Prevalence and genotypes of Rotavirus among children under 5 years presenting with diarrhoea in Moshi, Tanzania: a hospital based cross sectional study

Diarrhoea is a main cause of morbidity and mortality in children under 5 responsible for approximately four billion cases and 1.1 million deaths annually [1]. In developing countries, it causes two million deaths each year [2]. The major causative organism responsible is Rotavirus which is responsible for one-third of hospitalizations with approximately 40% mortality [3].

Rotavirus mainly affects children, with young children below five years being the most vulnerable. The prevalence of rotavirus is highest among children from 13 to 24 months (57.6%) followed by those less than 1 year (46.3%) and lowest among children aged 1 to 5 years [4].

Rotavirus is an RNA virus and a member of the Reoviridae family [5]. It has groups ranging from A to G with Group A having multiple strains and causing majority of childhood infections [6]. The Rotavirus has 14G and 11P serotypes which are critical to vaccine developments; the serotypes are the vaccine targets for stimulating a protective immune response [7]. G1P[8] is the commonest circulating serotype accounting for two-thirds of infection [8].Strain diversity has been reported and is greatest in Africa and Asia due to mixed infections and proximity to domestic animals shedding the virus [8]. There are also changes in genotypes circulating due to seasonal variation [9]. In Africa the most prevalent strains were G1P [8] and G3P [8, 10].

Rotavirus vaccine (Rotarix) was introduced into the Tanzania Expanded Programme of Immunization in January 2013. Studies have shown vaccine effectiveness to be 98% in high income countries as compared to 59% in middle income countries [11]. In Malawi data showed the vaccine to be 64% effective in rotavirus negative individuals [12]. In the pre-rotavirus vaccine era, rotavirus diarrhoea resulted in 220,000 annual hospital admissions, 1.8 million healthcare visits and 7.1 million episodes of diarrhoea among children living in developed countries [13]. In the United States, there were an estimated 20–60 deaths, 55,000 to 75,000 hospitalizations and 410,000 outpatient visits annually [14]. In the post vaccination era, it is estimated that approximately 36% of diarrhoea hospitalizations in children less than 5 years in countries where rotavirus surveillance is being carried out, can be attributed to rotavirus infection [3].

Prior to introduction of vaccine in Tanzania, the prevalence of Rotavirus in Dar-es-Salaam was 32.5% [15]. Previous studies have reported the emergence of new strains such as G9 and others like G12, G4 and G8 in Dar-es-Salaam [15]. Despite this there have not been any published studies to date looking at the circulating serotypes in the Kilimanjaro region both in the pre-or post-vaccine era. Because of this we fail to understand if the diarrhoea cases being observed are due to different virus strains or the same strain targeted by vaccine. We also fail to comment on the circulating serotypes. This study aimed to determine the prevalence of Rotavirus among children under 5 in Moshi, Tanzania post vaccination and the circulating serotypes in the region. Kilimanjaro Region is one of Tanzania’s 31 administrative regions. It has 7 districts with the regional capital being Moshi municipality. Moshi is situated on the lower slopes of Mount Kilimanjaro and covers about 59² km (23² m) with a population of 184,292 per the 2012 census.

The overall prevalence of Rotavirus in our study was 26.4% (73/277); 29/73 (39.7%) in children aged less than 12 months, 34.2% (25/73) in children aged 13-24 months and 21.9% (16/73) among children older than 24 months.

This study focused on determining the prevalence and genotypes of Rotavirus among children under five presenting with diarrhoea in four hospitals/health centres in Moshi Municipality in Northern Tanzania. We found the prevalence of Rotavirus to be 26.4% with children less than 24 months of age being most affected. The genotypes seen in children who had Rotavirus were G1, G4, G8, G9 and G12. Those who had completed Rotavirus vaccine had G1 serotype but also displayed the other strains.

