Impact evaluation of a small ruminant flock health intervention on animal health, welfare, and farmer income in Ethiopia

Abstract

This study evaluated the impact of a flock health intervention addressing key animal health and welfare challenges in small ruminants in Ethiopia through scheduled deworming and vaccination, complemented by flock health management training for smallholder farmers in two districts: Doyogena and Angollelana Tara. The study employed a randomized controlled trial, which was conducted for one year from May 1, 2023 - April 30, 2024, to measure the intervention’s effectiveness on animal health, animal welfare and farmer income. Post-intervention results showed lower morbidity (21% vs. 33% in Doyogena; 13% vs. 19% in Angollelana Tara) and mortality rates (6% vs. 9% in both districts) in treatment groups compared with control groups. However, only the reduction in morbidity in Doyogena was statistically significant (P < 0.05); the other differences lacked strong statistical support (P > 0.05). Most of the health and welfare indicators such as FAMACHA score, DAG score, body condition score, and hair coat condition recorded during the intervention showed improvement in the treatment group as compared with the control group with strong evidence (mostly P < 0.05). Gastrointestinal strongyle prevalence and worm burden decreased in treatment groups in both districts with strong evidence (P < 0.05) in Doyogena district. In Doyogena, prevalence decreased from 64% to 37%, and worm burden dropped from 500 to 290 EPG in the treatment group, compared with an increase in the control group (74% to 98% prevalence; 468 to 824 EPG worm burden). Economically, treatment flocks generated a higher annual gross margin than control flocks in both districts, a difference that was statistically significant in Doyogena (P < 0.05). Overall, the intervention led to substantial improvements in animal health, welfare, and farmer profitability, with the potential for expansion to similar production systems.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12917-026-05397-9.

Introduction

Ethiopia is home to a large sheep and goat population, estimated at approximately 95 million [1]. However, the productivity of this sector remains low due to several factors, including prevalence of animal diseases [2]. A national assessment of mortality in small ruminants showed an annual mortality of up to 34% which varied from flock to flock [3]. Mortality rate in lambs and kids is notably high amounting as much as 40% [4]. A study identified respiratory diseases as the highest concern among small ruminant diseases, followed by gastro-intestinal parasites in crop-livestock mixed system in Ethiopia [5]. A community-based disease control intervention was carried out in selected districts in the crop livestock mixed system between 2018 and 2021 to mitigate the impact of these diseases [6]. The main elements of this intervention included vaccinating against respiratory diseases, deworming against gastro-intestinal tract (GIT) parasites and community conversations to enhance awareness and share knowledge regarding flock health management [6].

During the 2018–2021 intervention, there was a notable decline in both morbidity and mortality rates, with morbidity rates exhibiting a decreasing trend throughout the implementation period [7]. Prevalence and worm burden of internal parasites also decreased in most of the intervention sites [8]. Generally, the intervention seemed effective in decreasing mortality and parasitic worm burden. However, it was not possible to conclusively attribute the improvements to the intervention because of the lack of a control group in the implementation design. The evaluation of an intervention needs to be done with an appropriate control to produce robust evidence. Animal health interventions should also be evaluated for their profitability as livestock are primarily kept as economic commodities to provide a means of livelihood, and an intervention may be highly effective but not profitable. Furthermore, besides financial impact, the impact of a health intervention on the welfare of animals must also be considered. To address these gaps, the present study evaluated the impact of implementing flock health intervention on the health and welfare of small ruminants, and farmer income in a randomized controlled design. This study aims to inform cost-effective strategies for improving small ruminant production in crop-livestock production systems by providing evidence on the benefits of flock health intervention.

