This Hospital-based study provides results on the magnitude of E. coli O157:H7 and associated factors in under-five children with diarrhea in Eastern Ethiopia. The magnitude of E. coli O157:H7 among the children was 15.3% (95% CI: 11.8–19.4) and associated with being rural residents, under cocked and raw vegetable consumers, history of dysentery type of diarrhea, number of children, mothers uneducated, owned livestock, and history of diarrhea in the household.
The prevalence of E.coli infection was in agreement with the prevalence report of study (14%) in Bahir-Dar town, Ethiopia [12] and it is comparable with study (20%) in Benin City, Nigeria [10]. However, it is higher than the report of a study conducted in central Ethiopia in 2017 [18] and southern Ethiopia [19]. It is also much higher than studies conducted in other African countries [20, 21]. The discrepancy might be explained by the difference in sample size, source of study population difference, method of sample collection and specificity of polymerase chain reaction.
Moreover, compared to hospitals based studies of the same age group with diarrhea, the current prevalence was higher than studies conducted in some Asian countries, as the reported prevalence was 1.14% in Iran [22], and 4.61% in India [23, 24]. Compared to the South American study, Argentina, the finding is also higher than the 10.1% report of Rivero et al. (2010) [25]. Generally, the relatively high prevalence of the organism obtained in the current finding might be because of the cosmopolitan and high level of animal-to human interaction in the study subjects than the above-mentioned countries, indicating E. coli O157:H7 to be an important diarrhea causing pathogen in the study population.
It was found that children from rural residents were about four times more likely to had E. coli O157:H7 compared to urban resident children. The finding is consistent with other studies conducted elsewhere [26, 27]. The positive association in the existence of high-density reservoir animals as a predictor of infection [28, 29] and the higher probability of rural residents to have animal contact there by the predictor of rural residence for childhood diarrhea in this age group [30, 31] might support the finding.
Consuming undercooked meat and raw vegetables is significantly associated with the prevalence of E. coli O157:H7. This finding is in line with the report of previous studies [32–34]. This could be evidenced by the high level of animal manure contamination of vegetables through either irrigation by untreated surface water or animal manure usage as a fertilizer [35]. As indicated in this study, animal manure usage as a fertilizer was significantly associated with infection to the organism. The organism was isolated from vegetable samples [36], can survive outside the host reservoirs [37]. There were lettuce associated large outbreaks raised in Sweden in the year 2005 [38]. However, there are some reports, which claim that, eating raw vegetables as a protective factor for the occurrence of E. coli O157:H7 infection [39].
The likelihood of E. coli O157:H7 was more than four times higher among children with bloody diarrhea relative to those who experience watery diarrhea. Previous study reported was more likely to be isolated from visibly bloody stool than without visible blood and was the pathogen most commonly isolated from visibly bloody stool that yielded a bacterial enteric pathogen [33, 40, 41].
Meanwhile the association of raw milk consumption and E. coli O157:H7 was not found statistically significant, although some authors described as consumption of raw milk is a risk factor for the occurrence of E. coli O157:H7 infection [42, 43]. However, the interpretation must be cautious because of the limitations posed by the small sample size (12.6%) of the children had consumed raw milk.
Age differences were not found independently associated with occurrence of E. coli O157:H7. However, Kargar and Homayoon found that children (< 2) years) of age were at highest risk of infection with E. coli O157:H7 [44]. Since the data related to exposure status of the child were not mentioned, it is difficult to justify. The contrary reported as the odds of detection proportion of E. coli O157:H7 were high in those children above the mean age in more than 11%. The discrepancy seen with the current finding might be attributed to the older child (1–17 years) age group in the later study, which will influence the social and cultural behavior [45].
Children from households with livestock were four times more likely to contract E. coli O157:H7 than households without livestock, which is consistent with the report of study on cattle owners [46] and study conducted in Harare, Zimbabwe [47]. Since ruminants had been identified as the major reservoir of E. coli O157:H7, there could be a cross-infection to the children through either by direct live animal contact or with animal manure. A matched case-control targeted particularly in children under three years of age in Germany also indicates nine times the risk of developing infection in those who touched ruminants including goats and sheep [14]. In the meantime, studies reported that contact with farm animal manure or cattle and living in or visiting a place with farm animals be well-established risk factors for the occurrence of E. coli O157:H7 infection [32, 48].
Children whose mothers had no formal education were three times more likely to have E. coli O157:H7 than mothers with formal education. It is possible to suggest that mother’s education level might have a positive influence in preventive measures of diarrhea such as household hygiene and sanitation. Similarly, poor level of household sanitation had been reported as predictor of E. coli O157:H7 infection [39]. Evidences strength of the presence of low awareness on hygiene and sanitation among low-income mothers and this could expose a child to several cross-contamination from one source to the other, indicative of poor personal hygiene [49].
Children from a household member with diarrhea prior to his/her diarrhea manifestation were four times more likely to contract E. coli O157:H7. The same finding was obtained by Rivas in their case-control study [39]. In addition, a child who was from a household having three to four under-five children had seven times higher odds of E. coli O157:H7 than a household with a single to two children. Household under-five year number is a well-studied determinant factor associated with childhood diarrhea through its effect on overall household sanitation and hygiene level [50–53]. Although the levels of contamination of the water sources are not well identified, drinking water treatment methods were not found associated with the occurrence of the organism. However, other researchers reported drinking contaminated water sources as a predictor for the occurrence of E. coli O157:H7 infection [36, 54–57].
However, the result obtained in this research work might be difficult to infer about the general source population because of the design. Besides, as study crosses sectional type, the absence and the existence of the organism may not declare the absence of other enteric pathogens as a cause of diarrhea.