Rotavirus Surveillance and Impact of Vaccination at the Gaoua Regional Hospital, Burkina Faso, from 2013 to 2022 ()
1. Introduction
Rotavirus infection represents a major public health issue, especially in Burkina Faso. Indeed global surveillance data show that nearly 40% of hospitalizations due to diarrhoea in children under 5 years of age are linked to rotavirus [1]. Thus, rotavirus is the leading cause of severe diarrhoea with dehydration in children under 5 years of age and ranks fourth as a cause of childhood mortality worldwide [2]-[4]. According to the World Health Organization (WHO) report, the annual number of childhood deaths due to rotavirus worldwide from 2013 to 2017 was between 122,000 and 215,000 [5]. In addition, 258 million cases and 128,500 deaths were from rotavirus in 2016. The burden is high in developing countries [6] [7]. In 2016, approximately 117 million episodes of rotavirus-associated diarrhoea occurred in sub-Saharan Africa, including approximately 104,000 deaths [8].
To strengthen the fight against rotaviruses, WHO is coordinating a global surveillance network that uses standardized case definitions and laboratory methods in sentinel hospitals to identify rotavirus cases in children with diarrhoea [9]. Burkina Faso established this surveillance network and expanded it to four sentinel sites in December 2013. However, although the burden of rotavirus disease has decreased considerably compared to 2000, its incidence remains high in Burkina Faso [10]. According to the 2020 statistical yearbook of the Ministry of Health and Public Hygiene, non-bloody diarrhoea is a reason for consultation in 690,422 children under 5 years of age in basic health facilities.
Few have analysed sentinel surveillance data on rotavirus in Burkina . The latest published was in 2017 by Bonkoungou et al. It did not determine the performance indicators for rotavirus surveillance. In addition, Burkina Faso has changed the type of vaccine it uses. It switched from Rotateq to Rotasiil in 2020. Our study therefore aims to analyse rotavirus epidemiological surveillance in order to determine the data prevalence of rotavirus, describe its epidemiological profile, determine the impact of vaccination, and determine the level of performance indicators for its system surveillance.
2. Methodology
2.1. Study Setting
Burkina Faso has four sentinel sites or rotavirus surveillance. The Gaoua Regional Hospital (CHRG) is in the southwest region of Burkina Faso. It initiated sentinel otavirus surveillance in December 2013 and has not experienced any interruptions. According to rotavirus database surveillance data from 2020 and 2021, the Gaoua sentinel site represents the site that has reported the most cases.
2.2. Concept of the Study
This was a cross-sectional study with the aim of analysing rotavirus surveillance data from the Gaoua sentinel site from October 2013 to June 2022.
2.3. Study Population and Sampling
The study population consisted of all cases of acute diarrhoea, notified to the CHRG in the rotavirus database, from October 2013 to June 2022.
Suspected case (or case of acute diarrhea): Any child (generally 0 to 5 years old, or 0 to 59 months) hospitalized for acute diarrhea (sudden onset of loose or watery stools, distinct from normal bowel movements, lasting less than 7 days) (Sentinel surveillance plan for rotavirus in Burkina Faso).
Inclusion criteria: All children aged 0 - 59 months meeting the sentinel case definition for acute diarrhoea and recorded in the national rotavirus surveillance database during the study period.
Exclusion criteria: No cases were excluded from the descriptive analyses. However, for analyses requiring laboratory confirmation, particularly the estimation of rotavirus prevalence, suspected cases without an ELISA test result (n = 784) were excluded because their rotavirus status could not be determined. Sampling was exhaustive, including all eligible cases, n = 2135.
2.4. Study Variables
Time variable: Quarter (1 - 4), date of onset of symptoms, date of consultation, date of ELISA test, date of ELISA test result, date of hospitalization, date of Exeat,
Person: Age (0 - 59 months), sex (male, female),
Performance indicators variables according to health surveillance in Burkina Faso (Directorate of Vaccination Prevention): Vaccination (Yes, No), staff capacity to photocopy booklets (Yes, No), ELISA test performance (Yes, No), and sample collected (Yes, No).
2.5. Data Collection
This was data collected in July 2022 from the acquired databases from the Ministry of Health and Public Hygiene using a data extraction grid. These data were previously collected by health workers from children aged 0 - 5 years, in the CHRG sentinel site, using rotavirus case investigation forms from October 2013 to June 2022. The child’s vaccination status was verified through examination of the vaccination record. They were entered into Epi Info 3.0 using the WHO rotavirus surveillance system database. For laboratory testing, stool samples were collected within 48 hours of admission and stored at 2˚C - 8˚C (if tested within 1 week) or −20˚C (if tested >1 week) before being transferred to the Burkina Faso National Public Health Laboratory by Poste Burkina, where rotavirus testing was performed. Samples were tested using a commercially available enzyme immunoassay (EIA; ProSpecT, Oxoid, Cambridge, UK). The specimens were stored at −70˚C at the National Public Health Laboratory of Burkina Faso.
