- Split View
-
Views
-
Cite
Cite
CECILIA NORDQVIST, ELISABETH WILHELM, KENT LINDQVIST, PREBEN BENDTSEN, CAN SCREENING AND SIMPLE WRITTEN ADVICE REDUCE EXCESSIVE ALCOHOL CONSUMPTION AMONG EMERGENCY CARE PATIENTS?, Alcohol and Alcoholism, Volume 40, Issue 5, September/October 2005, Pages 401–408, https://doi.org/10.1093/alcalc/agh175
- Share Icon Share
Abstract
Aims: Emergency care patients have an overrepresentation of risky drinkers. Despite the evidence on the effectiveness of a short feedback on screening or self-help material, most studies performed so far have required considerable time from staff and thus been difficult to implement in the real world. The present study evaluates the effect of the screening and whether simple written advice has any additional effect on risky drinking. Methods: An alcohol screening routine was implemented among injury patients in a Swedish emergency care department. Over 12 months, two cohorts were invited to answer an alcohol screening questionnaire in the waiting room. In the first 6 months, 771 patients were screened without any written advice (cohort A) and in the following 6 months, 563 were screened and in addition received simple written advice about sensible drinking (cohort B). None of the patients received one-to-one feedback. Six months after the screening, a follow-up interview by telephone explored the changes in drinking. Results: In cohort A 182 (24%) of the patients were defined as risky drinkers and in cohort B 125 (22%). Reached at follow-up after 6 months were 81 (44%) risky and 278 (47%) non-risky drinkers in cohort A, and 40 (32%) risky and 220 (50%) non-risky drinkers in cohort B. The number of patients with heavy episodic drinking decreased significantly in cohort A from 76 (94% of the risky drinkers) to 49 (59%). In cohort B a similar change was seen from 37 (92%) to 27 (68%). Only in cohort B, was a significant increase in readiness to change drinking habits seen [from 3 (8%) to 9 (23%)]. The reduction in heavy episodic drinking was comparable with previous reports from more extensive interventions. However, at the time of follow-up, drinking among non-risky drinkers at baseline had increased. When considering the greater numbers of non-risky drinkers, the total consumption in the study group increased during the study period. Conclusions: Owing to the reported difficulties of integrating more time-consuming alcohol interventions in emergency departments, it is suggested that at least screening for drinking should be implemented as routine in emergency departments. More research is needed in order to establish the optimal balance between effective alcohol intervention, and acceptable time and effort requirement from staff.
(Received 10 November 2004; first review notified 8 March 2005; in revised form 25 April 2005; accepted 18 May 2005)
INTRODUCTION
Several studies have suggested emergency departments as a suitable setting for prevention of alcohol problems (National Institute on Alcohol Abuse and Alcoholism, 1993; Zink and Maio, 1994; Maio et al., 1995; Wright et al., 1998; Gentilello et al., 1999; Hadida et al., 2001; Charalambous, 2002; Hermansson, 2002). According to a number of studies, emergency care patients have an overrepresentation of risky drinkers with either high weekly consumption or heavy episodic drinking (Conigrave et al., 1991; Romelsjö et al., 1993; Barnett et al., 1998; Peters et al., 1998; Charalambous, 2002), especially among injury patients (Raffle, 1989; Cherptiel, 1996; Barnett et al., 1998; Brooker et al., 1999; Gentilello et al., 1999; Roche et al., 2001). For example, in a Swedish study, Forsberg et al. (2003) found that 29% of patients in an emergency department were risky drinkers. The share of risky drinking in the general Swedish population is ∼15% for men and 5% for women (Bergman and Källmén, 2002).
In recent years, various forms of relatively simple alcohol preventive interventions have been introduced under the heading ‘brief intervention’ in primary health care as well as in emergency departments. Babor and Higgins-Biddle (2001) describe brief intervention as a short screening followed by education as a primary preventive measure for a large cohort of non-dependent consumers. Studies using different forms of brief intervention have consistently shown that shorter interventions, lasting a few minutes, are as effective as longer ones lasting from 20 min to several hours (Nilssen, 1991; WHO and Brief Intervention Study Group, 1996; McIntosh et al., 1997; Ockene et al., 1999; Moyer et al., 2002). The value of written self-help material, often used as a part of brief alcohol interventions, was reviewed in a metaanalysis by Apodaca and Miller (2003) who found that self-help materials can accomplish some reduction in the alcohol consumption of patients without a one-to-one consultation and constitutes a cost-effective intervention for the large number of risky drinkers.
The number of brief intervention studies in emergency care departments is still very small. Despite the evidence on the effectiveness of a short feedback on screening or self-help material, most studies have required considerable time from ordinary or research staff. In the studies published so far from emergency care settings, the initial screening procedure required 1–20 min. The follow-up intervention took 15 min to 2.5 h (Krishel, 1996; Bernstein et al., 1997; Wright et al., 1998; Gentilello et al., 1999; Monti et al., 1999; Forsberg et al., 2000; Hungerford et al., 2000, 2003). Consequently, brief alcohol interventions have been difficult to implement in a real world emergency setting mostly owing to the unrealistic demands on staff (Peters et al., 1998; Brooker et al., 1999; Charalambous, 2002). Thus, in order to implement a routine intervention in an emergency department, where the patient contact is short, there is a need for a simpler procedure that only requires a limited time effort from the staff (Hungerford et al., 2000, 2003; Charalambous, 2002). More studies are needed to establish the minimal level of alcohol intervention necessary to accomplish a reasonable change in alcohol consumption among risky drinkers.
