Microscopy detected 48 infected children: 34, or 70.8% of the positive cases, were from Chongwe, while 14, or 29.2%, were from Kafue. The distribution of microscopy-positive cases differed significantly between districts (χ² = 4.84; P < 0.05). Quantitative PCR showed the same pattern, with reported prevalence of 73.9% in Chongwe and 28.9% in Kafue. The qPCR difference was also significant (χ² = 13.52; P < 0.01). Together, the two laboratory approaches identified a greater burden of detected urogenital schistosomiasis in Chongwe among the children studied.

No significant association was observed between sex and infection status. Infection was, however, significantly associated with age across the study population. The investigators did not report an association between sex, whereas the association with age indicates that infection status was not evenly distributed across the children’s age groups in this sample.

The findings document ongoing transmission of urogenital schistosomiasis in the surveyed peri-urban districts, but the study’s design and geographic scope limit what can be inferred. A cross-sectional assessment provides a snapshot at one point rather than measuring how prevalence changed over time. It also does not establish why Chongwe had higher detected prevalence than Kafue, and the results cannot be assumed to represent all school-age children in Lusaka or Zambia. The researchers said the updated epidemiological information could help identify where surveillance and control efforts need strengthening and support planning and implementation within the World Health Organization’s neglected tropical disease agenda.

Subject of Research: Prevalence of urogenital schistosomiasis among school-age children in two peri-urban districts of Lusaka, Zambia

Article Title: Schistosoma haematobium infection among school age children in two peri-urban areas of Lusaka, Zambia: A cross-sectional study

Article References: Mutengo, M. M., Mwansa, J., Chisanga, K., Zulu, E., Mulunda, N. R., Muchinga, J., Rodríguez, E., Sánchez, S., Castro, L., Perteguer, M. J., Carmena, D., & Sotillo, J. (2026). Schistosoma haematobium infection among school age children in two peri-urban areas of Lusaka, Zambia: A cross-sectional study. PLOS Neglected Tropical Diseases, 20(10), e0014013. https://doi.org/10.1371/journal.pntd.0014013

Image Credits: AI Generated

DOI: 10.1371/journal.pntd.0014013

Keywords: schistosomiasis, Schistosoma haematobium, school-age children, Zambia, Lusaka, Chongwe, Kafue, cross-sectional study, qPCR, urine microscopy, parasite surveillance, Schistosoma

The study was a secondary analysis of data collected for the Africa Health Collaborative in Lusaka, Kafue, Kitwe and Ndola. It combined a digital questionnaire completed by 859 people with 64 key-informant interviews and five focus group discussions involving educators, employers, regulators, professional associations and health workers. Participants came from nine tertiary training institutions and four health facilities offering internships or early-career placements. Programmes represented medicine, nursing, public health, environmental health, pharmacy, clinical medicine, physiotherapy, midwifery and related primary health care disciplines.

Employment readiness was operationalised narrowly but transparently: respondents had to report that they felt prepared for the labour market and say that they had prepared a curriculum vitae. On that basis, 357 participants were classified as employment-ready and 502 were not. Readiness was reported by 27% of respondents aged 18–24, 53% of those aged 25–29, 75% of those aged 30–34 and 66% of respondents aged 35 or older. The descriptive pattern persisted in multivariable logistic regression, with adjusted odds ratios of 2.05 for ages 25–29, 4.31 for ages 30–34 and 3.28 for ages 35 or older, each compared with the youngest group.

Education and career status were also strongly associated with readiness. Respondents with undergraduate education or higher had nearly three times the odds of readiness compared with diploma holders, with an adjusted odds ratio of 2.91 and a 95% confidence interval of 1.89–4.48. Volunteers and early-career professionals had 17.55 times the odds of readiness compared with students, with a 95% confidence interval of 8.56–36.00, while interns had 2.97 times the odds, with a 95% confidence interval of 1.39–6.38. These are associations from a cross-sectional model and do not establish that internships or volunteering caused later employment or objectively improved competence.

District differences were observed, although the analysis cannot explain their causes. Compared with Kitwe, participants in Kafue had 3.22 times the adjusted odds of reporting readiness and those in Lusaka had 2.49 times the odds. The estimate for Ndola was also elevated, at 2.11, but did not meet the study’s threshold for statistical significance. The model reported a pseudo-R-squared of 0.2019, and a Hosmer–Lemeshow test did not indicate poor calibration, with a p-value of 0.9191. The survey also found that blended-learning participants were more likely to report readiness than those receiving exclusively in-person instruction, although only three respondents attended purely online programmes, making that category too small for a robust comparison.

Competency results exposed a gap between what participants knew and how confident they felt about applying their skills. Most respondents described themselves as very knowledgeable about their general training area, but their mean confidence across six technical domains was only 2.21 on a four-point scale. Communication scored highest at 2.42, followed by critical thinking at 2.26. Networking was lowest at 2.08, while technical skills scored 2.15 and leadership 2.19. The six-item scale had acceptable internal consistency, with Cronbach’s alpha of 0.78. These measures were self-reported, but interviews helped identify concerns that some training placed too much emphasis on theory and provided too little supervised practice with real patients.

Interviews and focus groups pointed to uneven training quality, weak regulation and insufficient clinical exposure as important barriers. Participants described concerns about training institutions prioritising enrolment and graduation over practical competence, alongside inadequate infrastructure and variable supervision. Some trainees reported short or poorly organised attachments and differences between classroom teaching and real workplace settings. Respondents also described anxiety about performing procedures, handling patients and entering an unfamiliar professional environment. These findings concern the sampled institutions and districts and cannot be assumed to describe every Zambian training programme or placement.

Employment was only one stage of the pathway examined: retention depended heavily on working conditions. Key informants cited staff shortages, heavy workloads, dilapidated facilities, limited equipment, low remuneration and a lack of incentives for difficult locations. Contract instability, including delays in absorbing volunteers and uncertainty over continued funding, added to the problem. Limited study leave, sponsorship and career progression restricted further education, while inadequate recognition and mentorship reduced morale. Participants described poor conditions and growth prospects as contributing to migration from public services; non-governmental organisations were viewed as offering more attractive support, including education opportunities and stronger work-life arrangements.

The study also found interest in employment beyond conventional clinical posts. Respondents most often preferred hospital facilities, selected by 80.7%, followed by non-governmental organisations at 66.5%, community settings at 40.5% and overseas work at 30.5%. They also identified opportunities in health education, community awareness, occupational health and safety, environmental management, entrepreneurship, private clinics, policy and sustainable development. The authors argue that primary health care should therefore be viewed as an employment ecosystem rather than a narrow set of clinical occupations, while stressing that career guidance, regulation, training quality and working conditions remain necessary for that ecosystem to support both services and workers.

Several limitations constrain what can be concluded. The cross-sectional design captures perceptions at one point and cannot establish causal relationships or track changes in readiness, job acquisition or retention over time. Employment readiness depends on self-reported confidence and possession of a curriculum vitae rather than standardized competency tests, employer evaluations or confirmed employment. Social-desirability and response biases may therefore influence the findings. The four districts were purposively selected because they host relevant training and employment institutions, the survey was concentrated in urban areas and the sample cannot fully represent remote rural health systems. The evidence supports closer attention to regulated training, supervised internships, mentorship, career progression and employment conditions, but longitudinal studies would be needed to show which pathways actually produce sustained jobs and better health services.

