**2.2 Effectiveness of brief interventions in reducing alcohol use—meta analysis and systematic reviews**

Convincing evidence exists about the effectiveness of BI for harmful alcohol users admitted to general hospital wards and in PHC settings. Wilk et al., studied 12 RTCs in which BI was given to heavy drinkers and found that heavy drinkers in the interventional group were twice as likely to moderate their drinking pattern after 6−12 months compared to the controlled group [48]. Ballesteros et al., did a study on efficacy of BIs on hazardous drinkers and included 13 studies. There was no clear evidence of a dose-effect relationship. Although indicating smaller effect sizes than previous meta-analyses, it does support the moderate efficacy of BIs [49]. Bertholet et al. [50], had a study on reduction of alcohol consumption by a brief intervention, which included 19 trials of 5639 individuals and it was found that that brief alcohol intervention was effective in reducing alcohol consumption at 6 and 12 months. McQueen et al. [51], did study on BIs for heavy alcohol users admitted to general hospital wards, which included 14 studies involving 4041 male participants and it was concluded that patients receiving BIs had a greater reduction in alcohol consumption compared to those in control groups at 6 and 9 months follow up, but it was not maintained at 1 year and had significantly fewer deaths. Sullivan et al. [52], did a study on metaanalysis of the efficacy of non-physician BIs for unhealthy alcohol use: implications for the patient-centred medical home including 13 studies and showed 1.7 times fewer standard drinks per week than control conditions. A meta-analysis on the effects on

mortality of BIs for problem drinking concluded that brief interventions may reduce mortality rates among problem drinkers by an estimated 23–26% [53].

### **2.3 BI and alcohol consumption in primary health care settings**

There is substantial evidence of the benefits of screening and BI for alcohol problems in PHC settings. BI was found to be effective at PHC setting in reduction of alcohol consumption and it is cost-effectively related to various problems associated with substance use [54–56]. Moreover, BIs have been found to be effective in both primary and secondary care settings for hazardous or harmful alcohol use when delivered under research conditions [57, 58]. Brief interventions have been shown to be cost-effective for hazardous drinkers whose alcohol use put them at risk of alcoholrelated problems, but who have few symptoms of alcohol dependence [24, 59]. Brief interventions have been used to encourage those with more serious dependence to engage or improve compliance with more intensive treatment [60]. Lock et al., had conducted a study on cluster RCT to test the effectiveness and cost-effectiveness of screening and BI for patients in PHC in which the intervention group was given 5–10-minute BI and standard advice was offered in the control group. However, ANOVA revealed no statistically significant difference between intervention and control patients at follow-up in alcohol use and economic benefits [61]. Chang et al., conducted an RCT to test the effectiveness of BI and the involvement of their partners in the PHC setting using T-ACE as screening tool and assessed the outcome measures in women with alcohol use, alcohol abstinence self-efficacy score andpartners' collateral report on the subjects' alcohol use. The intervention group received a 25-minute BI by either a nurse or doctor and the control group as usual care. It was found that alcohol use declined in both groups and BI was more effective in women group [62]. Ockene et al., made a study to compare the efficacy of BI in PHC setting with the control group. A 5–10 minutes patient-centred BI found significant reductions in alcohol consumption [63]. Similarly, Goodall et al. [64], reported that two brief sessions in the intervention group showed significantly greater reductions in the frequency of alcohol use variables.

In a community-based study in North India, a sample was followed for 3 months in which 90 male subjects (20−45 years) with an AUDIT score between 8 and 24 consented to participate and were allocated alternatively to the BI or simple advice (SA) protocols. The study showed significant differences across interventions, with a decrease in severity of dependence in the last 30 days, composite ASI (Addiction Severity Index) scores and improvement in physical and psychological quality of life. However, the result was not sustained for a longer duration and the author claimed that booster sessions were needed [65].