In 2001 we implemented a procedure with screening and simple written advice performed by ordinary staff in a Swedish emergency care department (Nordqvist et al., 2004). The procedure requires only a couple of minutes of the staff's working time. To our knowledge, only three other projects have studied screening in emergency care settings performed by ordinary staff, without extra resources, with varying results (Krishel, 1996; Wright et al., 1998; Brooker et al., 1999).
The aim of this study was to evaluate whether screening without one-to-one feedback and screening with simply written advice without one-to-one feedback are sufficient to initiate a self-regulation process concerning risky drinking among emergency care patients.
METHODS
Setting and study population
An alcohol screening routine was commenced at the emergency department at Motala County Hospital in southern Sweden serving a population of 80 000. The screening was performed by all medical secretaries and nurses who take turns working in the triage room. The study comprised 1 year of screening and written advice starting in April 2001.
The study population included all patients aged 16–70 visiting the emergency department for an injury from 1 April, 2001 to 31 March, 2002. Patients judged by staff to be too seriously injured were excluded (Nordqvist et al., 2004). In total 1895 patients, nearly all persons in the target group, were approached. Of these, 370 declined to participate or were excluded because the questionnaire lacked complete answers to the AUDIT-C questions. Abstainers (n = 191) were also excluded. Thus, a total of 1334 patients were included in the further analysis (Fig. 1).
During the first 6 months, when screening was performed without written advice or feedback, 771 eligible patients, 491 males (59%) and 280 females (41%), were included comprising cohort A. In the next 6 months, with screening and simple written advice, 563 eligible patients, 343 males (61%) and 220 (39%) females, were included comprising cohort B (Fig. 1).
Screening instrument
The screening instrument was a 10-item pen and paper questionnaire beginning with AUDIT-C (Bush et al., 1998; Gordon et al., 2001; Rumpf et al., 2002) which measures the frequency and quantity of consumption and frequency of heavy episodic drinking. Then, an additional seven questions with five answering alternatives explored satisfaction with drinking habits (from ‘totally satisfied’ to ‘not satisfied at all’), readiness to change drinking habits (from ‘never considering’ to ‘daily considering’), actual change during the last year (from ‘stopped’ to ‘increased considerable’), if the patient thought that the injury was alcohol related (from ‘had not been drinking’ to ‘yes, absolutely’), attitude to being asked about drinking (from ‘very negative’ to ‘very positive’), where the injury took place (town), and whether the patient agreed to a follow-up telephone interview (yes or no).
Screening procedure
After verbal informed consent, ordinary staff handed out the questionnaire in the triage room to all injury patients between 16 and 70 years of age. Patients completed the questionnaire in the waiting room and returned it to the staff, who were available to help if necessary. No one-to-one feedback on the screening result was offered. The procedure took a couple of minutes of the staffs' time and 1–2 min for the patient to fill out. After a screening period of 6 months, the staff started to hand out simple written advice about sensitive drinking, in addition to the screening questionnaire. The information was a brochure with recommendation not to drink >7 glasses per week for women and 10 for men (the limit for low risk). The patients were advised to spread the consumption over the week and have at least 2 days without drinking. Increased risk was set to 8–13 glasses per week for women and 11–18 glasses for men. Still no one-to-one feedback on the screening was offered.
Follow-up procedure
Regardless of the drinking status, all patients, except abstainers, were asked after completing the screening questions, to participate in a follow-up telephone interview 6 months after the screening. Those who accepted (∼50% in all groups) were considered as the final study group and contacted by telephone by the first author, who has previous experiences in interviewing and by a project assistant for a period. The same questionnaire as in the initial screening was used, slightly modified. There was no attempt to try to find out if the interviewee had had any contact with an alcohol agency or received treatment elsewhere. The follow-up interview took ∼5 min. Lost to follow-up was recorded when the person had moved and the interviewer could not find the new telephone number or when the person was not reached by telephone after five attempts. Then the questionnaire was mailed to the patient. About 25% of these mailed questionnaires were returned and included in the study. Lost to follow-up were 65 (8%) persons in cohort A and 52 (9%) in cohort B. Of the persons reached by telephone only two denied interview. In cohort A, 81 risky drinkers (44%) and 278 non-risky drinkers (47%) were followed up and in cohort B, 40 risky drinkers (32%) and 220 non-risky drinkers (50%) (Fig. 1).
Measures
The study evaluated alcohol consumption among injury patients comparing drinking status at baseline with the status at 6 months for cohorts A (screened) and B screened and simply advised. The three AUDIT-C questions and one question about motivation to change drinking habits were included in the analysis. The instrument used, the AUDIT-C, has been found to be a useful routine screening instrument because of its simplicity (Bush, 1998; Wallace, 2001). The rate of false positive cases has proved to be high though, when the score system is used (Bush, 1998; Aertgeerts, 2001; Gordon, 2001). Therefore, we did not use the scores, but calculated the consumption by multiplying questions one and two, according to the precise values in the ranges as stipulated by Säppä et al. (1995) as discussed in another study (Nordqvist, 2004).
Four measures of alcohol consumption were evaluated: frequency of drinking, number of drinks on a typical day, mean weekly alcohol consumption, and frequency of heavy episodic drinking.
