Strongyloidiasis in Ethiopia: systematic review on risk factors, diagnosis, prevalence and clinical outcomes

Background Strongyloidiasis is a gastrointestinal infection caused by the parasitic nematode Strongyloides stercoralis. It is estimated to infect up to 370 million people globally and is predominately found in tropical and subtropical areas of socioeconomic disadvantage. Main body This systematic literature review identified studies published in the last ten years on the risk factors, diagnosis, prevalence and/or clinical outcomes of strongyloidiasis in Ethiopia. The prevalence of S. stercoralis ranged from 0.2 to 11.1% in adults, 0.3% to 20.7% in children, 1.5% to 17.3% in HIV positive adults and 5% in HIV positive children. The identified studies primarily used microscopy based techniques that potentially underestimated the prevalence four fold compared with serology and PCR. Strongyloidiasis in children presents a particularly significant issue in Ethiopia as children often presented with anemia, which is associated with impaired mental and cognitive development. The most significant risk factor for strongyloidiasis was HIV status and although other risk factors were identified for helminth infections, none were statistically significant for S. stercoralis specifically. Several studies detected S. stercoralis in dogs and non-biting cyclorrhaphan flies. However, future research is needed to explore the role of these reservoirs in disease transmission. Conclusions This review demonstrated that strongyloidiasis is an overlooked and neglected disease in Ethiopia. There is a need for a systematic approach using a combination of molecular and serology based diagnostic methods to ascertain the true incidence and burden of strongyloidiasis in Ethiopia. Further research is also needed to break the cycle of transmission by identifying environmental reservoirs, risk factors and exploring the potential for zoonotic transfer. Electronic supplementary material The online version of this article (10.1186/s40249-019-0555-3) contains supplementary material, which is available to authorized users.


Background
Strongyloidiasis is caused by infection with the parasitic nematode worm, Strongyloides stercoralis, S. fuelleborni or S. fuelleborni kelli [1,2]. Symptoms of infection range from asymptotic to non-specific gastrointestinal complaints [2,3] and distinctive form of cutaneous larva migrans, larva currens [4]. The infection can remain undetected and undiagnosed for many decades [5,6]. However, when an infected person undergoes steroidal or immunosuppressive treatment, the worm infection undergoes hyperinfection, leading to enormous numbers of the parasite, which can then disseminate and move to other organs, a condition that is almost always fatal [7].
Strongyloidiasis has been assumed to be a disease associated with tropical or subtropical regions, and is recognised as a neglected tropical disease. However, it has been argued that strongyloidiasis should be described as a disease of disadvantage, as it is primarily a disease from developing countries and from poorer areas within developed countries [8]. Estimates of infection rates globally range from 50-100 million [3,4] to more than 300 million [9]. The lower estimate is likely to be an underestimate as the disease is often undiagnosed, either because it is not looked for [3,10] or because of difficulties with diagnosis [2,11].
Ethiopia's economy is one of the fastest growing in the world [12] although it is also one of the poorest countries in Africa [13]. The Ethiopian health care sector is currently underfunded by both global and regional standards [14] and housing and other environmental health hardware components in many areas is substandard in both rural and urban areas [15,16]. The lack of health care and poor environmental health conditions combine to make the country susceptible to parasitic infections, including strongyloidiasis. Here we evaluate the prevalence, risk factors and clinical outcomes for strongyloidiasis in Ethiopia reported in the literature using a systematic approach.

Search strategy
The databases Scopus (n = 46) and Web of Science (n = 44) were searched for articles written in English over the last ten years containing the keywords Strongyloides OR strongyloidiasis OR S. stercoralis OR S. fuelleborni AND Ethiopia OR Ethiopian. Figure 1 presents the systematic approach to article inclusion or exclusion. Articles were screened by reading titles and abstracts and initially excluded if they did not refer to specifically to S. stercoralis or if they were review articles. Article were further screened by reading abstracts and full articles and excluded if they did not investigate the prevalence of strongyloidiasis, or describe clinical presentations or potential environmental sources. As the aim of this systematic review was to provide an overview of the state of knowledge relating to strongyloidiasis in Ethiopia, all articles which met the inclusion criteria were included regardless of any flaw in study design. Limitations or bias identified are discussed in the discussion section.