Subject of Research: Determinants of youth employment readiness in Zambia’s primary health care workforce

Article Title: Youth employment in the primary health care ecosystem in Zambia: barriers and opportunities

Article References: Mwila, G., Lakati, A., Sakanga, V., Bulanda, D., Mkandawire, L., Omogi, J., Banda, L., Gatumia, B., Abdi, O., Agukoh, A., Matiang’i, M., & Masibo, P. (2026). Youth employment in the primary health care ecosystem in Zambia: barriers and opportunities. Frontiers in Health Services, 6, Article 1869977. https://doi.org/10.3389/frhs.2026.1869977

Image Credits: AI Generated

DOI: 10.3389/frhs.2026.1869977

Keywords: Zambia, youth employment, primary health care, health workforce, employment readiness, internships, training quality, workforce retention, Youth, employment, primary, health

The study combined nationally representative Demographic and Health Survey data from Zambia 2024, Mozambique 2022–2023 and Nepal 2022. Researchers included women with complete dietary, mental health and covariate information: 13,951 participants from Zambia, 13,183 from Mozambique and 7,442 from Nepal. The weighted prevalence of depressive symptoms was 5.7%, while 6.4% of participants had anxiety symptoms and 3.9% met thresholds for both. The analysis accounted for the surveys’ sampling weights, clusters and strata, allowing the pooled estimates to represent the target populations rather than simply treating all respondents as equally representative.

Dietary diversity was measured with a 24-hour recall covering the previous day and night. Women received a score from zero to 10 according to whether they consumed foods from ten groups, including grains and roots, pulses, nuts and seeds, dairy, flesh foods, eggs, dark green leafy vegetables, vitamin A-rich fruits and vegetables, other vegetables and other fruits. The study classified consumption of fewer than five groups as inadequate diversity and five or more as adequate, following the minimum dietary diversity for women, or MDD-W, indicator. Although higher scores generally suggest a greater likelihood of meeting micronutrient needs, the measure is not a complete assessment of diet quality, nutrient quantities or habitual eating patterns.

Depressive symptoms were assessed with the nine-item Patient Health Questionnaire-9, with a score of at least 10 defining symptoms, while the seven-item Generalized Anxiety Disorder-7 scale used the same threshold for anxiety symptoms. Participants scoring at least 10 on both scales were classified as having co-occurring depressive and anxiety symptoms. The scales assess symptoms over the preceding two weeks; because the surveys did not include clinical diagnostic interviews, the researchers could not independently assess criterion validity against clinical diagnoses. Country-specific weighted Cronbach’s alpha coefficients were 0.81, 0.92 and 0.86 for depression in Zambia, Mozambique and Nepal, respectively, and 0.79, 0.92 and 0.86 for anxiety.

The inverse associations persisted as the models incorporated progressively broader sets of possible confounders. Adjustment for country alone produced odds ratios of 0.62 for depressive symptoms, 0.59 for anxiety symptoms and 0.52 for co-occurring symptoms. Models accounting for age, marital status, urban or rural residence, education, media exposure, household wealth and employment yielded odds ratios of 0.70, 0.64 and 0.59, respectively. The fully adjusted estimates were 0.69 with a 95% confidence interval of 0.58–0.82 for depressive symptoms, 0.63 with an interval of 0.54–0.73 for anxiety symptoms and 0.58 with an interval of 0.47–0.72 for co-occurring symptoms. All three confidence intervals excluded 1, indicating statistically significant associations in the pooled models.

Food-group models offered a more complicated picture than a simple message that every category of produce behaves in the same way. After correction for multiple comparisons, consumption of other vegetables, pulses, eggs, flesh foods, other fruits and dairy was associated with lower odds of all three mental health outcomes. For other vegetables, the adjusted odds ratios were 0.55 for depressive symptoms, 0.65 for anxiety symptoms and 0.47 for co-occurring symptoms. Pulses had an odds ratio of 0.62 for anxiety symptoms, while the corresponding estimate for other vegetables was 0.65. Vitamin A-rich fruits and vegetables were associated with lower odds of anxiety and co-occurring symptoms but not depression after false-discovery-rate correction. Dark green leafy vegetables were the exception, with higher odds of depressive symptoms, anxiety symptoms and co-occurring symptoms, yielding odds ratios of 1.42, 1.28 and 1.67, respectively.

That exception should not be interpreted as evidence that leafy greens cause mental distress. The food-group models were separate, mutually unadjusted analyses, so each estimate describes consumption of one category in the context of the broader diet and participants’ characteristics rather than the effect of adding or removing a specific food. A single 24-hour recall may also not capture usual intake, day-to-day or seasonal variation, or portion size. The observed leafy-green association could reflect differences in the foods eaten alongside them, socioeconomic patterns or other factors not captured by the models, while the large number of comparisons also creates a continuing risk of chance findings despite statistical correction.

Subgroup analyses suggested that the overall relationship might not be identical across every setting. The inverse association was generally present across age, education and child-death categories, but statistical interactions were detected for residence in relation to depressive and co-occurring symptoms, wealth status in relation to depression, and marital status in relation to co-occurring symptoms. For example, the adjusted odds ratio for depressive symptoms was 0.63 among urban women and 0.82 among rural women, although the confidence interval for the rural estimate included 1. Among married women, adequate diversity was associated with lower odds of co-occurring symptoms, with an odds ratio of 0.49, compared with 0.79 among single women. These comparisons were exploratory and should not be treated as evidence that the relationship differs causally between groups.

Sensitivity analyses broadly supported the main result but also exposed its complexity. Separate models for each country generally yielded inverse associations, although their magnitudes and statistical significance varied. Modeling dietary diversity as a continuous score with natural cubic splines found evidence of nonlinearity for all three outcomes, each with a P value for nonlinearity below 0.001. This means the relationship cannot be summarized confidently as the same change in odds for every additional food group. Results were also similar after excluding the 2,181 participants who were currently pregnant, an important check because the MDD-W indicator was not developed specifically to assess micronutrient adequacy during pregnancy.

Several limitations constrain what can be concluded. The cross-sectional design measures diet and symptoms around the same period, so it cannot determine whether dietary patterns preceded the symptoms, whether poor mental health influenced eating, or whether both reflected shared conditions. Household food insecurity was not measured, and potentially relevant factors such as trauma, menopausal status and attitudes toward help-seeking were unavailable, leaving possible residual confounding. Self-report can introduce recall and reporting bias, and one day of dietary recall may not represent usual intake, seasonal variation or portion size. The findings therefore demonstrate adjusted associations rather than a dietary intervention effect.

