Prevalence and Factors Associated with Rotavirus Infection among Vaccinated Children Hospitalized for Acute Diarrhea in Mwanza City, Tanzania: A Cross Sectional Study

Introduction: Rotavirus infection is a leading cause of severe diarrhea cul-minating to dehydration among children under five years of age. Under-standing trends and factors that could assist towards devising effective pre-ventive strategies of Rotavirus infection beyond vaccination is crucial. Objec-tives: This study was done in an attempt to determine the prevalence and associated factors of Rotavirus infection among vaccinated children aged between 6 weeks and 24 months admitted with acute diarrhea Mwanza, Tanzania. Material and Methods: Across sectional study involving vaccinated children aged 6 weeks to 24 months was conducted in three selected hospitals from July 2017 to January 2018. Socio-demographic and other relevant clinical information were collected using a standardized data collection tool adopted from WHO Rotavirus surveillance tool. Rotavirus infection from the stool detected using an enzyme immunoassay. Data were analyzed using STATA version Results: A total of 301 vaccinated children with acute diarrhea with a median age of 12 8 - 17] were enrolled. Nine and 292 had received one dose and two doses of Rotavirus vaccine, respectively. The prevalence of Rotavirus infection was 74 (24.6%) [95% CI: 20.0 - 29.8]. Independent predictors of Rotavirus infection were: dry season (OR 6.9; 95% CI: 2.9 - 16.0; p < 0.001), 3 ≥ children indwelling in the same house (OR 2.1; 95% CI: 1.1 - 4.2; p = 0.043) and vomiting (OR 3.6; 95% CI 1.1 - 12.6; p = 0.045). Children with Rotavirus infection had a significantly shorter hospital stay than those without Rotavirus infection (3 [2 - 4] days versus 3 [3 - 5] days; p = 0.0297). Conclusions: The prevalence of Rotavirus infection has declined among vaccinated children in Mwanza, Tanzania with significant decrease in the hospital stay. Dry seasons, three or more children indwelling in the same house and vomiting were independent predictors of Rotavirus infection. There is a need to sustain the coverage of rotavirus vaccination in low-income countries in order to significantly reduce associated morbidity and mortality.


Introduction
Diarrhea remains the second most common cause of death among children below five years of age worldwide [1]. Most of these deaths are due to severe dehydration, with the majority of deaths occurring in low and middle-income countries [1]. Globally, Rotavirus is the commonest cause of severe infantile diarrhea resulting in dehydration and prolonged hospital stay among children below five years of age [2] [3]. Development of a safe and effective Rotavirus vaccine has been a priority since Rotavirus disease cannot be eliminated through the improvement of water and sanitation [4]. The World Health Organization (WHO) authorized the Rotavirus vaccine for infants to be incorporated in all national immunization programme with strong emphasis in the countries where diarrhea disease cause more than 10% of death in children below five years of age [5].
In Tanzania, Rotavirus vaccine, a live attenuated human monovalent [G1P8] vaccine (Rotarix), was introduced in the National immunization program in January 2013. Despite the Rotavirus vaccine implementation, acute diarrhea with severe dehydration cases is still reported in this setting. The prevalence of Rotavirus infection prior and after vaccine implementation has been continuously studied, however, the trends such as seasonality, severity of disease, hospital stay and associated factors of Rotavirus infection have not been adequately studied among vaccinated children [6] [7] [8]. In a view of that, this study was done to determine the prevalence of Rotavirus infection and factors associated with acute diarrhea among vaccinated children aged 6 weeks to 2 years admitted in three hospitals in the city of Mwanza. The information from this study is important in assessing the trend of Rotavirus infection in this vaccination era. D. Mahamba et al.

Study Design and Settings
This was a hospital based cross sectional study involving 301 vaccinated children which was conducted from July 2017 to January 2018 in the city of Mwanza, Tanzania in three hospitals (Nyamagana District Hospital, Sekou Toure Regional Referral Hospital and Bugando Medical Centre).

Sample Size, Sampling and Inclusion Criteria
Sample size of the study was calculated using Kish Leslie formula [9] using the prevalence of 20.7% from Temu et al. [6]. The minimum sample size obtained was 250 children. All children who received at least one dose of Rotavirus vaccine as evidenced by RCH card aged 6 weeks to 24 months admitted for treatment of acute diarrhea within seven days duration irrespective of the other illnesses, were serially enrolled until the sample size was attained.