Materials and methods

Study sites

The study was conducted in two districts: Doyogena district in Kembata administrative zone of Central Ethiopia region and Angollelana Tara district in North Shewa administrative zone of Amhara region (Fig. 1). The districts are among known sheep rearing areas within the crop livestock mixed system in the country [9, 10]. Small ruminants are usually kept in mixed herds with cattle on communal grazing areas in Angollelana Tara. In Doyogena, individual flocks are tethered during crop cultivation season of the year and mixed with other flocks in communal grazing areas for the rest of the year. The flocks in both districts are sheltered in barns at night, which helps protect them from predators and adverse weather. In both districts sheep was the dominant species of small ruminant flocks accounting about 90% of the study flocks. Animal health services in both districts are predominantly given by the public sector, and this include providing clinical services at stationed animal heath posts and clinics, and extension services by livestock extension agents. Small ruminant farmers were sampled from Hawora kebele from Doyogena and Cheki kebele from Angollelana Tara districts. Kebele refers to a subdivision of a district that is the smallest administrative unit in Ethiopia.

Fig. 1.

Map of regional states of Ethiopia showing the study sites, produced by authors using QGIS software [11]

The intervention and its implementation

The intervention was a package of flock health preventative measures focusing on helminth parasites and respiratory diseases of small ruminants. These include strategic deworming of small ruminants for control of GIT helminths, strategic vaccination against respiratory diseases including ovine pasteurellosis and sheep and goat pox, and participatory training for farmers on flock health and welfare management. Albendazole 300 g (Shialben 300 mg bolus, Ashish life science PVT limited, India) in Doyogena and Ivermectin 1% (Ivermectin 1%, Hebei Kexing Pharmaceutical Co., Ltd, China) in Angolla Tara were used for deworming. The choice of drugs was based on availability in the local market in the respective study sites during the study period. Ovine pasteurellosis and sheep and goat pox vaccines from Ethiopia’s National Veterinary Institute were used for vaccination.

Strategic deworming and vaccination calendar were prepared for each study district (See Additional file 1). The research team prepared the calendars in collaboration with animal health staff in the districts who provided local epidemiological factors in the form of seasonal climate variation, and market and pasture related animal movement.

The intervention package was implemented in the treatment flocks through public-private partnerships (PPP) for one year from May 1, 2023-April 30, 2024, while the control flocks were left for existing veterinary services and practices. In the PPP arrangement, private animal health service providers with demonstrated professional competence were selected by each district veterinary office and community leaders. These private animal health service providers were engaged to implement flock health preventive measures according to the calendar developed for the intervention study (Additional file 1). The district veterinary offices, as public partners, facilitated the intervention by mobilizing flock owners during scheduled activities and monitored the quality of service given by the private partners. The study project supported farmers with flock health service fees through a voucher system. More information on the PPP arrangement is available in reference [12]. The service fee subsidy from the project was to show the value of the intervention for farmers so that they would pay by themselves after the project’s lifetime.

The flock health and welfare management training focused on five areas including helminths and anthelminthics, infectious disease and vaccination, newborn management, animal welfare (including the nutrition and environmental domains of animal welfare [13]) and antimicrobial/anthelmintic resistance. The training was delivered to the intervention farmers at the start of the intervention, using a collaborative and experiential learning approach. A training material with key messages was co-developed with local research partners, who helped tailor the content to farmers’ needs and local contexts [14]. These partners also facilitated gender-responsive training sessions at accessible community centres and at times convenient for participants. The training served as a critical enabler and integral part of the technical intervention. The animal welfare topics included in the training were chosen to address priority issues identified by key stakeholders during program design and previously reported farmers’ welfare priorities [15]. Pre- and post-training assessment was made based on a random sample of trainees to evaluate the effectiveness of the training in addressing identified knowledge, attitude and practice gaps. The pre-training survey was administered before the training, and the post-training assessment was done immediately after its completion. Thirty-six out of 71 farmers in Doyogena and 10 out of 50 farmers from Angollelana Tara were involved in the training assessments, randomly chosen at certain intervals based on their seating during the training. The assessment survey covered knowledge, attitude and practice domains included in the training and was administered using KoboToolbox [16]. The MS Word version of the survey is provided in Additional file 2.