Prevalence
A total of 40 publications published in the last 10 years were identified that investigated the prevalence of strongyloidiasis, described clinical presentations or potential environmental sources. Table 1 presents the 27 publications that examined the prevalence of strongyloidiasis in   [33]. This study did not investigate the general population but rather patients from Gambo Rural Hospital presenting with diarrhea. The omission of this study increased the overall prevalence in healthy adults to 2.5%. It is also important to note that in all of these studies investigating HIV negative adults, Strongyloides was detected using microscopy based stool examination techniques, which is likely to significantly underestimate the prevalence due to false negative results.
In HIV negative children the prevalence of strongyloidiasis ranged from 0.3% to 20.7% and the overall combined prevalence from 8 studies was 8.6% (298 positive out of 3453 children). However, there were significant variations in incidence depending on the diagnostic technique used. The overall prevalence in children using microscopy based stool examination techniques was 4% (from seven studies, two of which concurrently used PCR or serology) ( Table 1); whereas the overall prevalence using PCR detection from stools or serology was 17.5% (three studies, two of which concurrently used microscopy). The two studies which tested using more than one detection method found 5% prevalence using microscopy compared with 17% using serology [19], and 4% using microscopy compared to 13% using PCR [26]. This suggests that microscopy based techniques could potentially underestimate prevalence by up to four fold compared with serology or PCR. This is consistent with the findings of Van Kesteren and Wojciechowski [19] that concurrently tested faeces using microscopy examination of stools and serology and found positives in 4.8% and 16.5% of patients respectively. Another study by Amor et al. reported detection of 3.5% by stool, 12.1% by Baermann and 13.4% by PCR [26]. The most common diagnostic technique by far was stool examination using the formalin-ether concentration technique, followed by the wet mount, Kato-Katz and Baermann with only one or two studies each using saline smear mount, sedimentation, agar, PCR or serology ( Table 1).
The highest prevalence seen using microscopy based stool examination was in HIV positive adults. The prevalence ranged from 1.5% to 17.3% and the overall combined prevalence was 11%. There was only one study of pediatric HIV patients which found the incidence to be 5% (2.5% in highly active antiretroviral therapy [HARRT] initiated and 6.9% in HARRT naïve patients) [31].

Risk factors
The most significant risk factor for strongyloidiasis identified was HIV status (Table 1). Four studies compared the prevalence of strongyloidiasis in HIV positive and HIV negative individuals and the incidence was 5-20 times higher in HIV positive individuals compared with HIV negative [20,41,43,45].
Two studies reported that the prevalence of strongyloidiasis in otherwise healthy individuals was slightly higher in males compared to females [17,18,33] and that higher incidence was seen in rural areas compared with urban areas [34] and in farmers [17,18]. Three studies investigated food handlers; however, the overall prevalence was 2.7%, which is not significantly higher than the general population [36,40,44].
Several studies identified additional risk factors associated with the prevalence of intestinal parasites. Not wearing shoes, not washing hands [21,34], not trimming fingernails, or having dirt under the nails [34] were statistically significantly associated (P < 0.05) with intestinal helminths (including Strongyloides). Eating unwashed/ raw fruit, open field defecation, and living in a rural area was statistically significantly associated (P < 0.05) with intestinal parasites (including Strongyloides and other parasites) [31]. However, there were no statistically significant risk factors associated with strongyloidiasis specifically. This could be a result of the small sample sizes due to the lower prevalence of strongyloidiasis. It also highlights the need for more systematic approaches to epidemiological studies investigating the prevalence and risk factors for strongyloidiasis. Table 2 presents studies describing clinical manifestations of strongyloidiasis in Ethiopian individuals. The most common clinical presentation was anemia, which was observed in immunocompetent adults and children and HIV positive children. Immunocompromised individuals presented with a range of symptoms and clinical manifestations including eosinophilia, fever, vomiting, hematemesis, diarrhea, abdominal pain, bacteremia, sepsis, cough, respiratory distress, chronic obstructive pulmonary disease, hypoxemia, diffuse alveolar hemorrhage and meningitis (Table 2). Two studies reported cases of hyperinfection, one of which was fatal [51,52]. Another study by Nadir & Zimhony [50] reported eight cases of strongyloidiasis in AIDS patients, seven of which were fatal. The diagnosis of strongyloidiasis in these cases was complicated by negative serology results but diagnosis was confirmed through PCR and microscopy. Negative serology results were also observed in two reported cases of strongyloidiasis in immunocompetent children and diagnosis was achieved through stool examination [47,48,50].  14-year-old Ethiopian girl adopted in Italy Severe anemia (hemoglobin 4.9 g/dL) and a lung nodule Stool examination (qPCR positive (primer information not provided) but negative for formalin-ether microscopy)

