Study Area
Cameroon, a country in Central Africa, has an area of 475650 km². According to the 3rd National Population and Housing Census, the estimated population of Cameroon in 2015 was at about 22179707 inhabitants. The annual crude birth rate in Cameroon is 35.4/1000 inhabitants, translating into about 882000 birth annually [17]. Accessibility to public health facilities is more evident for the wealthiest segments of the people, like those living in urban areas (52%). The health sector is structured into primary, secondary, and tertiary levels [14].
The Littoral region of Cameroon which comprises of 26 health districts and 212 health facilities with an estimated total population of over 3 million inhabitants and almost 80000 birth cohort. This study was conducted in seven health centres: the district hospital of Nkongsamba, Bare sub-district health centre, Eboumbeng sub-district health centre, Eboumbeng integrated health centre, Bonangoh integrated health centre, Nlongko'o sub-district health centre, and Bare integrated health centre. Regional Hospital of Nkongsamba represents a secondary hospital of about 320000 inhabitants and also served as the reference hospital for all the other study sites in the rural communities. The rest of the study sites are part of the primary healthcare level. The site selection procedure for this study was as follows: firstly, we randomly the selected Littoral region among four areas that comprise sentinel sites for influenza-like illnesses based on regional health retrospective data. Secondly, we chose two health districts (Melong and Nkongsamba) among 26 health districts of the region. Thirdly, we selected at random the seven study sites from the list of 45 health centres in the two health districts respecting the selection criteria of being situated at most 5km from the reference laboratory and possessing a well-functioning refrigerator (Figure 1). The study sites are representative of the healthcare system in Cameroon.
Sample size
In order to calculate the total number of patients required for estimating the prevalence of RSV precisely, we use normal approximation to binomial distribution. Specifically, first we assume that the true prevalence of RSV in outpatient and admitted patients is 50%, and there is a possibility of 3% non-response. Subsequently, in order to obtain a 95% confidence interval for RSV prevalence with a 10% margin of error, we will require 100 patients. The choice of 50% as the true prevalence was motivated by the fact that a prevalence of 50% requires the most number of patients. That is, it is a conservative choice. Consequently, if the true prevalence is either smaller or larger than 50%, then our margin of error will only be smaller than 10%, but never more than it.
The stratified sampling proportional allocation strategy was used to allocate the sample size to each study site. Each study site was considered a stratum and the size was based on the number of pediatric ARI cases of the year 2017 (Fig 1).
Fig 1. Study site selection and sample size allocation
Study design, target population, and Sampling
This multicentric cross-sectional study was conducted within six months, from March to September 2018. The primary objective of the study was to determine the prevalence of RSV infection in children less than two years of age with symptoms of ARI as measured by RSV ELISA IgM. Secondary objective was to assess the risk associated with various sociodemographic, medical and environmental characteristics and the probability of testing positive for RSV IgM. These characteristics and their categories are defined in Table 1.
A case of ARI was defined as illness fulfilling age-specific clinical inclusion criteria with onset within seven days in a child aged less than two years. Acute respiratory infection was defined as an illness presenting with one or more of the following symptoms: fever, cough, earache, nasal congestion, rhinorrhea, sore throat, vomiting after coughing, wheezing, and labored, rapid, or shallow breathing. Children who had respiratory symptoms lasting more than 14 days because RSV infection may have been acquired in the health facility during the perinatal period, who had neutropenia from chemotherapy, had been hospitalized elsewhere within four days or were newborns who had been hospitalized since birth were not enrolled.
The study was approved by a national ethical review committee. Regulatory approval was not sought as this was not an interventional drug or vaccine study. Mothers or caretakers of the children were invited to participated in the study, and after obtaining written informed consent children were enrolled. Study relevant data such as clinical signs and symptoms, sociodemographic, medical and environmental information were entered by study nurses into a questionnaire (Table 1). Nurses of each study health facility were previously trained in data collection and documentation using a pilot-tested questionnaire.
Venous blood of 2-ml was collected from all enrolled patients centrifuged and plasma transferred into 1-ml cryotubes containing virus transport medium. Plasma was stored at 4-8°C at the collection site for a maximum of 48 hours. Laboratory technicians were trained prior to study start in blood sample collection, transportation, and storage procedure as per a written Laboratory manual. Samples were then transported, maintaining the cold chain using the triple packaging system to the testing site and stored at -80°C pending testing. The Abcam ELISA (ab108766 anti-RSV IgM Human, Abcam, Cambridge, United Kingdom) testing procedure: A 96-well plate was precoated with Respiratory syncytial virus antigens to bind cognate antibodies. 100 µL of controls or diluted sample were added into appropriate wells and incubated at 37ºC. Following washing, a horseradish peroxidase (HRP) labeled anti-Human IgM conjugate was added to the wells, which binds to the immobilized Respiratory syncytial virus-specific antibodies. 3,3′,5,5′-Tetramethylbenzidine (TMB) was then catalyzed by the HRP to produce a blue substrate that changes to yellow after adding an acidic stop solution. The yellow coloration is directly proportional to the amount of Respiratory syncytial virus IgM sample captured in plate.
Data collection and analysis
Data documented in the questionnaires and laboratory results were recorded and stored in a spread sheet using MS Excel. The statistical analysis was performed using Stata, version 11.0 (StataCorp, College Station, TX).
The primary outcome - RSV prevalence - was determined by proportion of RSV IgM positive children, calculated by dividing the total number of RSV IgM positive children by the total number of children tested for RSV IgM. The corresponding 95% confidence interval was obtained using normal approximation to binomial distribution.
A second analysis assessed the association between the clinical characteristics and their categories as defined in Table 1 and the probability of testing positive for RSV IgM. Each clinical characteristic is categorical. Hence, to analyze them we utilized fisher’s exact test. The null hypothesis in each test was that the odds of obtaining a positive RSV test in each of the two groups of a clinical characteristic were equal. The alternative hypothesis was that the odds were unequal. A p-value of < 0.05 was considered statistically significant.