2.6. Data Management and Analysis
Data processing consisted of merging the database of the last 10 years and identifying missing, outlier, and inconsistent data. Excel ® 2016 and Epi Info 7.2.2.6 were used to analyse the data. The proportions, mean, median, and impact measurement (preventive fraction) were calculated. Rotavirus prevalence was calculated using only children who underwent ELISA testing, with positive cases in the numerator and the total number tested in the denominator. This approach ensures methodological rigor, as untested children cannot be classified virologically and therefore are excluded from prevalence estimates. Vaccination coverage rate: number of vaccinated cases in the total study population. Patient diagnosis time: is defined as the time interval between the date of onset of symptoms and the date of the first medical consultation (admission date). System diagnosis time is the time interval between the date of the first medical consultation and the date of diagnosis (date of delivery of results). Vaccine efficacy was calculated as follows: (1 − OR) × 100 (%). The odds ratio was first calculated as a crude OR using a 2 × 2 contingency table comparing vaccination status between ELISA-positive cases and ELISA-negative controls through univariable logistic regression (liberal significance threshold of p ≤ 0.20). In the multivariable logistic regression analysis (liberal significance threshold of p ≤ 0.05), adjustments were made for age, year, and season; therefore, the reported OR reflects these adjustments and was used to estimate vaccine effectiveness.
2.7. Ethical and Regulatory Considerations
We obtained authorisation to access the data from the Secretary General (SG) of the Ministry of Health and Public Hygiene (MSHP). The confidentiality of the data was respected.
3. Results
3.1. Demographic Characteristics of Participants
A total of 2135 suspected cases of infection in rotavirus children aged 0 - 59 months were reported. The most common age group was children aged 0 - 11 months, accounting for 44.96% of cases. Children aged 12 - 23 months and 24 - 59 months represented 39.11% and 15.93% respectively. Males represented 55.50%, with a sex ratio of 1.56 males to females.
3.2. Prevalence of Rotaviruses
Among the 2135 suspected cases included in the study, 784 (36.7%) did not undergo ELISA testing and were therefore excluded from the prevalence analysis because laboratory confirmation of rotavirus infection was unavailable. The prevalence analysis was consequently performed on the remaining 1351 children, among whom 366 tested positive, yielding a rotavirus prevalence of 27.10%.
3.3. Case of Rotavirus Infection by Year
The first quarter of each year from 2014-2022 constitutes the peak periods, except in 2019 when no positive cases were reported in the first quarter. The year 2014 had the highest number of positive cases with a total of 85 cases in the first quarter. Figure 1 shows the distribution of confirmed cases by year and by quarter.
Legend: T = Quarter.
Figure 1. Distribution of confirmed rotavirus cases by quarter and by year at the Gaoua CHR, 2013-2022.
3.4. Characteristics of Rotavirus Diarrhoea Cases
In our study, the most affected age group was children aged 0 - 11 months, with a proportion of 48.09%, followed by those aged 12 - 23 months, which was 44.81%. The male gender represented 60.93%.
3.5. Diagnosis and Hospitalization Times for Cases of Rotavirus Diarrhoea
The average time for diagnosis in the health system is the longest, at 27 days. This large difference between the mean (27 days) and the median (5 days) reflects a highly right-skewed distribution of diagnostic delays. Most patients were diagnosed within a few days, but a small number of cases experienced prolonged delays, which increased the mean (Table 1).
Table 1. Distribution of deadlines.
Variable |
Average |
Minimum |
Maximum |
Median |
Patient diagnosis time (days) |
3 |
0 |
10 |
3 |
Health system diagnostic time (days) |
27 |
0 |
664 |
5 |
Hospitalization time (days) |
4 |
0 |
18 |
4 |
Health system performance indicators for suspected cases
The test rate completion was the lowest indicator at 63.23%, according to Table 2.
Table 2. Performance indicators of the rotavirus surveillance system in the CHRG, Burkina Faso, 2013-2022.
Variables |
Staff |
Proportions (%) |
Rate of samples collected |
2098 |
98.26 |
One-dose vaccination coverage |
1711 |
80.14 |
Photocopied heath record |
1759 |
82.38 |
Percentage of tests performed |
1351 |
63.27 |
Health system performance indicators for suspected cases from 2013 to 2022: vaccination coverage, ELISA test completion rate
No ELISA tests were performed in 2022. Vaccination coverage was higher at 93% (see Figure 2).
Figure 2. Distribution of children vaccinated and the number of tests performed in the CHRG per year, from 2013-2022.
3.6. Impact of Vaccination
Vaccination coverage among subjects with rotavirus diarrhoea with at least one dose of vaccine was 68.58%.
In our study, vaccinated children without rotavirus diarrhoea were the most represented group, with 761 subjects. The adjusted odds ratio (ORa), controlling for age, year, and season, was 0.644, CI [0.469 - 0.884], p = 0.0065, representing a vaccine efficacy of 36%. See Table 3.
Table 3. Vaccine efficacy (VE) against rotavirus diseases.
