Cystic Echinococcosis: An Impact Assessment of Prevention Programs in Endemic Developing Countries in Africa, Central Asia, and South America

Background . Cystic echinococcosis (CE), caused by the tapeworm species, Echinococcus granulosus sensu stricto (G1), is one of many primary neglected zoonoses worldwide. Within endemic developing countries, CE has multiple e ﬀ ects on animal and human health and well-being. To address such e ﬀ ects, veterinary and human medical sector collaboration on prevention program delivery is essential. To begin preliminary evaluations of county speci ﬁ c prevention programs, a critically appraised topic (CAT) was conducted. It sought to answer: What impact do CE prevention programs have on human and animal disease prevalence, in populations living in endemic developing countries within Africa, Central Asia, and South America? Methodology . The aim was to assess the ability of prevention and control program outputs to produce measurable di ﬀ erences in health, social, and economic outcomes (e.g., improved access to medical services, positive behavioral change, or reduced treatment costs, respectively). Included articles were obtained using prede ﬁ ned inclusion/exclusion criteria from the four databases (CAB Abstracts and Global Health; the National Library of Medicine (PubMed); ScienceDirect; and WHO Institutional Repository of Information Sharing (IRIS)). The articles were appraised using three checklists: the Royal College of Veterinary Surgeons (RCVS), the Critical Appraisals Skills Programme (CASP), and the Joanna Briggs Institute checklists. Results . Ten articles were selected. Geographically, 20% of studies were conducted in South America, 30% in Africa, and 50% in Central Asia. For de ﬁ nitive hosts, dogs, CoproELISA antigen testing, before and after Praziquantel (PZQ) de-worming, was a primary focus. For humans, who are intermediate hosts (IH), disease surveillance methods, namely ultrasound (US), were commonly assessed. Whilst for sheep, also acting as IH, disease prevention methods, such as the EG95 livestock vaccine and de-worming farm dogs, were evaluated. Common to all studies were issues of program sustainability, in terms of regular human US screening, dog de-worming, and annual sheep vaccination. This was attributed to transient and remote human or animal populations; limited access to adequate roads or hospitals; few skilled health workers or veterinarians; an over-reliance on communities to administer preventatives; and limited resources. Conclusion . Despite variations in result validity and collection periods, useful comparisons of CE endemic countries produced key research and program recommendations. Future research recommendations included testing the signi ﬁ cance of multiple program outcomes in relation to prevalence (e.g., the social outcome: behavioral change), further research on the impact of livestock vaccinations, and the CE transmission role of waterways and sanitation. Program recommendations included calculating and distinguishing between stray versus owned dog populations; formal representation of internal and external stakeholder interests through institutional organization; establishing sustainable guidelines around the frequency of PZQ and vaccination administration; improved veterinary-human medical training and resource sharing; and combined prevention methods and multiple canine disease management.


Introduction
Globally each year, cystic echinococcosis (CE) causes an estimated 19, 300 deaths, and a loss of US$3 billion to treatment costs, especially within the livestock industry [1]. Within the multi complex species (Echinococcus granulosus sensu lato), E. granulosus sensu stricto (G1) causes the highest prevalence and widest global distribution of human cases ( [2], p. 2).
Dogs are the definitive hosts, infected by ingesting prey or raw meat containing metacestode cysts (larvae) [3]. Although dogs remain mostly asymptomatic, they excrete proglottids or infective eggs that are ingested by multiple mammalian intermediate hosts (IH), such as sheep, cattle, wildlife, and humans [4]. IH can remain asymptomatic for years, until growing cysts rupture or cause complications in adjacent organs ( [5], p.12).
For many living in low socioeconomic and/or remote areas within endemic countries, recommended medical diagnostic methods and preventatives, such as Praziquantel (PZQ) (5 mg/kg) for dogs [6] or the EG95 livestock vaccine [2,7], are not always readily accessible. Access to social services, such as healthcare, can act as a development capability [8,9]. That is, it presents an opportunity to enhance one's well-being or acts as a constraint, which contributes to vicious cycles of poverty [10][11][12]. Cost-effective preventatives, such as public health education programs, and veterinary and human medical collaboration [13][14][15][16][17] have been suggested to overcome resource constraints. Public health education campaigns and multi-stakeholder collaboration are essential to producing social outcomes, such as behavioural change, that can reduce disease incidence. Examples of positive behaviours include: not feeding dogs raw meat, hand washing after handling dogs, and preventing canine access to common livestock, wildlife, and human environments [18][19][20].
CE control and prevention programs have been broadly identified in endemic regions of Africa, South America, and Central Asia [16]. Since 2019, the progress is evident in countries, like Mongolia, where a multi-stakeholder program provided PZQ de-worming for dogs and human ultrasound (US) screening [21]. However, few studies have analyzed the effectiveness of prevention or control programs, in terms of specific outcomes. Most focus on human medical and surgical treatments or conducting cohort or case control studies to identify common risk factors. Thus, the aim of this critical appraisal was to conduct a program impact assessment to provide evidence-based recommendations for current public and animal health programs.

