- Email: [email protected]
DIAL: A Telephone Brief Intervention for High-Risk Alcohol Use With Injured Emergency Department Patients
INJURY PREVENTION/ORIGINAL RESEARCH
DIAL: A Telephone Brief Intervention for High-Risk Alcohol Use With Injured Emergency Department Patients Michael J. Mello, MD, MPH Richard Longabaugh, EdD Janette Baird, PhD Ted Nirenberg, PhD Robert Woolard, MD
From the Injury Prevention Center at Rhode Island Hospital, Department of Emergency Medicine, Warren Alpert Medical School of Brown University, Providence, RI (Mello, Baird, Nirenberg, Woolard); and Brown University Center of Alcohol and Addictions, Providence, RI (Longabaugh, Nirenberg).
Study objective: Brief interventions for high-risk alcohol use for injured emergency department (ED) patients have demonstrated effectiveness and may have a more pronounced effect with motor vehicle crash patients. We report on 3-month outcome data of a randomized controlled trial of injured patients, using a novel model of telephone-delivered brief interventions after ED discharge. Methods: ED research assistants recruited adult injured patients who screened positive for high-risk alcohol use and were to be discharged home. After discharge, participants received by telephone an assessment of alcohol use and impaired driving and then were randomized to treatment (n⫽140) or standard care (n⫽145). Treatment consisted of 2 sessions of brief interventions done by telephone, focusing on risky alcohol use. At 3 months, both groups had an assessment of alcohol use and impaired driving. Results: Two hundred eighty-five patients were randomized and had a baseline mean Alcohol Use Disorders Inventory Test (AUDIT) score of 11.0 (SD⫽7.4). Three-month follow-up assessments were completed on 273 (95%). Mean AUDIT score decreased in both the treatment (mean change⫽⫺3.4; 95% confidence interval [CI] ⫺4.5 to ⫺2.3) and standard care group (mean change⫽⫺3.2; 95% CI ⫺4.2 to ⫺2.2). Measures of impaired driving decreased for the treatment group (mean change⫽⫺1.4 95%; CI ⫺3.0 to 0.2) compared with standard care group (mean change⫽1.0; 95% CI ⫺0.9 to 2.9; P⫽.04; d⫽0.31). Participants were stratified post hoc into 3 groups by baseline alcohol problem, with the treatment effect only being in the highest-scoring group (d⫽.30). Conclusion: Telephone brief interventions decreased impaired driving in our treatment group. Telephone brief intervention appears to offer an alternative mechanism to deliver brief intervention for alcohol in this at-risk ED population. [Ann Emerg Med. 2008;51:755-764.] 0196-0644/$-see front matter Copyright © 2008 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2007.11.034
INTRODUCTION Alcohol has a well-documented role in injury occurrence.1 There is a particularly strong connection between alcohol use and motor vehicle crashes, with alcohol involved in approximately 41% of fatal motor vehicle crashes and 256,000 persons being injured in alcohol-related motor vehicle crashes.2 The resultant economic costs of impaired driving are staggering and estimated to be more than $50 billion.3 Thus, impaired driving is a public health problem with significant human suffering and societal cost that necessitates further efforts at control. Emergency departments (EDs) are an opportune setting for identifying persons using alcohol in a risky manner, especially in Volume , . : June
those presenting for injury4 or from a motor vehicle crash.5 Injury has also been identified as a motivator to change alcohol use.6 In patients admitted to a trauma service, a brief intervention for alcohol has been found to reduce injury recidivism7 and subsequent impaired driving arrests.8 With injured patients being treated in an ED and discharged to home, Longabaugh et al9 found that those with hazardous alcohol use who received brief intervention for alcohol while in the ED and a follow-up booster brief intervention within 2 weeks reduced alcohol-related negative consequences and alcohol-related injuries more than those who received only standard ED care. This effect was not dependent on whether the patients had consumed alcohol before their injury. Annals of Emergency Medicine 755
Telephone Brief Intervention for High-Risk Alcohol Use Editor’s Capsule Summary
What is already known on this topic There is some evidence that brief motivational interventions in the emergency department (ED) can decrease risky behaviors, but it is difficult to consistently deliver these messages in the chaotic ED environment. What question this study addressed Can a brief telephone intervention delivered after an ED visit to patients with risky alcohol use decrease alcohol consumption and related risky behavior? What this study adds to our knowledge In this 285-patient randomized controlled trial, the brief telephone intervention had little effect on self-reported risky alcohol use but had an effect by slightly decreased self-reported impaired driving at 3 months in the treatment group. How this might change clinical practice The teachable moment can extend beyond the ED visit, and interventions may not need to be conducted during the visit to be effective.
Furthermore, the treatment effect was greatest in those patients being treated in the ED after a motor vehicle crash.10 All of these studies7-10 used a brief intervention that was based on the technique of motivational interviewing.11 Thus, injured ED patients, motor vehicle crash patients in particular, appear to be a group to target for screening and a brief intervention for alcohol problems. Although organizations have advocated for brief intervention for alcohol to be done within the clinical practice of the ED,12,13 barriers exist that have prevented its universal adoption. In the busy environment of the ED, diagnosis and treatment of acute medical illness and injury regularly usurp prevention efforts. ED staff may perceive this as not being part of their clinical care of the injured patients or may not have the skills or additional resources to address alcohol use problems. For patients, the ED may not be the ideal environment for a brief intervention because it is a noisy, chaotic, stressful, and uncomfortable setting. While in the ED, the patient is likely to be fatigued, in pain, and possibly under the influence of alcohol, which may affect the efficacy of the intervention. The “teachable moment” perhaps does not only exist in the ED, but opportunities to offer an intervention for alcohol use may extend to a period beyond the initial ED visit. This view has led us to now explore whether brief intervention for alcohol would be effective with ED patients after discharge from the ED. A practical mechanism for reaching patients after discharge from the ED would be using the telephone. Telephone interventions have demonstrated efficacy as a modality to change behavior. Programs that include a telephone intervention component have 756 Annals of Emergency Medicine
Mello et al addressed varied health issues, including alcohol use in a primary care setting,14 management of chronic diseases,15 depressive symptoms,16,17 agoraphobia,18 nutrition,19 and smoking cessation.20-22 Midanik et al23 have also found, in comparing telephone and face-to-face interviews assessing alcohol-related harm, that the telephone survey yielded significantly higher rates of self-reported alcohol-related harm compared with the in-person survey, possibly because of increased anonymity with telephone surveys. Although only some of these telephone interventions are motivational interviewing– based brief intervention and none use an ED population, the existing barriers to brief intervention being done in the ED and the practicality of the telephone necessitates telephone-delivered brief intervention testing with an ED population. We do not know which ED patients would benefit from a brief intervention. If brief intervention is found to be effective with certain subgroups of ED patients, limited resources could be concentrated with those groups. Previous research has demonstrated a more pronounced effect of brief intervention with ED motor vehicle crash patients than other injured patients.10 This finding, along with the prevalence of alcohol use problems in motor vehicle crash patients,5 yields utility for this subgroup to be examined for treatment effect with a telephone brief intervention for alcohol. Also, we lack data on the number of alcohol problems among injured ED patients that provide the floor and ceiling for brief intervention to be effective. It may be that brief intervention has a differential effect on ED patients that depends on their alcohol problem severity. Using the Alcohol Use Disorders Inventory Test (AUDIT)24 screening data from a treatment-seeking sample, Donovan et al25 found that classifying patients according to 4 AUDIT score zones produced a linear relationship between the zone and measures of alcohol problem severity. Thus, AUDIT score zones appear to be a reliable instrument to divide patients into groups based on the number of alcohol problems and could be used to examine whether ED patients’ alcohol problem severity predicts their response to brief intervention. Goal of This Investigation With the telephone being used successfully in counseling and having potential advantages for delivering brief intervention with ED patients, we decided to alter the locus of brief intervention for ED patients. The primary aim of the present study is to test the effect of brief intervention given by telephone soon after the patient’s ED visit and followed by a “booster” brief intervention telephone session 2 weeks later on the extent of hazardous drinking and frequency of impaired driving at a 3-month follow-up. The secondary aims are to test whether the effectiveness of brief intervention given by telephone is moderated by the patient’s initial alcohol problem severity and to test whether ED motor vehicle crash patients are more responsive to brief intervention given by telephone than are other injured ED patients. Volume , . : June
Mello et al
MATERIALS AND METHODS Study Design DIAL, Decreasing Injuries from ALcohol, is a randomized clinical trial using a 2-group design for injured ED patients discharged to home. All participants were recruited while in the ED, but enrollment, assessment, randomization, and all interventions were done by telephone in the days after discharge. The treatment group received a brief intervention about their high-risk alcohol use during that same telephone call and a subsequent brief telephone booster brief intervention session. The control group received standard care for their ED visit, with no study interventions. Both groups received by telephone an assessment of alcohol use and impaired driving before randomization and at 3 months. All protocols were reviewed and approved by the institutional review boards at the study institutions. DIAL was funded by the Centers for Disease Control and Prevention’s National Center for Injury Prevention and Control and registered with clinicaltrials.gov. Setting Participants were recruited from the ED of a large, urban, academic, Level I trauma center and 2 smaller community hospital EDs in nearby suburban settings. Selection of Participants Participants were recruited in the ED by research assistants during preselected times representing all shifts and days of the week, from November 2003 to June 2006. The research assistants were scheduled for more shifts in the trauma center ED than both community hospitals combined. All noncritically injured adult (18 years or older) patients were screened during their ED visit with a health behaviors survey that included questions on smoking, exercise, stress, and amount of alcohol use. Those using alcohol at risky levels were eligible for the study. Risky alcohol use was defined using National Institute on Alcohol Abuse and Alcoholism quantity frequency guidelines (either more than 14 drinks/week for men and more than 7 drinks/week for women or 5 or more drinks/occasion for men and 4 or more drinks/occasion for women).26 Additional eligibility requirements included being discharged from the ED to home, having a telephone, English speaking, not suicidal, and not in police custody. In the first year of the study, we recruited only motor vehicle crash patients. To increase enrollment for the remainder of the study, both motor vehicle crash patients and other injured patients were recruited. This broadens the sample’s representation of injured ED patients and permitted us the opportunity to test whether motor vehicle crash patients had a differential treatment compared with other injured patients. Consent for participation in the study occurred in the ED. On approaching patients who screened positive for the alcohol criteria and met other eligibility criteria, the purpose of the study was explained to the patient. Using a scripted recruitment approach, the research assistant explained the schedule of telephone contacts, what would happen during the assessment portion of the telephone call, and how randomization was Volume , . : June
Telephone Brief Intervention for High-Risk Alcohol Use conducted. It was also explained to the patient before signing of the consent form that, if randomized into the treatment group, the participant would be invited at that time to engage with the trained interventionists in a brief conversation about the answers they had given during the assessment. It was further described that this conversation would be about alcohol use and driving behaviors. The patient was informed that the treatment group would also have a booster telephone call scheduled within 14 days of the first call. All participants were compensated for assessments done at both baseline and 3-month follow-up, for a total of $70. We based the sample size on the results of a previous randomized clinical trial at our institution on the effectiveness of face-to-face brief intervention delivered in the ED.9 Analysis of the data from that study demonstrated that brief intervention resulted in an effect size of d⫽0.34 for reducing alcohol-related injuries among patients who had been injured by a motor vehicle crash.10 Using Cohen’s procedure for calculating sample size, with a power of 0.80 and an ␣ of 0.05, we needed to recruit 114 participants per treatment condition to evaluate differences in treatment means. We estimated an attrition rate of 20%, also based on the previous ED research.9 This yielded a total sample recruitment of 143 participants per treatment condition. Interventions Counselors initiated telephone calls to participants who were recruited in the ED within 5 days after their discharge. Patients contacted by telephone were offered enrollment in the study. Those enrolled received an assessment of their alcohol use and impaired driving. After assessment, participants were randomized by computer assignment to either treatment or standard care group. Participants in the treatment group received a brief intervention of approximately 30 minutes’ duration. Two weeks later, they received an additional 15-minute brief intervention booster session. The primary goal of brief intervention is the mobilization of the participant’s own resources to bring about changes needed to reduce alcohol-related risky behaviors. The counselor’s role is to be reflective, to provide an atmosphere that will support and enhance the participant’s motivation for change and that will lead the participant to initiate and persist in behavioral change efforts.11 Counselors were master’s- or doctoral-level staff trained and certified in motivational interviewing by a certified trainer and practitioner of motivational interviewing, who developed a script on introducing the main themes of motivational interviewing when delivered by telephone. Intervention fidelity was maintained through weekly clinical supervision sessions with a licensed clinical psychologist with experience in MI. Those participants allocated to standard care condition received no study interventions. None of the 3 study EDs had clinical protocols for routine standardized alcohol screening or brief interventions with patients to be discharged. Acute intoxicated patients are kept in the ED until sober in all study EDs and may be referred to outside treatment facilities at the discretion of the ED clinician. Annals of Emergency Medicine 757
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 1. Impaired driving scale items.