The patients were classified as risky or non-risky drinkers, defined as weekly volume alcohol consumption above the recommended limit and/or heavy episodic drinking. The recommended cut-off level for weekly volume consumption was set according to Rydberg et al. (1993), as ≥80 g for females and ≥110 g for males. This cut-off has been used before in Sweden (Bergman and Källmén, 2002; Hermansson, 2002). Heavy episodic drinking was defined as six glasses or more (one glass = 12 g alcohol) at one occasion at least once a month (Bergman et al., 1998). The question about readiness to change alcohol consumption was assumed to be positive if the person considered a reduction at least once a month.
Statistical analysis
SPSS version 11.5 was used to compare data at screening and at 6 months follow-up. Wilcoxon's signed rank test and the Mann–Whitney test were used to evaluate if the changes were significant. For group comparisons, chi-squared test or the unpaired t-test was used. In the follow-up analysis, males and females were not separated owing to the small numbers in the subcohorts.
The study was approved by the Ethics Committee of Linköping University.
RESULTS
Drinking at baseline
The two cohorts, A with 771 patients (screened only) and B with 563 (screened plus simple written advice), were similar in mean age, 36 years for males and 40 years for females. About 10% more men were included in both cohorts.
At baseline, 182 (24%) of patients in cohort A were classified as risky drinkers mostly owing to heavy episodic drinking. Thus, a total of 8% of the males and 2% of the females drank more than the recommended weekly volume limit and 30% of the males and 8% of the females were engaged in heavy episodic drinking at least once a month. Readiness to change was stated by 8% of males and 2% of females. In cohort B, 125 (22%) were classified as risky drinkers. A total of 6% of the males and 5% of the females drank more than the recommended weekly volume limit. Heavy episodic drinking was seen in 30% of the males and 6% of the females. Readiness to change was stated by 7% of the males and 3% of the females.
Comparison of patients followed up with those not followed up
In the following analyses males and females are considered together owing to the relatively small number of risky drinkers among females.
There were no significant differences in cohorts A or B in any drinking variable or readiness to change at baseline between risky drinkers followed up and those who declined to participate in the follow-up or were lost to follow up. Readiness to change was stated by 19% at baseline among those followed up and 16% (NS) among those not followed up in cohort A, whereas in cohort B, 8% of those followed up and 17% (NS) of those not followed up reported readiness to change drinking habits.
Non-risky drinkers followed up also did not display any differences in any drinking variable at baseline compared with those not followed up. Readiness to change was stated by 2% among both those followed up and those not followed up in cohort A and by 2% among those followed up in cohort B compared with 4% of those not followed up.
Comparisons between the cohorts revealed that readiness to change at baseline was higher among risky consumers reached at the follow-up in cohort A, 19% compared with 8% among risky consumers reached at the follow-up in cohort B (NS). The mean weekly alcohol consumption at baseline among risky drinkers reached at the follow-up in cohort A was 92 g compared with 106 g in cohort B (NS). All other drinking variables, mean age, and sex distributions were similar at baseline among risky drinkers reached at follow-up when comparing the two cohorts.
Changes in drinking pattern among risky drinkers
Cohort A. After 6 months there was one significant change in drinking pattern among those who had been risky drinkers at baseline. The proportion of heavy episodic drinking decreased by 34%. Although half of the heavy episodic drinkers maintained such a drinking pattern, more than one-third indulged in heavy episodic drinking less often, in most cases to a level beneath the cut-off for heavy episodic drinking, whereas a few were drinking heavy episodic more often than before (Tables 1 and 2). Thus, at baseline 76 (94%) of the patients were classified as heavy episodic drinkers compared with 49 (59%) at follow-up (Table 2). Although some patients decreased their frequency of drinking, others increased. The number of drinks on a typical day increased for as many as it decreased (Table 1). Half of cohort A decreased the mean consumption, a reduction from 92 g per week at baseline to 87 g at follow-up (NS). The proportion of patients with risky weekly volume consumption remained the same, 23 (28%) compared with 21 (26%). The readiness of the patients to change their consumption decreased from 15 (19%) to 11 (14%) out of 81 patients (Table 2).
. | Percentage of risky drinkers . | . | . | |||
---|---|---|---|---|---|---|
. | More . | Alike . | Less . | |||
Cohort A (n = 81) | ||||||
Frequency of drinking | 11 | 69 | 20 | |||
Numbers of drinks on a typical day | 36 | 28 | 36 | |||
Mean weekly alcohol consumption | 31 | 20 | 49 | |||
Individuals above the level for risky weekly consumptiona | 14 | 71 | 15 | |||
Frequency of heavy episodic drinkingb | 12 | 49 | 38 | |||
Readiness to change | 16 | 65 | 19 | |||
Cohort B (n = 40) | ||||||
Frequency of drinking | 18 | 80 | 2 | |||
Numbers of drinks on a typical day | 15 | 45 | 40 | |||
Mean weekly alcohol consumption | 28 | 42 | 30 | |||
Individuals above the level for risky weekly consumptiona | 15 | 72 | 13 | |||
Frequency of heavy episodic drinkingb | 8 | 58 | 35 | |||
Readiness to change | 23 | 59 | 18 |
. | Percentage of risky drinkers . | . | . | |||
---|---|---|---|---|---|---|
. | More . | Alike . | Less . | |||
Cohort A (n = 81) | ||||||
Frequency of drinking | 11 | 69 | 20 | |||
Numbers of drinks on a typical day | 36 | 28 | 36 | |||
Mean weekly alcohol consumption | 31 | 20 | 49 | |||
Individuals above the level for risky weekly consumptiona | 14 | 71 | 15 | |||
Frequency of heavy episodic drinkingb | 12 | 49 | 38 | |||
Readiness to change | 16 | 65 | 19 | |||
Cohort B (n = 40) | ||||||
Frequency of drinking | 18 | 80 | 2 | |||
Numbers of drinks on a typical day | 15 | 45 | 40 | |||
Mean weekly alcohol consumption | 28 | 42 | 30 | |||
Individuals above the level for risky weekly consumptiona | 15 | 72 | 13 | |||
Frequency of heavy episodic drinkingb | 8 | 58 | 35 | |||
Readiness to change | 23 | 59 | 18 |
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses on any one occasion at least once a month.
