Strongyloides stercoralis infection in Ethiopia: systematic review and meta-analysis on prevalence and diagnostic methods

The genus Strongyloides is one of the soil-transmitted helminths that infect humans worldwide (Olsen et al., 2009). Strongyloides stercoralis and S. fuelleborni are the only two species that infect humans. Strongyloides stercoralis infection is prevalent across many areas of tropics and subtropics (Schar et al., 2013), whereas most S. fuelleborni human infections are prevalent in Africa (Schad et al., 1989). Strongyloides infection is a common problem in communities with poor personal hygiene, poor environmental sanitation and open defecation practicing areas (Abrescia et al., 2009).

The detection of larvae in stool is the major identification stage of the parasite (Siddiqui et al., 2001). The direct saline microscopy (DSM) is a very simple and rapid diagnostic method (Nielsen et al., 1987); however, it has poor sensitivity in S. stercoralis detection (Requena-Méndez et al., 2013). This is due to the fact that low parasite load and irregular larval excretion (Montes et al., 2010), and chronic low-intensity S. stercoralis infection (Schar et al., 2013) limit the sensitivity of traditional methods. As a result, misdiagnosis and underreporting of S. stercoralis infection by DSM is a common phenomenon.

Although better detection rate of S. stercoralis is obtained using one of the following: Baermann concentration technique (BCT), stool culture, Polymerase Chain Reaction (PCR), or a combination of these methods (Campo-Polanco et al., 2018), their limitations to apply as a routine diagnostic method in Ethiopia is a big challenge. This situation forced the health institutions to employ DSM method for the diagnosis of Strongyloides infection. As a result, under diagnosis and underreporting of the true prevalence of S. stercoralis infection in Ethiopia is a major problem (Terefe et al., 2019). Thus, the aim of this systematic review and meta-analysis was to provide an overview of the prevalence of Strongyloides infection by country and regional label and by diagnostic methods used in Ethiopia.

The PubMed, Google Scholar, and Science direct databases and Addis Ababa University repository were searched for articles written in English during the year 2010 to 2020 containing the keywords: “Strongyloidiasis” AND “Ethiopia” OR “Strongyloides AND “Ethiopia” OR “Strongyloides stercoralis” AND “Ethiopia” OR “Soil-transmitted helminths” AND “Ethiopia”. The electronic data search of studies was conducted from January to 30 June 2020. Identification, screening, checking the eligibility and the inclusion of the relevant literatures were done following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) (Fig 1). Articles were first screened to remove duplication. And then the articles were also screened by reading titles and abstracts and initially excluded if they did not specifically refer to S. stercoralis or if they were review articles. Finally, the articles were further screened by reading the full articles an excluded if they did not investigate the prevalence of Strongyloides infection.

Fig. 1

Inclusion criteria: All studies conducted in Ethiopian populations checking stool samples and diagnosed with DSM, Kato-Katz (KK), formol ether concentration techniques (FECT), BCT, culture, PCR or a combination these diagnostic techniques and got positive result for at least one individual among the study participants were included. Inclusion of literatures only from PubMed, Google Scholar, Science direct databases and Addis Ababa repository was the limitation. To minimize the risk of bias, publication bias assessment was done across studies.

Exclusion criteria: All articles dealing with Strongyloides infection in animals, soil, foreigners as study subjects in Ethiopia, non-stool samples, duplications, review articles, case studies, cohort studies and articles conducted before the year 2010 were excluded.

The data was extracted independently from each study and the pooled prevalence of S. stercoralis in Ethiopia, and across regions and by average prevalence by diagnostic method were also computed.

The meta-analysis was also performed using comprehensive meta-analysis 2.2 software (Biostat Inc., Englewood, NJ, USA). The pooled prevalence rate of S. stercoralis at country was calculated using a random-effect model at 95 % confidence interval (CI). In the subgroup analysis, the pooled prevalence and forest plot of S. stercoralis in the regions and diagnostic methods was also calculated. Additionally, separate meta-analyses were performed to compare the effect of diagnostic methods in the detection of S. stercoralis. Heterogeneity between studies in region and across diagnostic methods was assessed using Cochran (Q)-value, P-value and I² and visual inspection of the funnel plot. The level of statistical significance for all tests was set at P < 0.05. Publication bias was checked by funnel plot.