Methodology
Four online databases were searched (Table 1). These included the following: CAB Abstracts and Global Health (1973-present), the National Library of Medicine (PubMed), ScienceDirect, and the WHO Institutional Repository of Information Sharing (IRIS) [22] and "Echinococcosis" webpage [11]. PubMed and CAB abstracts produce comprehensive veterinary and human medical research [23][24][25], while the ScienceDirect database produces full text, peer reviewed literature on healthcare [26]. Additionally, the WHO resources were searched, as it is one of the leading international organizations conducting CE programs in endemic regions.
2.1. Screening Process. Within the ScienceDirect database, the journals were selected from two "Subject Areas": "Veterinary Science and Veterinary Medicine" and "Medicine and Dentistry." Further refinement was achieved by applying inclusion/exclusion based upon relevance to the topics of "parasitology," "public health," "zoonoses," and "developing countries" [26]. Additionally, within CAB abstracts, the word "veterinary" and phrases "one health" and "animal health" originally resulted in the exclusion of several relevant articles. This became evident when conducting repeat searches. Thus, these terms were excluded in the final search (Table 1) to improve sensitivity, as lower specificity could be easily corrected. For example, predefined inclusion/exclusion criteria (Table 2), synonyms, Boolean operators, parentheses, and truncation (CAB Abstracts only) were used to increase result specificity.
Furthermore, articles were refined using an ordered selection process ( Table 3). The fifth exclusion criteria "non-English articles" (Table 2) presents a methodological limitation, as sensitivity is reduced. However, the aim was to avoid timely and possible inaccurate article translations. Additionally, across all databases, the words "diagnostic" and "tests" were included in search terms; as many prevention programs encompassed diagnostic tests to monitor program impact. However, studies [27] that conducted generalized evaluations of individual treatment or diagnostic tools were excluded [28].
Finally, articles were sorted based upon relevance. The article title and abstract were reviewed for key words e.g., "echinococcus," "prevention," and/or "control." If relevance could not be determined, the full texts, namely, the discussion and methodology, were reviewed. Integrated prevention program assessments of multiple diseases [31][32][33] were excluded. Individual disease analysis is important, as each disease has its own complex transmission pathway and diagnostics. However, there was one exception [34], because CE program outputs and impacts were readily distinguished from other diseases.
Despite the described replicable criteria, methodological limitations may arise from inaccurate subjective application of inclusion/exclusion criteria or study design categorization. For example, controlled clinical trials were excluded because of a high number of false positive search results. Nevertheless, PubMed still produced a non-randomized controlled trial that fit the inclusion criteria. Differences in categorization of this study existed between this CAT, article authors, and the database.

Geographical Selection.
Africa, South America, and Central Asia are listed as CE endemic regions [16], consisting of low to middle income countries [35][36][37]. The included studies within these regions were mostly in countries that ranked below 50/188 on the United Nations Development Programme (UNDP) multidimensional poverty index [38]. For example, Kyrgyzstan (142) and Kenya (143) are two of the lowest ranked [38], but countries like China were included, as it represents 40% of the world's CE caused disability-adjusted life years (DALYs) ( [39], p.138).

Critical Appraisal Method.
Applying a triangulation method, which is not reliant upon one set of appraisal criteria, reduced the probability of information bias. Checklists

Study
Designs. This CAT focused on evaluating program outcomes and impacts, versus disease causality and risk factors, commonly tested by several excluded cohort and case control studies. Compared to cross-sectional studies and case control studies, cohort studies are relatively expensive, which is often not suitable in low socioeconomic contexts. Nevertheless, cohort studies are superior in terms of monitoring disease incidence rates and controlling against confounding demographic or signalment variations, by consistent respondent follow-up. Whilst case control studies are usually more cost effective, within a low socioeconomic context, access to consistent human and animal health data may be limited. Respondant follow-up was a challenge across most included studies, as sample human or animal populations were often transient [42,43,51] and resided in remote areas [15,34,43], with limited access to adequate roads, hospitals, or resources [2,6]. These challenges make selection of a control group practically difficult and may account for the few experimental studies encountered.

