnaire, were considered “stable” according to the clinician caring for the patient (i.e., did not have a life-threatening condition), had a normal mental status, and were able to give autonomous consent. Mental status was assessed by the patient’s ability to demonstrate alertness and orientation to person, time, and place.
Patients were excluded if they had no phone or were homeless; had an altered mental status secondary to pain, distress, medications, brain injury, or acute psychiatric illness; were too ill to be interviewed (e.g., major trauma or medical illnesses); were prisoners; or reported that they had not ridden in a car in the past 30 days. As needed, the treating physician was asked to confirm the patient’s level of orientation.
Details on study methodology appear elsewhere.5 Briefly, from February 2006 to May 2006, research staff systematically approached adult ED patients (age ≥ 21 years) by room-to-room assessment for study eligibility. Patients were issued a confidential, self-administered screening form that included a question regarding overall SBU (“Think about the times you’ve ridden in a car in the past 30 days. Overall how often did you wear a seat belt?”“Always,”“More than half the time,”“About half the time,”“Less than half the time,”“Never”). Patients who answered other than “always” were invited to participate. After providing written informed consent, participants completed a detailed intake questionnaire (see Data Supplement S1, available as supporting information in the online version of this paper) and were randomized into a control group (CG) that received a brochure with general injury prevention information or into an intervention group (IG) that received a brief intervention designed to increase SBU. The intervention—incorporating classic motivational interviewing techniques—was adapted from an intervention found to be successful in mitigating unhealthy alcohol use (http://www.ed.bmc.org/sbirt/). At 3- and 6-month follow-up, research staff blinded to group assignment contacted participants using a scripted telephone survey to determine changes from baseline in self-reported SBU.
As noted above, we used a single-item query on overall SBU to determine study eligibility. To control for crash-related injury, we included the question, “Is your ED visit today related to a car crash?” Participants completed an intake questionnaire, which included a nine-item series of questions regarding situation-specific SBU, scored on a five-point Likert scale (Data Supplement S1, available as supporting information in the online version of this paper).5 We constructed this nine-item SBU measure via a focus group of public health researchers and emergency physicians. We defined self-reported SBU as the average of all the responses to the nine-item measure. Participants were asked if a recent debate in the state legislature on strengthening belt use legislation had influenced their decisions to wear their safety belts. Second, participants were asked if a recent “Click It or Ticket” enhanced police enforcement campaign had influenced their decisions to wear their safety belts.
After completing the baseline intake questionnaire, we employed blocked randomization using permuted blocks of sizes four and six within each sex stratum to assure balance by group assignment.
To compare changes in SBU from 3 to 6 months among groups, we conducted repeated-measures analysis of covariance on 3- and 6-month measures of SBU scores, using the baseline value of SBU as a covariate. As an overall measure of SBU, we averaged each subject’s 3- and 6-month SBU score and dichotomized this average into categories of <4 and ≥4. To assess predictors of dichotomous SBU, we conducted bivariate analyses with randomization group, demographic, and other factors using chi-square tests and Student’s t-tests as appropriate. Randomization, age, sex, and independent variables with bivariate p-values of <0.10 were entered into a final multiple logistic regression model. A p-value of <0.05 was used to determine statistical significance in all analyses.
The baseline sample characteristics appear elsewhere.5 Overall, 292 participants (68% of those eligible) were enrolled in the study; 145 were randomized to the CG, and 147 to the IG. At the 6-month follow-up, 66% (96 in each group, n = 192) completed both the 3- and the 6-month follow-up phone surveys. Among those who completed both follow-ups, there were no statistical demographic differences by group assignment. Participants who completed both follow-ups were older than those who did not (median age = 36.5 years vs. 32.5 years, p = 0.001).
At baseline, the IG (n = 147) and CG (n = 145) had similar mean SBU scores. At the 6-month follow-up, the mean overall SBU score among those in the IG was higher than that for those in the CG (3.6 vs. 2.9, p < 0.01; Table 1). Accordingly, the change from baseline in mean overall SBU score was higher among the IG than the change from baseline for those in the CG (0.84 vs. 0.29, p < 0.01). Additionally, all nine situation-specific SBU scores among the IG were significantly higher than the CG (all scores p < 0.05). These differences in SBU scores were sustained from the 3-month interval (mean overall change from 3- to 6-month SBU scores = IG−0.02 vs. CG –0.06, all p values >0.05).
