Impact of the Sachet Alcohol Ban on Alcohol Availability in Uganda

Abstract

Objective:

Alcohol use continues to be a major public health problem in Uganda. This study sought to estimate the availability of sachet alcohol among retail food and beverage establishments in two Ugandan districts (Kampala and Jinja) before and after enactment of the national sachet alcohol ban.

Method:

Independent observations in 100 establishments were conducted by a pair of trained observers during the peak hours of 7 P.M. to 1 A.M. Using a standardized checklist, observers recorded indications of alcohol availability before and after enactment of the sachet alcohol ban.

Results:

Availability of (a) alcohol and (b) sachet alcohol were significantly affected by the ban. Before the ban, 69% of all establishments sold alcohol; there was a significant reduction in alcohol availability after enactment of the ban to 43% of the establishments (p < .001). This reduction was observed in off-premise establishments (p < .001), but not in on-premise establishments (p = .710). Additionally, before the sachet alcohol ban, 52% of all establishments sold sachet alcohol; however, there was a significant reduction in sachet availability after enactment of the ban (1.4%, p < .001).

Conclusions:

Legislation banning the manufacture and sale of sachet alcohol has the potential to reduce sachet availability. Future studies should examine changes in alcohol consumption following the sachet alcohol ban.

Alcohol misuse is a major global health threat, is associated with more than 200 disease and injury conditions (World Health Organization [WHO], 2018), and has both significant social and economic negative impacts on individuals and society. The WHO estimates that approximately 3 million deaths worldwide resulted from harmful use of alcohol in 2016 (WHO, 2018). Globally, the African region has the highest age-standardized alcohol-attributable burden of disease and injury (WHO, 2018). In addition, the western, southern, and central sub-Saharan regions of Africa have been associated with the greatest age-standardized alcohol-attributable rates for disability-adjusted life-years (DALYs; Shield et al., 2020). More specifically, alcohol consumption has been found to be the leading risk factor for DALYs amongst African males ages 15–24 years (Gore et al., 2011). In 2012 alone, alcohol was responsible for 6.4% of all deaths and 4.7% of all DALYs lost in Africa (Ferreira-Borges et al., 2016).

Uganda's consumption of alcohol is among the highest on the African continent (WHO, 2014). In a 2011 global assessment of alcohol-related negative consequences among drinkers by country, it was revealed that acute medical consequences, personal consequences, and broader social consequences were of the highest prevalence in Uganda (Graham et al., 2011). According to the WHO, 7.1% of the Ugandan population age 15 years and older had alcohol use disorder in 2016 (WHO, 2018). Nalwadda and colleagues (2018) studied alcohol use in rural Uganda by interviewing men in primary care clinics and a community sample in the Kamuli district using the Alcohol Use Disorders Identification Test. They found that 4.1% of the sample in the community study and 5.8% in the facility study tested positive for alcohol use disorders (Nalwadda et al., 2018). Kabwama and colleagues (2016) completed a secondary data analysis and estimated that 9.8% of the adult population had an alcohol use–related disorder.

The results of high levels of alcohol use disorders are devastating in terms of adverse health outcomes—youth who drink have increased odds of contracting HIV, of being fatally injured in a vehicular crash, of committing or being a victim of rape or other violence, or of incarceration (Graham et al., 2011; Rahav et al., 2006; Swahn et al., 2018; Tumwesigye & Kasirye, 2006; Tumwesigye et al., 2012). In addition, the individual-level cognitive and behavioral effects of alcohol are worth mentioning—alcohol suppresses the hippocampus, which is involved in making memories (Ryabinin, 1998).

The alcohol industry investments in the African region include corporate social responsibility activities as well as major marketing campaigns, signaling that Africans are a targeted and developing demographic for the alcohol industry (Babor et al., 2015). African governments have responded by initiating public health measures designed to thwart the effectiveness of alcohol industry marketing campaigns (Ferreira-Borges et al., 2015). One such public health measure is the ban on sachet alcohol sales in Uganda.