Methods of Measurement Both groups received the same assessments by telephone at baseline and at 3 months. A research assistant blinded to treatment condition contacted participants at 3 months for follow-up assessment. Risky alcohol use was measured by AUDIT,24 a 10-item instrument screening for hazardous or harmful drinking over the last 3 months (scores range from 0 to 40) and impaired driving behavior using the Impaired Driving Scale. The Impaired Driving Scale (Figure 1) is a 6-item scale that is based on 4 questions in the Drinking and Driving Scale27,28 and 2 from the Behavioral Risk Factor Surveillance System29 that measures impaired driving. The Impaired Driving Scale measure has a Cronbach ␣ of 0.95. Other assessments done at baseline and 3 months included other high-risk driving behaviors (ie, seatbelt use and the frequency of speeding).
Mello et al Driving Scale) between baseline and 3-month assessment within groups. Before data analysis, the raw scores from these measures were analyzed and appropriate transformations were conducted where necessary. Raw scores on the items of the Impaired Driving Scale were converted to t scores to standardize items and reduce the effects of items with different time metrics.30 An intention-to-treat analysis plan was used for all participants randomized in the study. A generalized linear model analysis of the effects of treatment, time, and a treatment-by-time interaction was used to analyze changes in 3-month AUDIT and Impaired Driving Scale scores. Where appropriate, means, standard deviations, or confidence intervals (CIs) at the 95% significance level are reported. To examine for moderation of treatment effect by extent of risky alcohol use, groups were divided into 3 zones by their baseline AUDIT score (zone 1 score less than 8 [n⫽122]; zone 2 score 8 to 15 [n⫽101]; zone 3 score greater than or equal to 16 [n⫽63]) and then analyzed for change in alcohol use and impaired driving between baseline and 3 months. Categorizing AUDIT into zones has been described previously in guidelines by the World Health Organization (WHO)24 and others.25 We created 3 zones by collapsing WHO’s third and fourth zones into a single highest-risk zone, which increased the size of the group for analysis and had the practical implication of the entire resultant zone (AUDIT greater than or equal to 16) being engaged in “hazardous alcohol use.”24 Post hoc tests of paired differences were conducted on any significant main and interaction effects to determine treatment differences among the treatment zone groups. Analysis examining the effects of cause of injury (motor vehicle crash or other injury cause) on difference in treatment effect for alcohol use and impaired driving was also conducted. Data are presented by outcome analysis (statistic and 95% CI) for the standard care and treatment group and by the interaction between treatment condition and baseline AUDIT zone.
RESULTS Data Collection and Processing Data were collected on paper by research assistants then transferred to an electronic database (Illume 2.2; DatStat, Seattle, WA). A double data entry system was used by 2 research assistants who would independently enter data before the database was electronically checked for inconsistencies. Inconsistent entries were then reviewed by the project manager and corrected to be consistent with data collection sheets. Data were electronically transferred to SAS version 9.1 (SAS Institute, Inc., Cary, NC) for analysis. Primary Data Analysis Descriptive analyses were conducted on all baseline data and contrasted treatment and standard care groups, as well as motor vehicle crash participants, with other injured participants. Because our study sought to reduce risky alcohol use and impaired driving, we then examined outcome measures of AUDIT summary score and impaired driving (Impaired 758 Annals of Emergency Medicine
Characteristics of Study Subjects From November 2003 to June 2006, 17,234 patients were treated in the EDs while research assistants were present. The majority of patients were ineligible for our study because they were there for treatment of a medical illness, were to be admitted to the hospital, or were too clinically unstable to be screened. During this period, there were 6,335 motor vehicle crash and other injured ED patients for the research assistants to screen for eligibility. Research assistants did not screen 249 patients in this group because the patients refused screening, were discharged before screening, or were otherwise unable to be located. The other 6,086 patients were screened to determine whether they met protocol eligibility and their willingness to participate. There were 1,329 patients who met protocol eligibility criteria, and 410 were interested and provided consent in the ED to receive a telephone call in the days after discharge for enrollment, assessment, and randomization. Of that group, 125 were not enrolled because we were unable to contact them Volume , . : June
Mello et al
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 2. Subject recruitment and disposition. MVC, Motor vehicle crash.
or they voluntarily withdrew. The study sample consisted of 285 participants contacted by telephone, who completed the baseline assessment and were randomized into treatment (n⫽140) or standard care (n⫽145) groups (see Figure 2). Participants were recruited from a Level I trauma center ED (n⫽253; 89%) and community hospitals’ EDs (n⫽31; 11%). A descriptive analysis of groups was conducted, examining sex, age, race, ethnicity, years of schooling, employment, injury type (motor vehicle crash or other), and baseline alcohol use (Table 1). The treatment and standard care groups were similar with respect to their baseline AUDIT and Impaired Driving Scale scores (Tables 2 and 3) and mechanism of injury (motor vehicle crash versus other injury). Telephone contact was successful on the first attempt in 50% of the study sample. More than 90% were successfully reached by the Volume , . : June
third attempt. The average time from ED discharge until assessment and randomization was 5 days (range 1 to 28 days; median⫽4 days). For those in the treatment group, the average length of time for the first brief intervention session was 31.2 minutes (SD⫽13.4; range 10 to 77 minutes; median⫽30 minutes); the average length of time for the booster brief intervention session was 18.8 minutes (SD⫽9.28; range 5 to 50 minutes; median⫽15 minutes). During the initial telephone call, 92% of participants completed the brief intervention immediately after completing the baseline assessment; the other 8% were recontacted later by telephone, on average 9 days after randomization, to receive the brief intervention. The booster brief intervention session was received by 69% (n⫽97) of participants in the treatment group. At 3-month follow-up, assessments were obtained on 272 (95%) of randomized participants. Annals of Emergency Medicine 759
Telephone Brief Intervention for High-Risk Alcohol Use
Mello et al
Table 1. Description of participant characteristics and baseline data. Demographic Characteristics
Treatment Group (nⴝ140)
Sex, % Age, y Race, % (White versus non-White) Ethnicity, % (Hispanic versus nonHispanic) Years of schooling, mean Employment,% (employed/ unemployed) Injury type, MVC, % (mvc versus other injury) Baseline alcohol use, quantity/ frequency Baseline alcohol use, most alcohol in the past month
SC Group (nⴝ145)
Eligible, Not Assessed*(nⴝ123)
Male, 61 Mean⫽28 (SD⫽9.8) White, 76 Non-Hispanic, 90
Male, 61 Mean⫽31 (SD⫽10.7) White, 69% Non-Hispanic, 84
Male, 62 Mean⫽30 (SD⫽11.2) White, 72 Non-Hispanic, 80
13.4 74
13.0 75
NA NA
51
52
59
Mean⫽11.7 drinks/wk (SD⫽12.4)
Mean⫽11.0 drinks/ wk (SD⫽12.3) Mean⫽7.8 drinks (SD⫽3.4)
Mean⫽8.9 drinks/wk (SD⫽14.7)
Mean⫽8.2 drinks (SD⫽3.2)
Mean⫽8.5 drinks (SD⫽4.9)
NA, Not available. *Patients recruited and consented in ED but not enrolled, not randomized, or do not have baseline assessment completed.
Table 2. Treatment effect on 3-month alcohol use. Outcome
Baseline Mean (nⴝ285)
Three-Month Mean (nⴝ273)
Change, Mean Change With 95% CI
SC: mean⫽10.8 T: mean⫽11.5 SC1: mean⫽5.2 T1: mean⫽5.2 SC2: mean⫽10.6 T2: mean⫽11 SC3: mean⫽23.1 T3: mean⫽22.5 SC1: mean⫽1.1 T1: mean⫽1.0 SC2: mean⫽2.0 T2: mean⫽2.0 SC3: mean⫽3.0 T3: mean⫽3.0
SC: mean⫽7.6 T: mean⫽8.1 SC1: mean⫽4.6 T1: mean⫽5.0 SC2: mean⫽7.1 T2: mean⫽8.1 SC3: mean⫽15 T3: mean⫽12.9 SC1: mean⫽0.8 T1: mean⫽0.8 SC2: mean⫽1.5 T2: mean⫽1.6 SC3: mean⫽2.4 T3: mean⫽1.9
SC: mean⫽⫺3.2 (⫺4.2 to ⫺2.2) T: mean⫽⫺3.4 (⫺4.5 to ⫺2.3) SC1: mean⫽⫺0.6 (⫺1.2 to 0) T1: mean⫽⫺0.2 (⫺1.1 to 0.7) SC2: mean⫽⫺3.5 (⫺4.7 to ⫺2.3) T2: mean⫽–2.9 (⫺4.3 to ⫺1.5) SC3: mean⫽⫺8.3 (⫺13.8 to ⫺4.8) T3: mean⫽–9.7 (⫺12.7 to ⫺6.7) SC1: mean⫽⫺0.4 (⫺0.6 to ⫺0.2) T1: mean⫽⫺0.2 (⫺0.5 to 0.1) SC2: mean⫽⫺0.6(⫺09 to ⫺0.3) T2: mean⫽⫺0.4 (⫺0.7 to ⫺0.1) SC3: mean⫽⫺0.6 (⫺1.1 to ⫺0.1) T3: mean⫽⫺1.1 (⫺1.7 to ⫺0.3)
AUDIT score Total AUDIT score in zone 1 Total AUDIT score in zone 2 Total AUDIT score in zone 3 Binge drinking in zone 1 Binge drinking in zone 2 Binge drinking in zone 3
SC, Standard care group; T, treatment group; SC1, standard care group in zone 1; T1, treatment group in zone 1; SC2, standard care group in zone 2; T2, treatment group in zone 2; SC3, standard care group in zone 3; T3, treatment group in zone 3. Binge drinking⫽6 or more drinks per occasion.