. | Percentage of risky drinkers . | . | . | |||
---|---|---|---|---|---|---|
. | More . | Alike . | Less . | |||
Cohort A (n = 81) | ||||||
Frequency of drinking | 11 | 69 | 20 | |||
Numbers of drinks on a typical day | 36 | 28 | 36 | |||
Mean weekly alcohol consumption | 31 | 20 | 49 | |||
Individuals above the level for risky weekly consumptiona | 14 | 71 | 15 | |||
Frequency of heavy episodic drinkingb | 12 | 49 | 38 | |||
Readiness to change | 16 | 65 | 19 | |||
Cohort B (n = 40) | ||||||
Frequency of drinking | 18 | 80 | 2 | |||
Numbers of drinks on a typical day | 15 | 45 | 40 | |||
Mean weekly alcohol consumption | 28 | 42 | 30 | |||
Individuals above the level for risky weekly consumptiona | 15 | 72 | 13 | |||
Frequency of heavy episodic drinkingb | 8 | 58 | 35 | |||
Readiness to change | 23 | 59 | 18 |
. | Percentage of risky drinkers . | . | . | |||
---|---|---|---|---|---|---|
. | More . | Alike . | Less . | |||
Cohort A (n = 81) | ||||||
Frequency of drinking | 11 | 69 | 20 | |||
Numbers of drinks on a typical day | 36 | 28 | 36 | |||
Mean weekly alcohol consumption | 31 | 20 | 49 | |||
Individuals above the level for risky weekly consumptiona | 14 | 71 | 15 | |||
Frequency of heavy episodic drinkingb | 12 | 49 | 38 | |||
Readiness to change | 16 | 65 | 19 | |||
Cohort B (n = 40) | ||||||
Frequency of drinking | 18 | 80 | 2 | |||
Numbers of drinks on a typical day | 15 | 45 | 40 | |||
Mean weekly alcohol consumption | 28 | 42 | 30 | |||
Individuals above the level for risky weekly consumptiona | 15 | 72 | 13 | |||
Frequency of heavy episodic drinkingb | 8 | 58 | 35 | |||
Readiness to change | 23 | 59 | 18 |
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses on any one occasion at least once a month.
. | Baseline . | 6 Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 92.1 | 86.9 | −5.3 | (−28.6 to 18.0) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 28.4 | 25.9 | −2.5 | (−16.2 to 11.2) | 5 | |||||
Proportion with heavy episodic drinking (%)b | 93.8 | 59.3 | −34.5 | (−46.4 to −22.6) | 100 | |||||
Proportion ready to change (%)c | 18.8 | 13.6 | −5.2 | (2−16.5 to 6.2) | 14 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 22.9 | 28.5 | +5.6 | (1.8 to 9.4) | 81 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 3.6 | +3.6 | (1.4 to 6.7) | 89 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 8.4 | +8.4 | (5.1 to 11.6) | 100 | |||||
Proportion ready to change (%)c | 1.8 | 6.9 | +5.1 | (1.7 to 8.4) | 84 |
. | Baseline . | 6 Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 92.1 | 86.9 | −5.3 | (−28.6 to 18.0) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 28.4 | 25.9 | −2.5 | (−16.2 to 11.2) | 5 | |||||
Proportion with heavy episodic drinking (%)b | 93.8 | 59.3 | −34.5 | (−46.4 to −22.6) | 100 | |||||
Proportion ready to change (%)c | 18.8 | 13.6 | −5.2 | (2−16.5 to 6.2) | 14 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 22.9 | 28.5 | +5.6 | (1.8 to 9.4) | 81 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 3.6 | +3.6 | (1.4 to 6.7) | 89 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 8.4 | +8.4 | (5.1 to 11.6) | 100 | |||||
Proportion ready to change (%)c | 1.8 | 6.9 | +5.1 | (1.7 to 8.4) | 84 |
Number of patients (risky drinkers, 81; non-risky drinkers, 278); percentage of male (risky drinkers, 87.7; non-risky drinkers, 52.2); mean age (years) (risky drinkers, 33.7; non-risky drinkers, 41.7).
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses on one occasion at least once a month.
Thinking at least once a month about changing drinking habits.
Extra bold type confidence interval = significant.