Researchers proposed several plausible but unproven explanations for the pattern. A more diverse diet may provide a broader range of proteins, B vitamins, minerals, antioxidants and essential fatty acids involved in neurotransmitter synthesis and synaptic signaling. Greater food-group diversity may also be linked to differences in gut microbial composition and gut–brain signaling, while serving as a marker of food security and living conditions that themselves influence psychological well-being. None of these mechanisms was directly tested, and the study did not measure food insecurity or establish any biological pathway. Longitudinal studies that track diet before symptoms emerge, together with randomized intervention trials, will be needed to determine whether changing dietary diversity can affect women’s mental health. For now, the results identify a measurable dietary pattern associated with fewer depressive and anxiety symptoms, rather than a proven preventive strategy.

Subject of Research: Cross-sectional associations between dietary diversity and depressive and anxiety symptoms among reproductive-age women in Zambia, Mozambique and Nepal

Article Title: Associations of dietary diversity with depressive symptoms, anxiety symptoms and their co-occurrence among reproductive-age women in three low- and middle-income countries: a cross-sectional analysis

Article References: Fang, J., Huang, L., & Fang, N. (2026). Associations of dietary diversity with depressive symptoms, anxiety symptoms and their co-occurrence among reproductive-age women in three low- and middle-income countries: a cross-sectional analysis. Frontiers in Public Health, 14, Article 1955678. https://doi.org/10.3389/fpubh.2026.1955678

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1955678

Keywords: dietary diversity, depressive symptoms, anxiety symptoms, co-occurring symptoms, women’s health, nutrition, cross-sectional study, Zambia, Mozambique, Nepal, Associations, dietary

The quantitative study examined how accurately petrol prices could be estimated from variables embedded in Zambia’s cost-plus framework. The Energy Regulation Board’s price build-up includes international petroleum costs, the kwacha-dollar exchange rate, excise duty, value added tax, transport costs, margins and regulatory fees. The dataset combined daily exchange-rate and international oil price observations with lower-frequency petrol price, tax and inflation data, producing approximately 5,387 daily-indexed observations.

Four algorithms were trained and evaluated on the same engineered feature set: five economic variables, calendar variables, petrol-price lags of one, seven and 30 days, and rolling means and standard deviations over three-, seven- and 14-day windows. The researchers used a chronological 70% training, 15% validation and 15% hold-out test split, with preprocessing fitted only on training data and temporal cross-validation used during development.

Ridge regression was substantially more reliable on the hold-out data than the competing nonlinear models. Its test R² of 0.9814 compared with 0.6693 for Support Vector Regression, −26.1367 for Random Forest and −1.5383 for LSTM. The nonlinear models fitted the training data closely but generalised poorly when test-period conditions differed from much of the historical sample, particularly after petrol prices increased structurally from 2021.

One-step-ahead Diebold–Mariano comparisons found significant differences between each machine-learning specification and a naive benchmark that carried the previous observed petrol price forward. The authors cautioned that these results measure conditional prediction accuracy, because contemporaneous exchange rates, oil prices, taxes and inflation were included. They are therefore not evidence that the model could predict future prices before those inputs were known.

Residual diagnostics showed that the Ridge specification did not capture all temporal dependence. Its Durbin–Watson statistic was 1.5355, and Ljung–Box tests rejected the null hypothesis of no autocorrelation at lags 1 and 10. Most residuals were small and centred around zero, while the largest deviations occurred near discrete regulatory price adjustments. Sensitivity analyses nevertheless found Ridge retained test R² values of about 0.98 in lagged-only specifications and when excise duty and VAT were excluded.

Explainable artificial intelligence results distinguished between structural and prediction-level influence. Coefficient-based analysis ranked the international oil price as the strongest predictor, while SHAP analysis identified the ZMW/US$ exchange rate as the largest contributor to individual predictions, followed by international oil price. Partial dependence plots showed positive, approximately linear relationships for those two market variables. Inflation had a weaker influence, while excise duty and VAT contributed limited day-to-day prediction variation after other features were accounted for; the study did not interpret that result as evidence that tax changes were unimportant to regulated pump prices.

Performance deteriorated substantially when the data were aggregated to monthly frequency, with Ridge test R² falling to −0.1816 in the monthly conditional specification and −0.1463 in the lagged-only version. The authors also noted that repeated values between policy updates increase the nominal observation count without adding equivalent new information. These limits, together with the use of contemporaneous inputs and a single-country dataset, constrain generalisation beyond Zambia’s regulated pricing conditions.

Subject of Research: Comparative machine-learning prediction and explainability analysis of petrol prices in Zambia’s regulated cost-plus pricing system

Article Title: Predicting petrol prices in Zambia using machine learning and explainable artificial intelligence under a regulated cost-plus pricing framework

Article References: Maliti, M., & Zimba, A. (2026). Predicting petrol prices in Zambia using machine learning and explainable artificial intelligence under a regulated cost-plus pricing framework. Discover Artificial Intelligence, 6(1), Article 1273. https://doi.org/10.1007/s44163-026-02356-z

Image Credits: AI Generated

DOI: 10.1007/s44163-026-02356-z

Keywords: Zambia, petrol prices, Ridge regression, machine learning, explainable AI, fuel regulation, cost-plus pricing, exchange rates, international oil prices, Predicting, petrol, prices

The Aphanisticus genus contains about 360 species distributed across warm-temperate and tropical parts of the Eastern Hemisphere, with 147 species previously known from the Afrotropical Region. Most are small, compact beetles with convex, nearly cylindrical bodies, but the genus also includes broader, oval or triangular forms. A few species groups possess conspicuously flattened elytra, the hardened forewings that cover and protect the flying wings. The new South African species, A. fencli, belongs to a newly proposed group that also includes A. gebhardti, A. lineolatus and A. planatus.

Researchers established the new A. gebhardti species group around a combination of characters unusual within Aphanisticus. Members have relatively large, elongated and flattened bodies, with the length typically exceeding the width by more than three times. Their heads are small and narrower than the front edge of the pronotum, while the frontovertex forms a shallow depression. The pronotum is moderately transverse and somewhat shield-shaped, and the elytra carry recognizable longitudinal striations. The group is currently known only from the Republic of South Africa, where all four recognized species have been recorded.

A. fencli is the most immediately conspicuous member of the group because its dorsal surface is bicolored: the head and pronotum are black with an indistinct bluish sheen, while the elytra are blue. Its body is strongly flattened and measures 3.95 to 5.95 millimeters long, with the holotype 4.5 millimeters long and 1.4 millimeters wide. The elytra are roughly 2.2 to 2.4 times as long as they are wide, and each bears a raised sutural ridge and another elevated interval. Fine longitudinal striae cover parts of the surface, becoming interrupted or tangled into short irregular marks farther from the midline.

The identification key developed in the study separates A. fencli from the three established members of its group by its blue elytra, black pronotum, rounded oculofrontal borders and less elongated body. A. lineolatus has a more regular pattern of elytral striae, deeper and smaller pronotal punctures and a body length of about 5.2 to 5.4 millimeters. The larger A. gebhardti and A. planatus are distinguished from one another by their proportions, coloration and the degree of projection along the inner margins of the eyes. The new species is known only from its type locality in Bulwer, KwaZulu-Natal, where the series was collected on November 11, 2001.