### **2.4 BI and Substance use in various settings**

Gryczynski et al., assessed the effectiveness of BI at 6-month follow-up at a rural health care centre. The screening was done with AUDIT and yes/no questions about past year's use of any illegal drug. Outcome measures were recorded as changes in self-reported frequency of illicit drug use, alcohol use and alcohol intoxication. Study showed that there was a greater magnitude of change in drinking behaviours and reductions in illicit drug use. While substantial, it did not differ significantly based on service variables [66]. Bertha et al., conducted a study in which screening, brief interventions and referral to treatment (SBIRT) were used in a wide variety of medical settings. The screening was done with AUDIT and Drug Abuse Screening Test (DAST)

### *Screening and Brief Intervention in Substance Use Disorders: Its Clinical Utility and Feasibility… DOI: http://dx.doi.org/10.5772/intechopen.107441*

and compared illicit drug use at intake and 6 months after drug screening and interventions. Study has shown that the intervention was feasible to implement, and the self-reported status at 6 months indicated significant improvements over baseline for illicit drug use and heavy alcohol use and also in functional domains [67]. Mitchell et al., had done pre−post analysis to assess the effectiveness of screening, brief interventions and referral to treatment (SBIRT) at 6-month follow-up at a schoolbased program. The screening was done with CRAFFT. It examined the outcomes of SBIRT services and compared the extent of change in substance use based on the intensity of intervention received. Participants receiving any intervention reported significant reductions in frequency of drinking to intoxication (*p* < 0.05) and drug use (*p* < 0.001) [68]. In another study done by Beintrein et al., in whichan RCT was conducted in inner-city teaching hospital outpatient clinics. Interventional group was given a brief motivational intervention and compared with the control group at 3 and 6 months follow-up. The intervention group was more likely to be abstinent than the control group for cocaine as well as heroin use with a reduction of cocaine level in the hair [69]. Similarly, Saunders et al. [70], also found that BI delivered to opiate users attending a methadone program to be effective in increasing participants' compliance with treatment and motivation to quit drug use, as well as reducing the number of reported drug-related problems and rate of relapse. Although there is growing evidence in support of BIs for a range of illicit substances, some studies have failed to find significant effects [71]. In a systematic review done by Young et al., on effectiveness of brief interventions as part of the SBIRT model for reducing the nonmedical use of psychoactive substances that identified 8836 records. They concluded that insufficient evidence exists as to whether BIs, as part of SBIRT, were effective or ineffective for reducing the use of substance and harm related to it [72].

Cannabis users generally had a low level of motivation to quit its use and have a concern about stigma to assess the treatment [73]. Despite all these the BIs have recently been developed for cannabis use in an attempt to address the gaps in treatment engagement, and a small number of studies have been conducted with promising results [74–77].

Stephens et al., in their first RCT, found two 90-minute individual sessions (comprising assessment, personalised feedback and advice) to be as effective as more extensive treatment and more effective than no treatment in reducing cannabis use and related problems [78]. Similarly, Walker et al., also found two sessions of motivational enhancement therapy delivered to adolescent cannabis users resulted in reduced cannabis use and fewer negative consequences at 12 months compared to a delayed-treatment control group [79].

In a simple single-group pre−post design, Denering and Spear [80] found screening and a brief 10−15 minute intervention delivered to college students resulted in reductions in the proportion of students reporting cannabis use at 6 months.

BIs for smoking cessation have also been found to be highly effective. A systematic review by Stead et al. [81], (included 42 clinical trials) conducted since 1972 found that brief advice to patients to quit smoking increased the likelihood of a cessation attempt, as well as abstinence at the 12-month follow-up with an additional benefit of more intensive advice on quit rates.