. | Baseline . | 6 Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 92.1 | 86.9 | −5.3 | (−28.6 to 18.0) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 28.4 | 25.9 | −2.5 | (−16.2 to 11.2) | 5 | |||||
Proportion with heavy episodic drinking (%)b | 93.8 | 59.3 | −34.5 | (−46.4 to −22.6) | 100 | |||||
Proportion ready to change (%)c | 18.8 | 13.6 | −5.2 | (2−16.5 to 6.2) | 14 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 22.9 | 28.5 | +5.6 | (1.8 to 9.4) | 81 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 3.6 | +3.6 | (1.4 to 6.7) | 89 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 8.4 | +8.4 | (5.1 to 11.6) | 100 | |||||
Proportion ready to change (%)c | 1.8 | 6.9 | +5.1 | (1.7 to 8.4) | 84 |
. | Baseline . | 6 Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 92.1 | 86.9 | −5.3 | (−28.6 to 18.0) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 28.4 | 25.9 | −2.5 | (−16.2 to 11.2) | 5 | |||||
Proportion with heavy episodic drinking (%)b | 93.8 | 59.3 | −34.5 | (−46.4 to −22.6) | 100 | |||||
Proportion ready to change (%)c | 18.8 | 13.6 | −5.2 | (2−16.5 to 6.2) | 14 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 22.9 | 28.5 | +5.6 | (1.8 to 9.4) | 81 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 3.6 | +3.6 | (1.4 to 6.7) | 89 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 8.4 | +8.4 | (5.1 to 11.6) | 100 | |||||
Proportion ready to change (%)c | 1.8 | 6.9 | +5.1 | (1.7 to 8.4) | 84 |
Number of patients (risky drinkers, 81; non-risky drinkers, 278); percentage of male (risky drinkers, 87.7; non-risky drinkers, 52.2); mean age (years) (risky drinkers, 33.7; non-risky drinkers, 41.7).
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses on one occasion at least once a month.
Thinking at least once a month about changing drinking habits.
Extra bold type confidence interval = significant.
Cohort B. In cohort B, there were two significant changes. Heavy episodic drinking decreased by 25%, reducing the numbers from 37 (92%) to 27 (68%). More than half of those who were heavy episodic drinkers at baseline did not change this pattern, but similar to cohort A one-third of those followed up drank heavy episodic less then, whereas a smaller group did it more often (Tables 1 and 3). Also, the number of patients who, at least once a month, considered changing their alcohol consumption increased significantly, from 3 (8%) to 9 (22%) out of 40 patients (Table 3). The frequency of drinking increased for more people than it decreased, whereas the situation was opposite concerning the number of drinks on a typical day (Table 1). Mean weekly consumption was 106 g at baseline and 99 g at follow-up. The number of patients above the recommended limits for risky weekly volume alcohol consumption was about the same, 12 (30%) at baseline and 13 (32%) at follow-up (Table 3).
. | Baseline 6 . | Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 105.8 | 99.4 | −6.4 | (−39.8 to 26.9) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 30.0 | 32.5 | +2.5 | (−17.8 to 22.8) | 4 | |||||
Proportion with heavy episodic drinking (%)b | 92.5 | 67.5 | −25.0 | (−41.7 to 8.3) | 81 | |||||
Proportion ready to change (%)c | 7.7 | 22.5 | +14.8 | (−0.006 to 30.2) | 50 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 23.5 | 32.9 | +9.4 | (4.2 to 14.5) | 94 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 5.0 | +5.0 | (2.1 to 7.9) | 92 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 14.7 | +14.7 | (10.0 to 19.4) | 100 | |||||
Proportion ready to change (%)c | 2.3 | 7.4 | +5.1 | (1.1 to 9.0) | 70 |
. | Baseline 6 . | Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 105.8 | 99.4 | −6.4 | (−39.8 to 26.9) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 30.0 | 32.5 | +2.5 | (−17.8 to 22.8) | 4 | |||||
Proportion with heavy episodic drinking (%)b | 92.5 | 67.5 | −25.0 | (−41.7 to 8.3) | 81 | |||||
Proportion ready to change (%)c | 7.7 | 22.5 | +14.8 | (−0.006 to 30.2) | 50 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 23.5 | 32.9 | +9.4 | (4.2 to 14.5) | 94 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 5.0 | +5.0 | (2.1 to 7.9) | 92 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 14.7 | +14.7 | (10.0 to 19.4) | 100 | |||||
Proportion ready to change (%)c | 2.3 | 7.4 | +5.1 | (1.1 to 9.0) | 70 |
Number of patients (risky drinkers, 40; non-risky drinkers, 220); percentage of male (risky drinkers, 90.0; non-risky drinkers, 52.7); mean age (years) (risky drinkers, 37.3; non-risky drinkers, 40.2).
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses at least once a month.
Thinking at least once a month about changing drinking habits.
Extra bold type confidence interval = significant.