Vaccination status |
Case |
Control |
HRa |
p-value |
VE (%) |
95% CI |
|
n |
% |
n |
% |
|
|
|
|
Not vaccinated |
115 |
31 |
224 |
23 |
1 |
|
- |
- |
Vaccinated (≥1 dose) |
251 |
69 |
761 |
77 |
0.644 |
0.0065 |
36 |
[12 - 53] |
4. Discussion
Prevalence
The prevalence of rotavirus infection in our series was 27.10%. It is similar to that of Parvathi Durga et al. in India, 27% [11]. Our prevalence was higher than that by Ahmed Mohamed Fidhow et al. in 2017 in Kenya, Doutchi Mahamadou et al. in 2019 in Niger, Noshen Basharat et al. in 2021 in Pakistan, and Anil Kumar Goel et al. in 2021 in India respectively 17%; 21.43%; 22%; 11.02% [12]-[15]. This similarity and these discrepancies with other studies could be explained by the confirmation test used, which was the ELISA test in our study and PCR in others. In Burkina Faso, rotavirus infection emains one of the main causes of acute diarrhoea in children.
Time for diagnosis in the health system
The average time for diagnosis in the health system is the longest, and the large difference between the mean and the median reflects a highly right-skewed distribution of diagnostic delays. Most patients were diagnosed within a few days, but a small number of cases experienced prolonged delays, which increased the mean. Based on surveillance data, tests are performed according to the availability of reagents, which may sometimes be out of stock. In 2022, no tests could be performed because the reference laboratory no longer had diagnostic reagents available. The availability of supplies, or even of rotavirus diagnostic tests, could improve the long diagnostic delay and even patient management.
Health system performance indicator: vaccination coverage
In our study, 68.58% of children with confirmed rotavirus infection had received at least one dose of vaccine, compared with 80.14% among suspected cases. Doutchi Mahamadou et al. in Niger in 2019 reported 71.43% [13]. This proportion of vaccinated children remains lower than the target rate assigned by the Directorate of Vaccination Prevention in Burkina Faso and those achieved between 2013-2022. Nevertheless, the vaccination coverage in our series concerns confirmed cases and may underestimate the rotavirus vaccination coverage in the Gaoua region. It follows from these results that rotavirus cases have low vaccination coverage.
Impact of vaccination
The preventive fraction or vaccine efficacy was 36% in our series. Aisleen Bennett et al. in Malawi in 2021 reported results almost similar to ours, 39% [16]. Results that are relatively higher than ours were reported than ours by Murunga et al. in 2020, Wang et al. in 2021, Erdem Gönüllü et al. in 2021 respectively of 61%, 84%, 75.1% [8] [17] [18]. These different results confirm the fact that vaccination against rotavirus has a protective effect against rotavirus infection. This discordance between these results could be explained by the type of vaccine used, the change of vaccine type during the year 2020, vaccination schedule adopted by the country, and the immune response of vaccinated individuals. However, vaccination is an effective strategy that must be optimized.
Study Limitations
The authors acknowledge several limitations of the study.
First, the single-center design limits the generalizability of the results due to limited external validity. Indeed, there are four sentinel sites for rotavirus surveillance in the country. Given that local practices and/or patient profiles differ from one region to another, selection bias may have arisen.
Furthermore, the absence of laboratory results for more than a third of suspected cases may have impacted reliability and validity. This could weaken the strength of the conclusions and limit the scientific scope of the research. To minimize this impact, some analyses, such as prevalence estimation, excluded these unconfirmed cases. However, this may have introduced measurement bias by underestimating or overestimating the prevalence of rotavirus.
In addition, our study, which used surveillance data, may have limitations related to incomplete records, measurement inaccuracies, and selection bias. Indeed, some suspected cases did not undergo biological confirmation, which may have reduced the statistical power of the tests and decreased the representativeness of the sample.
Finally, the cross-sectional nature of the study does not allow us to conclude a causal relationship due to the lack of temporal data, the risk of reverse causality, and the presence of probable confounding factors that could not be controlled.
However, we acknowledge the study’s strength in being the first to not only determine the prevalence of rotavirus and describe its epidemiological profile, but also to evaluate the impact of vaccination and determine the performance indicators of its surveillance system.
5. Conclusion
The prevalence of rotavirus remains high at the Gaoua Regional Hospital Centre with children aged 0 - 11 months being the most affected age group. The first trimester of the year was the peak period corresponding to the dry season. Males were the most affected. Vaccination coverage was low among confirmed cases. Furthermore, although this vaccination is initiated in Burkina Faso at 2 months of age as part of the Expanded Program on Immunization (EPI), children aged 0 to 11 months remain the most affected. Cohort studies could examine the effect of the timing of vaccine introduction on rotavirus prevention in children under 5 years of age.
Acknowledgements
We would like to express our sincere thanks to the REDISSE project for its financial and technical support, particularly for funding the BFELTP training program, the BFELTP for its technical support, as well as the Directorate of Vaccination Prevention for providing data.