On May 30, 2019, a new law was enacted that deemed the manufacture and sale of sachet alcohol illegal (Daily Monitor, 2019). Once ubiquitous throughout Uganda, sachet alcohol is a soft pouch-style plastic baggie typically containing 100 ml of average 42% alcohol by volume, or 33 g of pure alcohol total. One sachet typically costs approximately 650 Ugandan shillings (about 18 U.S. cents).

In a recent study, Otim and colleagues (2019) examined the health risks associated with sachet alcohol consumption in Uganda. Findings revealed that no amount of alcohol tested was safe to ingest because of the presence of heavy metals, specifically arsenic, lead, and chromium (Otim et al., 2019). These health risks are of great concern, as alcohol use has greatly increased amongst adolescents and individuals living in poverty (Masika, 2019; Uganda Alcohol Policy Alliance, 2014). After evaluating alcohol use among secondary school students in Uganda, Masika (2019) found that 35.0% had consumed sachet alcohol within the last 12 months, an alarming percentage when current projections estimate that 69.86% of Uganda's total population is under age 25 (World Population Review, 2021).

To date, only one other study has examined the implementation of a sachet ban on the African continent. In Malawi, the adolescent perceptions of a 2012 sachet ban and its effectiveness were examined, thus focusing on attitudes surrounding the ban rather than the actual impact of reducing alcohol consumption (Salimu & Nyondo-Mipando, 2020).

The legislation to reduce alcohol problems in Uganda should be monitored and evaluated, and the extended time-frame before the enactment of the ban provided a unique opportunity for alcohol policy evaluation. The aim of this study was to estimate the availability of sachet alcohol among retail food and beverage establishments in two districts, Kampala (urban setting) and Jinja (peri-urban setting), in Uganda to evaluate these establishments’ adherence to the national law banning sachet alcohol sales at times before and after the ban.

Method

Study area, design, and sampling

This was a repeated cross-sectional study that involved collection of observational data at food and beverage retail establishments along two streets—one in an urban setting in the Kampala district and the other in a peri-urban setting in the Jinja district. In Uganda, a district is the largest functional administrative unit. Current 2020 population estimates show that Kampala has a population size of 1,680,600 and that Jinja has a population size of 515,100 (Uganda Bureau of Statistics, 2020). Kampala is the capital of Uganda and lies in the central part of the country. Administratively, it is divided into 5 subcounties, 76 parishes, and 3,290 villages (Land Conflict Mapping Tool, 2020a). Jinja district lies in the southeast region of Uganda, 54 miles (87 km) northeast of Kampala. Administratively, the district is divided into 4 counties, 11 subcounties, 69 parishes, and 696 villages (Land Conflict Mapping Tool, 2020b).

An urban and a peri-urban setting were selected to achieve a greater generalizability of the study findings. In Kampala, a section of Tank-Hill Road in the Kabalagala area was selected. In Jinja, a segment of Jinja-Kamuli Road in the Mafubira area was selected. Although Mafubira is one of the busiest areas in Jinja, we assigned it a “peri-urban” classification because it is sparsely populated in comparison to other Ugandan cities. From those selected streets, every retail food and beverage establishment was observed. The streets were selected to fit a criterion of the adequate number of potential establishments (at least 40).

We observed any place (e.g., a marketplace-style stand, building, or open-air service area) in the defined area where one could pay to obtain either food and/or beverage. Because alcohol availability is not outwardly visible before entering an establishment, all establishments serving food or beverage were included for observation regardless of whether they sold alcohol. Although establishments that remained in operation and in the same location across waves were rated multiple times, we did not aspire to compare individual establishments across waves (especially because many of the retail establishments did not have names, making it difficult to confidently know which places were the same). Rather, we sought to understand the overarching, less granular landscape of alcohol availability in the defined areas. On-premise was simply defined as an establishment that allowed consumption of alcohol onsite, regardless of whether it also sold alcohol “to-go.” Off-premise was defined as an establishment requiring that alcohol sold at the establishment be consumed elsewhere, offsite. For off-premise establishments, we also observed empty discarded sachets found either on the ground or in a trash receptacle as evidence of sachet use before data collectors entered the establishment.