Journal of Zoological Systematics and Evolutionary Research
In addition, ethical issues arise if an animal or human, with little access to services, is denied preventative treatment to be assigned to a control group. Indeed, Yang et al. [43] concluded that it would be ethically negligent to select a control site, after the government identified that CE was endemic in Northwest China ( [43], p.357). Thus, in China [15,43], Mongolia [15], Argentina [2,6], and Kenya [42], many included studies utilized non-randomized convenience sampling. The majority used cross-sectional study designs, as they are relatively cost effective and practical at sampling populations at different points in time.
Furthermore, quasi-experimental studies were suitable in comparing a pre-and post-program impact. The sample population is ideally exposed to the same program output PubMed and WHO IRIS only: "full text" and "abstract" selected to exclude articles with incomplete access 3. Document type: (i) CAB: book, book chapter, thesis, conference proceedings, correspondence, bulletin, editorial, bulletin article, miscellaneous, and annual report (ii) PubMed: books and documents (iii) WHO: publications, technical documents, advisory committees, meeting minutes, governing bodies documents, and regional director documents (iv) Science Direct: exclusion of both document and study types combined under one category: "Article Types" 4.
(i) CAB: 13 "Research Areas" (Plant Sciences, Reproductive Biology, Marine Freshwater Biology, Meteorology Atmospheric Sciences, Anesthesiology, Forestry, Sports Science, Film Radio Television, Fisheries, Archaeology, Arts Humanities Other Topics, Materials Science, and Paleontology) All searched using "echinococcosis," "cyst," "tapeworm," "control," or "prevention," before final exclusion. (ii) Science Direct: 8 "Subject Areas" (Immunology and Microbiology; Agricultural and Biological Sciences; Biochemistry, Genetics and Molecular Biology; Pharmacology, Toxicology, and Pharmaceutical Science; Environmental Science; Chemistry; Neuroscience; and Engineering) 5. Language: non-English 6. Study type: (i) PubMed: clinical study, clinical trial, controlled clinical trial, meta-analysis, clinical trial veterinary, and systemic review (ii) Manual exclusion for CAB, WHO, and ScienceDirect Reviews: systemic reviews (exception for program and country specific reviews) and narrative reviews Experimental studies: crossover trials, controlled in vitro laboratory or clinical trials (based on individualized treatment and diagnostic methods, not country or program specific) Observational studies: case control studies, opinion pieces, and meta-analysis 7. Relevance based upon "title," "abstract," or "full text" (if exclusion could not be made from reviewing article title or abstract) Inclusion: 1. Date: 2015-2021 2. PubMed and WHO IRIS only: "full text" and "abstract" selected 3. Document type: (i) CAB: journal article, conference paper, and journal issue (ii) PubMed: journal article (iii) WHO ("Communities and Collections" in IRIS: Headquarters (Journal Bulletin, Journal articles); Regional office for Africa (Country Offices, Journal articles); Regional office for the Americas (Journal articles and newsletters, publications); Regional office for Southeast Asia (Country Offices, Regional Journals, Regional Publications, Sub-Committee on Policy and Programme Development and Management); and Regional office for the Western pacific (Regional Office for the Western Pacific Publications) (iv) Science Direct: "article types:" review articles, research articles, data articles, case reports, mini reviews, and other 4.
(i) CAB: all but 13 excluded "Research Areas" (ii) Science Direct: "Subject Areas": "Veterinary Science and Veterinary Medicine," and "Medicine and Dentistry" 5. Language: English 6. Study type: (i) PubMed: comparative study, evaluation study, observational study, observation study veterinary, and review (ii) No selection option for CAB, WHO, and Science Direct. Thus, manual inclusion Observational studies: cohort, cross-sectional, before and after study, and interrupted time series Descriptive studies: case series, program, and country-specific observational studies 7. Relevance based upon title, abstract, or full text review: (i) Addresses research question (ii) Focused on surveillance, control, or prevention programs; social, health, or economic outcomes (e.g., behavioral change and access to medical services); impact (e.g., disease prevalence) (iii) Disease: Echinococcus granulosus species (iv) Population: Endemic developing countries in Central Asia, Africa, or South America

Sample Size.
Four studies did not calculate sample size and/or include confidence intervals (CI) [39,42,43,46], and some provided limited information about sample collection methods [15,39]. This may be attributed to issues of accessing remote and/or transient communities. In Yu et al.'s [39] study, all data was centrally controlled by the National Ministry of Health, China, and no information about sample collection was provided. Consideration of how political or organizational agendas align with program outcomes is important, due to the centralized control of statistical data on CE prevalence. Nevertheless, four studies did calculate sample size [2,6,34,45]. For one before/after study [2], each expected proportion (prevalence rate) for humans, dogs, and sheep was treated independently. However, calculation based upon paired data (discordant pairs) would have been suitable, as one group of animals or humans was paired to two different prevalence values, at the start and end of the program. In addition, studies with limited population size [42,43] could have utilized a finite population sample calculation [52].
Notably, a major issue with sample size calculation was captured in the 8/10 studies that did not specify an estimated stray dog population or sample size. This meant that only owned dogs were treated [2,6,39,42] or generalized terms, such as "free roaming" dogs [34] or "dog management" [15], were used. Only three studies [43,45,46] clearly distinguished and treated stray dogs with PZQ. However, administration methods were not standardized, and population size was also not calculated by Yang et al. [43].
Additionally, one study [45] utilized convenience sampling of dogs caught by the local dog catcher. Although ethics approval was obtained, the number (n = 38) of dogs euthanised for necropsy was not justified using a predefined sample size calculation. This is essential to minimise and validate the number of stray dogs necessary to test for significant differences in prevalence. Convenience sampling also reduced external validity, as selected stray dogs may have only been representative of a small area.
Furthermore, one before/after study [34] revealed the benefit of Lot Quality Assurance Sampling (LQAS) to evaluate the quality of health care programs [53]. Compared to other sampling methods, LQAS allows randomized analysis of a small community sample size, but not individual village level analysis. For countries with small, but remote or widely distributed communities, utilizing a sample size, such as nineteen, is ideal and has been proven to minimize type a and b errors [34,54]. The LQAS enables identification of community areas that fall below average in achieving a specific program target. Indeed, Van Kesteren et al. [34] traced a transition from poor PZQ dosing coverage in 8/10 villages to improvements in reaching dosing targets (p.3). Nevertheless, LQAS may be logistically costly if researchers are required to travel to multiple program areas.