Design, sampling and data collection

A randomized controlled trial (RCT) study design was used for evaluation of the intervention. Small ruminant farmers from selected kebeles were randomized into treatment and control groups. The list of small ruminant farmers in the kebeles were obtained from the respective kebele administration offices. All small ruminant farmers (or flocks) in the selected kebeles were used for the trial. In Hawora kebele, 144 flocks were randomized either into the treatment or control groups resulting in 72 flocks in each of the treatment and control groups. Similarly, in Checki kebele, 100 small ruminant farmers were randomised into treatment and control groups with 50 flocks in each group. Considering a reduction in health problem (crude morbidity) as main outcome variable, the number of animals in the selected flocks (72 flocks * average flock size of 3 = 216 small ruminants in Doyogena, and 50 flocks * average flock size of 16 = 800 small ruminants in Angollelana Tara) were sufficient to detect a difference of 10% or more prevalence due to intervention with a power of 80% [17]. Randomization was done using excel random function after the list of farmers was recorded in an excel sheet.

Baseline data were collected based on farmers’ recall of flock dynamics (births, offtakes, intakes, deaths), health and welfare, production inputs and outputs, and associated costs and income for one year prior to the beginning of the intervention using Kobo Toolbox [16]. The same data were collected when the one-year (May 1, 2023-April 30, 2024) intervention period was ended both in the treatment and control flocks. For practical reason, it was not possible to blind both farmers and data collectors. Faecal samples were collected from a subsample of randomly selected animals in the study flock immediately before each deworming and two weeks after deworming by local research partners. The target number for subsamples was a minimum of 50 animals, but the final numbers were variable due to farmers’ availability during sampling. Faecal samples from animals in control flocks were collected at the same time as treatment flocks. The faecal samples before and after deworming were examined for helminth (GIT strongyle) status using faecal flotation technique and for faecal egg count using McMaster faecal egg counting method at local research laboratories. The subsampled animals both from treatment and control flocks were also visually inspected during faecal sampling for some general health and welfare indicators including body condition, FAMACHA (Faffa Malan chart) scoring, Daginess, hair coat status (shiny/rough), injury/lesion on different parts of the body, tail status (docked or normal), general signs of ill health (dullness and inappetence) and for specific diseases targeted by the intervention such as respiratory diseases (coughing and nasal discharge), and sheep and goat pox. The body condition, FAMACHA and Daginess were quantitively scored following standard procedures (Additional file 3). Although the flock health and welfare management training included both animal-based and resource-based indicators of animal welfare [18], this study focused only on animal-based indicators, primarily within the health domain, which were evaluated through visual inspections of selected individual animals.

Data analysis

Outcome variables analysed to assess impact of the intervention included crude morbidity and mortality, general health and welfare indictors, prevalence and burden of GIT strongyle parasites, and income from small ruminants as described below.

Crude morbidity and mortality were calculated as the number of sick and dead animals by any cause during the intervention period divided by average number of animals in the flock (average flock size). The average flock size was determined by adding the flock size at the beginning and at the end of the sampling period and then dividing by two.

General health and welfare indicators were calculated as the number of animals affected by a specific condition divided by the total number of animals observed or examined for that condition. For variables which were recorded quantitively such as body condition, FAMACHA and DAG scores, average scores were calculated for all animals examined for these conditions. Prevalence of GIT strongyles was calculated as the number of sampled animals tested positive for one or more strongyle eggs divided by the total number of animals sampled. Worm burden was summarized by mean of eggs per gram of faeces. Income from small ruminant was determined by calculation of the annual gross margin for small ruminant as follows:

The effectiveness of the implemented flock health package was seen as the change in the outcome variables in the treatment flocks compared to the control flocks, as well as before and after treatment in the treatment flocks. For the outcome variables for which reliable before and after treatment data was obtained, difference in difference (DID) analysis was conducted using generalized linear models. In these models, outcome variables were used as dependent variables and the interaction of experimental group (control vs. treatment) and time (before vs. after intervention) were used as an independent variable [19]. A direct regression of the experimental groups on the outcome variable was performed for variables that have only post intervention data. Although the randomization would be expected to balance potential confounders, some recorded potential confounders such as housing and feeding management, years of farming experience and flock composition (sheep versus sheep and goat mixed) were considered as potential confounders in the analyses. The housing and feeding management were uniform among flocks in each site and were left out from the confounding analysis. Years of farming experience and flock composition were included as covariates in the flock level regression models such as in crude morbidity, crude mortality and gross margin analyses to adjust their potential confounding effect. Weighting was used in the analysis for flock/household level variables such as mean annual gross margin, flock size being the weighting factor. The impact of the training on knowledge improvement was evaluated by comparing average farmer knowledge scores pre- and post-training using independent t test. All statistical tests were analysed in R [20].