Data collection tool

The Sachet Alcohol Observation Checklist was constructed to gather evidence of sachet alcohol availability or use inside the establishment and evidence of sachet alcohol availability or use outside. The term availability was used to describe alcohol that was available to be purchased from a given establishment. The checklist is based on the Tobacco Observation Checklist used in similar environmental-level substance use policy evaluation research (Nesoff et al., 2017). Most items (e.g., patron characteristics and establishment domains) of the original tobacco observational assessment were retained, and additional observational items related to the sachet ban were added for this research. The Sachet Alcohol Observation Checklist consisted of various establishment characteristics, including size, cover charge, outdoor seating/space, and patron density. All checklist items were restricted to observations available in the public domain—activities and observations made by the layperson who entered the establishment. The Sachet Alcohol Observation Checklist is freely available to any user via Epicollect under the project title “Sachet Alcohol in Uganda” (https://five.epicollect.net/project/sachet-alcohol-in-uganda).

Data collection

Data before the enactment of the sachet alcohol ban were collected on January 26 and 27, 2019. After enactment of the ban, data were collected on November 8 and 9, 2019. During both waves of data collection, observations occurred between 7 P.M. and 1 A.M.

Data collectors worked in pairs. To conduct the observational study without attracting attention to data collection, all data collectors were Ugandan. The data collectors were trained to walk around the establishment together, observing behaviors in publicly accessible areas, including the exterior of the establishment, bathrooms, and hallways. The observers entered data into smartphone devices equipped with the Epicollect form containing the Sachet Alcohol Observation Checklist. In addition to completing the observational checklist, data collectors made field notes regarding information not captured by the checklist, as needed. The data were uploaded from smartphones to the Epicollect server after every single day of data collection.

Data quality procedures

All data collectors were required to attend a training session that included policies and procedures for data collection, review of the items and their definitions in the Sachet Alcohol Observation Checklist, and instructions for using Epicollect. Additional training and quality control were provided through meetings following each data collection.

During data cleaning, a simultaneous testing approach was used, and any discrepancies in observations were reconciled in favor of the presence of an observed indicator; in other words, if one observer noted an indicator, it was deemed as present.

Statistical analyses

Analyses were performed using Stata 15 statistical software (StataCorp LP, College Station, TX). Descriptive analyses, such as frequencies and percentages, were performed for establishment and sachet alcohol availability characteristics. The Wilcoxon rank-sum test was used to assess differences in observed sachet alcohol availability and consumption in establishments before and after enactment of the sachet alcohol ban. All probability values (p values) were considered significant if less than .05.

Results

Relevant results are presented in Table 1. Before the ban, 69% of all establishments (n = 100; 60 in Kampala and 40 in Jinja) sold alcohol; there was a significant reduction in alcohol availability after enactment of the ban to 43% of establishments (p < .001). This significant reduction was observed among off-premise establishments (p < .001), but not for on-premise establishments (p = .710).

TABLE 1.

Distribution of observed sachet alcohol availability and consumption characteristics before and after enactment of the sachet ban