The study also clarified a long-standing naming problem within the group. A lectotype was designated for A. gebhardti from a specimen collected in Cape Town, and another lectotype was selected for A. lineolatus. Comparisons showed that A. semicoeruleus, described by Théry in 1930, and A. lineolatus, described by Obenberger in 1928, represent the same species. The name A. lineolatus is therefore retained, and A. semicoeruleus is recorded as its junior synonym. The researchers concluded that the specimen used by Théry had probably been drawn from Obenberger’s original type series.

A. franki represents a different form within the genus, with a convex, subcylindrical body rather than the flattened profile characteristic of the A. gebhardti group. Described from the West Okawango South Delta in Botswana, it measures 4.7 to 5.5 millimeters long and has an olivaceous-green surface with occasional, indistinct bronzy reflections. Its head is broad, the eyes are large and convex, and the pronotum is nearly regularly rounded, with two shallow transverse elevations across its disc. Each elytron ends in a distinct lateral tooth, one of the traits used to distinguish it from the superficially similar A. mitratus of Madagascar.

A. snizeki, the third new species, is known from a single female specimen collected near Kapiri Mposhi in Zambia’s Central Province. It is 5.05 millimeters long and 1.4 millimeters wide, with a bluish-green back carrying an indistinct violet reflection and a black underside with a faint bronzy sheen. Its body is convex and subcylindrical, and the pronotum is broadly heart-shaped. The elytra are about 2.55 times as long as they are wide, are depressed along the raised suture toward the rear and terminate in sharp triangular teeth. Unlike A. franki, A. snizeki has a deeply concave frontovertex and more sharply angled posterior corners on the pronotum.

The authors placed A. snizeki in comparison with A. coeruleipennis, another Southern African species with a metallic appearance. A. coeruleipennis is smaller and more robust, with a body only about 3.15 millimeters long, and it has a dark reddish-violet pronotum paired with blue-violet elytra. The identification traits of A. snizeki therefore depend on the combination of its larger, more convex body, bluish-green coloration, more pronounced surface reticulation and distinct elytral teeth. Because the new species is known from a female holotype, its male anatomy remains unknown, limiting direct comparison with the illustrated aedeagi of the newly described A. fencli and A. franki.

All three species remain known only from their type localities: Bulwer in South Africa, the West Okawango South Delta in Botswana and the vicinity of Kapiri Mposhi in Zambia. The research relied mainly on extensive material collected in recent years by Czech entomologists, supplemented by specimens held in major natural-history and research collections across Europe, Asia and South Africa. Microscopic examination, ocular measurements and focus-stacked macro photography were used to assess the fine punctures, ridges, grooves and surface textures that distinguish closely related species. The resulting additions show that even a comparatively well-studied jewel-beetle genus retains hidden diversity, particularly in Africa’s undercollected landscapes.

Subject of Research: Taxonomic discovery of three new Aphanisticus jewel-beetle species in Southern Africa

Article Title: Three new species of buprestid genus Aphanisticus Latreille, 1810 from Southern African countries ( Coleoptera , Buprestidae ), with nomenclatural and taxonomic notes

Article References: Kalashian, M. Y., & Kubáň, V. (2026). Three new species of buprestid genus Aphanisticus Latreille, 1810 from Southern African countries (Coleoptera, Buprestidae), with nomenclatural and taxonomic notes. ZooKeys, 1287, 201-211. https://doi.org/10.3897/zookeys.1287.200030

Image Credits: AI Generated

DOI: 10.3897/zookeys.1287.200030

Keywords: Aphanisticus, jewel beetles, Buprestidae, new species, Southern Africa, taxonomy, entomology, species discovery, Botswana, Zambia, South Africa, Three

The logic behind differentiated service delivery is straightforward but powerful. Traditional antiretroviral therapy programs require patients to visit a health facility frequently, often monthly, for refills and check-ups. For stable clients, this is inefficient for the health system and burdensome for the patient, particularly in rural areas where a single clinic visit can consume an entire day and significant transport money. DSD models reorganize care around the needs of distinct patient groups, allowing clinically stable clients to receive their medication through less intensive channels, such as community distribution points, while reserving intensive facility-based care for those who need closer monitoring. The approach was established with the explicit aim of improving patient outcomes while simultaneously increasing the efficiency of the health system itself.

To evaluate how these models are actually working on the ground, the research team adopted a case study design. They conducted in-depth interviews with individuals receiving antiretroviral therapy through a DSD model, and key informant interviews with program implementers across multiple administrative levels. All interviews were guided by semi-structured interview guides, audio recorded, and transcribed verbatim. Participants were selected through purposive sampling, a technique that deliberately identifies information-rich cases—in this case, people with direct, varied experience of the models. The team then applied thematic analysis, organizing the findings according to the five dimensions of the RE-AIM framework: Reach, Effectiveness, Adoption, Implementation, and Maintenance. This structure allowed the researchers to assess not just whether the models work, but who they serve, how they are adopted, and whether they can endure.

On the question of reach, the study delivered a sobering but important finding: the overall reach of the DSD models is shaped significantly by cost and distance to the health facility. In other words, the very barriers that differentiated care is designed to reduce can also determine who manages to access it in the first place. For patients living far from distribution points or unable to afford transport, the theoretical benefits of a more flexible model may remain out of reach. This finding underscores a persistent tension in implementation science: an intervention that looks equitable on paper can reproduce existing inequalities in access unless geography and economics are explicitly engineered into its design.

Effectiveness, by contrast, emerged as a clear strength. The researchers found that the DSD models improved adherence to antiretroviral therapy, a critical outcome given that consistent medication-taking suppresses viral load and prevents both disease progression and onward transmission. The models also produced an unexpected systems benefit: they allowed health facility staff to plan client visits systematically. Rather than contending with unpredictable daily surges of patients, facilities could anticipate demand, allocate staff more rationally, and devote scarce clinician time to complex cases. In this sense, differentiated service delivery functions as a kind of demand-management technology for the health system, redistributing routine tasks so that human resources can be deployed where they matter most.

Adoption—the willingness of facilities and communities to take up the models—was strengthened by deliberate local adaptations. The study documented modifications that facilitated adoption, including the use of information, education, and communication materials to explain the models to clients and communities, and the active participation of traditional and community leaders. These local champions play a crucial role in contexts where trust in formal health institutions may be uneven and where stigma around HIV remains a barrier to engagement. By involving leaders who command community credibility, program implementers helped legitimize the new care pathways and encourage uptake among people who might otherwise have stayed away from clinic-based services.

The implementation dimension of the framework asked whether the models were delivered as intended, and here the findings were encouraging. The DSD models were successfully implemented as planned, supported by regular monitoring and ongoing training. This fidelity between design and execution is notable in large public health programs, where drift between policy and practice is common. Yet the study also surfaced a forward-looking concern: implementers recommended capacity building and expansion of human resources to maintain the models over time. Without a pipeline of trained staff and adequate workforce numbers, the systematic visit planning and monitoring that make the models work could erode, threatening the durability of the gains.