Main Results An ANOVA model with omnibus F tests was conducted to examine the effects of time and treatment on 3-month AUDIT and impaired driving. No significant difference was found between treatment and standard care groups at 3-month follow-up on mean AUDIT scores and mean change in AUDIT scores (Table 2). There was an omnibus effect found for the model with 3-month impaired driving as the outcome variable (F(7, 537)⫽8.78; P⬍.001). As can be seen in Table 3 and Figure 3A, there was a lower report of impaired driving between baseline and 3 months in the treatment group (mean⫽49.1; mean change⫽–1.4; 95% CI ⫺3.0 to 0.2) but not the standard care group, in which there was an increase in reported 3-month impaired driving (mean⫽50.8; mean change⫽1.0; 95% CI ⫺0.9 to 2.9). Using Cohen’s d, the difference between the 3-month Impaired Driving Scale scores for the standard care and treatment groups corresponds to an effect size of 0.31. Because there was an overall effect of treatment and baseline AUDIT strata on 3-month impaired driving scores, further 760 Annals of Emergency Medicine
analysis was conducted (Figure 3B). This post hoc analysis of paired differences showed that when alcohol use was stratified into the 3 AUDIT zones, there was a significant effect of treatment group by baseline AUDIT zone (Table 3). The participants in the highest zone (AUDIT score greater than or equal to 16) receiving treatment had a lower 3-month impaired driving score (mean⫽52.3; mean change⫽⫺6.4; 95% CI ⫺13.2 to 0.4) than participants in the highest zone who were in the standard care group who reported on average increasing impaired driving (mean⫽57.8; mean change⫽5.1; 95% CI ⫺4.3 to 14.5). No significant effects were found in zones 1 or 2 (Table 3). We found there was no difference between those who were injured as a result of a motor vehicle crash and those injured in other ways for mean changes in Impaired Driving Scale scores over time (motor vehicle crash mean change⫽⫺0.10; 95% CI ⫺1.6 to 1.8; other injury mean change⫽⫺0.3; 95% CI ⫺2.4 to 1.8). To better understand the effects that brief intervention had on impaired driving behaviors (Impaired Driving Scale scores), Volume , . : June
Mello et al
Telephone Brief Intervention for High-Risk Alcohol Use
Table 3. Treatment effect on 3-month impaired driving measures. Outcome IDS IDS in zone 1 IDS in zone 2 IDS in zone 3 % Reporting Impaired Driving in Zone 1 % Reporting Impaired Driving in Zone 2 % Reporting Impaired Driving in Zone 3
Baseline Mean (nⴝ285)
Three–Month Mean (nⴝ273)
Change Mean or % with 95% Confidence Interval (CI)
SC: mean⫽49.8 T: Mean⫽50.5 SC1: mean⫽47.3 T1: Mean⫽47.4 SC2: mean⫽51 T2: Mean⫽48 SC3: Mean⫽52.7 T3: Mean⫽58.7 SC1: 34 T1: 39 SC2: 55 T2: 38 SC3: 56 T3: 65
SC: mean⫽50.8 T: Mean⫽49.1 SC1: mean⫽47.9 T1: Mean⫽47.9 SC2: Mean⫽50.4 T2: Mean⫽48.4 49.0) SC3: Mean⫽57.8 T3: Mean⫽52.3 SC1: 41 T1: 28 SC2: 55 T2: 49 SC3: 48 T3: 42
SC: Mean⫽1.0 (⫺0.9 to 2.9 T: Mean⫽⫺1.4 (⫺3.0 to 0.2) SC1: Mean⫽0.6 (0.2 to 1.0) T1: Mean⫽0.5 (0.2 to 0.8) SC2: Mean⫽⫺0.7 (⫺2.4 to 1.0) T2: Mean⫽0.3 (⫺0.3 to 0.9) SC3: Mean⫽5.1 (⫺4.3 to 14.5) T3: Mean⫽⫺6.4 (⫺13.2 to 0.4) SC1: 7% (⫺3.0 to 17.0%) T1: ⫺11%% (⫺19 to ⫺3.0%) SC2: 0% (⫺4 to 4.0%) T2: 11% (2.0 to 20.0%) SC3: ⫺8% (⫺30 to 14.0%) T3: ⫺23% (⫺37 to ⫺9.0%)
IDS, Impaired Driving Scale.
we dichotomized the data into participants reporting impaired driving behaviors versus not reporting any impaired driving behaviors (ie, Impaired Driving Scale⫽0). The percentage reduction of self-reported impaired driving episodes over time for each AUDIT zone group is shown in Table 3. The greatest reported reduction in impaired driving episodes (23%) was observed with those participants in AUDIT zone 3 who received brief intervention. For motor vehicle safety other than alcohol impairment, at the 3-month follow-up there was no main effect found for the percentage of participants reporting always wearing their seatbelts (standard care percentage of change⫽6%; 95% CI ⫺2% to 14%; treatment percentage of change⫽7%; 95% CI 0% to 14%) or for change in percentage of reporting speeding at more than 10 miles per hour above the posted limit (standard care percentage of change⫽⫺4%; 95% CI ⫺8% to 0%; treatment⫽⫺7%; 95% CI ⫺15% to 1%).
LIMITATIONS The most significant limitation of our study was the large number of patients who were excluded from it, in part because of the many ED patients who were ineligible for our protocol because they were presenting for treatment of illnesses and not an injury. Future research in this area should address the efficacy of brief intervention for noninjured ED patients. Despite our having a certificate of confidentiality from National Institutes of Health and assuring patients of the confidentially of the research project, some motor vehicle crash patients may not have wanted to participate because of concerns of legal implication surrounding impaired driving. Additionally, we identified 109 eligible patients in the ED who expressed interest in being in the study, but we were unable to contact them, and another 16 voluntarily withdrew. These patients did not differ significantly from our study sample in age (mean age⫽30 years; 95% CI⫽28 to 32 years); sex (men⫽62%), race (white⫽72%), injury type (58% motor vehicle crash), or, for the 76 who provided screening Volume , . : June
information, on weekly alcohol quantity and frequency of intake (mean⫽8.9; 95% CI 5.5 to 12.0). Another limitation is that the sample size needed to conduct the secondary analyses of the interaction of alcohol use zone with treatment was not part of the original sample size calculation. Future studies would need to verify our findings by estimating and testing with an appropriate sample size for measuring these effects. On average, participants received the telephone assessment within 5 days (range 1 to 28 days; median⫽4 days) of being recruited into the study in the ED. Participants randomized into the treatment condition were on average involved for 16 days to incorporate the baseline brief intervention and the subsequent booster brief intervention (range 5 to 49 days; median⫽15 days). Because the 3-month assessment was scheduled from the date of the baseline telephone assessment, there was on average less of a lag between active study involvement and the 3-month assessment for the treatment group, which is often a feature of randomized controlled trials in which the assessment-only group is involved with the study protocol for a shorter period. The impact that differential time from completing either the assessment-only contact or the assessment and intervention contact had on the treatment outcomes is most likely minimal, as demonstrated by the lack of difference between standard care and treatment groups for Impaired Driving Scale, except with AUDIT zone 3. Not all patients in the treatment group received both the initial brief intervention and the follow-up booster brief intervention. Our 69% completion rate for the subsequent booster session is not surprising, though, for this non– treatment-seeking population. Other studies of face-to-face brief intervention with injured ED patients also have reported booster session completion rates similar to this, with failure to attend having a small but negative effect on treatment outcome.9 Our use of an intention-to-treat analysis, including all those randomized to brief intervention regardless of Annals of Emergency Medicine 761
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 3. All-sample treatment and control. Each vertical line represents a study participant. The line goes from baseline IDS to 3-month IDS. Lines that go up from baseline reference line indicate an increase in the impaired driving scale during that period.
completion of the booster session, may understate the magnitude of the effect found.