A total of 207 studies identified from PubMed, Google scholar, Science direct databases and Addis Ababa University repository. Forty three studies were screened and recorded after duplications removed. Finally, 43 studies were eligible after full text assessment and included in qualitative analysis (Fig 1).

Prevalence of Strongyloides stercoralis

A total of 43 studies having S. stercoralis reports which full-filled the inclusion criteria and a total of 78,959 study participants were involved. The overall prevalence of S. stercoralis among study participants was 1.82 % [1437/78959] (Table 1).

No	First Authors	Year of Pub	Region	Participant history	Sample size	No SS cases	Prevalence (95%CI)	Diagnostic method
1	Hailu T	2020	Amhara	Sch	844	127	15.05 [12.74-17.68]	FECT,STST,BCT,APC
2	Aramendia AA	2020	Amhara	>5 years	792	441	55.68 [52.14-59.17]	FECT,BCT,PCR
3	Getaneh F	2020	Amhara	Patient	67	2	3.0 [0.82-10.25	DSM, KK
4	Kuti KA	2020	Oromia	FH	198	8	4.04 [2.06-7.77]	DSM, FECT
5	Tsegay B	2020	SNNPR	Children	622	12	1.93 [1.11-3.34]	DSM,FECT
6	Menjetta T	2019	SNNPR	UN/student	13,679	41	0.30 [0.22-0.41]	DSM
7	Gemech A	2019	SNNPR	Prisoner	320	18	5.63 [3.59-8.72]	DSM, FECT
8	Alemu G	2019	SNNPR	Sch	351	7	1.99 [0.97-0.41]	DSM, FECT
9	Alemu G	2018	SNNPR	HIV	220	4	1.82 [0.71-4.58]	DSM, FECT
10	Gebretsadik D	2018	Amhara	HIV	223	1	0.45 [0.02-2.86]	DSM, FECT
11	Hailegebriel T	2018	Amhara	Sch	382	5	1.31 [0.48-3.21	FECT
12	Teklmariam D	2018	Oromia	Sch	280	4	1.43 [0.46-3.87]	FECT,KK
13	Mengist HM	2017	Oromia	Pregnant	372	1	0.27 [0.01-1.73]	DSM,FECT
14	Eshetu T	2017	Amhara	HIV	223	8	3.59 [1.68-7.21]	FECT
15	Feleke DG	2017	Tigray	Patient	7,663	47	0.61 [0.45-0.82]	DSM,FECT
16	Alemu M	2017	Tigray	Patient	427	8	1.87 [0.87-3.80]	DSM,KK
17	Hailegebriel T	2017	AA	Patient	351	43	12.25 [9.22-16.09]	DSM,FECT, BCT Culture
18	Abdi M	2017	Amhara	Sch	408	3	0.74 [0.25-2.15]	FECT
19	Derso A	2016	Amhara	Pregnant	348	6	1.72 [0.79-3.70]	FECT
20	Amor A	2016	Amhara	Sch	396	82	20.71 [17.01-24.97]	FECT,BCT, PCR
21	Shimlis T	2016	SNNPR	HIV	491	22	4.48 [2.90-6.81]	DSM,FECT
22	Shiferaw MB	2015	Amhara	Patient	464	5	1.08 [0.40-2.65]	DSM,FECT
23	Aleka Y	2015	Amhara	Patient	277	1	0.36 [0.06-2.01]	DSM,FECT
24	Gedle D	2015	SNNPR	HIV	305	5	1.6 4[0.70-3.78]	DSM, FECT
25	Ramos JM	2014	SNNPR	Patient	32,191	92	0.29 [0.24-0.35]	DSM
26	Mekonnen B	2014	AA	St/dweller	355	19	5.35 [3.45-8.20]	DSM, FECT, KK
27	Mamo H	2014	Amhara	Prisoner	236	6	2.54 [1.10-5.71]	DSM,FECT
28	Eriso F	2014	SNNPR	Sch	710	142	20.0 [17.16-23.17]	BCT
29	Mahmud MA	2013	Tigray	Sch	600	5	0.83 [0.31-2.05]	DSM,FECT,KK
30	Adamu H	2013	Oromia	HIV	378	1	0.26 [0.01-1.69]	DSM,FECT
31	Bayessa C	2013	SNNPR	Patient	6,342	73	1.15 [0.92-1.44]	DSM, FECT
32	Abera B	2013	Amhara	Sch	778	27	3.47 [2.40-5.00]	FECT, KK
33	Zeynudin A	2013	Oromia	HIV	91	6	6.59 [3.05-13.64]	DSM,FECT
34	Abate A	2013	Amhara	Patient	410	1	0.24 [0.04-1.36]	DSM, FECT
35	King JD	2013	Amhara	Children	2,338	5	0.21 [0.09-0.49]	FECT
36	Fekadu S	2013	SNNPR	HIV	343	12	3.50 [2.01-6.02]	DSM, FECT
37	Teklemariam Z	2013	Harari	HIV	371	15	4.04 [2.46-6.56]	DSM,FECT
38	Wogayehu T	2013	SNNPR	All age	858	51	5.94 [4.55-7.73]	DSM,FECT
39	Huruy K	2011	Amhara	Patient	384	12	3.13 [1.80-5.39]	DSM
40	Legese L	2010	Tigray	Sch	386	1	0.26 [0.05-1.45]	KK
41	Nyantekyi LA	2010	SNNPR	Children	288	2	0.69 [0.19-2.49]	FECT, KK
42	Getaneh A	2010	SNNPR	HIV	384	27	7.03 [4.88-10.04]	DSM, FECT, BCT
43	Belyhun Y	2010	SNNPR	Kid+ Mother	1,813	39	2.15 [1.58-2.93]	FECT
	Total				78,959	1,437	1.82 [1.73-1.92]

