Study design
This research utilised a quantitative, cross-sectional, survey research design, in which participants were screened on one occasion utilising the AUDIT-C measure. Data collection occurred in Alexandra township, Johannesburg, South Africa, across all four quarters of 2021. The administration of the study coincided with the global COVID-19 pandemic and the national government’s consequent alcohol bans and travel restrictions. South Africa was one of the few countries to implement an alcohol ban intermittently as an emergency response [9].
The data collection process was managed by HIV South Africa (HIVSA), a not-for-profit organisation that has been in existence since 2002. HIVSA is a key strategic partner of the Gauteng Provincial Department of Health and the Gauteng Provincial Department of Social Development. Given HIVSA’s existing work in response to HIV and associated socioeconomic and health issues, the organisation was considered an appropriate partner to assist in the administration of the screening tool. HIVSA was responsible for recruiting, supervising and managing the lay counsellors who administered the AUDIT-C tool and offered brief intervention with or without referral. The lay counsellors received training over an initial three-day period followed by two additional one-day booster sessions. The training was facilitated by two registered counselling psychologists who were both experienced in providing brief intervention services. The training included the project’s aims, the study protocol, and the scoring of the AUDIT-C tool, followed by role-play exercises focusing on the brief advice and referral processes.
Data were collected in person using paper and pencil. The brief advice provided was premised on the Feedback, Responsibility, Advise, Menu for change, Empathy, and enhancing Self-efficacy (FRAMES) intervention framework [18], which focuses on feedback and advice components. Feedback comprised explaining the screening score to participants. Participants who received a low-risk score (AUDIT-C score of 1–4) were advised that while there is no completely safe level of alcohol use, their screening score suggested that they were drinking in a way that was less likely to result in harm to themselves or others. However, these participants were cautioned to keep track of their drinking to ensure that it did not gradually increase. Participants who received a moderate-risk score (AUDIT-C score of 5–7) were advised of the risk of experiencing harm from their drinking. Advice comprised explaining the best way to reduce the risk of harmful alcohol use, which focused on reducing the amount of alcohol consumed. Participants who received a high-risk score (AUDIT-C score of 8 +) were also advised that they were at risk of experiencing harm from their drinking. The AUDIT-C tool, in addition to the recommended intervention, took between 10 and 30 min to administer, depending on each participant’s score and the form of intervention needed.
Recruitment and the administration of the AUDIT-C tool took place across four sites. This included two primary healthcare clinics, Community HIV Testing Services, and an Orphans, Vulnerable Children and Youth Community-based Organisation (OVCY CBO).
The primary healthcare settings comprised the River Park and East Bank clinics. Both clinics provide primary healthcare services, in addition to other specialised services. Recruitment occurred in the primary care sections of both primary healthcare settings. The recruitment process was based on convenience sampling in both primary healthcare settings, and participants were not recruited based on their reasons for visiting the clinics. The recruitment process took place in the public communal areas of the primary health care settings, the screenings took place in cordoned-off health desk areas.
The Community HIV Testing Services, as well as the OVCY CBO, aligned with some of the interventions managed by HIVSA. Potential participants were then recruited through mobile testing services, home visits, and recruitment in informal community settings. The recruitment and screening processes across the mobile testing services and home visits took place in private settings.
Referral sites
The River Park Clinic is a public primary healthcare facility that includes the River Park Community-based Substance Abuse Treatment Centre, which opened in 2018. The overall intent of the centre, as a division of the clinic, is to provide outpatient rehabilitation services to substance users and their families, and as such, the River Park Clinic was also nominated as a referral site for individuals identified by the AUDIT-C screening process as needing alcohol use counselling or treatment. The East Bank Clinic provides psychiatric and psychological services to individuals across different life stages, in addition to primary health care services, and was therefore also nominated as a referral site. The South African National Council on Alcoholism (SANCA) was also a referral partner, although not a screening site.
Sample
Participant selection criteria included individuals aged 18 years or older who were living and/or working in Alexandra township during the research.
Measure
This study implemented the AUDIT-C, a shortened version of the 10-question AUDIT instrument, comprising 3-item alcohol consumption questions. The AUDIT and AUDIT-C have been recognised as valid instruments to screen for possible unhealthy alcohol use [20, 21]. The AUDIT includes three domains relating to the level of alcohol consumption, evidence of dependence and harm from drinking. The shortened AUDIT-C is considered to perform as well as the full AUDIT in primary healthcare settings [21]. The AUDIT-C has also been implemented in different Southern African settings, at times indicating greater sensitivity to risk than other instruments [19, 20].
For the administration of the AUDIT-C tool, 12 g of pure alcohol was considered a standard drink, as typically defined in South Africa [22, 23]. Additionally, the imagery of a standard drink and the descriptive language were adjusted to ensure relevance to Alexandra township. This included references to 30 ml spirits as a ‘tot’ and a 330 ml beer as a ‘dumpie’. A South African-specific beverage, ‘traditional home-brewed beer called umcombotsi’, was also included. The alcohol percentage of traditional home-brewed beer/umcombotsi cannot be determined since it is subject to various home-brewing recipes [24]. The latter was particularly relevant during the COVID-19 pandemic, given the full and partial restrictions on alcohol sales.
The study protocol also required that the lay counsellors indicate whether brief advice or a referral was provided and whether the referral was accepted or declined. The respondents’ age and sex were also recorded. Research suggests that the optimal cut-off score of the AUDIT-C may vary across contexts, populations and countries. The AUDIT-C score, as per AUDIT, therefore, allows for an adjustment of the screening threshold for particular settings. The scoring should therefore be influenced by national and cultural standards as assessed by clinicians [20]. This study implemented a score of 5 as the threshold for a positive AUDIT-C score because it has been shown to be an optimal cut-off score by select studies [21, 25, 33]. Additionally, this cut-off score has been found to be appropriately sensitive while also preventing high false-positive rates [21, 25]. The latter was considered of importance in the already resource-constrained context. Each AUDIT-C question is scored 0 to 4 points, resulting in a total score ranging from 0 to 12 [26]. A total score ranging from 0 to 4 is considered low-risk alcohol consumption, a score ranging from 5 to 7 is regarded as moderate risk and a score of 8 to 12 signals high risk alcohol consumption behavior. Participants with a score of 5 and above are recommended to receive brief advice, whereas both brief advice and a referral are recommended for participants with a score of 8 and above. In the current study, individuals scoring 1 to 4 also received a brief intervention as a preventive measure.
Data entry and analysis
Data entry staff were employed to capture survey responses in Microsoft Excel. Data-cleaning processes included removing records where demographic or AUDIT-C-specific responses were missing or captured incorrectly.
Age was categorised to allow for a nonlinear association between age and the AUDIT-C score. The associations of gender, age (categorised) and AUDIT-C score were determined by the chi-squared test. The odds ratio (OR) of each study variable, rejection of brief advice and rejection of referral was determined using binomial regression. Reference categories were determined based on sample size and clinical relevance (place of screening: largest site; age: youngest age category; gender: female (because males were thought to be at higher risk of rejection of advice/referrals); AUDIT-C score: 5–8 for brief advice rejection (lowest AUDIT score for which brief advice is routinely given) and 8 for referral rejection (lowest AUDIT score for which referral is routinely given). Variables that were significant at the univariate level were included in a multivariable binomial regression analysis. Nonsignificant variables were sequentially removed from the multivariable model. Data analysis was carried out using STATA version 8. A 1% significance level was used.
Source link