. | Baseline 6 . | Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 105.8 | 99.4 | −6.4 | (−39.8 to 26.9) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 30.0 | 32.5 | +2.5 | (−17.8 to 22.8) | 4 | |||||
Proportion with heavy episodic drinking (%)b | 92.5 | 67.5 | −25.0 | (−41.7 to 8.3) | 81 | |||||
Proportion ready to change (%)c | 7.7 | 22.5 | +14.8 | (−0.006 to 30.2) | 50 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 23.5 | 32.9 | +9.4 | (4.2 to 14.5) | 94 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 5.0 | +5.0 | (2.1 to 7.9) | 92 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 14.7 | +14.7 | (10.0 to 19.4) | 100 | |||||
Proportion ready to change (%)c | 2.3 | 7.4 | +5.1 | (1.1 to 9.0) | 70 |
. | Baseline 6 . | Months . | Difference . | Diff CId . | Power % . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 105.8 | 99.4 | −6.4 | (−39.8 to 26.9) | 6 | |||||
Proportion with risky weekly volume consumption (%)a | 30.0 | 32.5 | +2.5 | (−17.8 to 22.8) | 4 | |||||
Proportion with heavy episodic drinking (%)b | 92.5 | 67.5 | −25.0 | (−41.7 to 8.3) | 81 | |||||
Proportion ready to change (%)c | 7.7 | 22.5 | +14.8 | (−0.006 to 30.2) | 50 | |||||
Non-risky drinkers | ||||||||||
Mean weekly alcohol consumption (g) | 23.5 | 32.9 | +9.4 | (4.2 to 14.5) | 94 | |||||
Proportion with risky weekly volume consumption (%)a | 0 | 5.0 | +5.0 | (2.1 to 7.9) | 92 | |||||
Proportion with heavy episodic drinking (%)b | 0 | 14.7 | +14.7 | (10.0 to 19.4) | 100 | |||||
Proportion ready to change (%)c | 2.3 | 7.4 | +5.1 | (1.1 to 9.0) | 70 |
Number of patients (risky drinkers, 40; non-risky drinkers, 220); percentage of male (risky drinkers, 90.0; non-risky drinkers, 52.7); mean age (years) (risky drinkers, 37.3; non-risky drinkers, 40.2).
Females ≥80 g/week; men ≥110 g/week.
Drinking ≥6 glasses at least once a month.
Thinking at least once a month about changing drinking habits.
Extra bold type confidence interval = significant.
Changes in drinking pattern among non-risky drinkers
Cohort A. Among those who were non-risky drinkers at baseline, significant changes were seen after 6 months in all alcohol consumption measures as well as in readiness to change. Thus, mean consumption increased from 23 to 29 g per week, 8% started to engage in heavy episodic drinking at least once a month and 4% passed the recommended weekly volume limit. However, readiness to change increased from 2 to 7% (Table 2).
Cohort B. Just as among non-risky drinkers in cohort A, all alcohol consumption measures as well as readiness to change were changed significantly. Mean consumption increased from 24 to 33 g per week. In addition, 15% started to engage in heavy episodic drinking and 5% in high weekly volume consumption. Readiness to change increased from 2 to 7% (Table 3).
When considering the greater number of non-risky drinkers the net effect in the study group was an increase in the alcohol consumption.
DISCUSSION
Brief alcohol interventions during emergency care have been difficult to disseminate into a real world setting mainly owing to organizational constraints, such as time (Charalambous, 2002; Hungerford and Pollock, 2003; Hungerford et al., 2003). The present study was designed such that ordinary nurses could easily include the screening routines into their current work situation and therefore implementation was considered to be realistic. Cohort A did not receive any verbal or written feedback, whereas cohort B received simple written advice about sensible drinking but still no one-to-one feedback. We had no control group but evaluated the effect of screening compared with simple written advice.
Since most previous studies have had different designs and outcome measures, the results are not directly comparable. Our screening procedure and simple advice was simpler than the screening and intervention in most previous studies and in some cases even more simple than the control group in those few studies that had a control group. For example, the control group called ‘standard care’ in a study by Longabaug et al. (2001), was assessed for 30–40 min. In Forsberg et al. (2000) the comparison group was assessed with three questionnaires in total lasting for 26 min.
Changes in drinking behaviour among risky drinkers
The main drinking behaviour among risky drinkers in our study was heavy episodic drinking comprising 30% of the male and 7% of the female eligible patients at baseline. In both cohorts, there was a significant reduction in heavy episodic drinking at follow-up (34% in cohort A and 25% in cohort B). Both fairly similar and greater reduction has been reported in some of the previous studies with more extended alcohol interventions. Also, in a number of previous studies a significant reduction in the frequency of heavy episodic drinking was seen independently of the intensity of the intervention. Thus, Longabaugh et al. (2001) reported a 12% reduction in the frequency of heavy episodic drinking among risky drinkers in both the ‘standard care’ group and the brief intervention group. Other studies have reported a more pronounced reduction in heavy episodic drinking than in our study. Forsberg et al. (2000) reported a 71% reduction in the frequency of heavy episodic drinking with no significant difference between two intervention methods: simple feedback on screening or 1–2.5 h extended counselling by a psychologist. Bernstein et al. (1997) showed a reduction of 64% in the frequency of heavy episodic drinking after a brief negotiating interview lasting 20 min. Thus, the reduction in heavy episodic drinking in our study was less than in most previous studies with more extended intervention. Adding simple written advice in cohort B did not increase the effect. In fact, cohort B, who received simple written advice, displayed less reduction in the frequency of heavy episodic drinking. That could be explained by the fact that risky consumers in cohort A who were followed up, were more willing to change at baseline than the risky consumers who were followed up in cohort B.