|Variable||Control (n = 96), Mean (±SD)||Intervention (n = 96), Mean (±SD)||p-value, 6-month Comparison*||p-value, Change From 3- to 6-month Comparison†|
|Intake||3 Months||6 Months||Intake||3 Months||6 Months|
|Highway||3.3 (1.5)||3.7 (1.4)||3.5 (1.5)||3.4 (1.3)||4.0 (1.1)||4.1 (1.1)||<0.01||0.07|
|Local streets||2.5 (1.2)||2.7 (1.3)||2.9 (1.3)||2.6 (1.1)||3.4 (1.3)||3.4 (1.3)||0.01||0.27|
|Daytime||2.5 (1.2)||2.8 (1.3)||2.8 (1.3)||2.7 (1.1)||3.6 (1.2)||3.5 (1.2)||<0.01||0.82|
|Nighttime||2.8 (1.3)||3.1 (1.3)||3.1 (1.4)||2.8 (1.2)||3.8 (1.2)||3.8 (1.2)||<0.01||0.83|
|As driver||2.9 (1.4)||3.4 (1.4)||3.3 (1.3)||3.1 (1.3)||3.9 (1.2)||3.8 (1.2)||0.01||0.48|
|As front seat passenger||2.8 (1.3)||3.3 (1.4)||3.2 (1.4)||2.9 (1.2)||3.9 (1.2)||3.8 (1.2)||<0.01||0.78|
|As back seat passenger||1.9 (0.9)||2.2 (1.4)||2.2 (1.3)||2.2 (1.1)||2.9 (1.4)||3.0 (1.5)||<0.01||0.67|
|Trips < 10 minutes||2.1 (1.1)||2.2 (1.3)||2.3 (1.4)||2.2 (1.2)||3.0 (1.4)||3.0 (1.4)||<0.01||0.52|
|Trips > 10 minutes||2.8 (1.3)||3.4 (1.3)||3.2 (1.4)||3.0 (1.2)||3.9 (1.1)||3.8 (1.2)||<0.01||0.49|
|Overall seat belt score||2.6 (1.1)||3.0 (1.1)||2.9 (1.2)||2.8 (1.1)||3.6 (1.1)||3.6 (1.1)||<0.01||0.67|
We averaged the SBU scores from the two intervals and used the combined mean scores for the bivariate and multivariate analyses. For these analyses, a SBU score of ≥4 (i.e., “more than half the time”) was used as an outcome. Overall, owning a car, being in the ED as a result of a car crash, and receiving the intervention were significantly related to having a SBU score of ≥4 at follow-up in bivariate analysis (data not shown). In the multivariate logistic regression model, being randomized to the IG (odds ratio [OR] = 2.81, 95% confidence interval [CI] = 1.42 to 5.55), car ownership (OR = 2.17, 95% CI = 1.10 to 4.29), increasing age (OR = 1.03, 95% CI = 1.00 to 1.06), and being in the ED due to a car crash (OR = 3.34, 95% CI = 1.26 to 8.80) were significantly associated with a SBU score of ≥4.
Tension exists between advocates of incorporating a public health agenda within ED practice and those who caution that such programs detract from the core function of emergency medicine.3,4 While several ED-based interventions have shown promise, many lack long-term efficacy.4 We previously demonstrated that an ED-based brief intervention increased self-reported SBU at 3 months compared to controls.5 In this study with further follow-up, we found that the increase in SBU score found at 3 months was sustained at 6 months.
Our findings are in accord with those of others. In one study, adolescents who received a brief ED-based intervention had a 47% increase in self-reported SBU compared to a CG.6 The authors of that study suggest that injured patients may be more sensitive to receiving a behavioral intervention due to the “teachable moment” postinjury. Unfortunately, our study was not a priori powered to examine the issue of effect modification. However, previous research suggests that a crash-related ED visit can influence positive behavior change (i.e., a “teachable moment”).7 In future work, injured patients could be oversampled to test such a theory. In contrast to findings from ED-based programs, the benefit of harm reduction counseling has not been established in the primary care setting.8
The SBU scores have no intuitive meaning in clinical practice. Nevertheless, any improvement in SBU may be important from a public health perspective.4 Also, self-reported belt use has been shown to overestimate actual belt use.9 However, a multi-item SBU measure may estimate actual belt use better than a single-item query.9 Finally, our findings may not be generalizable to the general population. ED patients may have higher injury-prone behaviors than others.10 This provided us the opportunity to intervene upon a risk-taking population with low SBU.
Compared with controls, the improvements in self-reported safety belt use scores that were reported at 3 months among participants receiving a brief intervention have been sustained at 6-month follow-up. A crash-related ED visit and car ownership were also significantly associated with increased safety belt use at follow-up. An ED-based intervention may enhance lasting safety belt use among risk-taking ED patients.
The authors acknowledge Ms. Kathy Shea for her administrative assistance in support of this project; Dr. Brian D. Johnston for his assistance in developing our measurement instrument; and Dr. Holly Hackman, Ms. Carlene Pavlos, Ms. Heather Rothenberg, Ms. Sarah Hughes, Ms. Caroline Hymoff, Ms. Cindy Rodgers, and Dr. Jonathan Olshaker for their helpful suggestions in the development and implementation of this project.