*AA = Addis Ababa, SNNPR = Southern Nations, Nationalities Peoples’ Region, Sch = School children, FH = Food handler, HIV = Human Immunodeficiency Virus, St = Street, SS = Strongyloides stercoralis, No = Number, Pub = Publication, UN = University

A relatively high prevalence (55.68 %) of S. stercoralis infection was recorded among participants age greater than five years (Aramendia et al., 2020) followed by (20.71 %) in schoolchildren of rural highlands of Amhara Regional State (Amor et al., 2016), (20.0 %) in SNNPR schoolchildren (Eriso F, 2014), (15.05 %) of schoolchildren in the Amhara Regional State (Hailu et al., 2020), and (12.25 %) in patients of health institution of Addis Ababa City (Hailegebriel et al., 2017) among studies conducted in Ethiopia (Table 1).

A very low prevalence of S. stercoralis; (0.21 %) in a community children (King et al., 2013), and 0.24 % in patients of Amhara Regional State (Abate et al., 2013), 0.26 % in schoolchildren of Tigray Regional State (Legese et al., 2010), and (0.26 %) in HIV cases in Oromia Regional State (Admasu H, 2013), and 0.29 % in patients (Ramose et al., 2014) was obtained from studies conducted in the country (Table 1).

In this review, the lowest prevalence of Strongyloides infection reported from a single study was 0.21 % by FECT (King et al., 2013), followed by 0.24 % by combination of DSM and FECT (Abate et al., 2013), 0.26 % by KK (Leegese et al., 2010) and by combining DSM and FECT (Adamu et al., 2013), and 0.27 % by DSM and FECT combination (Mengist et al., 2017) (Table 1).

Regarding regional reports relatively high prevalence of S. stercoralis 55.68 % (Aramendia et al., 2020) and 20.21 % (Amor et al., 2016), was reported using a combination of diagnostic methods in Amhara Regional State followed by 20.0 % among Schoolchildren in SNNPR (Eriso H, 2014), and 12.25 % among patients in Addis Ababa (Hailegebriel et al., 2017) (Table 1).

Among studies used single diagnostic methods, high prevalence (20.0 %) of S. stercoralis was recorded by BCT among schoolchildren in SNNPR (Eriso F, 2014) followed by 3.59 % S. stercoralis prevalence by FECT in HIV cases in the Amhara Regional State (Eshetu T, 2017) and 3.13 % prevalence by DSM among patients in Amhara Regional State (Huruy et al., 2011).