The second drinking outcome measure in our study was the percentage of patients with weekly volume consumption above recommended limits. We did not find any significant change at the 6-month follow-up among those classified as risky drinkers at baseline. In cohort A, mean consumption decreased non-significantly by 5 to 87 g per week and decreased by 7 g in cohort B to 99 g per week. This stands in contrast to a number of previous studies on various forms of brief interventions in emergency departments where the effects on weekly consumption were reported to be significant. Thus, Gentilello et al. (1999) found a significant difference in weekly average consumption between an intervention group and a control group at a 12-month follow-up. In the study by Forsberg et al. (2000), weekly consumption decreased from 133 to 96 g at 6 months but increased to 106 g after 12 months in two intervention groups. Wright et al. (1998) found a 65% reduction in average weekly volume intake after a brief intervention, from 240 to 136 g per day. Hungerford et al. (2000) reported a reduction in alcohol consumption in 68% of 23 risky drinkers reached for a follow-up after a 15–20 min counselling session and in another study, a reduction in weekly volume intake was reported in 62% of 519 risky drinkers after a similar intervention (Wright et al., 1998; Hungerford et al., 2003). Average weekly volume consumption was not calculated in either of Hungerford's studies. The change is based on a reduction in AUDIT scores. One explanation for the lack of change in our study is that the mean alcohol consumption was relatively low at baseline in our study groups, ∼90 g per week, and thus lower than in all the other studies.
The frequency of alcohol intake was the third outcome measure used in our study. Bernstein et al. (1997) found a 56% reduction in frequency of drinking after a brief negotiating interview. In our study 18% of patients in cohort B drank more often after intervention and 2% less often, whereas in cohort A, 11% drank more and 20% less frequently. Still the majority drank as often as before. Forsberg et al. (2000) found a similar small increase in frequency of drinking. Thus, at 6-month follow-up, 32% drank more and 16% less frequently after a brief or extended intervention.
Although the numbers are low, the readiness to change increased significantly, by 15% in cohort B who received simple written advice. Most patients did not alter their readiness to change but 23% were more ready and 18% less ready to change. Cohort A seemed to have become less ready to change, 16% were more ready and 19% less ready. However, at baseline, followed up risky drinkers in cohort A were more ready to change than followed up risky drinkers in cohort B (19% compared to 8%). At the follow-up, cohort A had decreased to 14% readiness to change whereas cohort B increased to 22%. This is in contrast to Forsberg et al. (2000) who showed that 16% were more ready and 33% were less ready to change at the 6-month follow-up. In the study by Hungerford et al. (2000), 43% of the 23 patients at the follow-up had become more ready to change, even though 59% of the patients were already prepared at baseline to set goals to reduce or stop drinking. One explanation for cohort A being less ready to change at follow-up is that they had reduced heavy episodic drinking more than cohort B and thus, were in less need for additional change in their drinking behaviour.
Changes among non-risky drinkers
Non-risky drinkers are seldom or never followed up in brief intervention studies in emergency departments. In our study, both risky and non-risky drinkers were included in the follow-up in order to compare the effect of the screening and written advice and also to highlight some of the normal changes in alcohol habits in the population. Alcohol consumption increased significantly among low consumers in both cohorts according to all measures used, some to the level of risky drinking. The brochure with simple advice about drinking levels given to cohort B could have encouraged the non-risky drinkers in that cohort to drink more. However, this does not explain the increase in consumption among non-risky consumers in cohort A where no written advice was offered. It is possible that filling out the questionnaire showed the patients that they were low consumers according to the answering options. An additional explanation is that the increase is owing to the general increase in drinking in Sweden or the normal variation in consumption in the population (Bergman and Källmén, 2002; Leifman and Trolldal, 2002). If so, the changes seen among the risky drinkers are even more encouraging.
Positive effects of only screening
Previous brief intervention studies in emergency care settings have used different kinds of screening and intervention measures (Krishel, 1996; Bernstein et al., 1997; Wright et al., 1998; Gentilello et al., 1999; Monti et al., 1999; Forsberg et al., 2000; Hungerford et al., 2000, 2003). Screening varies from one questionnaire with no further assessment, to interview or several questionnaires with further assessment. The interventions were more extensive than in our study, often using motivational interviewing and sometimes referral to treatment elsewhere. Although our screening procedure seemed to have a somewhat less effect on heavy episodic drinking than in most previous studies, the reduction was not neglectable. The additional written advice to cohort B did not increase the effect on heavy episodic drinking but might have increased the readiness to change. Thus, our results support the self-regulating theory described by Agostinelli et al. (2004) indicating that when people are given an opportunity to reflect on their drinking habits, a spontaneous self-monitoring process can lead to problem recognition. The positive effect of only screening without any feedback is also supported by some previous studies (Anderson and Scott, 1992; WHO and Brief Intervention Study Group, 1996). In the study by Anderson and Scott (1992), a control group offered only screening showed a marginally less reduction in alcohol consumption compared with an intervention group. Also Monti et al. (1999) failed to show any difference in the reduction of alcohol consumption in young people, between a cohort receiving 35–40 min of motivational interviewing and a cohort receiving 5 min of 'standard care'. Other studies from emergency care displayed a more mixed result. In the WHO study a control group, receiving no advice, showed no significant difference between intervention and control groups among female patients but among males, the reduction was significantly higher in the intervention group (WHO and Brief Intervention Study Group, 1996). Gentillello et al. (1999) showed a similar reduction in drinking after 6 months in a control group as in the intervention group. After 12 months, however, the control group drank as much as before the screening whereas the reduction in the intervention group was stable.
