Evaluating Biosecurity Measures and Farmer Behavioural Change Strategy to Mitigate Antimicrobial Resistance and Climate Impacts in Poultry Farms in Nepal: A Cluster-Based Intervention Study

The farms in this arm will receive both technical and behavioural interventions. As a part of behavioural intervention, the farmers will receive weekly training and guidance sessions covering rational AMU practices, waste management, hygiene, and biosecurity practices. Delivery will be via. verbal communication, supplemented with IEC materials. Tailored ad hoc guidance will also be offered to improve existing practices. As a part of technical intervention, farmers will be oriented and encouraged to manually spray disinfectant (Virkon®-S at 10 gm/L) twice a week at 6 am on the bedding. Given that the application will occur during shed uptime, poultry flocks will be directly exposed to disinfectant agents, so disinfectants should minimize adverse health effects and minimize AMR selective pressure. Thus, Virkon®-S, a peroxygen compound, would be the ideal.

Technical and Behavioural Interventions
Farm selection and study design
The intervention study will include only 288 farms. The number of farms is calculated using Kelsey’s method with Fleiss’s continuity correction.
The intervention will help evaluate the combined effects of a technical intervention (biosecurity measures) coupled with behavioural intervention (farmer training) on AMU, AMR, CO2 emissions, ammonia levels, and poultry health outcomes. The randomized controlled trial (RCT) design will be applied in two arms.
Control arm: No intervention will be provided during the poultry cycle. However, routine farm monitoring and data collection will be performed as per the protocol.
Intervention arm: The farms in this arm will receive both technical and behavioural interventions. As a part of behavioural intervention, the farmers will receive weekly training and guidance sessions covering rational AMU practices, waste management, hygiene, and biosecurity practices. Delivery will be via. verbal communication, supplemented with IEC materials. Tailored ad hoc guidance will also be offered to improve existing practices. As a part of technical intervention, farmers will be oriented and encouraged to manually spray disinfectant (Virkon®-S at 10 gm/L) twice a week at 6 am on the bedding. Given that the application will occur during shed uptime, poultry flocks will be directly exposed to disinfectant agents, so disinfectants should minimize adverse health effects and minimize AMR selective pressure. Thus, Virkon®-S, a peroxygen compound, would be the ideal.

2026-02-01

2026-04-30

Effectiveness of disinfectant and farmer training in poultry farms in mitigating AMR and climate change

● Changes in knowledge, attitudes, and practices (KAP) of improving poultry health.
● Effect of the intervention of biosecurity measures and farmer training on AMU, AMR, CO2, ammonia levels, and poultry health.
●Understanding implementation process of the biosecurity measures and farmer training interventions using the RE-AIM framework.

A mixed-methods, randomized control trial approach combining quantitative and qualitative data collection will be employed. The farm observation tool will be utilized to monitor and evaluate the intervention's effectiveness. Additionally, process monitoring will be conducted to document fidelity, the quality of implementation, and identify challenges and success factors through observations, note-taking, and verbal communication.

Study Area and Study Population
The study will be conducted in the Bagmati Province in Nepal. Bagmati Province is the largest producer of poultry products in the country, producing approximately 113,432 tonnes of poultry meat and 899 million eggs in the last year.15 Bagmati Province houses a 30 million poultry population, which is 45.23% of the total poultry population of Nepal. The list of the broiler farms of small- and medium-scale will be prepared with data from the Ministry of Agriculture and Livestock Development (MoALD), Department of Livestock Services (DLS), Veterinary Standards Drugs Regulation Laboratory (VSDRL), Central Veterinary Laboratory (CVL), Nepal Poultry Federation, and poultry feed manufacturers and suppliers.
Sample Size and Sampling Technique
In 2015, there were 1,341 commercial small- and medium-scale broiler farms in Bagmati province.9
Sample size (n)= [DEFF*Np(1-p)]/ [(d2/z21-α/2*(N-1)+p*(1-p)] where,
Finite population (N)= 1,341,
95% Confidence level (z)= 1.96,
Margin of error (α)= 5%,
Estimated proportion (p)= 0.5 (most conservative value).
Design effect (DEFF)= 1.2
Plugging in the values, the calculated sample size is 359. Anticipating a 10% non-response rate as a buffer, the total sample size would be 395 (≈ 400 farms). The farm mapping and baseline KAP survey will be conducted in Kathmandu, Chitwan, and Makwanpur districts of Bagmati province. The random sampling of the farms will be done proportionately among districts. The lead farmer of the farm will be the study participants for the KAP survey, as they are the key decision-makers and/or implementers of the practices inside the farm.
Survey Tool and Data Collection Technique
The questionnaires will be developed by consulting stakeholders from MoALD, DLS, and VSDRL. The farm mapping questionnaire will include the questions relating to the basic farm characteristics (flock stock, flock density, shed type, floor type, ventilation system), litter management practices, any traditional feed practices, water supply system, feed types and suppliers, chicks suppliers, vaccination status, climate-smart practices, geolocation (GPS), number of cycle/years, willingness to participate in intervention study, farmer’s sociodemographic characteristics (including gender), and farmer qualification/experience. The baseline KAP survey questionnaire will comprise different sections, viz., basic farm characteristics, farmers’ understanding of climate-related stressors, AMU practices, existing biosecurity practices, and their willingness to adopt mitigation measures. The data will be collected using KoboToolbox.
Pre-Testing of Baseline KAP Survey Tool
Prior to the survey, the questionnaire will be pre-tested among 40 farmers from the Lalitpur District. The questionnaire will be refined based on the pre-test result and expert consultation. The farms included in the pre-testing will be excluded from the study.
Data analysis of the baseline KAP survey
The Kendall rank correlation coefficients of KAP scores will be used to assess the strength and direction of association between variables.
Farm selection for intervention study
The farms will be allocated into clusters based on the shared baseline characteristics from farm mapping. The clusters will represent homogeneous strata. This will ensure minimalistic farm-level confounding biases.

Stakeholder Workshop for Co-designing IEC Materials and Intervention Strategy
Stakeholder Workshop
A one-day consultative workshop will be conducted. The participants will include poultry farmers, feed industries, and stakeholders from DLS, CVL, VSDRL, MoALD, Quality, Standards and Regulation Division (QSRD), and Ministry of Health and Population (MoHP).
IEC material development and co-designing intervention study
Based on preliminary KAP findings, the IEC materials will be developed with a focus on rational AMU, implementing basic biosecurity measures, and climate-smart practices. The IEC materials will include social media clips, leaflets, posters, AMU logbooks, flock weekly management checklists, flock health logbooks, and farm logbooks.
IEC material distribution
The IEC materials will be distributed among the farms in close coordination with CVL, DLS, VSDRL, and MoALD.

Randomization will be done by random number method using computer (MS Excel).

Farm

No

The calculated sample size is 288, split equally into two arms.

288

Using stratified randomization, the farms from the clusters will be allocated equally across two intervention arms (144 arms per arms: control and intervention).

Assumptions for sample size calculation for Randomized Controlled Trials (RCT): Two-sided significance level (1-alpha): 95 Power (1-beta, % chance of detecting): 80 Ratio of sample size, control/intervention: 1 Odds Ratio: 2