Variables and Outcomes
Socio-demographic data and relevant clinical information were collected using a structured pre-tested questionnaire adopted from WHO Rotavirus surveillance tool. Diarrhea was defined according to the WHO guidelines as passage of three or more loose, liquid or watery stools within 24 hours period [10]. Duration of illness, frequency of diarrhea, consistency of stool, history of fever, vomiting, history of Rotavirus vaccination were recorded followed by clinical examination to elicit hydration and nutritional status. All children were admitted and managed according to respective standard hospital guidelines. The duration of hospital stay for each child was noted.

Laboratory Procedures
All samples were collected and analyzed at the Bugando Medical Centre laboratory which is one of the WHO sentinel surveillance sites for Rotavirus gastroenteritis using enzyme-linked immunosorbent assay (The ProSpecT Rotavirus Microplate kit, Oxoid Ltd., UK) as per manufacturer's instructions.

Data Analysis
The data were analyzed using STATA software version 13. Continuous data were summarized using median with interquartile range (IQR) while categorical data were summarized using proportions. To determine the factors associated with Rotavirus infection univariable logistic regression followed by multivariable logistic regression models were performed. Factors that were statistically significant on univariable analysis were subjected to multivariable logistic regression model. Odd ratios with their respective 95% confidence interval were noted and a p value of less than 0.05 was considered statistically significant. To compare the median hospital stay after initiation of hospital management between vaccinated children with and without Rotavirus infection admitted for acute diarrhea,

Socio-Demographic Characteristics of Study Participants
A total of 301 children with acute diarrhea admitted in three hospitals were recruited in the study with a median age of 12 [IQR: 8 -17] months, of these 40 (13.3%), 184 (61.1%) and 77 (25.6%) were from Nyamagana District Hospital, Sekou Toure Regional Referral Hospital and Bugando Medical Centre, respectively. Majority of children, 166/301 (55.2%) were males and most of the children enrolled, 285/301 (94.7%) were from urban areas ( Table 1).

Clinical Findings among Vaccinated Children with Acute Diarrhea
The median duration of diarrhea was 3

Prevalence of Rotavirus Infection among 301 Children with Acute Diarrhea
Of

Factors Associated with Rotavirus Infection
On univariable logistic regression analysis, the factors associated with Rotavirus

Hospital Stay of Vaccinated Children with Acute Diarrhea
Children with Rotavirus infection had a significant shorter hospital stay than those without Rotavirus infection (3 [2 -4] days versus 3 [3 -5] days; p = 0.0297), Table 4 and Table 5.

Discussion
This study has observed the prevalence of rotavirus infection of 24.6% which is significantly low compared to 49.4% which was observed in the same setting in pre-vaccinated era [7]. Similar findings have been realized in a recent country wide surveillance in Tanzania [12]. These findings suggest that there is seasonal variation following the implementation of rotavirus vaccine [8].
This study showed that vomiting was significantly associated with Rotavirus infection, which is similar to the previous studies done in pre-vaccine era in ru- which ultimately induces vomiting reflex, nausea and vomiting [18]. This study has found that vaccinated children indwelling with three or more fellow children in the same house are significantly more likely to acquire Rotavirus infection compared to those indwelling with less than three fellow children. This finding is similar to the study done in Chiapas, Mexico in which children indwelling in a house with seven or more people were significantly more likely to acquire Rotavirus infection [19]. This is attributed by the fact that the mode of spread of Rotavirus infection is largely person-to-person transmission, hence as the number of children increases in the house likewise transmission rate increases.  [7]. However, in this study, it was found that children with Rotavirus infection had a shorter hospital stay compared to those without Rotavirus infection. This observation is similar to the study done in Accra, Ghana in which children without Rotavirus stayed longer in the hospital than those with Rotavirus infection [15]. This could be explained by the fact that Rotavirus vaccination reduces the severity of diarrhea and hence children recover quickly and ultimately the length of hospital stay is shortened. This confirms that Rotavirus vaccine has reversed the trend of hospitalization.
One of the major limitations of this study is that the study was not done for entire year so the interpretation for Rotavirus seasonality is somehow limited.

Ethical Approval
The ethical approval to conduct the study was sought from the joint Catholic University of Health and Allied Sciences/Bugando Medical Centre Research and Ethics Review Committee with ethical clearance number CREC/208/2017. Parental/guardian consent was obtained for each child prior recruitment.

Availability of Data and Materials
All data have been included in this manuscript.

Funding
The research was funded by the University of Dodoma, Tanzania and World Health Organization (WHO). Funders had no role in execution of this study.

Author Contributions
Conceived and designed the study: DM, AH, SEM and MMM. Data collection and laboratory testing: DM, FM, DRM, ECB, ENK, AH, SEM and MMM. Analyzed the data: DM, AH, SEM and BRK. Wrote the paper: DM and DRM. Edited and reviewed critically the manuscript: AH, SEM and MMM. All authors read and approved the final manuscript.