Whether screening per se has an effect on drinking habits is still an unanswered question that probably will be very difficult to answer. However, adding simple written advice do not add any benefit to the possible effect of screening with regards to reducing risky drinking.
Methodological considerations
We did not have a control group but evaluated the effect of screening compared with screening and simple advice. A weakness in our study is that we did not study the longer term effect of our screening and simple written advice. However, the number of patients included in the follow-up, 619 (of which 121 were risky drinkers), are a strength in our study and also that there was no difference in drinking pattern at baseline between those followed up and those declining to participate. The positive results shown in other studies are often based on a small number of follow-ups, from 12 to 87 patients (Krishel, 1996; Wright et al., 1998; Monti et al., 1999; Hungerford et al., 2000) although some had higher numbers, from 165 to 519 (Forsberg et al., 2000; Longabaugh et al., 2001). For example Krishel et al. (1996) showed that out of 12 followed up, 5 drank less and 1 was on treatment.
CONCLUSIONS
Our results imply that simply screening, without one-to-one feedback might be enough to accomplish a reasonable change in heavy episodic drinking, the main risky drinking behaviour in our study, over a 6-month perspective. However, since we did not have a real control group that was not screened, which is a methodological dilemma, our results only suggest a possible effect of screening per se. Also, we do not know if the reduction in risky drinking was a result of the trauma experienced. Adding written advice without any one-to-one feedback does not add any benefit in terms of reducing risky drinking, but could enhance motivation to change, although this might be dependent on the level of motivation at baseline. The intervention concept in our study is a simple routine that easily could be implemented in the absence of a more extensive routine. Other studies with more pronounced effect on reduction in drinking have demanded considerable time from ordinary staff and other resources, such as external staff. Consequently, this kind of alcohol intervention is difficult to disseminate into the daily routine of emergency departments. Our concept, with a more modest influence on drinking habits, has the potential to continue as a routine with ordinary staff. However, this could provide a base for implementation of a more effective, but probably more time-consuming intervention in the long term. For example, a natural development could be implementation of a computerized screening offering personalized feedback to the patients with or without one-to-one feedback (Karlsson and Bendtsen, 2004). Further studies are needed in order to evaluate whether such a concept or other new concepts are acceptable for the staff and effective in a real world setting. More research is also needed in order to establish whether all patients need one-to-one feedback and if not, how it is possible to differentiate between the various needs in different patients groups with respect to risky drinking.
We are grateful to the staff of Motala Emergency Department who accomplished the routine screening and to the patients who participated. Thanks to Marika Holmqvist for statistical help and to Lotta Strömberg for executing part of the follow-up interviews. Trygg-Hansa Research Foundation, Ester Johansson Memorial Fund, Milan Valverius' Fund and Per Eckerberg's Fund financed the study.
REFERENCES
Aertgeerts, B., Buntinx, F., Ansoms, S. et al. (
Agostinelli, G., Floyd, T., Grube, J. et al. (
Anderson, P. and Scott, E. (
Apodaca, T. and Miller, W. (
Babor, T. F. and Higgins-Biddle, J. C. (
Barnett, N., Spirito, A., Colby, S. et al. (
Bergman, H. and Källmén, H. (
Bergman, H., Källmén, H. and Rydberg, U. (
Bernstein, E., Bernstein, J. and Levenson, S. (
Brooker, C., Peters, J., McCabe, C. et al. (
Bush, K., Kivlahan, D., McDonell, M. et al. (
Charalambous, M. (
Cherptiel, C. (
Conigrave, K. M., Burns, F. H., Reznik, R. B. et al. (
Forsberg, L., Ekman, S., Halldin, J. et al. (
Forsberg, L., Halldin, J. and Wennberg, P. (
Gentilello, L., Rivara, F., Donovan, D. et al. (
Gordon, A. J., Maistro, S. A., McNeil, M. et al. (
Hadida, A., Kapor, N., Mackway-Jones, K. et al. (
Hermansson, U. (
Hungerford, D. and Pollock, D. (
Hungerford, D., Pollock, D. and Todd, K. (
Hungerford, D., Williams, J., Frurbee, P. et al. (
Karlsson, A. and Bendtsen, B. (
Krishel, S. (
Leifman, H. and Trolldal, B. (
Longabaugh, R., Woolard, R., Nirenberg, T. et al. (
Maio, R., Shope, J., Blow, F. et al. (
McIntosh, M. C., Leigh, G., Baldwin, N. J. et al. (
Monti, P., Colby, S., Barnett, N. et al. (
Moyer, A., Finney, J., Swearingen, C. et al. (
National Institute on Alcohol Abuse and Alcoholism (
Nilssen, O. (
Nordqvist, C., Johansson, K. and Bendtsen, P. (
Ockene, J. K., Adams, A., Hurley, T. G. et al. (
Peters, J., Brooker, C., McCabe, C. et al. N. (
Raffle, P. (
Roche, A., Watt, K., McClure, R. et al. (
Romelsjö, A., Alberts, K. and Andersson, R. (
Rumpf, H.-J., Hapke, U., Meyer, C. et al. (
Rydberg, U., Thakker, K. D. and Skerfving, S. (
Seppä, K., Mäkelä, R. and Sillanaukee, P. (
Wallace, P. (
WHO and Brief Intervention Study Group (
Wright, S., Moran, L., Meyrick, M. et al. (