Results

Morbidity and mortality

The crude morbidity and mortality rates during the intervention period were lower in the treatment groups than the control groups in both study districts. However, regression analyses showed that the evidence for the difference was strong (P < 0.05) only for morbidity in Doyogena district (Table 1). The inclusion of years of farming experience and species composition as potential confounder revealed that none of them have strong association with the outcome variables (the morbidity and mortality rates) in both sites (P > 0.05) and have little impact on the strength of the association between the experimental group and these outcome variables. The crude morbidity and mortality observed in Doyogoena was notably high both in control and treatment groups, with a 6-month morbidity and mortality rates exceeding 20% and 5%, respectively.

Table 1.

Impact of the intervention on crude morbidity and mortality as evaluated by linear regression

	Experimental group	Number of flocks	Average flock size	Average *morbidity rate	Average *mortality rate
Doyogena	control	63	2.82	0.33	0.09
	Treatment	64	2.96	0.21 (p = 0.03)	0.06 (p = 0.66)
Angollelana Tara	control	24	16.31	0.19	0.09
	Treatment	35	17.78	0.13 (p = 0.15)	0.06 (p = 0.31)

*Morbidity and mortality are per 6 months for Doyogena and annual for Angollelana Tara

Flock health and welfare indicators

A subsample of animals was individually observed for some health and welfare indicators during the intervention period. In almost all indicators, the difference in outcome measures, comparing before and after the intervention, between the treatment and control groups was higher after the intervention than before the intervention in the treatment group indicating the positive impacts of the intervention (Table 2).

Table 2.

Impacts of the intervention on some selected health and welfare indicators (analysed by logistic regression for the binary outcome variables and linear regression for quantitative outcome variables)

Indicator variables	Before intervention		After intervention		Before intervention difference (C1-T1)	After intervention difference (C2-T2)	Difference in difference (C1-T1) -(C2-T2)
	Control (C1)	Treatment (T1)	Control (C2)	Treatment (T2)
1Binary variables (% positive)
Doyogena district	N = 42	N = 83	N = 77	N = 115	-	-	-
haircoat (rough)	17	28	41	28	-11(P = 0.15)	13(P = 0.04)	-24(P = 0.02)
Ocular Secretion (yes)	31	29	48	29	2(P = 0.90)	19(P = 0.012)	-17(P = 0.17)
Body lesion	5	3	10	2	2(P = 0.54)	8 (P = 0.02)	-6(P = 0.32)
Head lesion	5	5	9	5	0(P = 0.92)	4(P = 0.37)	-4 (P = 0.57)
Ectoparasite/itching	24	23	38	23	1(P = 0.96)	15(P = 0.02)	-14 (P = 0.20)
Leg injury	14	9	13	10	5(P + 0.39)	3(P = 0.22)	2(P = 0.92)
Respiratory disease	33	38	66	32	-5(P = 64)	34(P < 0.01)	-39(P = 0.02)*
Shoat pox	19	9	24	7	10(P = 0.13)	17(P = 0.00)	-7(0.41)
Other diseases	45	62	45	30	-17(P = 0.05)	15(0.04)	-32(P < 0.01)
Quantitative variables (mean score)
Doyogena district	N = 42	N = 83	N = 77	N = 115	-	-	-
Body condition score	3.05	3.04	3.25	3.25	0.01 (p = 0.94)	0.00(P = 0.56)	0.01 (P = 69)
FMACHA score	2.43	2.84	2.96	2.75	-0.41(P < 0.01)	0.21(P = 0.02)	-0.62(P < 0.01)
DAG score	0.95	1.55	2.24	1.98	-0.60(P < 0.01)	0.26 (P < 0.01)	-0.86(P < 0.01)
Angollelana Tara district	N = 10	N = 41	N = 12	N = 24
Body condition score	2.70	3.07	3.08	3.12	-0.37(P = 0.27)	-0.04(P = 870)	-0.33(P = 0.45)
FMACHA score	2.70	2.24	1.83	2.20	0.46(p = 0.22)	-0.37(p = 13)	0.83(P = 0.07))
(DAG score	0.20	0.60	0.91	0.25	-0.4(P = 0.15)	0.66(P = 0.04)	-1.06(P = 0.10)*