DISCUSSION Previous research31 has demonstrated the effectiveness of brief intervention delivered in an ED setting. Outcomes have included a decrease of negative consequences from alcohol use with injured ED patients9 and a decrease in reported impaired driving after a brief intervention delivered in the ED to alcohol-using adolescents.28 Our study expands on these results by demonstrating decreased impaired driving after a telephone brief intervention on hazardous alcohol use with adult injured ED patients. As in earlier works,9,32-35 decreases in alcohol use were found in treatment and standard care groups in our study. It has been postulated for ED brief intervention studies that this could be an effect of the injury causing the ED visit, an intervention effect of 762 Annals of Emergency Medicine
Mello et al doing the baseline assessment, or regression to the mean of alcohol use (though less likely with randomized controlled trial design36). Despite this decrease in overall alcohol use found in both groups, it is only the group receiving the brief intervention that had a decrease in impaired driving behaviors. This was demonstrated by the mean score on the Impaired Driving Scale decreasing in the treatment group, as well as more patients in the treatment group (7%) moving to the no-impaired-driving behaviors (Impaired Driving Scale⫽0) than the standard care group (⬍1%). Although an overall treatment effect of 7% might seem small, in the context of the public health problem of impaired driving, with an estimated 120 million events annually,37 it would be a substantive improvement. There were no observed effects of the intervention on other measured high-risk driving behaviors (ie, seatbelts, speeding). Considering that our brief intervention specifically addressed highrisk alcohol use, it is not surprising that the brief intervention decreased only alcohol-impaired driving. Our additional finding that those in the zone with greatest alcohol use problems had the most benefit is enlightening. In this group, our telephone brief intervention for alcohol decreased those with impaired driving behaviors by 23%. Although the brief intervention did not significantly decrease overall alcohol use in this zone’s treatment group compared with the standard care group, it clearly did have an effect on this zone’s impaired driving behaviors. This effect was not found in those zones with fewer alcohol use problems. Although this may be due to the lower number of impaired driving episodes in these groups at baseline or a short follow-up period, it may also suggest that brief intervention is not effective with these groups with lower alcohol problem severity in addressing impaired driving. Both of these findings suggest that there may be a select alcohol use profile for ED patients in which there is a benefit from brief intervention. Although we have previously reported that a motor vehicle crash may cause an ED brief intervention to have more of a treatment effect,10 we did not find this in the present study. This finding may be due to a difference in the delivery of the brief intervention (by telephone), the shorter follow-up period, or the comparison measures used. The previous reported finding was based on 12month alcohol-related injuries, and here we report on 3-month alcohol use and impaired driving only. Further study will determine whether the previous effect of a motor vehicle crash will occur at 12 months in reducing alcohol-related injuries when brief intervention is delivered by telephone. Our study found that a brief intervention for alcohol use could be delivered successfully to ED patients by telephone in the days after their discharge from the ED. Though screening and brief intervention is being advocated by the American College of Emergency Physicians12 and American College of Surgeons,38,39 the mechanism of how screening and brief intervention should be invoked into clinical ED practice is not clearly established. In the busy ED environment with increasing demands on clinical staff, it is important to find a practical mechanism for brief intervention delivery. DIAL offers a potential model for this to occur with utilizing a minimal Volume , . : June
Mello et al amount of ED clinicians’ time. Although a research assistant was present in the ED to identify patients, a future model of this research could be translated to have ED staff screen and identify patients who are injured and using alcohol in risky ways. Clinicians could refer identified risky drinkers to receive a follow-up telephone call for a brief intervention delivered by a trained counselor in the days after discharge. Telephone brief intervention also offers the advantage of providing a brief intervention by a qualified counselor in a patient’s native language. A counselor who is fluent in the desired language and trained in motivational counseling could be located in any part of the country, call the patient, and conduct the brief intervention in the native language of the patient. Telephone brief intervention also has potential advantages to EDs that vary in size and resources in providing the possibility of sharing prevention resources across health systems. In our study, we recruited from a large trauma center, as well as two local community hospitals, with all interventions being delivered by the same staff based at our research offices. The proximity of injury and intervention may create a “teachable moment,” giving the counselor an opportunity to help patients explore this association and motivate them to plan for ways in which to reduce or eliminate negative consequences from alcohol in the future. Although the ED visit is more proximate to the injury, it may not necessarily be the environment to best deliver the intervention. During a telephone brief intervention given soon after an ED visit for injury, the patient is more likely to be sober and alert while still having the emotional and physical reaction to the injury he or she had when it happened. That is, the teachable moment should still be in place. It may not be a “moment” but a period that persists beyond the ED visit. More research is needed to explore this hypothesis and examine what the optimal period is for an intervention after an ED visit. In summary, this study found that a brief intervention for alcohol could be successfully delivered by telephone to ED patients soon after discharge. Furthermore, the intervention decreased impaired driving, and this effect was greatest for those with more severe alcohol problems. Telephone brief intervention appears to offer an alternative mechanism to deliver brief intervention to this at-risk population and can assist with decreasing impaired driving. Supervising editor: Debra E. Houry, MD, MPH Author contributions: MJM, RL, TN, and RW conceived and designed the study. MJM obtained funding. All authors participated in the conduct of the study. JB provided statistical assistance. MJM drafted the article, and all authors contributed substantially to its revisions. MJM takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article, that might create any potential conflict of interest. See the Manuscript Submission Agreement in this issue for examples Volume , . : June
Telephone Brief Intervention for High-Risk Alcohol Use of specific conflicts covered by this statement. The study was funded by the Centers for Disease Control and Prevention’s (CDC’s) National Center for Injury Prevention and Control (R49/CCR1232280; Mello, principal investigator) and is registered at clinicaltrials.gov (NCT00457548). Disclaimer: The contents of this publication are solely the responsibility of the authors and do not necessary represent the official views of the CDC. Publication dates: Received for publication May 31, 2007. Revisions received September 23, 2007; November 7, 2007; and November 14, 2007. Accepted for publication November 27, 2007. Available online April 23, 2008. Presented in part at the 2007 American College of Emergency Physicians Research Forum, October 2007, Seattle, WA. Reprints not available from authors. Address for correspondence: Michael J. Mello, MD, MPH, Injury Prevention Center at Rhode Island Hospital, Department of Emergency Medicine, 592 Eddy Street, Claverick 2, Providence, RI 02903; 401-444-2685, fax 401-444-2249; E-mail [email protected]. REFERENCES 1. World Health Organization. Alcohol and Injury in the Emergency Department: Summary of the Report From the WHO Collaborative Study on Alcohol and Injuries. Geneva, Switzerland: World Health Organization; 2007. 2. National Highway Traffic Safety Administration. Motor vehicle traffic crash fatality counts and estimates of people injured for 2006. September 2007. Available at: http://www.nhtsa.gov/ portal/site/nhtsa/menuitem.6a6eaf83cf719ad24ec86e10dba 046a0. Accessed November 1, 2007. 3. Blincoe L, Seay A, Zaloshnja E, et al. The Economic Impact of Motor Vehicle Crashes 2000. Washington, DC. NHTSA 2002. DOT HS 809 446. 4. Becker B, Woolard R, Nirenberg TD, et al. Alcohol use among subcritically injured emergency department patients. Acad Emerg Med. 1995;2:784-790. 5. Maio RF, Waller PF, Blow FC, et al. Alcohol abuse/dependence in motor vehicle crash victims presenting to the emergency department. Acad Emerg Med. 1997;4:256-262. 6. Longabaugh R, Minugh PA, Nirenberg TD, et al. Injury as a motivator to reduce drinking. Acad Emerg Med. 1995;2:817-825. 7. Gentilello LM, Rivara FP, Donovan DM, et al. Alcohol interventions in a trauma center as a means of reducing the risk of injury recurrence. Ann Surg. 1999;230:473-480; discussion 480-483. 8. Schermer CR, Moyers TB, Miller WR, et al. Trauma center brief interventions for alcohol disorders decrease subsequent driving under the influence arrests. J Trauma. 2006;60:29-34. 9. Longabaugh R, Woolard RE, Nirenberg TD, et al. Evaluating the effects of a brief motivational intervention for injured drinkers in the emergency department. J Stud Alcohol. 2001;62:806-816. 10. Mello MJ, Nirenberg TD, Longabaugh R, et al. Emergency department brief motivational interventions for alcohol with motor vehicle crash patients. Ann Emerg Med. 2005;45:620-625. 11. Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change. New York, NY: Guilford Press; 2002. 12. American College of Emergency Physicians. Alcohol screening in the emergency department (Policy #400346, Approved January 2005). Available at: http://www3.acep.org/practres.aspx?id⫽ 29074. Accessed November 1, 2007.
Annals of Emergency Medicine 763
Telephone Brief Intervention for High-Risk Alcohol Use
Mello et al
13. National Highway Traffic Safety Administration. Identification and Referral of Impaired Drivers Through Emergency Department Protocols. National Technical Information; 2002. DOT HS 809 412. 14. Fleming MF, Mundt MP, French MT, et al. Brief physician advice for problem drinkers: long-term efficacy and benefit-cost analysis. Alcohol Clin Exp Res. 2002;26:36-43. 15. Balas EA, Jaffrey F, Kuperman GJ, et al. Electronic communication with patients. Evaluation of distance medicine technology. JAMA. 1997;278:152-159. 16. Mohr DC, Likosky W, Bertagnolli A, et al. Telephone-administered cognitive behavioral therapy for the treatment of depressive symptoms in multiple sclerosis. J Consult Clin Psychol. 2000;68: 356-361. 17. Simon GE, Ludman EJ, Tutty S, et al. Telephone psychotherapy and telephone care management for primary care patients starting antidepressant treatment: a randomized controlled trial. JAMA. 2004;292:935-942. 18. Swinson RP, Fergus KD, Cox BJ, et al. Efficacy of telephoneadministered behavioral therapy for panic disorder with agoraphobia. Behav Res Ther. 1995;33:465-469. 19. Resnicow K, Jackson A, Wang T, et al. A motivational interviewing intervention to increase fruit and vegetable intake through black churches: results of the Eat for Life trial. Am J Public Health. 2001;91:1686-1693. 20. Zhu SH, Stretch V, Balabanis M, et al. Telephone counseling for smoking cessation: effects of single-session and multiple-session interventions. J Consult Clin Psychol. 1996;64:202-211. 21. Lando HA, Hellerstedt WL, Pirie PL, et al. Brief supportive telephone outreach as a recruitment and intervention strategy for smoking cessation. Am J Public Health. 1992;82:41-46. 22. Curry SJ, McBride C, Grothaus LC, et al. A randomized trial of self-help materials, personalized feedback, and telephone counseling with nonvolunteer smokers. J Consult Clin Psychol. 1995;63:1005-1014. 23. Midanik LT, Greenfield TK, Rogers JD. Reports of alcohol-related harm: telephone versus face-to-face interviews. J Stud Alcohol. 2001;62:74-78. 24. Babor TF, Higgins-Biddle JC, Saunders J, et al. The Alcohol Use Disorders Identification Test. Guidelines for Use in Primary Care. 2nd ed. World Health Organization; 2001. 25. Donovan DM, Kivlahan DR, Doyle SR, et al. Concurrent validity of the Alcohol Use Disorders Identification Test (AUDIT) and AUDIT zones in defining levels of severity among out-patients with
26.
27. 28.
29.
30. 31.
32.
33.
34.
35.
36. 37.
38.
39.
alcohol dependence in the COMBINE study. Addiction. 2006;101:1696-1704. National Institute on Alcohol Abuse and Alcoholism. The Physician’s Guide to Helping Patients With Alcohol Problems. Washington, DC: DHHS; 1995. NIH Publication No. 95-3769. Donovan JE. Young adult drinking-driving: behavioral and psychosocial correlates. J Stud Alcohol. 1993;54:600-613. Monti PM, Colby SM, Barnett NP, et al. Brief intervention for harm reduction with alcohol-positive older adolescents in a hospital emergency department. J Consult Clin Psychol. 1999;67:989-994. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System survey questionnaire. US Department of Health and Human Services, Centers for Disease Control and Prevention. Available at: http://www.cdc.gov/brfss/index.htm. Accessed November 1, 2007. Gravetter F, Wallnau L, Essentials of Statistics for the Behavioral Sciences. 5th ed. Belmont, CA: Thomson; 2005. Nilsen P, Baird J, Mello M, et al. A systematic review of emergency care brief alcohol interventions for injury patients. Journal of Substance Abuse Treatment. In press. Gentilello L, Rivara F, Donovan D, et al. Alcohol interventions in a trauma center as a means of reducing the risk of injury recurrence. Ann Surg. 1999;230:473-483. Dauer AR, Rubio ES, Coris ME, et al. Brief intervention in alcoholpositive traffic casualties: is it worth the effort? Alcohol Alcohol. 2006;41:76-83. Sommers MS, Dyehouse JM, Howe SR, et al. Effectiveness of brief interventions after alcohol-related vehicular injury: a randomized controlled trial. J Trauma. 2006;61:523-531; discussion 532-533. Neumann T, Neuner B, Weiss-Gerlach E, et al. The effect of computerized tailored brief advice on at-risk drinking in subcritically injured trauma patients. J Trauma. 2006;61:805814. Streiner DL. Regression toward the mean: its etiology, diagnosis, and treatment. Can J Psychiatry. 2001;46:72-76. Dellinger AM, Bolen J, Sacks JJ. A comparison of driver- and passenger-based estimates of alcohol-impaired driving. Am J Prev Med. 1999;16:283-288. American College of Surgeons Committee on Trauma. Resources for Optimal Care of the Injured Patient: 2006. American College of Surgeons Committee on Trauma; 2006 Gentilello LM. Alcohol and injury: American College of Surgeons Committee on Trauma Requirements for Trauma Center Intervention. J Trauma. 2007;62(6 suppl):S44-45.
Did you know? You can personalize the new Annals of Emergency Medicine Web site to meet your individual needs.
Visit www.annemergmed.com today to see what else is new online!