Program Outcomes and Impact.
It is helpful to distinguish between program outputs, outcomes, and impacts when evaluating programs. Outputs can be defined as "the goods or services produced by programs…while outcomes are defined as the impact on social, economic, or other indicators arising from the delivery of outputs" [55]. An adapted definition of impact is when a program outcome "helps solve the problem that inspired actors to create" ( [56], p.460) it.
Studies focused on health outputs and outcomes (e.g., access to medical or veterinary services) and their impact, in terms of disease prevalence. One study assessed economic outcomes (e.g., accumulated financial costs), and a few studies [34,39] evaluated social outcomes (e.g., positive behavioral change in response to public health campaigns). While a core impact of prevention programs is to resolve the issue of rising disease prevalence, most studies failed to test the significance of prevalence changes in correlation to multiple social, health, and economic outputs and associated outcomes [2, 6, 34, 42, 43]. (iii) Result reliability: CoproELISA tests confirmed by two different methods: WB in first two study periods and PCR in the last (iv) Limited sensitivity of US in detecting pulmonary cysts (v) Sample size calculation: prevalence treated as a single independent proportion for each sample population (dogs and children). Comparing two independent proportions may be more suitable, as the two populations were treated as interdependent (vi) Few details about community survey method (e.g., verbal, written, and electronic) or question type (open, closed, and MCQ) (vii) CoproELISA results not instantaneous. Thus, control measures may be delayed Confounding variables (i) Did not distinguish between owned and stray dogs (ii) Non-standardized PZQ administration: owner and rural health workers (iii) Researchers not blinded to de-worming administration method (e.g., if farmers or rural health workers de-wormed dogs) (iv) List "sanitary education" as a program output, but do not measure behavioral outcomes 7 Journal of Zoological Systematics and Evolutionary Research  between 1985 and 2012. Continuing trend of higher prevalence in females, across multiple age groups.
Although a relatively smaller sample size, prevalence (>50 years) has not changed significantly when compared to other age groups. Attributed to persistent infection and lower life expectancy before program (iv) Gender: CE prevalence per 1000 males or females between 1985 and 2010-2011 or 2011-2012, statistically significant P = 0:0083 ð , P < 0:05Þ (v) Behavioral outcomes: appropriate disposal of offal and dog population control. Health education programs targeted to women, who spent most time at home with dogs, linked to behavioral changes and decreased CE prevalence. Effects of health education conclusively slow and rarely effective alone (vi) Infrastructure: construction of abattoirs (vii) Statistically significant changes to prevalence linked to both gender and age, using Pearson χ2 (P < 0:001 ) and ordinal Somers' d tests (P < 0:01) (viii) US proved superior results (higher sensitivity) compared to serology ELISA. CE cysts detected in 198 patients using US vs. 76 using serology ELISA More specifically, studies identified health education campaigns as program outputs [2,6,34,39,42,43,50], but only three [2,34,42] measured campaign outcomes (e.g., social outcome: behavioral change). For example, Arezo et al. [6] identified sanitary education (e.g., adequate disposal of infected offal) as a program output, but it was not analyzed with respect to specific outcomes. Van Kesteren et al. [34] went further to measure behavior as a social outcome, in terms of dog owner PZQ administration. While semi-structured questionnaires measured CE disease knowledge, it was not linked to a specific program output (e.g., public health education) or tested for significant correlations to specific program impacts (e.g., decreased CE prevalence).
Additionally, a study in China [45] linked communities' level of knowledge to achieving high PZQ dosing rates in the previous year (p.5). However, significant correlations of deworming behavior to a specific program output (e.g., public health campaign) or prevalence, were not tested. Thus, the final impact could not be concluded. Similarly, Yu et al. [39] measured behavioral change, by measuring pre and post-program de-worming coverage, but did not test for significant correlations to a program output or impact.
Solomon et al.'s [42] study went further, in terms of evaluating a specific health education campaign, which was targeted to women in Kenya, who spent most of the time at home with dogs. Prevalence reduction was attributed to health education producing positive behavioral change, such as appropriate offal disposal. Given that education programs targeted women and statistically significant changes in prevalence were linked to gender (Pearson χ2, P < 0:001; ordinal Somers' d tests, P < 0:01) ( [42], p.591), this correlation seems valid. However, confounding variables, such as literacy rates, previous reductions in dog population size, and increased PZQ treatment, introduced confounding bias.