One of the most striking recommendations from the study concerns inclusion. The researchers proposed that future models should be created with a specific focus on meeting the needs of people with disabilities. This is a dimension of differentiated HIV care that is often overlooked, and it highlights how reach and equity must be considered together. A model that reduces visits for the general population of stable clients may inadvertently exclude clients who face physical, sensory, or cognitive barriers to accessing community distribution points or who require accommodations that the current designs do not anticipate. Building disability-inclusive features into the architecture of DSD models from the outset, rather than retrofitting them later, would be a significant step toward genuinely universal access to therapy.

Taken together, the RE-AIM analysis suggests that differentiated service delivery in Zambia is a promising strategy for improving the delivery of antiretroviral services, and the authors conclude that the findings highlight the importance of rolling out these models to cater to more people throughout the country. But the study is equally clear-eyed about what national scale-up will require. The models will only extend their reach if the financial and geographic barriers identified by patients are addressed, if community engagement through local leaders and communication materials continues, if the health workforce grows to sustain systematic planning and monitoring, and if future designs explicitly incorporate the needs of people with disabilities. In a country where so many people depend on lifelong therapy, the lesson of this research is that the next phase of the HIV response depends as much on implementation science—on understanding how programs actually behave in the world—as on the medicines themselves.

Subject of Research: Implementation of Differentiated Service Delivery models for antiretroviral therapy in Zambian health facilities, evaluated using the RE-AIM framework

Article Title: Implementation of DSD models in selected health facilities in Zambia: a RE-AIM framework analysis

Article References: Mulavu, M., Mulawa, M. J., Sitali, D., Jacobs, C., Zulu, J. M., Hamoonga, T. E., Chavula, M. P., Zyambo, C., Matenga, T. F. L., Kaonga, P., Musukuma, M., Sampa, M., Phiri, H., Hikabasa, H., & Mweemba, O. (2026). Implementation of DSD models in selected health facilities in Zambia: a RE-AIM framework analysis. BMC Public Health. https://doi.org/10.1186/s12889-026-28789-w

Image Credits: AI Generated

DOI: 10.1186/s12889-026-28789-w

Keywords: HIV, antiretroviral therapy, Differentiated Service Delivery, Zambia, RE-AIM framework, implementation science, public health, health systems, adherence, health facilities, community health, task shifting

The Greater Kafue Ecosystem comprises Kafue National Park and nine surrounding Game Management Areas, covering approximately 66,000 square kilometers. Between 1984 and 2024, the proportion of natural habitat in the region declined from approximately 98.31% to 83.96%, a transformation that has been spatially heterogeneous across the landscape. Some Game Management Areas have retained over 96% of their natural cover, while others have fallen below 40%, indicating that local factors interact with regional pressures to produce distinct outcomes. This variability highlights the need for targeted interventions rather than uniform conservation policies across the entire ecosystem.

To address this complexity, the researchers employed a Random Forest classifier, an ensemble learning algorithm known for its ability to handle high-dimensional data and capture non-linear relationships. The model was trained on a binary response variable derived from Landsat satellite classifications for 1984 and 2024, which distinguished between natural cover types such as forest and grassland, and anthropogenic types such as cropland and built-up areas. Fourteen predictor variables were evaluated, including population density, distance to settlements, mean annual rainfall, elevation, and the proportion of cultivated land within a one-kilometer neighborhood.

The model demonstrated strong predictive performance on an independent, spatially blocked test set, achieving an accuracy of 0.77 and an Area Under the Curve (AUC-ROC) of 0.83. The F1-score of 0.81 and Cohen’s Kappa of 0.53 further confirmed the model’s reliability, with ten-fold spatial cross-validation showing stable results. A higher false-positive rate compared to the false-negative rate indicated a precautionary bias, which is particularly suitable for conservation screening where identifying potential risk areas is prioritized over missing them.

Among the fourteen predictors, the percentage of cultivated land emerged as the dominant factor, accounting for 33.6% of the total variable importance. This was followed by mean annual rainfall at 12.0% and distance to settlements at 11.89%. The analysis revealed that socio-economic factors collectively contributed approximately 47.3% of the predictive power, significantly outweighing climatic, topographic, and conservation status variables. This finding aligns with broader evidence that smallholder agricultural expansion is the primary pathway of habitat conversion in sub-Saharan African woodlands.

The study identified specific cut-points where the predicted probability of habitat loss changes sharply. For cultivated land, a threshold of approximately 10% cover within a local area marked a significant increase in susceptibility, with loss probability rising from 11.3% below this level to 79.9% above it. Similarly, areas within 4 kilometers of settlements showed a high-risk influence zone, with a 75.1% probability of loss compared to 25.7% beyond that distance. These statistical cut-points serve as operational early-warning indicators for conservation managers, although they are not validated ecological thresholds.

Spatial susceptibility mapping revealed pronounced contrasts across the ecosystem. The Game Management Areas of Bilili, Namwala, and Mumbwa were identified as the principal high-susceptibility zones, with Bilili showing the highest risk at 94.5% of its area classified as high or very high susceptibility. In contrast, Kafue National Park remained predominantly low risk, with 95.4% of its area classified as very low susceptibility. This spatial differentiation supports a strategy of targeted monitoring and preventive action in the most vulnerable Game Management Areas, while recognizing that the national park remains relatively stable.

The researchers also conducted a sensitivity analysis to address potential endogeneity in the cultivated-land predictor, which is temporally close to the 2024 outcome. When the model was refitted without this variable, performance declined but remained well above chance, indicating that the predictive skill is not solely dependent on this single factor. However, the dominance of socio-economic predictors was conditional on the inclusion of cultivated land, with climatic and proximity variables becoming more prominent in its absence. This nuance is crucial for interpreting the results and designing robust conservation strategies.

The study acknowledges several limitations, including the cross-sectional nature of the analysis, which prevents the establishment of temporal causation. Most socio-economic predictors represent conditions around 2020–2021, while the response integrates conversion over four decades. Additionally, the model’s reliance on a companion Landsat classification means that thematic and change-detection errors propagate into the results. Despite these constraints, the study provides a valuable methodological template for applying machine learning to conservation challenges in other African multi-use landscapes.

By integrating the Human-Environment Interaction and Driver-Pressure-State-Impact-Response frameworks with interpretable machine learning, the study offers a comprehensive view of habitat loss drivers in the Greater Kafue Ecosystem. The findings underscore the critical role of agricultural expansion and accessibility in shaping conservation outcomes and provide actionable metrics for prioritizing interventions. As the ecosystem continues to face pressures from population growth and land-use change, this spatially explicit approach can help guide effective and targeted conservation efforts to protect its rich biodiversity.

Subject of Research: Drivers of habitat loss in Zambia's Greater Kafue Ecosystem

Article Title: Drivers of forest loss and habitat degradation in Zambia’s Greater Kafue Ecosystem

Article References: Mulenga, G., Phiri, D., Simwanda, M., & Nyirenda, V. R. (2026). Drivers of forest loss and habitat degradation in Zambia’s Greater Kafue Ecosystem. Discover Forests, 2(1), Article 81. https://doi.org/10.1007/s44415-026-00143-3

Image Credits: AI Generated

DOI: 10.1007/s44415-026-00143-3

Keywords: habitat loss, Random Forest, Zambia, Greater Kafue Ecosystem, agricultural expansion, conservation planning, remote sensing, Miombo woodlands, spatial analysis, biodiversity, Drivers, forest

The research team, led by Vincent R. Nyirenda of the Copperbelt University together with colleagues from the park’s management authority and a South African environmental management centre, surveyed 206 households across the human-dominated landscape surrounding the park. Liuwa Plain is no conventional wilderness: established in 1972 on the site of a 19th-century royal hunting reserve created by the Lozi king Lubosi Lewanika, it is today jointly managed by African Parks, Zambia’s Department of National Parks and Wildlife, and the Barotse Royal Establishment. Roughly 24,000 people live in the surrounding Liuwa constituency, most of them poor and heavily dependent on the floodplain’s fisheries, reeds, papyrus, and raffia palms for their livelihoods.