In comparison to other studies done around the world; our findings were similar to studies done in Mali by Ouermi et al. [16] and in Tanzania, Mwanza by Temu et al. [17] which found prevalence’s of 22.7 and 20.7%, respectively. This similarity could be attributed to the fact that they had the same study duration and both were carried out in hospital settings. The former studies were both pre-vaccine studies which were carried out in Africa. The low prevalence in Mali could be due to their small sample size. Similarities were also seen in developed countries, in Japan a study done by Phan et al. [18] found a prevalence of 14.9%.

Differences were observed in Yemen in 2014 in a study done by Badani et al. [19] with a prevalence of 45.9% and in 2010 Nakawesi et al. [20] in Uganda found a prevalence of 45.4%. Both studies were hospital based and done before the Rotavirus vaccine was introduced in their respective regions. Gachanja et al. [21] in 2016 in Arusha, Tanzania reported a prevalence of 37% which is higher compare to the observed from our study. This study was done after introduction of Rotavirus vaccine as our study, however, the higher prevalence reported could be attributed by the fact that Arusha is a mainly pastoral community and living near livestock is a risk factor for acquiring Rotavirus as there are also bovine strains which can affect humans.

In our study we found the genotypes in 41 of the children who had Rotavirus were G1, G4, G8, G9 and G12.We also found genotype G1, G4, G9 and G12 was in all the children who had completed Rotavirus vaccine, except G8 where one child was not vaccinated. Our findings were similar to those by Kiulia et al. [22], 2008 in Kenya who found G1, G8 and G9 to be the circulating strains. This similarity could be attributed to both studies being done in the same geographical setting and same study population despite the former study being a pre-vaccine study. Both Moyo et al. [15] in Tanzania in (2014) and Ansari et al. [23] in Nepal in (2013) found genotypes similar to ours. This similarity could be attributed to the fact that the study included all children under 5 like ours and are known to be high risk group for acquiring Rotavirus as well as having used similar detection methods. Moyo et al. [15] in Tanzania was the first to describe pre-vaccine serotypes. Our study is of value as it’s the first to be done in the region. We cannot comment if the prevalence of these strains has increased or not since this is the first study to be done looking at the strain patterns in Kilimanjaro region. We now have baseline information to monitor trend and pattern these other subtypes.

Differences were noted in studies by Badani et al. [19] from Yemen in 2014, Sanchez-Uribe et al. [24], 2014 from Mexico who found G2P [4] and G9P [4]. This difference is due to the fact that both studies were able to do P genotyping unlike our study hence were able to report combination genotypes that existed in their respective region. Both G and P typing are critical for vaccine development [6]. Li et al. [25], 2014 in China found G1, G2, G3 and G9 and in which some strains were different from our study. This difference could be attributed to the fact that his study was done in the pre-vaccine era and hence no baseline serotypes were known making comparability difficult. Moyo et al. in 2007 [26] in Tanzania reported G1, G3 and G9 to be the serotypes presented with G1 being dominated which is similar to this study; however I did not find G3 in this study. This difference could be due to the fact it was a pre-vaccine study and done only in one season hence no room for changing of serotypes due to selection pressure exerted by vaccine or the seasonal effect of Rotavirus. However, looking at Moyo et al. study and our study as both done in Tanzania, we can say that majority of pre-and post vaccine strains are still dominating in Tanzania.

Rotavirus is responsible for severe gastroenteritis in children under 5 years with a prevalence of 26.4% post vaccine era in Moshi, Tanzania. Most affected age group being children under 24 months of age. The dominated genotype was G1 53.8% with all children with this strain being received vaccine. Strain diversity was noted mostly in children who had completed rotavirus vaccine.

There is a need for continuous surveillance by relevant authorities of circulating serotypes as well further studies on vaccine efficacy. Long term assessment of vaccine potency, efficacy and circulating strain is needed so as to inform on the need for new vaccine or introduction of other genotype vaccines in the vaccination program.