Confounding Variables.
Numerous studies did not measure behavior as an outcome of public health education outputs, which essentially introduced confounding bias. Consistent positive behavioral change is essential to minimize CE transmission. The behavior can potentially enhance or constrain the effects of animal or human health outcomes and, ultimately, the impact of disease prevalence. For example, feeding dogs infected offal is an established transmission pathway [3] that can constrain the effects of de-worming dogs. Indeed, Van Kesteren et al. [45] concluded that health education had the potential to decrease CoproELISA prevalence, by inciting positive behavioral change, such as increasing PZQ administration. Whilst Larrieu et al. [2] concluded that programs with a combined education component enhanced the positive effects of canine anthelmintic treatment and sheep vaccination coverage ( [2], p.5).
In addition to behavioral change, only one study [46] accounted for seasonal climatic variations. It is essential, as at 4°C, E. granulosus eggs have a lifespan of ≥300 days compared to 2-14 days at 37-39°C ( [3], p.438). Indeed, Amarir et al. [46] reported that calendar time and location had significant effects (P < 0:001) on CE prevalence in stray and owned dogs in Morocco (p.440).
Finally, there were evident disparities in skill levels when conducting diagnostic tests or administering preventive treatments, such as PZQ [6,34,42,43,46]. Van Kesteren et al. [34] acknowledged the reality that leaving dog owners  were primary definitive host (v) Dogs played important cultural and productive roles as shepherds of livestock. In ethnic communities in Western China, Buddhist religion forbids the killing of any animals (vi) Dogs' proximity to humans and water sources increased transmission (vii) Barriers to testing roaming stray dogs due to inaccessibility of remote and widely distributed terrain Limitations: (i) No sample size calculation for sheep (ii) No sample size calculation or population size for owned or stray dogs (iii) No detail about sample collection methods (iv) Mentioned community public health education, but no measures of program outcomes or correlation to significant differences in prevalence (v) PZQ administration method not standardized for domestic vs. stray dogs (vi) Each village appointed one resident to de-worm dogs and deliver health education. No mention of how resident was selected or trained (vii) Post-mortem: visual assessment could introduce subjective bias if variations in assessor skill or method (viii) The finding of increased incidence as sheep age was not clearly supported by the study's cited reference [44] 10 Journal of Zoological Systematics and Evolutionary Research  (3) [2] reported that the calculated prevalence rates from arecoline tests were not statistically significant ðP value = 0:08, P > 0:05Þ compared to CoproELISA tests (P = 0:04, P < 0:05). Previous research has revealed that when the prevalence remains high, CoproELISA sensitivity, alone, may be used for accurate CE diagnosis [47]. Nevertheless, arecoline was utilized, as both a program output (e.g., a treatment or diagnostic method) and a research method [2,42,43,46].
3.6.2. Human Intermediate Hosts: Diagnostic Ultrasound and Treatment. Assessing the impact of program surveillance, using US, was the primary focus of the four studies that measured human CE prevalence [2,6,39,42]. In contrast to dogs, it was generally concluded that CoproELISA serology was less sensitive than US, although US had limited sensitivity in detecting pulmonary cysts [6,42]. For example, 198 CE cases were US identified compared to 76 using serology ELISA ( [42], p.588). Larrieu et al. [2] substituted serology (double diffusion 5, ELISA) tests for US, due to higher sensitivity.
Surveillance measures, such as US, are essential to identifying rising incidence or prevalence rates for endemic diseases, like CE. In CE endemic countries, such as Mongolia, when programs focused more on human surgical treatment than preventative dog management, they have resulted in under reporting and under diagnosis of CE cases ( [15], p.64). While complete CE eradication is difficult in endemic countries, an over emphasis on postinfection control and treatment measures may inevitably lead to missed opportunities for early disease prevention. This may lead to increased demand and costs for surgical or medical treatments.
Additionally, one study acknowledged patient hesitation to undergo surgical treatment [42]. However, no study  (i) Failed to deliver 3 vaccination doses to 40% of sheep who escaped or failed to present at vaccination times (ii) Selection bias: non-randomized sample selection of children from one school (iii) Sample size calculations for each species treated prevalence separately. However, calculation based upon paired data (discordant pairs) for each species may be more suitable for two prevalence values in a before and after study Confounding variables: (i) No clear distinguishment between owned and stray dogs. Only reference to "roaming dogs" (ii) Older, unvaccinated sheep, purchased from areas outside vaccination zones, may limit program outputs (iii) Sheep housed in mixed ages and sexes may have affected transmission and prevalence rates (iv) Variation between health workers or dog owner PZQ administration (v) Program outcomes not tested (e.g., health education effects on behavioral change) (vi) Goats excluded from vaccination program, as low prevalence in goats (unpublished study) and little local slaughter. However, one CE infected dog lived on a property that exclusively housed goats 12 Journal of Zoological Systematics and Evolutionary Research  [2,6,7]. This would likely result in higher post-operative complication rates. Thus, initiating early prevention methods has potential benefits of reducing prevalence rates and associated treatment costs and/or medical complications.  (ii) Open-ended questions delivered verbally in Kyrgyz, by native speaker (Bermet Mytynova) (iii) Answers to "CE causes" classified as "correct," "incorrect," or "partially correct." Correct answers based 15 Journal of Zoological Systematics and Evolutionary Research 3.6.3. Program Sustainability. A major finding across program evaluations was the unsustainability of six weekly PZQ treatments. Larrieu et al. [2] explained that programs using dog de-worming alone, often failed globally due to logistical constraints of sustaining 100% coverage; up to eight times per year, in remote areas (p.6). In Kyrgyzstan, it was concluded that six weekly PZQ intervals were not practical, due to funding and human resource constraints ( [34], pp. 9,16). Two studies recommended three to four monthly intervals to reduce canine and livestock prevalence rates to ≤1% within 10-15 years ( [45], p. 6, [34]). Nearby, in northwest China, six weekly PZQ treatments were also    Ministry of Health, hospitals, zoonotic health centers, and veterinary departments Primary qualitative research (interview responses, n = 15 ): (i) 79.3% not aware of the national plan for infectious disease control (ii) 44.8% stated limited funding was a challenge for control (iii) Two participants claimed that the government and international agencies' research funding reduced monetary constraints (iv) 58.6% concluded that there are no CE field control efforts (v) 75.9% received no training associated with CE in the last 5 years (e.g., respondents from the WHO Mongolia Office, Mongolian government sectors, local hospitals, veterinary institutes, and laboratories) (vi) Two clinical doctors stated diagnosis was based upon experience vs. following WHO-Informal Working Group on Echinococcosis (IWGE) ultrasound cyst staging Proposed challenges and solutions: (i) 8 proposed more government engagement (ii) 10 identified low public awareness (iii) 17 identified insufficient capacity (iv) 22 proposed establishing a national strategy, which encompassed routine disease surveillance and technical support from China Group discussions: (i) Limited nationwide disease surveillance and associated distribution mapping (ii) Poor management of stray dogs (iii) Inadequate diagnostic tools for humans and dogs (iv) Praziquantel and albendazole viewed as an obstruction to control (v) High drug costs Recommendations: (i) Mongolia led, China supported, bilateral cooperation. China prepares formal documents and aids in developing technical guidelines and standards Journal of Zoological Systematics and Evolutionary Research unsustainable, as there were issues obtaining dogs' weight, and with dosing logistics (e.g., dogs disliked taste, funding, remoteness, and skilled worker availability). Moving to Northern African, in Morocco, it was concluded that only two monthly PZQ de-worming intervals effectively controlled infective egg shedding in stray and owned dogs ( [46], p. 441). Furthermore, Yu et al. [39] attributed failed prevalence reduction in China, and the autonomous regions of Mongolia and Tibet, to the unsustainable de-worming of domestic canines, and controlling "wild canines" (p.2), such as foxes. Although studies [34,42,45,50] identified the sylvatic cycle, no program outputs were discussed, apart from surveying community knowledge of wildlife transmission pathways [45].
To address sustainability and a lag in vaccination flock effects, Larrieu et al. [2] and Van Kesteren et al. [45] suggested a combined program, which includes sheep vaccination, canine PQZ de-worming, and health education. Qian et al. [15] also recommended integrating CE with other neglected canine zoonotic diseases, such as rabies, to improve efficiency and reduce costs. Although excluded from this review, integrated zoonotic disease programs have potential cost-effective benefits, in terms of access to multiple health technologies ( [7,57], p.18, [58]). Indeed, it has been calculated that it would cost 30% more per dog treated separately for rabies, cystic echinococcosis, and visceral leishmaniasis, compared to an integrated program ( [31], p.7).