To map the invisible wiring of this economy, the researchers treated every entrepreneur as a node in a network and drew a tie between two people whenever one reported a relationship with the other. They then quantified classic network measures—density, closeness, and betweenness—and used community detection algorithms to reveal clusters of overlapping relationships. The picture that emerged was strikingly sparse. Network density was a mere 0.7 percent, with reciprocity at just 2.7 percent, which the authors attribute to inadequate cooperation, lack of trust, negative attitudes, and the disruptive rhythm of a highly seasonal floodplain system. Across the 206 individuals surveyed, the total number of connections amounted to only 590, with individuals maintaining between one and eight ties each.

Yet the sparse network carried rich social information. The largest clusters of ties occurred among male entrepreneurs, who both fostered alliances and engaged in competition—most of the recorded rivalry involved men. Women, by contrast, tended to form alliances, and the study found that nearly every participant relied on at least one inter- or intra-group alliance for their enterprise. Individuals with many ties were almost exclusively in cooperative relationships rather than competitive ones, suggesting that well-connected entrepreneurs occupy a distinct structural position in the local economy. The ties also clustered by enterprise type: fishing was commonly mixed with weaving, timber, or beekeeping, and the Lozi ethnic group, which made up 75 percent of respondents and largely held primary or secondary education, dominated the relationship structure with a marked concentration of intra-ethnic ties.

One of the most consequential findings concerns livelihood diversification. Households with fewer alliances showed a greater tendency to shift away from fishing—the historic economic mainstay of the region—into alternative enterprises such as weaving, caterpillar harvesting, beekeeping, and mushroom production. Those endowed with stronger alliance networks remained fish-dependent. The authors interpret this as evidence that social connectivity and economic strategy are deeply intertwined: diversification appears partly as an adaptive response when relational capital is thin, while abundant alliances anchor households in traditional resource use. Gender and household size also shaped the pattern, with women from households of two to nine members showing more enterprise diversity than men, whose diversification was concentrated in households of four to nine members.

When the researchers asked entrepreneurs what drove their enterprises, income generation topped the list by a statistically significant margin, followed by traditional practices embedded in the strong cultural setting governed by the Barotse Royal Establishment, and then information sharing within existing social ties. Financial capital, the analysis showed, contributed significantly more than social, cultural, and natural capitals combined. Local market availability, water transport, and scarce credit facilities reinforced the financial driver, while the availability of resources such as fish, reeds, and timber played a smaller supporting role.

The barriers were equally revealing. Climate change impacts and poor market access together formed the most significant constraint, outweighing any individual factor. Seasonal flooding restricts mobility, damages infrastructure, and raises transport costs, while human-wildlife conflicts involving crocodiles, hyenas, and lions create insecurity that discourages participation in livelihood activities. Governance-related barriers—weak institutional support, lack of transparency among some leaders, inequitable resource access, and poor conflict management—followed closely, alongside socio-economic frictions such as non-cooperation, poverty, and lack of trust, and technical shortcomings including insufficient skills and inappropriate fishing methods.

Against these obstacles, the study identified three bridging pillars for building resilient enterprises: adaptive capacity, good governance, and traditional practices. Adaptive capacity—encompassing cooperation, trust, self-motivation, functional cooperatives, peace, negotiation skills, and good customer relations—proved statistically more significant than governance factors, which in turn outperformed traditional practices. The governance pillar included communication among entrepreneurs, government agencies, conservation NGOs, and traditional leaders, along with accountability, skills transfer, transparency, and transport infrastructure. Traditional practices, rooted in cultural values and long-standing co-adaptation to resource use, rounded out the triad.

The findings carry a double edge for conservation policy. Strong kinship-based networks can promote trust, reciprocity, and collective action, but the same bonding social capital can privilege insiders—dominant clans, influential families, politically connected fishers—while excluding migrants, women, and poorer households. The authors note that kinship loyalty and patron-client relationships may even shield rule-breakers, weakening fisheries enforcement and allowing inappropriate fishing methods to persist. In this sense, social cohesion functions simultaneously as a governance asset and a source of regulatory failure, a nuance the authors argue distinguishes Liuwa from wildlife-tourism-focused studies elsewhere in African transfrontier conservation areas.

For the researchers, the practical message is that strengthening adaptive capacity, collaborative governance, and social connectivity within shared landscapes is critical for enterprise resilience—and that conservation planning aimed at the Sustainable Development Goals, from poverty reduction to climate action and life on land, must account for the relational fabric of communities, not just the ecological resources they harvest. Because the study captured only a single seasonal snapshot of a dynamic floodplain system, the authors call for continuous, longitudinal monitoring to track how social networks and enterprise performance shift under changing socio-ecological conditions, ensuring that interventions remain tuned to the living architecture of relationships on which Liuwa’s nature-based economy ultimately depends.

Subject of Research: Social network drivers of nature-based livelihood enterprises in a Zambian protected landscape

Article Title: Sociocentric relationships drive nature-based enterprises in the human dominated Liuwa Plain National Park

Article References: Nyirenda, V. R., Phiri, D., Namukonde, N., Kalolekesha, H., & Nkhata, B. A. (2026). Sociocentric relationships drive nature-based enterprises in the human dominated Liuwa Plain National Park. Frontiers in Sustainable Resource Management, 5, Article 1871815. https://doi.org/10.3389/fsrma.2026.1871815

Image Credits: AI Generated

DOI: 10.3389/fsrma.2026.1871815

Keywords: social network analysis, nature-based enterprises, Liuwa Plain National Park, Zambia, livelihood diversification, community conservation, gender, climate change, governance, social capital, sustainable development goals, floodplain fisheries

The study is grounded in the Social Ecological Model, a framework that holds that individual behavior is shaped not only by personal knowledge but also by interpersonal, community, institutional, and policy-level factors. In the context of plastic waste, this means that household recycling is not simply a matter of personal willingness but depends on enabling conditions such as segregation bins, educational campaigns, incentive programs, and regulatory enforcement. The researchers argue that understanding these multiple interacting levels is essential for designing interventions that actually work, rather than relying on awareness campaigns alone.

Plastic pollution is a particularly acute problem in the region. Sub-Saharan Africa’s annual plastic importation and production are estimated at 117.6 million metric tons, with roughly 80 percent ending up as solid waste. In Ghana alone, about 8.2 billion sachets of drinking water are consumed each year, and improper disposal has resulted in approximately 86 percent of the country’s plastic waste clogging stormwater drains, rivers, and streams, with an estimated 250,000 metric tons eventually reaching the Atlantic Ocean. Sierra Leone imported 9 million kilograms of plastic in 2019 alone, 92 percent of which came from Ghana, illustrating the regional nature of the problem.