764 Annals of Emergency Medicine
Volume , . : June
DIAL: A Telephone Brief Intervention for High-Risk Alcohol Use With Injured Emergency Department Patients Michael J. Mello, MD, MPH Richard Longabaugh, EdD Janette Baird, PhD Ted Nirenberg, PhD Robert Woolard, MD
From the Injury Prevention Center at Rhode Island Hospital, Department of Emergency Medicine, Warren Alpert Medical School of Brown University, Providence, RI (Mello, Baird, Nirenberg, Woolard); and Brown University Center of Alcohol and Addictions, Providence, RI (Longabaugh, Nirenberg).
Study objective: Brief interventions for high-risk alcohol use for injured emergency department (ED) patients have demonstrated effectiveness and may have a more pronounced effect with motor vehicle crash patients. We report on 3-month outcome data of a randomized controlled trial of injured patients, using a novel model of telephone-delivered brief interventions after ED discharge. Methods: ED research assistants recruited adult injured patients who screened positive for high-risk alcohol use and were to be discharged home. After discharge, participants received by telephone an assessment of alcohol use and impaired driving and then were randomized to treatment (n⫽140) or standard care (n⫽145). Treatment consisted of 2 sessions of brief interventions done by telephone, focusing on risky alcohol use. At 3 months, both groups had an assessment of alcohol use and impaired driving. Results: Two hundred eighty-five patients were randomized and had a baseline mean Alcohol Use Disorders Inventory Test (AUDIT) score of 11.0 (SD⫽7.4). Three-month follow-up assessments were completed on 273 (95%). Mean AUDIT score decreased in both the treatment (mean change⫽⫺3.4; 95% confidence interval [CI] ⫺4.5 to ⫺2.3) and standard care group (mean change⫽⫺3.2; 95% CI ⫺4.2 to ⫺2.2). Measures of impaired driving decreased for the treatment group (mean change⫽⫺1.4 95%; CI ⫺3.0 to 0.2) compared with standard care group (mean change⫽1.0; 95% CI ⫺0.9 to 2.9; P⫽.04; d⫽0.31). Participants were stratified post hoc into 3 groups by baseline alcohol problem, with the treatment effect only being in the highest-scoring group (d⫽.30). Conclusion: Telephone brief interventions decreased impaired driving in our treatment group. Telephone brief intervention appears to offer an alternative mechanism to deliver brief intervention for alcohol in this at-risk ED population. [Ann Emerg Med. 2008;51:755-764.] 0196-0644/$-see front matter Copyright © 2008 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2007.11.034
INTRODUCTION Alcohol has a well-documented role in injury occurrence.1 There is a particularly strong connection between alcohol use and motor vehicle crashes, with alcohol involved in approximately 41% of fatal motor vehicle crashes and 256,000 persons being injured in alcohol-related motor vehicle crashes.2 The resultant economic costs of impaired driving are staggering and estimated to be more than $50 billion.3 Thus, impaired driving is a public health problem with significant human suffering and societal cost that necessitates further efforts at control. Emergency departments (EDs) are an opportune setting for identifying persons using alcohol in a risky manner, especially in Volume , . : June
those presenting for injury4 or from a motor vehicle crash.5 Injury has also been identified as a motivator to change alcohol use.6 In patients admitted to a trauma service, a brief intervention for alcohol has been found to reduce injury recidivism7 and subsequent impaired driving arrests.8 With injured patients being treated in an ED and discharged to home, Longabaugh et al9 found that those with hazardous alcohol use who received brief intervention for alcohol while in the ED and a follow-up booster brief intervention within 2 weeks reduced alcohol-related negative consequences and alcohol-related injuries more than those who received only standard ED care. This effect was not dependent on whether the patients had consumed alcohol before their injury. Annals of Emergency Medicine 755
Telephone Brief Intervention for High-Risk Alcohol Use Editor’s Capsule Summary
What is already known on this topic There is some evidence that brief motivational interventions in the emergency department (ED) can decrease risky behaviors, but it is difficult to consistently deliver these messages in the chaotic ED environment. What question this study addressed Can a brief telephone intervention delivered after an ED visit to patients with risky alcohol use decrease alcohol consumption and related risky behavior? What this study adds to our knowledge In this 285-patient randomized controlled trial, the brief telephone intervention had little effect on self-reported risky alcohol use but had an effect by slightly decreased self-reported impaired driving at 3 months in the treatment group. How this might change clinical practice The teachable moment can extend beyond the ED visit, and interventions may not need to be conducted during the visit to be effective.
Furthermore, the treatment effect was greatest in those patients being treated in the ED after a motor vehicle crash.10 All of these studies7-10 used a brief intervention that was based on the technique of motivational interviewing.11 Thus, injured ED patients, motor vehicle crash patients in particular, appear to be a group to target for screening and a brief intervention for alcohol problems. Although organizations have advocated for brief intervention for alcohol to be done within the clinical practice of the ED,12,13 barriers exist that have prevented its universal adoption. In the busy environment of the ED, diagnosis and treatment of acute medical illness and injury regularly usurp prevention efforts. ED staff may perceive this as not being part of their clinical care of the injured patients or may not have the skills or additional resources to address alcohol use problems. For patients, the ED may not be the ideal environment for a brief intervention because it is a noisy, chaotic, stressful, and uncomfortable setting. While in the ED, the patient is likely to be fatigued, in pain, and possibly under the influence of alcohol, which may affect the efficacy of the intervention. The “teachable moment” perhaps does not only exist in the ED, but opportunities to offer an intervention for alcohol use may extend to a period beyond the initial ED visit. This view has led us to now explore whether brief intervention for alcohol would be effective with ED patients after discharge from the ED. A practical mechanism for reaching patients after discharge from the ED would be using the telephone. Telephone interventions have demonstrated efficacy as a modality to change behavior. Programs that include a telephone intervention component have 756 Annals of Emergency Medicine
Mello et al addressed varied health issues, including alcohol use in a primary care setting,14 management of chronic diseases,15 depressive symptoms,16,17 agoraphobia,18 nutrition,19 and smoking cessation.20-22 Midanik et al23 have also found, in comparing telephone and face-to-face interviews assessing alcohol-related harm, that the telephone survey yielded significantly higher rates of self-reported alcohol-related harm compared with the in-person survey, possibly because of increased anonymity with telephone surveys. Although only some of these telephone interventions are motivational interviewing– based brief intervention and none use an ED population, the existing barriers to brief intervention being done in the ED and the practicality of the telephone necessitates telephone-delivered brief intervention testing with an ED population. We do not know which ED patients would benefit from a brief intervention. If brief intervention is found to be effective with certain subgroups of ED patients, limited resources could be concentrated with those groups. Previous research has demonstrated a more pronounced effect of brief intervention with ED motor vehicle crash patients than other injured patients.10 This finding, along with the prevalence of alcohol use problems in motor vehicle crash patients,5 yields utility for this subgroup to be examined for treatment effect with a telephone brief intervention for alcohol. Also, we lack data on the number of alcohol problems among injured ED patients that provide the floor and ceiling for brief intervention to be effective. It may be that brief intervention has a differential effect on ED patients that depends on their alcohol problem severity. Using the Alcohol Use Disorders Inventory Test (AUDIT)24 screening data from a treatment-seeking sample, Donovan et al25 found that classifying patients according to 4 AUDIT score zones produced a linear relationship between the zone and measures of alcohol problem severity. Thus, AUDIT score zones appear to be a reliable instrument to divide patients into groups based on the number of alcohol problems and could be used to examine whether ED patients’ alcohol problem severity predicts their response to brief intervention. Goal of This Investigation With the telephone being used successfully in counseling and having potential advantages for delivering brief intervention with ED patients, we decided to alter the locus of brief intervention for ED patients. The primary aim of the present study is to test the effect of brief intervention given by telephone soon after the patient’s ED visit and followed by a “booster” brief intervention telephone session 2 weeks later on the extent of hazardous drinking and frequency of impaired driving at a 3-month follow-up. The secondary aims are to test whether the effectiveness of brief intervention given by telephone is moderated by the patient’s initial alcohol problem severity and to test whether ED motor vehicle crash patients are more responsive to brief intervention given by telephone than are other injured ED patients. Volume , . : June
Mello et al
MATERIALS AND METHODS Study Design DIAL, Decreasing Injuries from ALcohol, is a randomized clinical trial using a 2-group design for injured ED patients discharged to home. All participants were recruited while in the ED, but enrollment, assessment, randomization, and all interventions were done by telephone in the days after discharge. The treatment group received a brief intervention about their high-risk alcohol use during that same telephone call and a subsequent brief telephone booster brief intervention session. The control group received standard care for their ED visit, with no study interventions. Both groups received by telephone an assessment of alcohol use and impaired driving before randomization and at 3 months. All protocols were reviewed and approved by the institutional review boards at the study institutions. DIAL was funded by the Centers for Disease Control and Prevention’s National Center for Injury Prevention and Control and registered with clinicaltrials.gov. Setting Participants were recruited from the ED of a large, urban, academic, Level I trauma center and 2 smaller community hospital EDs in nearby suburban settings. Selection of Participants Participants were recruited in the ED by research assistants during preselected times representing all shifts and days of the week, from November 2003 to June 2006. The research assistants were scheduled for more shifts in the trauma center ED than both community hospitals combined. All noncritically injured adult (18 years or older) patients were screened during their ED visit with a health behaviors survey that included questions on smoking, exercise, stress, and amount of alcohol use. Those using alcohol at risky levels were eligible for the study. Risky alcohol use was defined using National Institute on Alcohol Abuse and Alcoholism quantity frequency guidelines (either more than 14 drinks/week for men and more than 7 drinks/week for women or 5 or more drinks/occasion for men and 4 or more drinks/occasion for women).26 Additional eligibility requirements included being discharged from the ED to home, having a telephone, English speaking, not suicidal, and not in police custody. In the first year of the study, we recruited only motor vehicle crash patients. To increase enrollment for the remainder of the study, both motor vehicle crash patients and other injured patients were recruited. This broadens the sample’s representation of injured ED patients and permitted us the opportunity to test whether motor vehicle crash patients had a differential treatment compared with other injured patients. Consent for participation in the study occurred in the ED. On approaching patients who screened positive for the alcohol criteria and met other eligibility criteria, the purpose of the study was explained to the patient. Using a scripted recruitment approach, the research assistant explained the schedule of telephone contacts, what would happen during the assessment portion of the telephone call, and how randomization was Volume , . : June
Telephone Brief Intervention for High-Risk Alcohol Use conducted. It was also explained to the patient before signing of the consent form that, if randomized into the treatment group, the participant would be invited at that time to engage with the trained interventionists in a brief conversation about the answers they had given during the assessment. It was further described that this conversation would be about alcohol use and driving behaviors. The patient was informed that the treatment group would also have a booster telephone call scheduled within 14 days of the first call. All participants were compensated for assessments done at both baseline and 3-month follow-up, for a total of $70. We based the sample size on the results of a previous randomized clinical trial at our institution on the effectiveness of face-to-face brief intervention delivered in the ED.9 Analysis of the data from that study demonstrated that brief intervention resulted in an effect size of d⫽0.34 for reducing alcohol-related injuries among patients who had been injured by a motor vehicle crash.10 Using Cohen’s procedure for calculating sample size, with a power of 0.80 and an ␣ of 0.05, we needed to recruit 114 participants per treatment condition to evaluate differences in treatment means. We estimated an attrition rate of 20%, also based on the previous ED research.9 This yielded a total sample recruitment of 143 participants per treatment condition. Interventions Counselors initiated telephone calls to participants who were recruited in the ED within 5 days after their discharge. Patients contacted by telephone were offered enrollment in the study. Those enrolled received an assessment of their alcohol use and impaired driving. After assessment, participants were randomized by computer assignment to either treatment or standard care group. Participants in the treatment group received a brief intervention of approximately 30 minutes’ duration. Two weeks later, they received an additional 15-minute brief intervention booster session. The primary goal of brief intervention is the mobilization of the participant’s own resources to bring about changes needed to reduce alcohol-related risky behaviors. The counselor’s role is to be reflective, to provide an atmosphere that will support and enhance the participant’s motivation for change and that will lead the participant to initiate and persist in behavioral change efforts.11 Counselors were master’s- or doctoral-level staff trained and certified in motivational interviewing by a certified trainer and practitioner of motivational interviewing, who developed a script on introducing the main themes of motivational interviewing when delivered by telephone. Intervention fidelity was maintained through weekly clinical supervision sessions with a licensed clinical psychologist with experience in MI. Those participants allocated to standard care condition received no study interventions. None of the 3 study EDs had clinical protocols for routine standardized alcohol screening or brief interventions with patients to be discharged. Acute intoxicated patients are kept in the ED until sober in all study EDs and may be referred to outside treatment facilities at the discretion of the ED clinician. Annals of Emergency Medicine 757
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 1. Impaired driving scale items.