Clinical manifestations
Serology positive (IgG antibodies) [46] 13 months old Ethiopian adoptee in Canada Eosinophilia and bloating Stool examination and serology (serology was negative -antibody information not provided) [

Environmental sources
There were four studies that detected S. stercoralis in potential environmental sources. Two studies detected S. stercoralis in dog faeces [55,56] and two in non-biting cyclorrhaphan flies [57,58], all of which used microscopy based stool examination techniques. A study conducted in Hawassa examined 448 dogs and found 31% were positive for S. stercoralis [56]. Another examined 384 pet dogs and 46 stray dogs in Bahir Dar and found 30% and 46% positive respectively [55]. A study in Woreta examined 6530 non-biting cyclorrhaphan flies and found the overall presence of S. stercoralis to be 1.7%. This included 12% of Chrysomya rufifacies, 16% Musca sorbens and 34% Lucilia cuprina. However, Musca domestica, Calliphora vicina, Chrysomya bezziana and Wohlfahrtia magnifica were all negative [58]. These findings support a previous study conducted in Addis Ababa which examined 9550 non-biting cyclorrhaphan flies and found that 0.1-0.2% of C. rufifacies and 0.6% of M. sorbens were positive for S. stercoralis, whereas L. cuprina, M. domestica, C. vicina, Sarcophaga spp., and Wohlfahrtia spp. were all negative. Interestingly, S. stercoralis positive flies were only found in defecating areas or garbage and were not found near the butchery or market [57].