The survey found that households across all three countries generate similar forms of plastic waste, dominated by sachet bags, plastic bottles, shopping bags, food packaging, and detergent packs. In Ghana, more than half of respondents reported generating all five waste types, reflecting the country’s high consumption of packaged goods. When asked how they dispose of plastic waste, the most common practice in all three countries was dumping it in bins, with burning as the second most frequent method. The persistence of open burning, the researchers note, has serious implications for air quality, greenhouse gas emissions, and human health, and reflects deficiencies in waste collection services and the absence of alternative disposal mechanisms.

One of the most striking findings concerns knowledge of plastic recycling symbols. The Society of the Plastics Industry introduced a coding system in 1988 that places a numbered triangle on most plastic products to indicate the type of polymer, ranging from PETE (polyethylene terephthalate) used in beverage bottles to HDPE, PVC, LDPE, PP, PS, and a miscellaneous category labeled OTHER. Despite the high educational attainment of the sample, with 96.3 percent of respondents holding tertiary qualifications, awareness of these symbols was generally low. Knowledge of the PETE symbol was comparatively higher, particularly in Zambia at 51.35 percent and Ghana at 34.45 percent, while Sierra Leone recorded just 12.28 percent. Awareness of PS and OTHER symbols was particularly poor, with roughly one-tenth of respondents reporting knowledge of these classifications.

Chi-square analysis revealed statistically significant differences among the three countries in knowledge of the PETE and HDPE symbols, suggesting that exposure to recycling activities and public awareness campaigns varies across national contexts. However, knowledge of PVC, LDPE, PP, PS, and OTHER symbols did not differ significantly, indicating that poor awareness of less common polymer classifications is a shared challenge across the region. The researchers interpret this as evidence that formal education does not necessarily translate into technical understanding of recycling systems, and that environmental literacy depends more on exposure to recycling information than on educational attainment alone.

The study also examined the role of economic incentives and institutional support. A significant majority of respondents, 77.31 percent, had never received rewards in the form of money or household goods in exchange for their plastic waste, and only 8.33 percent had been given dustbins by authorities for plastic waste segregation. Furthermore, 85.19 percent of respondents had never received visits from regulators or officials to educate them about waste segregation. These findings point to weak institutional engagement and largely absent incentive mechanisms within household recycling systems across all three countries.

Despite these barriers, respondents expressed broad support for potential interventions. Overall, 74.54 percent believed that measures such as household goods exchange programs, provision of segregation buckets, and periodic visits by officials would effectively increase household waste segregation. Notably, the Kruskal-Wallis H test found no statistically significant difference among the three countries in perceptions of intervention effectiveness, suggesting that households across different national contexts share broadly similar views on what would help. The researchers cite previous work in Mozambique, where providing sorting bins and conducting periodic home visits significantly boosted household waste segregation, as evidence that such approaches can succeed when implemented.

The study’s authors conclude that improving household plastic waste segregation in Ghana, Sierra Leone, and Zambia requires integrated interventions that combine public education, provision of segregation facilities, incentive schemes, and strengthened regulatory support. They argue that policies aimed at improving plastic waste management should be tailored to country-specific conditions rather than adopting a uniform approach across Sub-Saharan Africa, given the variations in socio-economic and institutional contexts observed. They also note that introducing deposit-refund systems, plastic buy-back schemes, and household goods exchange programs could increase the supply of recyclable materials and strengthen the circular economy in the region.

The research contributes to a growing body of literature on household plastic waste management in developing countries and provides empirical evidence to support the design of sustainable waste management policies. By demonstrating that household behavior is shaped by interactions between individual awareness and institutional support systems rather than by personal attitudes alone, the findings underscore the need for coordinated efforts addressing behavioral, institutional, and infrastructural barriers simultaneously if Sub-Saharan Africa is to transition from a linear take-make-dispose model toward a functioning circular plastic economy.

Subject of Research: Household plastic waste segregation, disposal practices, and recycling awareness in Ghana, Sierra Leone, and Zambia

Article Title: Comparative assessment of household plastic waste segregation, disposal practices, and recycling awareness in Ghana, Sierra Leone, and Zambia

Article References: Afeliga, E., Essuman, M. A., Hamidu, I., Asamoah, K., Munyenyembe, H., Odis, A. I., & Lavalie, E. D. M. (2026). Comparative assessment of household plastic waste segregation, disposal practices, and recycling awareness in Ghana, Sierra Leone, and Zambia. Frontiers in Sustainability, 7, Article 1816953. https://doi.org/10.3389/frsus.2026.1816953

Image Credits: AI Generated

DOI: 10.3389/frsus.2026.1816953

Keywords: plastic waste, recycling, household segregation, Sub-Saharan Africa, circular economy, waste management, Ghana, Sierra Leone, Zambia, recycling symbols, economic incentives, Social Ecological Model

The study formed part of the Combating Antimicrobial Resistance and Antimicrobial Residues in Zambian Poultry project, known as CAZAAP, which was established in 2022 with an ambitious target: reduce antimicrobial use and residues by at least 30 percent in intensively managed broiler systems over three years. To evaluate progress, the researchers surveyed the veterinary workforce in four districts chosen for their intensive poultry production: Chongwe and Chilanga in Lusaka Province, and Ndola and Kitwe in the Copperbelt Province. Because the eligible population of veterinary personnel serving these districts was small, the team invited every one of them to participate, enrolling 50 respondents at baseline in March and April 2024 and retaining 49 complete pairs at endline in December 2025. The cohort included registered veterinary surgeons holding bachelor-level qualifications and veterinary paraprofessionals, the certificate- and diploma-holding frontline staff who head the operational units called veterinary camps and who provide the first line of animal health services to farmers.

Between the two surveys, the CAZAAP project delivered a package of stewardship interventions. In August 2025, all veterinary professionals from both provinces attended week-long training workshops covering prudent prescribing, resistance awareness, and compliance with national guidelines. Every paraprofessional who took part in the survey also served as a trainer in farmer field schools, undergoing antimicrobial stewardship training before delivering instruction to farmers themselves. In November 2025, the project launched and disseminated poultry Standard Treatment Guidelines, with all survey participants present at the event. The researchers deliberately used the identical questionnaire at both time points, administered by the same team of six trained enumerators using the Kobo Collect digital platform with built-in validation, so that any observed differences would reflect genuine changes in respondents’ answers rather than changes in the instrument.

The quantitative results tell a story of measurable improvement in knowledge depth. At baseline, every respondent reported knowing what antibiotics are, but only 68 percent in the Copperbelt and 81 percent in Lusaka could provide a correct definition. By endline, those figures had climbed to 94 percent and 88 percent respectively. Correct definitions of antimicrobial resistance rose from 63 percent to 78 percent in the Copperbelt and from 44 percent to 75 percent in Lusaka. Awareness of antibiotic residues, a concept central to food safety because residues in meat and eggs can trigger allergic reactions in consumers, disrupt gut microbiota, and expose humans continuously to low levels of antibiotics that select for resistant bacteria, showed particularly marked gains: correct definitions jumped from 39 percent to 84 percent in the Copperbelt and from 33 percent to 80 percent in Lusaka. Understanding of withdrawal periods, the mandatory interval between the last antibiotic dose and slaughter or egg collection, also improved, reaching 85 percent correct in the Copperbelt and 88 percent in Lusaka at endline.