### **2.5 The ASSIST-linked brief intervention**

Spear et al. [82], did a study on substance abuse screening and BI in a mental health clinic and concluded that administration of the ASSIST in a campus mental health clinic was feasible and brought an opportunity for discussion related to substance use. Humeniuk et al., did an international RCT to evaluate the effectiveness of ASSIST-linked BI for illicit drugs (cannabis, cocaine, ATS and opioids). Participants were recruited from PHC settings in four countries (Australia, Brazil, India and the United States of America) and were randomly allocated to an intervention or waitlist control group at baseline and the groups were followed up after 3 months. A total of 731 participants were recruited from a variety of PHC settings for the international study (Australia *n* = 171; Brazil *n* = 165; India *n* = 177 and United States of America *n* = 218). Participants were aged between 16 and 62 years. It was concluded that the ASSIST-linked BI was effective in getting participants to reduce their substance use and risk as supported by feedback from at 3 months follow-up [83]. Zibe-Piegel and Boerngen-Lacerda did research work from city hall in a southern city of Brazil representative sample of employees (*n* = 1310), 144 individuals in risky use and 139 dependents on tobacco, alcohol and/or other substances where ASSIST-linked BI was used during 3-month follow-up. It showed a significant reduction in ASSIST scores and was feasible in workplace to prevent hazardous/ harmful substance use without prejudice or stigma, enabling earlier detection, intervention and treatment referral [46]. Assanangkornchai et al., demonstrated the implementation, acceptability and uptake of the screening and BI program based on the ASSIST to help decrease substance misuse in primary care in Thailand. Here 5931 patients were screened with the ASSIST. Of these, 29.6% and 3.4% were in the moderate and high-risk groups, respectively and were offered BI or other treatments. The ASSIST detected many substance users capable of benefiting from the intervention. The program was well received by patients and staff and suggested as a model for introducing similar procedures into developing countries [84]. Saitz et al., did a study to test the efficacy of two brief counselling interventions for unhealthy drug use (any illicit drug use or prescription drug misuse). A total of 528 adult primary care patients were randomised into three groups after screening with ASSIST scores greater than or equal to 4. A brief negotiated interview (10- to 15-minute structured interviews) and an adaptation of motivational interviewing (30- to 45-minute intervention based on motivational interviewing with a 20- to 30-minute booster) and compared with no brief intervention. There were no significant effects of brief negotiated interviews or an adaptation of motivational interviewing on self-reported measures of drug use and its consequences. These results did not support widespread implementation of illicit drug use and prescription drug misuse screening and brief intervention [85]. Loretta et al., provided preliminary evidence of the effectiveness of ASSIST-linked BI in a college mental health clinic where 453 students (ages 18–24) participated in the evaluation and completed baseline and 6-month follow-up interviews. Study showed a slight reduction in the rates and number of days (in the prior 30 days) of binge drinking and marijuana use and it was concluded that routine screening and BI procedures in a mental health setting may reduce problematic substance use among college students [86]. Pengpid et al., did RCT including screening and concurrent BI of conjoint hazardous or harmful alcohol and tobacco use in hospital outpatients in Thailand. Results of the interaction (group × time) effects indicated that there were statistically significant differences between the three study groups [tobacco only intervention, alcohol only intervention and the polydrug use (alcohol and tobacco) integrated intervention groups] over the 6-month follow-up on the ASSIST tobacco score and past week tobacco use abstinence. The result show reduction in scores in all six outcome parameters (Alcohol ASSIST score, low alcohol risk score, past week tobacco abstinence or low alcohol risk score and past week tobacco abstinence and

*Screening and Brief Intervention in Substance Use Disorders: Its Clinical Utility and Feasibility… DOI: http://dx.doi.org/10.5772/intechopen.107441*

low alcohol risk score) [87]. Lasebikan and Ola did a study to determine whether screening, BI and referral for treatment (RT) can reduce the prevalence of tobacco use in rural and semi-rural settings in Nigeria. Participants received a single ASSISTlinked BI and RT at entry, and a booster ASSIST BI and RT at 3 months. It shows that BI with booster sessions at 3 months had a significant effect on tobacco use in people living in community and suggested the need for promotion of such program [88].