Discussion
This systematic review revealed that the prevalence of S. stercoralis in HIV negative Ethiopians ranged from 0.2 to 11.1% in adults and 0.3% to 20.7% in children and the calculated overall prevalence was 0.5-2.5% and 8.6% in adults and children respectively. This is comparable with other studies from Sierra Leone, Côte d'Ivoire and Sudan, which found the prevalence of S. stercoralis to be less than 5% [59][60][61]. However, it is lower compared with other studies from Angola, Nigeria and Ghana, which found the incidence of strongyloidiasis to be above 15% [25,62,63]. The differences in prevalence observed in different studies may be due to different laboratory techniques, study populations, geographical factors, or economic status. In studies exploring the prevalence in children a common clinical presentation identified was anemia (Table 1). This suggests that undiagnosed strongyloidiasis in children may have significant long term consequences as iron deficiency and anemia in early childhood is associated with impaired cognition and learning ability [64]. This presents an issue of significant public health, economic and social concern for Ethiopia as infection in children can adversely affect physical, mental, education and overall societal development [65]. In this review, HIV positive individuals represented the highest risk group for S. stercoralis infection in Ethiopia. This finding supports several previous studies conducted across the globe [66,67]. It is interesting to note that the increased risk of strongyloidiasis in HIV positive individuals is not associated with an increase incidence of disseminated or hyperinfection [68]. Only one case of hyperinfection in a 36-year-old HIV positive individual was identified in this review [52]. However, a case report of strongyloidiasis in eight AIDS patients illustrated the severity of this disease in immunocompromised individuals with seven out of eight cases resulting in fatalities [50]. A complication of strongyloidiasis in HIV positive and immunocompromised individuals is the observation that serology can be falsely negative [50,69]. Previous global studies have also identified human T-cell lymphotropic virus type 1 Infection (HTLV-1) and alcoholism as risk factors for strongyloidiasis [70]; however, this was not seen in the Ethiopian studies. Although the absence of these risk factors could be attributed to a lack of a systematic approach to epidemiological surveillance of strongyloidiasis in Ethiopia.
There was a significant discrepancy in the prevalence observed in Ethiopia based on the diagnostic methods used in a study. Comparisons between studies found that the microscopy based techniques could potentially underestimate the prevalence by four fold compared with serology or PCR. As the majority of studies identified in this review used microscopy based techniques the true incidence of strongyloidiasis could be much higher. This supports a recent study that suggests that S. stercoralis infection could be overlooked and neglected in Ethiopia [27]. Similarly, a review of the global perspectives of strongyloidiasis indicated that the disease was currently underestimated in many countries, but despite this underestimation, prevalence is steadily increasing [71].
In Ethiopia, a higher prevalence of S. stercoralis was recorded in farmers and in rural populations. These findings are supported by studies conducted in Cambodia and China which identified rural populations to be at greater risk for strongyloidiasis [72,73]. This could be due to greater environmental contamination by the larvae of S. stercoralis and conditions supporting their survival. However, there is a need for environmental sampling to confirm this assumption and to identify the main environmental reservoirs. Global studies have also indicated that tropical and subtropical settings in areas of economic disadvantage provide ideal conditions for transmission dynamics of S. stercoralis [71,74]. This highlights the needs for research identifying the best practice in managing these environmental systems to break the life cycle and transmission dynamics of this parasite.
There were several studies which demonstrated the incidence of strongyloidiasis in Ethiopian food handlers was not greater than that of the general public. This supports the findings of a study in Malaysia which detected S. stercoralis in water samples that were used in the production of different vegetables (pegaga, kesum and water spinach) suggesting that vegetables and herbs may be a source of strongyloidiasis in this region [75].
Two studies demonstrated the prevalence of S. stercoralis in Ethiopian dogs [55,56]. This is noteworthy given recent studies that have demonstrated the potential zoonotic transmission of S. stercoralis from dogs to humans [76,77]. This incidence of S. stercoralis in Ethiopian dogs should inform future strategies for the control of strongyloidiasis. This is particularly relevant to the argument that mass drug administration schemes to treat human should consider concomitant treatment of dogs [78]. The two studies that detected S. stercoralis in nonbiting cyclorrhaphan flies [57,58] provide further insight into the potential transmission and environmental reservoirs of this parasite. This association with flies found in defecating areas and garbage supports the argument that strongyloidiasis is associated with improper sanitation hardware [74] although there is a need for further research to explore survival and transmission from these sources. It is also significant when considering the argument made by Gove [79] almost thirty years ago that the most effective approach to controlling Strongyloides infection is to control the environmental sources.
The complied information on clinical manifestations suggests that strongyloidiasis commonly presents with anemia, diarrhea, vomiting and eosinophilia. It provides a body of evidence for control campaigns or clinical management schemes to critically consider high-risk groups, such as HIV positive children, who may benefit from routine testing and/or de-worming of helminth infection [80]. Currently there are no government programs specifically targeting the surveillance, treatment and intervention of strongyloidiasis in Ethiopia. However, the Federal Ministries of Health and Education in collaboration with different international organizations deliver deworming treatments to children to tackle soil-transmitted helminthiases (intestinal worms) and schistosomiasis (bilharzia). These treatments programs are being scaled up each year and have the target of reaching 26.1 million children annually (80% of all at-risk children) by 2020 [81]. However, to be able to evaluate the effectiveness of these deworming programs there is a need for a systematic approach using a combination of molecular and serology based diagnostic to ascertain the true incidence and burden of strongyloidiasis. Currently there is not enough information to inform medical practitioners and public health policy advisors. For example, there are limited reports from the Eastern and South Western parts of Ethiopia, although there are many hospital-based studies were reported from northern and southern part of the country which were mainly conducted by universities in the nearby areas. In addition to this, there are very few reports investigating the role of domestic animals and environmental reservoirs in the transmission of this disease.

Conclusions
This review identified that strongyloidiasis is a potentially overlooked and neglected disease in Ethiopia. Currently, there is limited information on the epidemiological situation of strongyloidiasis across the country and the studies that are available are based on microscopy techniques that underestimate the true incidence. The incidence of strongyloidiasis in children presents a significant issue for Ethiopia as infection is associated with impaired mental and cognitive development, affecting the education and societal development of an individual. In order to break the cycle of disease transmission future research is needed to identify the environmental reservoirs and routes of exposure. This includes the potential zoonotic capacity of Strongyloides from dogs and cats and information about survival within different reservoirs.