Methods of Measurement Both groups received the same assessments by telephone at baseline and at 3 months. A research assistant blinded to treatment condition contacted participants at 3 months for follow-up assessment. Risky alcohol use was measured by AUDIT,24 a 10-item instrument screening for hazardous or harmful drinking over the last 3 months (scores range from 0 to 40) and impaired driving behavior using the Impaired Driving Scale. The Impaired Driving Scale (Figure 1) is a 6-item scale that is based on 4 questions in the Drinking and Driving Scale27,28 and 2 from the Behavioral Risk Factor Surveillance System29 that measures impaired driving. The Impaired Driving Scale measure has a Cronbach ␣ of 0.95. Other assessments done at baseline and 3 months included other high-risk driving behaviors (ie, seatbelt use and the frequency of speeding).
Mello et al Driving Scale) between baseline and 3-month assessment within groups. Before data analysis, the raw scores from these measures were analyzed and appropriate transformations were conducted where necessary. Raw scores on the items of the Impaired Driving Scale were converted to t scores to standardize items and reduce the effects of items with different time metrics.30 An intention-to-treat analysis plan was used for all participants randomized in the study. A generalized linear model analysis of the effects of treatment, time, and a treatment-by-time interaction was used to analyze changes in 3-month AUDIT and Impaired Driving Scale scores. Where appropriate, means, standard deviations, or confidence intervals (CIs) at the 95% significance level are reported. To examine for moderation of treatment effect by extent of risky alcohol use, groups were divided into 3 zones by their baseline AUDIT score (zone 1 score less than 8 [n⫽122]; zone 2 score 8 to 15 [n⫽101]; zone 3 score greater than or equal to 16 [n⫽63]) and then analyzed for change in alcohol use and impaired driving between baseline and 3 months. Categorizing AUDIT into zones has been described previously in guidelines by the World Health Organization (WHO)24 and others.25 We created 3 zones by collapsing WHO’s third and fourth zones into a single highest-risk zone, which increased the size of the group for analysis and had the practical implication of the entire resultant zone (AUDIT greater than or equal to 16) being engaged in “hazardous alcohol use.”24 Post hoc tests of paired differences were conducted on any significant main and interaction effects to determine treatment differences among the treatment zone groups. Analysis examining the effects of cause of injury (motor vehicle crash or other injury cause) on difference in treatment effect for alcohol use and impaired driving was also conducted. Data are presented by outcome analysis (statistic and 95% CI) for the standard care and treatment group and by the interaction between treatment condition and baseline AUDIT zone.
RESULTS Data Collection and Processing Data were collected on paper by research assistants then transferred to an electronic database (Illume 2.2; DatStat, Seattle, WA). A double data entry system was used by 2 research assistants who would independently enter data before the database was electronically checked for inconsistencies. Inconsistent entries were then reviewed by the project manager and corrected to be consistent with data collection sheets. Data were electronically transferred to SAS version 9.1 (SAS Institute, Inc., Cary, NC) for analysis. Primary Data Analysis Descriptive analyses were conducted on all baseline data and contrasted treatment and standard care groups, as well as motor vehicle crash participants, with other injured participants. Because our study sought to reduce risky alcohol use and impaired driving, we then examined outcome measures of AUDIT summary score and impaired driving (Impaired 758 Annals of Emergency Medicine
Characteristics of Study Subjects From November 2003 to June 2006, 17,234 patients were treated in the EDs while research assistants were present. The majority of patients were ineligible for our study because they were there for treatment of a medical illness, were to be admitted to the hospital, or were too clinically unstable to be screened. During this period, there were 6,335 motor vehicle crash and other injured ED patients for the research assistants to screen for eligibility. Research assistants did not screen 249 patients in this group because the patients refused screening, were discharged before screening, or were otherwise unable to be located. The other 6,086 patients were screened to determine whether they met protocol eligibility and their willingness to participate. There were 1,329 patients who met protocol eligibility criteria, and 410 were interested and provided consent in the ED to receive a telephone call in the days after discharge for enrollment, assessment, and randomization. Of that group, 125 were not enrolled because we were unable to contact them Volume , . : June
Mello et al
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 2. Subject recruitment and disposition. MVC, Motor vehicle crash.
or they voluntarily withdrew. The study sample consisted of 285 participants contacted by telephone, who completed the baseline assessment and were randomized into treatment (n⫽140) or standard care (n⫽145) groups (see Figure 2). Participants were recruited from a Level I trauma center ED (n⫽253; 89%) and community hospitals’ EDs (n⫽31; 11%). A descriptive analysis of groups was conducted, examining sex, age, race, ethnicity, years of schooling, employment, injury type (motor vehicle crash or other), and baseline alcohol use (Table 1). The treatment and standard care groups were similar with respect to their baseline AUDIT and Impaired Driving Scale scores (Tables 2 and 3) and mechanism of injury (motor vehicle crash versus other injury). Telephone contact was successful on the first attempt in 50% of the study sample. More than 90% were successfully reached by the Volume , . : June
third attempt. The average time from ED discharge until assessment and randomization was 5 days (range 1 to 28 days; median⫽4 days). For those in the treatment group, the average length of time for the first brief intervention session was 31.2 minutes (SD⫽13.4; range 10 to 77 minutes; median⫽30 minutes); the average length of time for the booster brief intervention session was 18.8 minutes (SD⫽9.28; range 5 to 50 minutes; median⫽15 minutes). During the initial telephone call, 92% of participants completed the brief intervention immediately after completing the baseline assessment; the other 8% were recontacted later by telephone, on average 9 days after randomization, to receive the brief intervention. The booster brief intervention session was received by 69% (n⫽97) of participants in the treatment group. At 3-month follow-up, assessments were obtained on 272 (95%) of randomized participants. Annals of Emergency Medicine 759
Telephone Brief Intervention for High-Risk Alcohol Use
Mello et al
Table 1. Description of participant characteristics and baseline data. Demographic Characteristics
Treatment Group (nⴝ140)
Sex, % Age, y Race, % (White versus non-White) Ethnicity, % (Hispanic versus nonHispanic) Years of schooling, mean Employment,% (employed/ unemployed) Injury type, MVC, % (mvc versus other injury) Baseline alcohol use, quantity/ frequency Baseline alcohol use, most alcohol in the past month
SC Group (nⴝ145)
Eligible, Not Assessed*(nⴝ123)
Male, 61 Mean⫽28 (SD⫽9.8) White, 76 Non-Hispanic, 90
Male, 61 Mean⫽31 (SD⫽10.7) White, 69% Non-Hispanic, 84
Male, 62 Mean⫽30 (SD⫽11.2) White, 72 Non-Hispanic, 80
13.4 74
13.0 75
NA NA
51
52
59
Mean⫽11.7 drinks/wk (SD⫽12.4)
Mean⫽11.0 drinks/ wk (SD⫽12.3) Mean⫽7.8 drinks (SD⫽3.4)
Mean⫽8.9 drinks/wk (SD⫽14.7)
Mean⫽8.2 drinks (SD⫽3.2)
Mean⫽8.5 drinks (SD⫽4.9)
NA, Not available. *Patients recruited and consented in ED but not enrolled, not randomized, or do not have baseline assessment completed.
Table 2. Treatment effect on 3-month alcohol use. Outcome
Baseline Mean (nⴝ285)
Three-Month Mean (nⴝ273)
Change, Mean Change With 95% CI
SC: mean⫽10.8 T: mean⫽11.5 SC1: mean⫽5.2 T1: mean⫽5.2 SC2: mean⫽10.6 T2: mean⫽11 SC3: mean⫽23.1 T3: mean⫽22.5 SC1: mean⫽1.1 T1: mean⫽1.0 SC2: mean⫽2.0 T2: mean⫽2.0 SC3: mean⫽3.0 T3: mean⫽3.0
SC: mean⫽7.6 T: mean⫽8.1 SC1: mean⫽4.6 T1: mean⫽5.0 SC2: mean⫽7.1 T2: mean⫽8.1 SC3: mean⫽15 T3: mean⫽12.9 SC1: mean⫽0.8 T1: mean⫽0.8 SC2: mean⫽1.5 T2: mean⫽1.6 SC3: mean⫽2.4 T3: mean⫽1.9
SC: mean⫽⫺3.2 (⫺4.2 to ⫺2.2) T: mean⫽⫺3.4 (⫺4.5 to ⫺2.3) SC1: mean⫽⫺0.6 (⫺1.2 to 0) T1: mean⫽⫺0.2 (⫺1.1 to 0.7) SC2: mean⫽⫺3.5 (⫺4.7 to ⫺2.3) T2: mean⫽–2.9 (⫺4.3 to ⫺1.5) SC3: mean⫽⫺8.3 (⫺13.8 to ⫺4.8) T3: mean⫽–9.7 (⫺12.7 to ⫺6.7) SC1: mean⫽⫺0.4 (⫺0.6 to ⫺0.2) T1: mean⫽⫺0.2 (⫺0.5 to 0.1) SC2: mean⫽⫺0.6(⫺09 to ⫺0.3) T2: mean⫽⫺0.4 (⫺0.7 to ⫺0.1) SC3: mean⫽⫺0.6 (⫺1.1 to ⫺0.1) T3: mean⫽⫺1.1 (⫺1.7 to ⫺0.3)
AUDIT score Total AUDIT score in zone 1 Total AUDIT score in zone 2 Total AUDIT score in zone 3 Binge drinking in zone 1 Binge drinking in zone 2 Binge drinking in zone 3
SC, Standard care group; T, treatment group; SC1, standard care group in zone 1; T1, treatment group in zone 1; SC2, standard care group in zone 2; T2, treatment group in zone 2; SC3, standard care group in zone 3; T3, treatment group in zone 3. Binge drinking⫽6 or more drinks per occasion.