## **2.6 The ASSIST-linked brief intervention at the workplace**

There are few published international studies about the implementation of a screening-linked BI using WHO's ASSIST screening scale in the workplace settings. There is a single published study from India conducted by Joseph et al. [89], on the feasibility of conducting the ASSIST-linked screening and BI from a tertiary hospital in north India (from this same institute). The study showed that it was feasible to use ASSIST for screening at the workplace to identify risk level substance use and to use ASSIST-BI for their brief intervention [90]. Joseph et al. [90], also studied the effect of ASSIST-linked BI and compared the mean pre and post-alcohol ASSIST scores in workplace settings for harmful drinking among class C employees of a tertiary hospital in north India. A sample of 39 workers with moderate and high-risk levels of alcohol use was identified by randomly screening 162 employees with ASSIST. Employees who were identified as moderate and high-risk drinkers by the ASSIST were given the BI as per WHO ASSIST-linked BI [90].A significant difference over 4 months (*p* < 0.001) was noticed where the mean ASSIST score reduced from 26.55 (pre-intervention) to 20.06 (post-intervention). There were also improvements in other variables like alcohol consumption, strong desire to use alcohol and health, social and legal problems due


### **Table 1.**

*Comparison of groups at baseline and follow-up on the basis of ASSIST using two-way repeated measure ANOVA.*

*problem of multiple comparisons. Adjusted alpha (α) = α/k (number of comparison). (0.05/3 = 0.016).*


*Risky use of substances was assessed with an application of ASSIST and thus subjects were categorised into different risk levels on the pattern of substance use. As per Table 2, the mean ASSIST score of tobacco users at baseline in the control group was 28.32 (±2.38) and ranged between 22 and 31. Most of the subjects were at high levels of risky use of tobacco (high level, n = 27 and moderate level, n = 3). In the intervention group, the mean ASSIST score was 29.27 (±2.75) and ranged between 22 and 36. Most of the subjects were at high levels of risky use of tobacco (high level, n = 29 and moderate level, n = 4). However, both the groups did not differ statistically on basis of ASSIST score and severity (p = 0.242), (p = 0.350), respectively.*

*In the same Table 2, the mean ASSIST score of alcohol users at baseline in the control group was 31.20 (±−3.4) and ranged between 24 and 38. Most of the subjects were at high levels of risky use of alcohol (high level, n = 26 and moderate level, n = 4]. In the intervention group, the mean ASSIST score was 32.67 (±2.65) and ranged between 26 and 37. Most of the subjects were at high levels of risky use of alcohol [high level, n = 30 and moderate level, n = 1]. However, both the groups did not differ statistically on basis of ASSIST score and risk level (p = 0.113), (p = 0.150), respectively.*

*In the same Table 2, the mean ASSIST score of cannabis users at baseline in the control group was 32.83 (±2.31) and ranged between 30 and 35. All the cannabis users were at high levels of risky use (high level, n = 6 and moderate level, n = 0). In the intervention group, the mean ASSIST score was 31.85 (±4.98) and ranged between 22 and 37. Here also most of the subjects were at high levels of risky use of cannabis (high level, n = 6 and moderate level, n = 1). However, both the groups did not differ statistically on basis of ASSIST score and risk level (p = 0.639), (p = 0.296), respectively.*

### **Table 2.**

*ASSIST score and risk level of randomised groups at baseline.*

to alcohol at follow-up (*p* < 0.001) [90]. In a recent study using randomised controlled trial design, to study the efficacy of ASSIST-linked BI where major objectives were to reduce risky substance use among class C male workers, enhance the progress of subjects through the stages of change and motivate the subjects to seek treatment [91]. The inferential analysis showed that participants receiving BI had a significant reduction of ASSIST scores for all risky use of substances compared with Control. Thus there was a significant reduction in the risk level of all categories of substance use in the intervention group compared with the control group.

The interaction effects in the stage of change indicate that the participants in the intervention group who were using tobacco had significantly changed their stage to action stage more than that of the control group. Similar significant changes were also noticed in the risky alcohol users of the intervention group compared with that of the control group. However, in the risky users of cannabis, the interaction effects indicate *Screening and Brief Intervention in Substance Use Disorders: Its Clinical Utility and Feasibility… DOI: http://dx.doi.org/10.5772/intechopen.107441*

that there was no significant change in the contemplation stage in both groups but significant changes were noticed in precontemplation and action stages in the intervention group compared with the control group.