¹these variables were not recorded in Angollelana Tara district

nce and parasite burden measured innce and parasite burden measured in

Helminth prevalence and burden

The impact of the intervention on GIT strongyle parasites was evaluated through a change in prevalence and parasite burden measured in eggs per gram of faces (EPG) (Table 3). In most cases, a higher prevalence and EPG were observed in the control groups than in the treatment groups. The DID analysis further showed that the evidence of the differences was strong in Doyogena district (P < 0.05). But the statistical evidence for the differences in the Angollelana Tara was weak (P > 0.05). Helminths parasite prevalence and EPG were generally low in Angollelana Tara even before deworming. In both districts, parasite prevalence was almost always greater than 50% but worm burden was mostly less than 500 EPG.

Table 3.

Pre- and post-deworming parasitic prevalence and strongyle worm load

GIT Parasitic status	Pre-deworming		Post-deworming		Pre-deworming difference (C1-T1)	Post- deworming difference (C2-T2)	Difference in difference (C1-T1) -(C2-T2)
	Control (C1)	Treatment (T1)	Control (C2)	Treatment (T2)
Doyogena district
First round	N = 42	N = 77	N = 42	N = 77	-	-	-
Prevalence (%)	74	64	98	37	10	61	-51(P < 0.01)
Average worm burden	468	500	824	290	-32	534	-566(P < 0.01)
Second round	N = 51	N = 47	N = 51	N = 47	-	-	-
Prevalence (%)	90	83	96	51	7	45	-38 (P = 0.01)
Average worm burden	774	387	1620	200	387	1420	-1033 (P < 0.01)
Angollelana Tara
First round	N = 9	N = 35	N = 9	N = 9
Prevalence (%)	89	78	89	36	0	42	-42 (P = 0.28)
Average worm burden	100	0	483	125	-383	-125	-258(P = 0.56)
Second round	N = 9	N = 19	N = 12	N = 29
Prevalence (%)	78	37	80	37	-2	0	-2 (0.90)
Average worm burden	483	125	362	158	121	-33	154(P = 0.82)

¹ EPG eggs per gram of faeces

Impact of flock health intervention on income from small ruminants

The primary benefit of the small ruminant health intervention is increased income for farmers, alongside improved animal welfare. Enterprise gross margin analysis was used to evaluate the impact of the intervention on farmers’ income from small ruminants. In both districts, the annual gross margin was higher in treatment groups than control groups and the difference was statistically significant (P < 0.01) in Doyogena (Table 4). However, in absolute terms, the gross margin from Doyogena district was negative both in the control and treatment farms. The inclusion of farming experience and species composition of flocks as potential confounders did not significantly influence the association between outcome variable (gross margin) and experimental group.

Table 4.

The impact of the intervention on annual gross margin (in USD*) from small ruminant flocks as evaluated by linear regression

	Experimental group	Number of flocks	Flock change in value	Offtake value	Revenue	Variable costs	Mean annual gross margina
Doyogena	control	53	-57.02	13.41	-43.60	53.33	-96.95
	treatment	62	-33.88	60.07	26.19	31.20	-4.99 (P = 0.001)
Angollelana Tara	control	18	554.78	211.73	766.52	220.05	546.46
	treatment	27	600.24	358.37	958.61	202.93	746.33 (p = 0.26)

* All incomes and costs were originally collected in Ethiopia Birr and converted to USD based on the 2023 annual average exchanged rate 57.8 USD to Ethiopian Birr. a. The mean annual gross margins is flock size weighted mean