Main Results An ANOVA model with omnibus F tests was conducted to examine the effects of time and treatment on 3-month AUDIT and impaired driving. No significant difference was found between treatment and standard care groups at 3-month follow-up on mean AUDIT scores and mean change in AUDIT scores (Table 2). There was an omnibus effect found for the model with 3-month impaired driving as the outcome variable (F(7, 537)⫽8.78; P⬍.001). As can be seen in Table 3 and Figure 3A, there was a lower report of impaired driving between baseline and 3 months in the treatment group (mean⫽49.1; mean change⫽–1.4; 95% CI ⫺3.0 to 0.2) but not the standard care group, in which there was an increase in reported 3-month impaired driving (mean⫽50.8; mean change⫽1.0; 95% CI ⫺0.9 to 2.9). Using Cohen’s d, the difference between the 3-month Impaired Driving Scale scores for the standard care and treatment groups corresponds to an effect size of 0.31. Because there was an overall effect of treatment and baseline AUDIT strata on 3-month impaired driving scores, further 760 Annals of Emergency Medicine
analysis was conducted (Figure 3B). This post hoc analysis of paired differences showed that when alcohol use was stratified into the 3 AUDIT zones, there was a significant effect of treatment group by baseline AUDIT zone (Table 3). The participants in the highest zone (AUDIT score greater than or equal to 16) receiving treatment had a lower 3-month impaired driving score (mean⫽52.3; mean change⫽⫺6.4; 95% CI ⫺13.2 to 0.4) than participants in the highest zone who were in the standard care group who reported on average increasing impaired driving (mean⫽57.8; mean change⫽5.1; 95% CI ⫺4.3 to 14.5). No significant effects were found in zones 1 or 2 (Table 3). We found there was no difference between those who were injured as a result of a motor vehicle crash and those injured in other ways for mean changes in Impaired Driving Scale scores over time (motor vehicle crash mean change⫽⫺0.10; 95% CI ⫺1.6 to 1.8; other injury mean change⫽⫺0.3; 95% CI ⫺2.4 to 1.8). To better understand the effects that brief intervention had on impaired driving behaviors (Impaired Driving Scale scores), Volume , . : June
Mello et al
Telephone Brief Intervention for High-Risk Alcohol Use
Table 3. Treatment effect on 3-month impaired driving measures. Outcome IDS IDS in zone 1 IDS in zone 2 IDS in zone 3 % Reporting Impaired Driving in Zone 1 % Reporting Impaired Driving in Zone 2 % Reporting Impaired Driving in Zone 3
Baseline Mean (nⴝ285)
Three–Month Mean (nⴝ273)
Change Mean or % with 95% Confidence Interval (CI)
SC: mean⫽49.8 T: Mean⫽50.5 SC1: mean⫽47.3 T1: Mean⫽47.4 SC2: mean⫽51 T2: Mean⫽48 SC3: Mean⫽52.7 T3: Mean⫽58.7 SC1: 34 T1: 39 SC2: 55 T2: 38 SC3: 56 T3: 65
SC: mean⫽50.8 T: Mean⫽49.1 SC1: mean⫽47.9 T1: Mean⫽47.9 SC2: Mean⫽50.4 T2: Mean⫽48.4 49.0) SC3: Mean⫽57.8 T3: Mean⫽52.3 SC1: 41 T1: 28 SC2: 55 T2: 49 SC3: 48 T3: 42
SC: Mean⫽1.0 (⫺0.9 to 2.9 T: Mean⫽⫺1.4 (⫺3.0 to 0.2) SC1: Mean⫽0.6 (0.2 to 1.0) T1: Mean⫽0.5 (0.2 to 0.8) SC2: Mean⫽⫺0.7 (⫺2.4 to 1.0) T2: Mean⫽0.3 (⫺0.3 to 0.9) SC3: Mean⫽5.1 (⫺4.3 to 14.5) T3: Mean⫽⫺6.4 (⫺13.2 to 0.4) SC1: 7% (⫺3.0 to 17.0%) T1: ⫺11%% (⫺19 to ⫺3.0%) SC2: 0% (⫺4 to 4.0%) T2: 11% (2.0 to 20.0%) SC3: ⫺8% (⫺30 to 14.0%) T3: ⫺23% (⫺37 to ⫺9.0%)
IDS, Impaired Driving Scale.
we dichotomized the data into participants reporting impaired driving behaviors versus not reporting any impaired driving behaviors (ie, Impaired Driving Scale⫽0). The percentage reduction of self-reported impaired driving episodes over time for each AUDIT zone group is shown in Table 3. The greatest reported reduction in impaired driving episodes (23%) was observed with those participants in AUDIT zone 3 who received brief intervention. For motor vehicle safety other than alcohol impairment, at the 3-month follow-up there was no main effect found for the percentage of participants reporting always wearing their seatbelts (standard care percentage of change⫽6%; 95% CI ⫺2% to 14%; treatment percentage of change⫽7%; 95% CI 0% to 14%) or for change in percentage of reporting speeding at more than 10 miles per hour above the posted limit (standard care percentage of change⫽⫺4%; 95% CI ⫺8% to 0%; treatment⫽⫺7%; 95% CI ⫺15% to 1%).
LIMITATIONS The most significant limitation of our study was the large number of patients who were excluded from it, in part because of the many ED patients who were ineligible for our protocol because they were presenting for treatment of illnesses and not an injury. Future research in this area should address the efficacy of brief intervention for noninjured ED patients. Despite our having a certificate of confidentiality from National Institutes of Health and assuring patients of the confidentially of the research project, some motor vehicle crash patients may not have wanted to participate because of concerns of legal implication surrounding impaired driving. Additionally, we identified 109 eligible patients in the ED who expressed interest in being in the study, but we were unable to contact them, and another 16 voluntarily withdrew. These patients did not differ significantly from our study sample in age (mean age⫽30 years; 95% CI⫽28 to 32 years); sex (men⫽62%), race (white⫽72%), injury type (58% motor vehicle crash), or, for the 76 who provided screening Volume , . : June
information, on weekly alcohol quantity and frequency of intake (mean⫽8.9; 95% CI 5.5 to 12.0). Another limitation is that the sample size needed to conduct the secondary analyses of the interaction of alcohol use zone with treatment was not part of the original sample size calculation. Future studies would need to verify our findings by estimating and testing with an appropriate sample size for measuring these effects. On average, participants received the telephone assessment within 5 days (range 1 to 28 days; median⫽4 days) of being recruited into the study in the ED. Participants randomized into the treatment condition were on average involved for 16 days to incorporate the baseline brief intervention and the subsequent booster brief intervention (range 5 to 49 days; median⫽15 days). Because the 3-month assessment was scheduled from the date of the baseline telephone assessment, there was on average less of a lag between active study involvement and the 3-month assessment for the treatment group, which is often a feature of randomized controlled trials in which the assessment-only group is involved with the study protocol for a shorter period. The impact that differential time from completing either the assessment-only contact or the assessment and intervention contact had on the treatment outcomes is most likely minimal, as demonstrated by the lack of difference between standard care and treatment groups for Impaired Driving Scale, except with AUDIT zone 3. Not all patients in the treatment group received both the initial brief intervention and the follow-up booster brief intervention. Our 69% completion rate for the subsequent booster session is not surprising, though, for this non– treatment-seeking population. Other studies of face-to-face brief intervention with injured ED patients also have reported booster session completion rates similar to this, with failure to attend having a small but negative effect on treatment outcome.9 Our use of an intention-to-treat analysis, including all those randomized to brief intervention regardless of Annals of Emergency Medicine 761
Telephone Brief Intervention for High-Risk Alcohol Use
Figure 3. All-sample treatment and control. Each vertical line represents a study participant. The line goes from baseline IDS to 3-month IDS. Lines that go up from baseline reference line indicate an increase in the impaired driving scale during that period.
completion of the booster session, may understate the magnitude of the effect found.