Characteristics	Number of establishments where characteristic was observed		P
	Pre-sachet ban n (%)	Post-sachet ban n (%)
Establishment size
Large (>2,000 sq. ft.)	8 (8.0)	0 (0.0)	.024
Medium (600-2,000 sq. ft.)	15 (15.0)	5 (5.0)
Small (<600 sq. ft.)	22 (22.0)	30 (30.0)
Very small (too small to enter)	55 (55.0)	65 (65.0)
Establishment type			.182
On-premise	61 (61.0)	70 (70.0)
Off-premise	39 (39.0)	30 (30.0)
Alcohol availability			<.001
Yes	69 (69.0)	43 (43.0)
No	31 (31.0)	57 (57.0)
Sale of alcohol at on-premise establishments			.710
Yes	32 (52.5)	39 (55.7)
No	29 (47.5)	31 (44.3)
Sale of alcohol at off-premise establishments			<.001
Yes	37 (94.9)	12 (40.0)
No	2(5.1)	18 (60.0)
Sachet alcohol availability			<.001
Yes	52 (52.0)	1 (1.0)
No	48 (45.0)	99 (1.0)
Sale of sachet alcohol at on-premise establishments			<.001
Yes	32 (52.5)	1 (1.4)
No	29 (47.5)	69 (98.6)
Sale of sachet alcohol at off-premise establishments			<.001
Yes	20 (51.3)	0 (0.0)
No	19 (48.7)	30 (100.0)
Evidence of prior sachet consumption at off-premise establishments			.002
Yes	11 (28.2)	0 (0.0)
No	28 (71.8)	30 (0.0)

Before the sachet alcohol ban, 52% of all establishments sold sachet alcohol; however, there was a significant reduction in its availability after enactment of the ban (reduced to 1.4%; p < .001). In addition, this significant reduction in sale of sachet alcohol was observed at both on-premise establishments (p < .001) and off-premise establishments (p < .001) after the ban. With regard to the off-premise establishments, prior evidence of sachet consumption was also reduced (p = .002) after implementation of the sachet alcohol ban.

Discussion

In 2019, the Ugandan government implemented two policies banning the manufacture and sale of a specific but ubiquitous alcohol product, sachet alcohol. This study aimed to determine the availability of sachet alcohol among food and beverage establishments in two Ugandan districts before and after the ban. Our findings reveal that there was a significant reduction (decreased to almost zero) in the availability of both alcohol and sachet alcohol after enactment of the ban. Only one establishment in the measured areas continued to sell sachet alcohol, thereby breaking the law. The ban prohibited alcohol manufacturers from packaging alcohol in sachets, which surely affected the availability of sachet alcohol on the consumer market (Daily Monitor, 2019). The decline in alcohol availability seems to have been driven by off-premise establishments, which may imply that sachet alcohols were the only form of alcohol that many off-premise establishments sold, likely because they are low cost. However, banning sachet alcohol may lead consumers to resort to other forms of low-cost alcohol, such as informal alcohol (often referred to as local brew), which may account for as much as 86% of all alcohol consumed in Uganda (WHO, 2018).

The sachet alcohol ban may have led to reduced sachet alcohol availability in retail food and beverage establishments. An inspection completed 3 months after enactment of the sachet alcohol ban by the Uganda National Bureau of Standards (Uganda's standard regulatory authority) found that 37 alcohol-manufacturing companies were still producing sachet alcohol in July 2019 (Gahene, 2019). Although we do not know whether 37 represents a large or small fraction of the number of companies that manufactured sachet alcohol before the ban, this number indicates that active enforcement of the ban may be necessary to prevent reemergence of sachet alcohol sales. Furthermore, this act of noncompliance indicates that an in-depth analysis of these 37 companies and their marketing activities needs to take place.

Our study, however, has several limitations. First, the sample of establishments used for this study may not be representative of retail establishments selling alcohol in Kampala and/or Jinja. Second, the observation of establishments on one street per district likely does not provide a fair representation of the establishments in a district. Third, because many of the retail establishments did not have names, we were unable to confidently know which places were the same between waves of data collection. Last, the Tobacco Observation Checklist/tool has not been validated in Uganda or any other low-to-middle-income country.

To build on this work, future studies should examine (a) urban/rural differences in sachet alcohol availability and (b) changes in alcohol consumption following the sachet alcohol ban. In addition, a second wave of data collection in the same locations could further monitor sustained compliance.

Acknowledgments

The authors thank Megan Mulheron for manuscript coordination and the research assistants for their tremendous efforts to collect the data. Data availability: The data set used to support the findings of this study is available from the corresponding author on request.

Conflict-of-Interest Statement

The authors declare no conflicts of interest regarding the publication of this article.