Attitudes toward stewardship were strongly positive across both provinces at both time points. Respondents overwhelmingly agreed that good biosecurity and hygiene can reduce antibiotic use, that vaccination of broilers can lower the need for antimicrobials, and that resistance is an important problem for both animal and public health. Every single respondent, in both surveys and both provinces, endorsed the necessity of a prescription before antibiotic sale. Yet the researchers observed a curious divergence over time: while the Copperbelt cohort maintained or strengthened its strong agreement on these items, Lusaka respondents shifted toward more moderate positions at endline, producing statistically significant provincial differences on several attitude items. The authors caution that these subgroup comparisons are exploratory, given that the study was powered only to detect large effects; with Lusaka numbering just 16 respondents, the minimum detectable difference between provinces was on the order of 38 to 41 percentage points.

Practice data revealed both encouraging shifts and stubborn gaps. The proportion of professionals reporting frequent use of antibiotics for viral infections, a biologically inappropriate practice since antibiotics have no effect on viruses, declined substantially in the Copperbelt, where those reporting they always prescribed for viral infections fell from 21 percent to just 3 percent. Prophylactic prescribing for disease prevention also dropped, with 88 percent of Copperbelt respondents and every Lusaka respondent reporting they never prescribed for this purpose at endline. Biosecurity practices improved as well: consistent use of farm-specific protective clothing rose from 41 percent to 73 percent in the Copperbelt, and footbath compliance was universal at both time points. However, laboratory diagnostics remained strikingly underused. By endline, 45 percent of Copperbelt respondents reported never sending samples for bacterial identification, a highly significant difference from Lusaka, where some improvement was seen. Access to treatment guidelines, though improved from a dismal 2.9 percent to 24 percent in the Copperbelt, remained unavailable to the majority of professionals in both provinces.

The qualitative strand of the study, three key informant interviews with senior veterinary professionals including a district coordinator in Chongwe, a District Veterinary Officer in Ndola, and a Senior Veterinary Officer for the Copperbelt Province, exposed the disconnect between headline awareness and applied competence. One informant estimated that functional knowledge of resistance among professionals ranged from only 40 to 60 percent, a figure that initially appears to contradict the survey’s near-universal awareness claims but converges with the survey when matched against the correct-definition rates rather than the binary awareness items. The Chongwe district coordinator captured the paradox starkly: professionals may claim to know more, but in practice they are at times the ones who perpetuate the emergence and transmission of resistance. Informants described empirical prescribing without diagnostic confirmation, repeated switching of antibiotics after treatment failure, and reliance on prior experience or peer advice rather than laboratory-guided therapy.

Beyond the professional workforce, the interviews painted a picture of a farmer ecosystem in which self-medication is widespread and antibiotics flow freely through weakly regulated channels. Poultry farmers commonly obtain drugs directly from agrovet shops without prescriptions, guided by peer recommendations or previous experience. One informant recounted a farmer who decided to buy and administer a particular drug before his animals got sick, based on the fact that his pigs had died the previous time. Withdrawal periods, despite over 90 percent of professionals reporting they educate clients on them, appear to be routinely ignored: informants described farmers as shocked to learn that withdrawal periods exist at all. Agrovet shops, driven by commercial targets, dispense whatever the farmer needs without asking for prescriptions, and veterinary drugs even enter the supply chain through general farmers’ shops. Economic pressures compound the problem, with farmers underdosing to save money and offloading stock onto the market under pressure, while the cost of veterinary consultations and laboratory testing discourages professional engagement.

The convergent mixed-methods integration, in which each quantitative result was juxtaposed with corresponding qualitative findings and assigned a fit category of convergence, divergence, or expansion, yielded a central meta-inference: stewardship challenges in Zambia’s poultry sector are driven less by professional awareness than by applied competence and the structural environment in which professionals operate. The qualitative strand expanded on the survey’s hints by supplying the mechanisms, unregulated agrovet dispensing, weak prescription enforcement, cold-chain and storage failures, and economic pressures on both farmers and professionals, that the survey could only gesture toward through items such as antibiotic selling and guideline access. An instructive divergence arose on withdrawal periods, where provider-reported education did not correspond to farmer comprehension or adherence, indicating a transmission and efficacy gap that message delivery alone cannot close.

The authors are careful to acknowledge the limitations of their pre-post design. Without a concurrent control group, the observed changes cannot be causally attributed to the CAZAAP interventions specifically; during the study period Zambia was simultaneously implementing its multisectoral National Action Plan on Antimicrobial Resistance, and resistance received sustained national and international attention. Repeated testing with the same questionnaire may itself have contributed to observed changes through familiarisation, and all practice data were self-reported, with no direct observation of prescribing behaviour or verification against dispensing records, creating a risk of social desirability bias that may have been heightened at endline after respondents had received stewardship training. The triangulation with qualitative interviews partly mitigates this, showing that stated awareness exceeded applied competence and that professional self-reports of improved practice coexisted with persistent inappropriate use in the wider farmer ecosystem. Nonetheless, the findings point toward a coordinated One Health response: finalising and systematically disseminating national veterinary treatment guidelines, improving diagnostic access and functional resistance surveillance, strengthening regulation and inspection of veterinary drug outlets, enforcing prescription requirements, and designing farmer-centred education around uptake and retention rather than message delivery alone. These measures are mutually reinforcing and unlikely to succeed in isolation, the authors conclude, highlighting that the binding constraints on antimicrobial stewardship in Zambia’s poultry sector are structural, not merely knowledge-based.

Subject of Research: Bridging knowledge and practice gaps in antimicrobial stewardship among veterinary professionals in Zambia’s poultry sector: a mixed-methods pre–post intervention study

Article Title: Bridging knowledge and practice gaps in antimicrobial stewardship among veterinary professionals in Zambia’s poultry sector: a mixed-methods pre–post intervention study

Article References: Chileshe, C., Mumba, C., Chazya, R., Ngosa, W., Mudenda, S., Chishimba, V., Sinyawa, T., Chawinga, K., Shawa, M., Mainda, G., Muchepa, I., Chizimu, J. Y., Nchima, G., Hang’ombe, B. M., Goma, F., Muleya, W., & Mudenda, N. B. (2026). Bridging knowledge and practice gaps in antimicrobial stewardship among veterinary professionals in Zambia’s poultry sector: a mixed-methods pre–post intervention study. Frontiers in Antibiotics, 5, Article 1898867. https://doi.org/10.3389/frabi.2026.1898867

Image Credits: AI Generated

DOI: 10.3389/frabi.2026.1898867

Keywords: Bridging, knowledge, practice, gaps, antimicrobial, stewardship, veterinary, professionals, Zambia, poultry, sector, mixed-methods