Discussion and Recommendations
4.1. Measure Multiple Program Outcomes. Cross-disciplinary research that focuses on measuring multiple social, health, and economic program outputs and outcomes, in correlation to changes in disease prevalence, is essential. Most studies concentrated on measuring disease prevalence without obtaining an understanding of the specific program outputs and outcomes that caused changes. Without this understanding, future programs may fail to reproduce successful outputs or improve upon existing ones.
Failure to assess the link between social outcomes, such as positive behavioral change, with disease prevalence may be a disciplinary issue, as behavioral studies are often confined to the psychological and social sciences. Measuring behavior, pre-and post-health education campaigns is essential. However, consistently measuring sanitary practices, such as hand hygiene, may be difficult in remote communities. Additionally, identifying if access to clean water is a location specific constraint is essential. This would also highlight an area in need of program development and funding.

Regular Training: Standardized Administration and
Measurement of Program Outputs (PZQ de-worming). Stratified sampling and, if funding permitted, engaging an external statistician to analyze data may have improved the methodological reliability of studies. For example, separating dogs based upon owner versus health worker PZQ administration would control for confounding bias and increase the (ii) China-Mongolia cross sector collaboration (e.g., disciplines: medicine, veterinary, parasitology, and epidemiology) and government departments (e.g., public health, quarantine, and animal health) to share technology, technical and project management skills, information, and resources (iii) Integrating CE control with other dog transmitted, neglected zoonoses (e.g., rabies), to improve efficiency and reduce costs (iv) Cross-sectional population survey to create a disease baseline (v) Dog management: de-worming and registration (vi) Financial support (vii) Public health education (viii) Strengthen existing surveillance systems (ix) Increase physical and economic access to affordable drugs Limitations:  [6]). Finally, Solomon et al. [42] specified that a health education campaign was initially delivered by organizational officers, but community members were subsequently trained (p.588). Common to all these studies was that resident skill set level was unclear and ongoing supportive training was not regularly provided. Variations in skill set compromised program sustainability and study reliability, due to the increased likelihood of incorrect PZQ administration, irregular dosing intervals, dog owner recall bias, and potential inaccuracies in health information. Additionally, an absence of regular training may extend to professional program workers. Qian et al. [15] revealed that 75.9% of surveyed respondents from the WHO Mongolia Office, Mongolian government sectors, local hospitals, veterinary institutes, and laboratories reported not receiving CE training in the last 5 years (p.63). Thus, all included programs would benefit from ongoing training support for both local communities and organizational program workers.