DISCUSSION Previous research31 has demonstrated the effectiveness of brief intervention delivered in an ED setting. Outcomes have included a decrease of negative consequences from alcohol use with injured ED patients9 and a decrease in reported impaired driving after a brief intervention delivered in the ED to alcohol-using adolescents.28 Our study expands on these results by demonstrating decreased impaired driving after a telephone brief intervention on hazardous alcohol use with adult injured ED patients. As in earlier works,9,32-35 decreases in alcohol use were found in treatment and standard care groups in our study. It has been postulated for ED brief intervention studies that this could be an effect of the injury causing the ED visit, an intervention effect of 762 Annals of Emergency Medicine
Mello et al doing the baseline assessment, or regression to the mean of alcohol use (though less likely with randomized controlled trial design36). Despite this decrease in overall alcohol use found in both groups, it is only the group receiving the brief intervention that had a decrease in impaired driving behaviors. This was demonstrated by the mean score on the Impaired Driving Scale decreasing in the treatment group, as well as more patients in the treatment group (7%) moving to the no-impaired-driving behaviors (Impaired Driving Scale⫽0) than the standard care group (⬍1%). Although an overall treatment effect of 7% might seem small, in the context of the public health problem of impaired driving, with an estimated 120 million events annually,37 it would be a substantive improvement. There were no observed effects of the intervention on other measured high-risk driving behaviors (ie, seatbelts, speeding). Considering that our brief intervention specifically addressed highrisk alcohol use, it is not surprising that the brief intervention decreased only alcohol-impaired driving. Our additional finding that those in the zone with greatest alcohol use problems had the most benefit is enlightening. In this group, our telephone brief intervention for alcohol decreased those with impaired driving behaviors by 23%. Although the brief intervention did not significantly decrease overall alcohol use in this zone’s treatment group compared with the standard care group, it clearly did have an effect on this zone’s impaired driving behaviors. This effect was not found in those zones with fewer alcohol use problems. Although this may be due to the lower number of impaired driving episodes in these groups at baseline or a short follow-up period, it may also suggest that brief intervention is not effective with these groups with lower alcohol problem severity in addressing impaired driving. Both of these findings suggest that there may be a select alcohol use profile for ED patients in which there is a benefit from brief intervention. Although we have previously reported that a motor vehicle crash may cause an ED brief intervention to have more of a treatment effect,10 we did not find this in the present study. This finding may be due to a difference in the delivery of the brief intervention (by telephone), the shorter follow-up period, or the comparison measures used. The previous reported finding was based on 12month alcohol-related injuries, and here we report on 3-month alcohol use and impaired driving only. Further study will determine whether the previous effect of a motor vehicle crash will occur at 12 months in reducing alcohol-related injuries when brief intervention is delivered by telephone. Our study found that a brief intervention for alcohol use could be delivered successfully to ED patients by telephone in the days after their discharge from the ED. Though screening and brief intervention is being advocated by the American College of Emergency Physicians12 and American College of Surgeons,38,39 the mechanism of how screening and brief intervention should be invoked into clinical ED practice is not clearly established. In the busy ED environment with increasing demands on clinical staff, it is important to find a practical mechanism for brief intervention delivery. DIAL offers a potential model for this to occur with utilizing a minimal Volume , . : June
Mello et al amount of ED clinicians’ time. Although a research assistant was present in the ED to identify patients, a future model of this research could be translated to have ED staff screen and identify patients who are injured and using alcohol in risky ways. Clinicians could refer identified risky drinkers to receive a follow-up telephone call for a brief intervention delivered by a trained counselor in the days after discharge. Telephone brief intervention also offers the advantage of providing a brief intervention by a qualified counselor in a patient’s native language. A counselor who is fluent in the desired language and trained in motivational counseling could be located in any part of the country, call the patient, and conduct the brief intervention in the native language of the patient. Telephone brief intervention also has potential advantages to EDs that vary in size and resources in providing the possibility of sharing prevention resources across health systems. In our study, we recruited from a large trauma center, as well as two local community hospitals, with all interventions being delivered by the same staff based at our research offices. The proximity of injury and intervention may create a “teachable moment,” giving the counselor an opportunity to help patients explore this association and motivate them to plan for ways in which to reduce or eliminate negative consequences from alcohol in the future. Although the ED visit is more proximate to the injury, it may not necessarily be the environment to best deliver the intervention. During a telephone brief intervention given soon after an ED visit for injury, the patient is more likely to be sober and alert while still having the emotional and physical reaction to the injury he or she had when it happened. That is, the teachable moment should still be in place. It may not be a “moment” but a period that persists beyond the ED visit. More research is needed to explore this hypothesis and examine what the optimal period is for an intervention after an ED visit. In summary, this study found that a brief intervention for alcohol could be successfully delivered by telephone to ED patients soon after discharge. Furthermore, the intervention decreased impaired driving, and this effect was greatest for those with more severe alcohol problems. Telephone brief intervention appears to offer an alternative mechanism to deliver brief intervention to this at-risk population and can assist with decreasing impaired driving. Supervising editor: Debra E. Houry, MD, MPH Author contributions: MJM, RL, TN, and RW conceived and designed the study. MJM obtained funding. All authors participated in the conduct of the study. JB provided statistical assistance. MJM drafted the article, and all authors contributed substantially to its revisions. MJM takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article, that might create any potential conflict of interest. See the Manuscript Submission Agreement in this issue for examples Volume , . : June
Telephone Brief Intervention for High-Risk Alcohol Use of specific conflicts covered by this statement. The study was funded by the Centers for Disease Control and Prevention’s (CDC’s) National Center for Injury Prevention and Control (R49/CCR1232280; Mello, principal investigator) and is registered at clinicaltrials.gov (NCT00457548). Disclaimer: The contents of this publication are solely the responsibility of the authors and do not necessary represent the official views of the CDC. Publication dates: Received for publication May 31, 2007. Revisions received September 23, 2007; November 7, 2007; and November 14, 2007. Accepted for publication November 27, 2007. Available online April 23, 2008. Presented in part at the 2007 American College of Emergency Physicians Research Forum, October 2007, Seattle, WA. Reprints not available from authors. Address for correspondence: Michael J. Mello, MD, MPH, Injury Prevention Center at Rhode Island Hospital, Department of Emergency Medicine, 592 Eddy Street, Claverick 2, Providence, RI 02903; 401-444-2685, fax 401-444-2249; E-mail [email protected]. REFERENCES 1. World Health Organization. Alcohol and Injury in the Emergency Department: Summary of the Report From the WHO Collaborative Study on Alcohol and Injuries. Geneva, Switzerland: World Health Organization; 2007. 2. National Highway Traffic Safety Administration. Motor vehicle traffic crash fatality counts and estimates of people injured for 2006. September 2007. Available at: http://www.nhtsa.gov/ portal/site/nhtsa/menuitem.6a6eaf83cf719ad24ec86e10dba 046a0. Accessed November 1, 2007. 3. Blincoe L, Seay A, Zaloshnja E, et al. The Economic Impact of Motor Vehicle Crashes 2000. Washington, DC. NHTSA 2002. DOT HS 809 446. 4. Becker B, Woolard R, Nirenberg TD, et al. Alcohol use among subcritically injured emergency department patients. Acad Emerg Med. 1995;2:784-790. 5. Maio RF, Waller PF, Blow FC, et al. Alcohol abuse/dependence in motor vehicle crash victims presenting to the emergency department. Acad Emerg Med. 1997;4:256-262. 6. Longabaugh R, Minugh PA, Nirenberg TD, et al. Injury as a motivator to reduce drinking. Acad Emerg Med. 1995;2:817-825. 7. Gentilello LM, Rivara FP, Donovan DM, et al. Alcohol interventions in a trauma center as a means of reducing the risk of injury recurrence. Ann Surg. 1999;230:473-480; discussion 480-483. 8. Schermer CR, Moyers TB, Miller WR, et al. Trauma center brief interventions for alcohol disorders decrease subsequent driving under the influence arrests. J Trauma. 2006;60:29-34. 9. Longabaugh R, Woolard RE, Nirenberg TD, et al. Evaluating the effects of a brief motivational intervention for injured drinkers in the emergency department. J Stud Alcohol. 2001;62:806-816. 10. Mello MJ, Nirenberg TD, Longabaugh R, et al. Emergency department brief motivational interventions for alcohol with motor vehicle crash patients. Ann Emerg Med. 2005;45:620-625. 11. Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change. New York, NY: Guilford Press; 2002. 12. American College of Emergency Physicians. Alcohol screening in the emergency department (Policy #400346, Approved January 2005). Available at: http://www3.acep.org/practres.aspx?id⫽ 29074. Accessed November 1, 2007.
Annals of Emergency Medicine 763
Telephone Brief Intervention for High-Risk Alcohol Use
Mello et al
13. National Highway Traffic Safety Administration. Identification and Referral of Impaired Drivers Through Emergency Department Protocols. National Technical Information; 2002. DOT HS 809 412. 14. Fleming MF, Mundt MP, French MT, et al. Brief physician advice for problem drinkers: long-term efficacy and benefit-cost analysis. Alcohol Clin Exp Res. 2002;26:36-43. 15. Balas EA, Jaffrey F, Kuperman GJ, et al. Electronic communication with patients. Evaluation of distance medicine technology. JAMA. 1997;278:152-159. 16. Mohr DC, Likosky W, Bertagnolli A, et al. Telephone-administered cognitive behavioral therapy for the treatment of depressive symptoms in multiple sclerosis. J Consult Clin Psychol. 2000;68: 356-361. 17. Simon GE, Ludman EJ, Tutty S, et al. Telephone psychotherapy and telephone care management for primary care patients starting antidepressant treatment: a randomized controlled trial. JAMA. 2004;292:935-942. 18. Swinson RP, Fergus KD, Cox BJ, et al. Efficacy of telephoneadministered behavioral therapy for panic disorder with agoraphobia. Behav Res Ther. 1995;33:465-469. 19. Resnicow K, Jackson A, Wang T, et al. A motivational interviewing intervention to increase fruit and vegetable intake through black churches: results of the Eat for Life trial. Am J Public Health. 2001;91:1686-1693. 20. Zhu SH, Stretch V, Balabanis M, et al. Telephone counseling for smoking cessation: effects of single-session and multiple-session interventions. J Consult Clin Psychol. 1996;64:202-211. 21. Lando HA, Hellerstedt WL, Pirie PL, et al. Brief supportive telephone outreach as a recruitment and intervention strategy for smoking cessation. Am J Public Health. 1992;82:41-46. 22. Curry SJ, McBride C, Grothaus LC, et al. A randomized trial of self-help materials, personalized feedback, and telephone counseling with nonvolunteer smokers. J Consult Clin Psychol. 1995;63:1005-1014. 23. Midanik LT, Greenfield TK, Rogers JD. Reports of alcohol-related harm: telephone versus face-to-face interviews. J Stud Alcohol. 2001;62:74-78. 24. Babor TF, Higgins-Biddle JC, Saunders J, et al. The Alcohol Use Disorders Identification Test. Guidelines for Use in Primary Care. 2nd ed. World Health Organization; 2001. 25. Donovan DM, Kivlahan DR, Doyle SR, et al. Concurrent validity of the Alcohol Use Disorders Identification Test (AUDIT) and AUDIT zones in defining levels of severity among out-patients with
26.
27. 28.
29.
30. 31.
32.
33.
34.
35.
36. 37.
38.
39.
alcohol dependence in the COMBINE study. Addiction. 2006;101:1696-1704. National Institute on Alcohol Abuse and Alcoholism. The Physician’s Guide to Helping Patients With Alcohol Problems. Washington, DC: DHHS; 1995. NIH Publication No. 95-3769. Donovan JE. Young adult drinking-driving: behavioral and psychosocial correlates. J Stud Alcohol. 1993;54:600-613. Monti PM, Colby SM, Barnett NP, et al. Brief intervention for harm reduction with alcohol-positive older adolescents in a hospital emergency department. J Consult Clin Psychol. 1999;67:989-994. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System survey questionnaire. US Department of Health and Human Services, Centers for Disease Control and Prevention. Available at: http://www.cdc.gov/brfss/index.htm. Accessed November 1, 2007. Gravetter F, Wallnau L, Essentials of Statistics for the Behavioral Sciences. 5th ed. Belmont, CA: Thomson; 2005. Nilsen P, Baird J, Mello M, et al. A systematic review of emergency care brief alcohol interventions for injury patients. Journal of Substance Abuse Treatment. In press. Gentilello L, Rivara F, Donovan D, et al. Alcohol interventions in a trauma center as a means of reducing the risk of injury recurrence. Ann Surg. 1999;230:473-483. Dauer AR, Rubio ES, Coris ME, et al. Brief intervention in alcoholpositive traffic casualties: is it worth the effort? Alcohol Alcohol. 2006;41:76-83. Sommers MS, Dyehouse JM, Howe SR, et al. Effectiveness of brief interventions after alcohol-related vehicular injury: a randomized controlled trial. J Trauma. 2006;61:523-531; discussion 532-533. Neumann T, Neuner B, Weiss-Gerlach E, et al. The effect of computerized tailored brief advice on at-risk drinking in subcritically injured trauma patients. J Trauma. 2006;61:805814. Streiner DL. Regression toward the mean: its etiology, diagnosis, and treatment. Can J Psychiatry. 2001;46:72-76. Dellinger AM, Bolen J, Sacks JJ. A comparison of driver- and passenger-based estimates of alcohol-impaired driving. Am J Prev Med. 1999;16:283-288. American College of Surgeons Committee on Trauma. Resources for Optimal Care of the Injured Patient: 2006. American College of Surgeons Committee on Trauma; 2006 Gentilello LM. Alcohol and injury: American College of Surgeons Committee on Trauma Requirements for Trauma Center Intervention. J Trauma. 2007;62(6 suppl):S44-45.
Did you know? You can personalize the new Annals of Emergency Medicine Web site to meet your individual needs.
Visit www.annemergmed.com today to see what else is new online!
764 Annals of Emergency Medicine
Volume , . : June