The interaction effect on quality of life shows that the participants receiving BI had significantly increased scores for all the domains of WHOQOL-BREF compared with that of the control group. Participants receiving BI were significantly more motivated to seek treatment compared to the control group.

### **2.7 Effect of the ASSIST BI on specific substance involvement score**

Two-way repeated measures ANOVA results show that there was a significant reduction of mean tobacco ASSIST scores over time among groups (*F* = 218.95, *p* < 0.001 and observed power 100%). There was also a significant reduction in mean scores among the groups. Moreover, there was a significant interaction effect and the


*ASSIST score was re-assessed after 3 months of follow-up. The mean ASSIST score of tobacco in the control group was 26.62 (±2.57) and ranged between 22 and 31 whereas in the intervention group it was19.29 (±3.26) and it was statistically significant (t = 9.913; p < 0.001). It means that most of the subjects from the intervention group were at a moderate level and none were at high level of risky use of tobacco (moderate risk, n = 31 and high risk, n = 0), and it was statistically significant as compared with the control group (moderate, n = 9 and high level, n = 23) [χ2 = 35.093; p < 0.001] (Table 3).*

*The mean ASSIST score of alcohol users in the control group was 27.66 (±4.03) whereas in the intervention group was 11.62 (5.87) and it was statistically significant [t = 11.831; p < 0.001]. It means that most of the subjects from the intervention group were at low and moderate levels and none were at high level of risky alcohol use (moderate risk, n = 12 and low risk, n = 17), and it was statistically significant as compared to the control group (moderate, n = 8 and high level, n = 19) [χ2 = 36.775; p < 0.001] (Table 3).*

*Similarly, none of the subjects were at a high-risk level of cannabis use in the intervention group as compared with the control group. Further, there was a statistically significant difference in risky use of cannabis in the intervention group as compared with the control group with respect to the mean ASSIST score and mean risk level of cannabis used [(t = 8.805; p < 0.001), (χ2 = 4.950; p = 0.026)], respectively (Table 3).*

**Table 3.**

*ASSIST score and risk level of randomised groups at follow-up.*

**Figure 1.** *Change in total ASSIST score for risky tobacco use.*

**Figure 2.** *Change in total ASSIST score for risky alcohol use.*

group receiving the BI at baseline had significantly lower mean tobacco ASSIST scores at follow-up compared with the control group (*F* = 104.34, *p* < 0.001 and observed power 100%) (**Tables 1**–**3** ). The result is shown graphically in **Figure 1**.

Similarly, statistical significance reduction of mean alcohol as well as cannabis ASSIST scores over time among the groups (*F* = 482.06, *p* < 0.001, observed power 100% and *F* = 92.87 *p* = 0.001, observed power 100%, respectively). There was also a significant reduction in mean scores among the groups using alcohol and cannabis. Moreover, there was a significant interaction effect and the group receiving the BI at baseline had significantly lower mean alcohol as well as cannabis ASSIST

*Screening and Brief Intervention in Substance Use Disorders: Its Clinical Utility and Feasibility… DOI: http://dx.doi.org/10.5772/intechopen.107441*

**Figure 3.** *Change in total ASSIST score for risky cannabis use.*

scores (*F* = 246.16, *p* < 0.001, observed power 100% and *F* = 25.11, *p* < 0.001, observed power 99.3%, respectively) (**Tables 1**–**3**). Results are shown graphically in **Figures 2** and **3**.

## **3. Conclusion**

BI has clear scientific principles in harm reduction, stage of change, motivational interview, simple to deliver and cost-effectiveness. It can use even in opportunistic setting by non-specialist professionals. It can be an extended service for an individual who needs help but not seeking treatment from specialised centres. Thus BI could be considered as part of clinician's responsibility, in addition as such prescribing medicine, ordering test, performing surgical procedures, filling medical forms, etc. It has favourable outcome as evidences show reduction and prevention of various substance-related consequences.