Distinguish Between Stray and Owned Dog Populations.
The OIE recommends distinguishing between owned and stray dogs to accurately calculate and trace population size [59]. Although the role of stray dogs in disease transmission was considered [2,15,42,43], 70% of studies did not clearly make this distinction and/or administer PZQ to stray dogs. Only 20% calculated stray dog population size [45,46]. The importance of this distinction was captured in Morocco, where photographic records tracked the number of stray dogs, who were 14 times as likely to be infected with E. granulosus compared to owned dogs (odds ratio = 14, 95% CI: 6-30; P < 0:001) ( [46], p.439).
While the population of owned dogs is more readily calculated, using registration records or household surveys, measuring stray dog populations is essential to minimizing CE transmission. Roaming stray dogs may access common livestock, wildlife, and human environments [18][19][20]. A single dog can be the source of infection for 30,000 ha ([3],  However, data was presented outside these dates (e.g., canine cases (2004-2014) and "mean human prevalence" (2008-2014) (x) No controls or significance testing for prevalence changes in correlation to program outputs (e.g., education programs) and outcomes (e.g., behavioral change: improved sanitation methods; PZQ deworming) 20 Journal of Zoological Systematics and Evolutionary Research Higher prevalence in stray dogs attributed to free access to condemned organs from slaughterhouses and weekly markets (v) Monthly risk was lowest in group A (2 monthly intervals) compared to B (3 monthly) and C (4 monthly intervals). Infection risk highest in group C (vi) 2 monthly PZQ intervals for owned and stray dogs can effectively control shedding of infective eggs (vii) Season significantly (P < 0:001) associated with prevalence; (a) Reduced risk of infection during second sampling period, as dry and warm summer conditions decrease environmental survival of CE eggs (b) During colder, winter months, higher risk of infection, due to extended lifecycle of eggs. Increase also attributed to increased livestock slaughter during winter (c) Interactions between time and dog type (stray or owned), and time and site not significant (P = 0:9, P > 0:05) Barriers: (i) Primary transmission cycle: stray dogs in urban areas; roaming or shepherding dogs in rural areas (ii) Dogs are kept as house and livestock guards, often in close contact with owners; especially women and children (high risk demographic) (iii) Home slaughtered livestock primary source of infection for owned dogs (iv) Condemned offal from slaughterhouses or weekly markets (souk) source of transmission for stray or roaming dogs 21 Journal of Zoological Systematics and Evolutionary Research p. 438). Although labor-intensive, the OIE [59] recommends using a refined marking method or a wildlife biology, markrecapture method, to calculate stray dog populations. The marking method entails accounting for daily differences in dogs' distribution (e.g., variations in weather, access to food, shelter, and human activity). Dogs are temporarily captured and marked using a standardized method (e.g., a distinctive collar, paint smudge, or ear tag). After a few days, plotting the daily number of marked dogs against the accumulated total of marked dogs produces a representative value for population size. The second wildlife biology method entails initially marking and releasing dogs within a defined area. The number of marked and unmarked dogs is then calculated by distant re-observation of the same area. To calculate a total population estimate, the number of dogs initially marked and released is multiplied by the total number of re-observed dogs [59]. This is all divided by the number of marked dogs identified during re-observation. Notably, there is an increased risk of zoonotic disease transmission to researchers, and ethical concerns, in terms of the stress caused during the capture and marking process.
4.4. Coproantigen ELISA Tests: An Efficient All Rounder for Canine Prevalence. As identified, Coproantigen ELISA was commonly used to diagnose prevalence in both stray and domestic dogs [2,6,34,45]. It has reported advantages, such as the ability to detect antigen 5-10 days post-infection and treatment, when >100 parasites are present ( [3], p. 437 [60]). In the context of endemic low to middle income countries, this test is suitable in terms of achieving adequate sensitivity and affordability. In addition, CoproPCR [45,46] was utilized to calculate canine CE prevalence, as a standa-lone verification test or sometimes coupled with WB [6] or arecoline tests [2]. Although more costly, and limited to research, CoproPCR tests offer a higher sensitivity and, unlike CoproELISA, can differentiate taeniid spp. from E. granulosus or E. multilocularis species ( [3], p. 437).

