JMIR Publications

ABSTRACT

Background:

Smallholder farmers receive educational interventions on safe pesticide handling to reduce exposure risks by governmental agencies, industries or Non-Governmental Organizations. However, existing educational interventions have limited effects on changing behaviors. Targeting psychosocial determinants of behavior change in educational interventions through theory- and evidence-based approaches may enhance their effectiveness.

Objective:

We aim at describing the intervention development and study design of a three-arm cluster-randomized controlled trial to assess the effects in improving safe pesticide handling and reducing pesticide exposure of (i) an existing educational intervention and (ii) a newly developed text messages intervention based on the behavior change approach: risks, attitudes, norms, abilities and self-regulation (RANAS).

Methods:

We enrolled 539 Ugandan smallholder farmers in 12 clusters (subcounties). The clusters each with 45 farmers were randomly allocated to the three arms: (i) educational intervention; (ii) educational intervention + RANAS-based text messages; or (iii) a control group. The educational intervention comprised a two-day workshop that targeted multiple aspects of safe pesticide handling while the text messages targeted use of personal protective equipment (PPE) and were based on the RANAS approach. For the intervention development in the current study, this approach includes identifying psychosocial determinants of PPE use at baseline and selecting behavior change techniques to target them in text messages. The primary outcomes of the study are (i) pesticide knowledge, attitudes and practices (KAP) scores indicating the performance along the educational intervention and (ii) frequency of PPE use. Secondary outcomes are the RANAS-based behavioral determinants of PPE use, the frequency of glove use, algorithm-based pesticide exposure-intensity scores, and signs and symptoms of pesticide poisoning. The outcomes were assessed in structured interviews before the intervention (baseline) and at 12-months follow-up. The effect of the interventions among the arms will be analyzed using the intervention arms and baseline measures as predictors and follow up measures as outcomes in linear multivariable mixed models, including the clusters as random effects. The mediating psychosocial determinants of the interventions will be assessed in multiple mediation models.

Results:

The study was conducted from 2020 to 2021: Baseline interviews were conducted in October 2020, the educational intervention was delivered in November 2020. RANAS-based text messages were developed based on the baseline data for relevant behavioral determinants of PPE use and sent between February and September 2021. Follow up interviews were conducted in October 2021. Overall, 539 farmers were enrolled in the study at baseline, and 45 farmers were lost to follow-up by the end of the study.

Conclusions:

The outlined study will contribute to a better understanding of the effectiveness and behavior change mechanisms of educational interventions by using an experimental, cluster-randomized study design to improve pesticide handling among smallholder farmers. Clinical Trial: ISCRN (43921)