'As soon as they can hold a glass, they begin taking alcohol': a qualitative study on early childhood substance use in Mbale District, Uganda.

BACKGROUND: Globally, substance use is a leading contributor to the burden of disease among young people, with far reaching social, economic and health effects. Following a finding of harmful alcohol use among 5-8-year-old children in Mbale District, Uganda, this study aims to investigate community members' views on early childhood substance use among children below the age of 10 years. METHODS: In 2016, we conducted eight focus group discussions with 48 parents and 26 key informant interviews with teachers, health workers, alcohol distributors, traditional healers, religious leaders, community leaders and youth workers. We used thematic content analysis. Four participants and two research assistants reviewed and confirmed the findings. RESULTS: Alcohol in everyday life: 'Even children on laps taste alcohol': Almost all participants confirmed the existence of and concern for substance use before age 10. They described a context where substance use was widespread in the community, especially intake of local alcoholic brews. Children would access substances in the home or buy it themselves. Those living in poor neighbourhoods or slums and children of brewers were described as particularly exposed. Using substances to cope: 'We don't want them to drink': Participants explained that some used substances to cope with a lack of food and resources for childcare, as well as traumatic experiences. This made children in deprived families and street-connected children especially vulnerable to substance use. Participants believed this was a result of seeing no alternative solution. CONCLUSIONS: To our knowledge, this is the first study to describe the context and conditions of childhood substance use before age 10 in Mbale District, Uganda. The study shows that community members attributed early childhood substance use to a social context of widespread use in the community, which was exacerbated by conditions of material and emotional deprivation. These social determinants for this practice deserve public health attention and intervention.

Does mhGAP training of primary health care providers improve the identification of child- and adolescent mental, neurological or substance use disorders? Results from a randomized controlled trial in Uganda.

BACKGROUND: Integrating child and adolescent mental health (CAMH) into primary health care (PHC) using the WHO mental health gap action program (mhGAP) is recommended for closing a mental health treatment gap in low- and middle-income countries, but PHC providers have limited ability to detect CAMH disorders. We aimed to evaluate the effect of PHC provider mhGAP training on CAMH disorder identification in Eastern Uganda. METHODS: Thirty-six PHC clinics participated in a randomized controlled trial which compared the proportion of intervention (n = 18) to control (n = 18) clinics with a non-epilepsy CAMH diagnosis over 3 consecutive months following mhGAP-oriented CAMH training. Fisher's exact test and logistic regression based on intention to treat principles were applied. (clinicaltrials.gov registration NCT02552056). RESULTS: Nearly two thirds (63.8%, 23/36) of all clinics identified and recorded at least one non-epilepsy CAMH diagnosis from 40 692 clinic visits of patients aged 1-18 recorded over 4 months. The proportion of clinics with a non-epilepsy CAMH diagnosis prior to training was 27.7% (10/36, similar between study arms). Training did not significantly improve intervention clinics' non-epilepsy CAMH diagnosis (13/18, 72.2%) relative to the control (7/18, 38.9%) arm, p = 0.092. The odds of identifying and recording a non-epilepsy CAMH diagnosis were 2.5 times higher in the intervention than control arms at the end of 3 months of follow-up [adj.OR 2.48; 95% CI (1.31-4.68); p = 0.005]. CONCLUSION: In this setting, mhGAP CAMH training of PHC providers increases PHC clinics' identification and reporting of non-epilepsy CAMH cases but this increase did not reach statistical significance.