Knowledge of small ruminant farmers about flock health and welfare

One of the components of the intervention was farmers’ training on flock health and welfare management. Although the training is expected to enhance overall flock performance in the long term, its immediate effect is increased farmer knowledge of animal health and welfare. This immediate effect of the training was assessed by evaluating the knowledge of a group of intervention farmers before and after training. Farmers had good knowledge even prior to the training as indicated by the average knowledge score of 80% (N = 10) in Angollelana Tara district and 54% (N = 36) in Doyogena district. After the training, baseline knowledge scores were improved to 90% and 80%, respectively, the improvements being statistically significant (P < 0.05) in both districts.

Discussion

This study adds to the evidence base on small ruminant health and welfare management by demonstrating measurable improvements in practices and outcomes. The preventive small ruminant flock health and welfare intervention which included strategic deworming and vaccination, and farmers training on flock health and welfare management generally made considerable improvements in animal health and welfare. This improvement also led to an increase in farmers’ income. This was evidenced by higher performance of treatment group flocks as compared with their control counterparts, which used existing animal health practices and services. Similar interventions were implemented in a previous study, yielding comparable reductions in morbidity, and parasite prevalence and burden [7, 8]. This study confirmed the effectiveness of the intervention with randomized control trial design which produces stronger evidence. Moreover, it showed the cost-effectiveness of the intervention, going beyond the previously documented technical effectiveness. The elements of the tested interventions are technically feasible and easily accessible and affordable for smallholder farmers. If a proper community awareness and training is provided to increase the uptake of this innovation, there is a big opportunity to improve the livelihood of farmers and the welfare of their animals. livelihood and reflects the intervention’s viability as a self-sustaining

However not all variables in both districts tested showed the expected statistically significant improvement which can be due to different reasons. In terms of crude morbidity and mortality, statistically significant improvement was observed only for morbidity in Doyogena district. Although training on flock health and welfare can improve the general flock health, technological interventions such as deworming and vaccination targeted a few selected problems, which were thought to be important, and hence the impact on crude morbidity and mortality could be variable, depending on burden of targeted diseases and relative burden of other diseases not affected by these measures. The statistical significance was also reduced by the small sample size used in the analysis. There was a significant difference in average flock size between the two study districts. Although this may not influence the overall outcome as the analysis was conducted at the flock level, the larger flock size in Angollelana Tara could impact the quality of information gathered from farmers. Specifically, recalling specific health events, such as morbidity, may be more challenging in larger flocks. The observed crude mortality in both districts in both the control and treatment groups was relatively lower than the nationally reported small ruminant mortality of 34% [3]. But still, they are high especially in Doyogena. At this site, flocks were tethered separately during cultivation season, which should help reduce disease issues compared to the Angollelana Tara site, where animals continuously mix in a free grazing area throughout the year. However, this trend was not reflected in the data, possibly because the effect of this management difference was masked by other factors. Despite significant improvement, crude morbidity and mortality rates are still high in Doyogena, with annual morbidity and mortality of above 40% and 10%, respectively in the treatment flocks, justifying greater efforts to reduce disease burden.

One of the major elements of the intervention was strategic deworming based on the epidemiological situation of each district. The pre-deworming prevalence of GIT strongyles was very high almost always higher than 50% and reaching up to 90%. This is in line with the high helminth prevalence observed in small ruminants in Ethiopia with national pooled prevalence of 75% [21] and mean prevalence of 54% reported around the same study sites [8]. This indicates that helminths pose significant challenges to small ruminant production systems in Ethiopia, warranting proper attention. The high pre-deworming prevalences were decreased by the intervention. This was supported by improvements in FAMACHA and DAG scores which are clinical indicators of parasitism. But the body condition score did not show significant improvement as it is impacted not only by health but also nutrition. The statistical evidence for the decrease in parasitic prevalence and burden in Angollelana Tara was weak which could be due to the small sample that only had statistical power to detect large differences. Despite high GIT strongyle prevalence, worm burden measured in EPG even pre-deworming was mostly within a range of mild infestation (< 500EPG) [22], and it was further reduced by the intervention.