Livestock
Intermediate Hosts: More Studies on Vaccination Efficacy. Only two studies [2,43] evaluated programs in reference to livestock vaccinations within endemic countries. Future longitudinal research is necessary to clearly establish the effectiveness of vaccination programs, in terms of multiple outcomes and significant differences in prevalence, not only in sheep, but other susceptible species, such as goats and cattle. However, relative cost constraints and the cited [2,45] unsustainability of booster and annual Limitations: (i) No sample size calculation (ii) Study design and methods unclear: a self-identified longitudinal study, which used odds ratio to identify risk factors, such as being a stray vs. owned dog. However, unlike case control studies that focus on preexisting disease cases, odds ratio seemed to be applied like a relative risk ratio within a prospective cohort study, to calculate disease incidence. This made reported measurements of incidence vs. prevalence confusing (iii) Ethical issue: arecoline can cause adverse reactions in young or old dogs and is generally prohibited in pregnant dogs. No signalment details for dog ages or sex (iv) Difficult to identify exact prevalence values (v) Pre-program prevalence ranges: unclear if this was an average across all three sites (vi) Could not source cited articles for fecel floatation or microscopic tests. However, use of fecel flotation to detect E. granulosus is not highly sensitive, as Taenia and Echinococcus eggs are morphologically indistinguishable using fecel float ( [47], p.123) (vii) PCR methodology described in reference to several secondary studies. Difficult to identify which parts of each methodology were utilized, especially in reference to PBS and DNA extraction methods from Mathis et al. [48] and Abbassi et al. [49] Confounding variables: (i) Different exposure times for groups A, B, and C, due to different administration interval periods (e.g., variable time spans from 4-8 months at three sites). Instead, having three different interval groups per site, sampled at the same time, would have controlled for seasonal variation (ii) Research bias: researchers or statisticians not blinded to de-worming intervals for each group (iii) Not clear if dog owners or skilled worker administered PZQ (iv) No specification about whether skilled or unskilled personnel collected fecel samples, analyzed them for CoproPCR, fecel flotation, or microscopy (v) Does not distinguish between stray versus owned dogs 22 Journal of Zoological Systematics and Evolutionary Research enforcement agencies, educators)" [59] is essential to producing program indicators that reflect multi-stakeholder interests. Saadi et al. [50] and Qian et al. [15] conducted multi-stakeholder analyses, using qualitative field research in Morocco, China, and Mongolia. Qian et al. [15] conducted in depth interviews with representatives from the WHO Mongolia Office, Mongolian government sectors, local hospitals, veterinary institutes, and laboratories. Participants identified the benefit of bilateral China-Mongolia cooperation, in terms of joint research and training. To optimize this benefit, Qian et al. [15] suggested interdisciplinary collaboration between the fields of medicine, veterinary science, parasitology, epidemiology, and government departments. Indeed, a key WHO recommendation is adopting a "One Health" approach that entails medical and veterinary collaboration [16]. Saadi et al.'s [50] study also concluded that stakeholder relationships promote the sharing and development of knowledge, resources, and new technologies (e.g., diagnostics or treatment methods) (p.6). Thus, sustained stakeholder collaboration proves an important means to ensuring program sustainability. More specifically, acknowledging the interest of multiple state and non-state stakeholders has been viewed as important in addressing a "democratic deficit" ( [61], p.353, [62], p. 778-9) within self-regulatory governance models. This "deficit" is linked to actors governing themselves without equally representing the interests of public, private, and civil society actors. Indeed, in Morocco, Saadi et al. [50] highlighted different levels of power and interest. Internal government stakeholders wielded the most power but considered CE programs a lower priority when compared to external actors (e.g., WHO, OIE, physicians or veterinarians) who possessed less political agency. This web of stakeholders transcends animal and human health spheres and highlights the difficulty in representing multiple interests.
A reported neglect of stakeholder interests ( [50], p.5) represents why measuring social outcomes is essential to program analysis. A primary social outcome indirectly identified by three studies ( [39], p.146; [42,43]) was cultural competency. Local cultural and religious beliefs may not always align with program measures. For example, Yang et al. [43] explained that within ethnic communities in North-Western China, Buddhist religion forbids the killing of any animals (p.358), which complicates dog population control measures. Indeed, Yu et al.'s [39] study, in China, Tibet, and Mongolia, also concluded that cultural acceptance of roaming stray dogs, over population control, is common ( [39], p.146).
Participatory versus top-down programs, which enable open communication, are essential to fostering mutual respect between medical and veterinary practitioners with communities. Achieving this entails avoiding an overreliance on divisional dichotomies (e.g., developed/developing, core/periphery, and modern/traditional) [63] to explain the world through one dominant western lens [64,65]. This simply reproduces a hegemonic form of medical knowledge that may disregard local knowledge. Local cultural beliefs and practices regarding dogs' roles in communities will continue to shape how people respond to prevention measures.
Indeed, for people of Turkana, Kenya, dogs played multiple roles, which included acting as cattle rustlers, protectors against wildlife, as pets, and family members [42]. Thus, achieving mutual understanding ultimately increases the chance of positive behavioral change, which can reduce CE transmission.
To formally represent multiple stakeholders, Saadi et al. [50] recommended creating a national central office, which contains representative stakeholders who prioritize CE programs. However, most internal stakeholders who wielded relatively higher power considered CE programs a low priority. Thus, including both external and internal state actors is recommended to overcome these power dynamics. For example, a veterinarian could represent animal health interests; a human medical physician for public health interests; a government official for state interests; a representative for breeders, production animal, and slaughterhouse worker interests; and a community member for cultural interests.

Conclusion
To conclude, ten studies situated across three major CE endemic regions (Central Asia, Africa, and South America) were selected using pre-defined inclusion/exclusion criteria from the four databases. Common research limitations included no sample size calculation and numerous confounding variables, which limited result validity. One major confounder was the absence of standardized PZQ administration methods for stray and domestic dogs. Nevertheless, the studies produced useful comparisons, in terms of program barriers faced in remote, low to middle income countries. Generally, more research and/or programs are recommended within African and South American regions.
More broadly, future research and program development are essential, in terms of focusing more on prevention versus diagnostics and treatment. Future research recommendations included the following: measuring the effects of program outputs, in terms of multiple program outcomes, especially the social outcome of behavioral change. Going further, to test these outcomes for significant correlation to CE prevalence is essential. Additionally, research on livestock prevention measures to clearly establish the practicality and benefit of program outputs, distinguish between stray versus owned dogs to calculate population sizes and clearly target program outputs, and identify the transmission role of waterways and sanitation, is essential.
Finally, the key program recommendations include the following: regular local community training to deliver sustainable program outputs (e.g., PZQ), and conduct data collection to monitor CE prevalence; organized representation of multi-stakeholder interests; clearer or standardized guidelines around PZQ and livestock vaccination administration frequency; programs that encompass multiple prevention methods (e.g., dog de-worming, public health education, and sheep vaccination) and integrated canine disease management; and enhanced veterinary-human medical training and resource sharing to improve program sustainability.

Data Availability
Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Conflicts of Interest
No conflict of interest to declare.

Funding
As this was a Critically Appraised Topic (CAT) review, no university funding was provided/necessary. However, The University of Sydney provided funding for publication costs.