Using random effect analysis, the pooled prevalence of S. stercoralis in Ethiopia was 2.1 % (95 %CI: 1.20 – 3.60). The heterogeneity was high (Q = 4264.8, I² = 99.0 %, P < 0.001) (Fig 2).

Fig. 2

The studies were distributed symmetrically about the combined effect size that showed the absence of publication bias (Fig 3).

Fig. 3

From 43 studies, 16 (37.21 %) were conducted in Amhara regional state followed by 15 (34.88 %) in SNNPR. The number of participants was high 58,917 (74.62 %) and 9,076 (11.49 %) in the SNNPR and the Tigray Regional State, respectively. The pooled prevalence of S. stercoralis was relatively high in the Addis Ababa City (8.78 %) followed by (8.54 %) in the Amhara Regional State among regions. Low prevalence S. tercoralis infection among regions was recorded in Tigray Regional State (0.67 %) followed by (0.93 %) in SNNPR (Table 2).

Table 2

Name of the region	Number of studies [N]	Total examined [N]	SS Positive [N]	Pooled prevalence (95%CI)
Addis Ababa City	2	706	62	8.78 [6.85 – 11.17]
Amhara	16	8,570	732	8.54 [7.96 – 9.16]
Harari	1	371	15	4.04 [2.36 – 6.72]
Oromia	5	1319	20	1.52 [0.96 – 2.38]
SNNPR	15	58,917	547	0.93 [0.85 – 1.01]
Tigray	4	9,076	61	0.67 [0.52 – 0.86]
Total	43	78,959	1,437	1.82 [1.73 – 1.92]

*SS = Strongyloides stercoralis

Using random effect analysis, the pooled prevalence of S. stercoralis across the regions was 2.6 % (95 %CI: 0.80 – 8.20). The heterogeneity was high (Q = 1808.2, I² = 99.7 %, P < 0.001) (Fig 4). In this review, 37 (86.05 %) of the studies were conducted by DSM, KK, FECT or a combination these methods. High prevalence 44.02 % rate of S. stercoralis infection was recorded with a combination of FECT, BCT and PCR and followed by 20 % with only BCT and 15.05 % S. stercoralis prevalence with a combining FECT, STST, BCT, and culture diagnostic methods (Table 3). A low prevalence of S. stercoralis was traced 0.26 %, 0.31 %, and 1.20 % by the respective KK, DSM and FECTs (Table 3).

Fig 4

Table 3

Diagnostic methods	No of studies [N]	Total examined [N]	S. stercoralis Positive [N]	Pooled prevalence (95%CI)
DSM	3	46,254	145	0.31 [0.26 – 0.36]
KK	1	386	1	0.26 [0.05 – 1.45]
FECT	6	5,512	66	1.20 [0.94 – 1.53]
BCT	1	710	142	20.0 [17.16 – 23.17]1
DSM+KK	2	494	10	2.02 [1.10 – 3.68]
DSM+FECT	20	20,535	296	1.44 [1.28 – 1.61]
FECT+KK	3	1,346	33	2.45 [1.72 – 3.46]
DSM+FECT+KK	2	955	24	2.51 [1.65 – 3.77]
DSM+FECT+BCT	1	384	27	7.03 [4.88 – 10.04]
FECT+BCT+PCR	2	1,188	523	44.02 [41.18 – 46.90]
DSM+FECT+BCT+CULTURE	1	351	43	12.25 [9.22 – 16.09]
FECT+STST+BCT+CULTURE	1	844	127	15.05 [12.74 – 17.68]
TOTAL	43	78,959	1,437

*DSM = Direct saline microscopy, FECT = Formol ether concentration technique, KK = Kato-Katz, STST = Spontaneous tube sedimentation technique, BCT = Baermann concentration technique, PCR = Polymerase chain reaction

The pooled prevalence of S. stercoralis across different diagnostic methods was 3.7 % (95 %CI: 1.10 – 11.70) using random effect analysis. The heterogeneity was high (Q = 4376.6, I² = 99.8 %, P < 0.001) (Fig 5).