In addition to evaluation of the technical effectiveness of the intervention, its cost effectiveness in terms of profitability for farmers was also evaluated. The mean annual gross margin from small ruminants was improved in the treatment groups as compared with the control groups. Similarly, veterinary services in community-based breeding program sites in Ethiopia have been reported to enhance the offtake rate of small ruminants and increase household income [23]. However, the evidence of improvement was statistically weak in Angollelana Tara which may again be an issue of lack of statistical power. The evidence related to positive financial impact is very important to demonstrate economic benefits of the intervention on farmers' livelihood and reflects the intervention’s viability as a self-sustaining, user-pays measure.

Improvements in welfare of small ruminants in this study were mostly in the domain of health. However, the farmer training covered other domains of animal welfare, specifically nutrition and environment. The mean body condition for the treatment groups did not change. Changes in body condition can take a long time to manifest affected by several factors. Feed investments can be more expensive and seasonal. Reproductive and health factors can also influence body condition [24].

Animal-based indicators related to the environmental domain of animal welfare did not change because of the intervention. Injury lesions, which can be an indicator of health, environmental and behavioral issues depending on context, were stable in the treatment groups. However, lesions on the head and body increased in the control groups after the intervention compared to before. Leg injuries, that are often the result of extended tethering, remained the same before and after intervention. Tethering can be behaviorally restrictive but often requires changes in housing/environment to be reduced. While improving sheds and housing for small ruminants was covered in the training, specific focus on reducing time spent tethered or improving tethering techniques were not included as a feature in the intervention.

The flock health training contributed to improved knowledge of farmers around flock health and welfare management which led to improved practices. The training supported the contextualization of technical information to local realities, addressed practical challenges faced by farmers, and reinforced critical messages through participatory and gender-sensitive approaches. This helped create confidence and motivation among farmers to apply the intervention package to improve flock health and welfare. Moreover, the mentoring and follow-up support integrated into the training design ensured that knowledge was translated into action, enhancing the likelihood of sustained behavioural change at the household and community levels.

A key challenge in this controlled trial was the failure of some trial flocks to participate in scheduled intervention activities. Some participants were completely lost to follow up during the study period, particularly in Angollelana Tara district, resulting in smaller sample size and statistical power. This is the main limitation of the study. However, the lost follow up was both from treatment and control groups and is not thought to bias the study, only reduce its power. Future studies should ensure early engagement of farmers, regular follow-up by local partners, and context-adapted scheduling to minimize loss to follow-up and maintain statistical power.

Conclusions

This study demonstrated that organizing and strategizing animal health technologies that are already in use by farmers along with providing practical training on flock health and welfare management can effectively improve animal health and welfare, and income from small ruminants. The flock health package evaluated in this study can be expanded to other regions of similar livestock production systems. Nevertheless, epidemiological situation of diseases can vary within a small geographic scope, and future interventions should be adapted to fit each unique context.

Authors’ contributions

WTJ conceived and designed the study, analyzed the data, and drafted the manuscript. ML conceived and designed the study and drafted the manuscript. TF implemented the study, collected the data, and performed the laboratory analysis. EY implemented the study, collected the data and performed the laboratory analysis. RD designed the study and drafted the manuscript. MD conceived and designed the study. TKJ conceived and designed the study. All authors reviewed the manuscript.

Funding

The study was conducted as part of the CGIAR Research Program on Livestock, a global partnership dedicated to achieving a food-secure future (www.cgiar.org). The preparation of this manuscript was supported by CGIAR’s Sustainable Animal Productivity for Livelihoods, Nutrition, and Gender Inclusion (SAPLING) initiative, and continued under the Sustainable Animal and Aquatic Foods Program (SAAF), and the contributions of donors and organizations supporting the Research Programs through the CGIAR Trust Fund.

Data availability

The data supporting the conclusion of this article are included within the article (and its additional file(s)).

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the International Livestock Research Institute’s Institutional Research Review Committee (ILRI-IREC2019-42/1). In addition, study participants have given informed consent for their animals to be used in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.