Fig 5

The true prevalence estimation of Strongyloides infection in Ethiopia is generally difficult due to application of very low sensitive diagnostic techniques and the presence of a few studies conducted with high sensitive diagnostic approaches so far in the country. The most widely used methods for the diagnosis of helminthic infections include DSM, FECT and KK. These methods are less sensitive for the detection of Strongyloides infection (Siddiqui et al., 2001; Buonfrate et al., 2015). Similarly, in this review, the authors on Strongyloides infection have clearly demonstrated that surveys conducted with these three methods mentioned above might provide untrustworthy prevalence reports among the peoples of Ethiopia.

The low distribution of Strongyloides infection in the current review might be explained by the fact that low sensitive diagnostic methods and small quantity (about 2 mg) of stool samples that have been used in DSM. For instance, single stool examined by DSM can give 70 % S. stercoralis false negativity (Siddiqui et al., 2001; Mirdha et al., 2009). The intermittent excretion nature (Burke et al., 1978) and low-intensity chronic infection of S. stercoralis (Schar et al., 2013) might also affect the true prevalence. In Ethiopia, highly sensitive diagnostic methods are not employed for Strongyloides infection and this might be due to their high cost and lack of awareness. As a result, almost all health institutions are still using low sensitive diagnostic methods for the clinical diagnosis of Strongyloides infection. This leads to under diagnosis and under-report of S. stercoralis infection throughout the country.

On the other hand, spontaneous tube sedimentation technique (STST) (Tello et al., 2012), BCT, stool culture and molecular (e.g. PCR) methods are more sensitive than DSM and FECT for the diagnosis of Strongyloides infection (Schar et al., 2013; Buonfrate et al., 2015). A combination of these methods in a single stool sample examination provides a higher detection rate of S. stercoralis infection (Aranzazu et al., 2016; Albonico et al., 2016; Hailu et al., 2020). Reports in this review showed that those studies conducted using a combination of more than one method provided a better Strongyloides infection detection rate in Ethiopia (Abera et al., 2013; Aranzazu et al., 2016; Tamirat et al., 2017; Hailu et al., 2020). However, the sensitivity of these tests is not perfect since they were performed on a single faecal specimen which might underestimate the true prevalence. Therefore, there is a need to define a standard protocol in diagnostic methods being used to detect S. stercoralis in Ethiopia, especially in health institutions. Such priority recommendations might be important for elaboration of mapping of S. stercoralis infection in the country.

In the current review, the overall prevalence of human S. stercoralis infection in Ethiopia was low (1.82 %). This result is lower than from previous reports 5.1 % among human immune-viruses (HIV) infected cases reported previously globally (Ahmadpour et al., 2019), and 20 % obtained from a large heterogeneity population and diagnostic methods in Latin America (Buonfrate et al., 2015). The high prevalence in the previous studies might be justified as both reviews include studies conducted by serological tests which are much more sensitive tests (Bisoffi et al., 2013). The study participants in the former study were also among HIV cases only. In addition, the variation in the ambient environment could favor the high prevalence of Strongyloides infection.

The prevalence of S. stercoralis infection was varied across regions of Ethiopia and relatively high prevalence of recorded in Ad-dis Ababa City and Amhara Regional State. This difference might be due to the difference in the diagnostic methods used, sample size and the health status of study participants. For instance, all the participants in the Addis Ababa City were street dwellers and HIV cases who are highly vulnerable to S. stercoralis infection.

Generally, the low prevalence of S. stercoralis in Ethiopia is due to absence of better sensitive diagnostic methods and the low attention given to S. stercoralis infection, unlike other soil-transmitted helminthes by policy makers. Based on this review, we encourage scholars to further work on the standardization of S. stercoralis test protocols and to advise policy makers for the inclusion of S. stercoralis in soil-transmitted helminths prevention and control package.

Limitation of this review: We used only PubMed, Google Scholar and Science direct databases and Addis Ababa University databases as a source of articles which might be the limitation of the current review.

Conclusions: This review confirmed that the prevalence of S. stercoralis is under-reported in Ethiopia due to the use of low sensitive diagnostic methods. Diagnostic methods including culture, BCT or PCR or a combination these methods give better detection rate of S. stercoralis infection. Therefore, there is a need to revise the current diagnostic methods of Strongyloides infection to have better sensitive diagnostic methods in the country. Further research is also desirable to break the transmission cycle and reduce the impacts of Strongyloides infection in Ethiopia.