Rubella IgM epidemiology in the pre-rubella vaccination era in Uganda

Background Control of Rubella and Congenital Rubella Syndrome using vaccination has shown great success in the America’s. Uganda is due to introduce the Rubella vaccine however the magnitude of transmission is not well documented. Therefore this study was done to determine IgM sero-prevalance for Rubella in order to help monitor vaccine effectiveness post introduction of the vaccine in routine vaccination programme. Methods A retrospective review of suspected measles cases data for the reporting period January 2007 to December 2016 in Uganda was Done. rubella IgM testing was done on 15,296 of the cases and the data was analyzed using STATA version 13. Results In total 15,296 cases were tested and 4255 (27.8%) tested positive and among females aged 15-49 years 88 out of 322 (27%) tested positive. The age distribution range was 0–80 years, rubella IgM positivity was reported in all the 15 regions of Uganda and throughout the ten year period in every month. Age group 5–15 years had OR 2.5 p-value < 0.001 of being rubella IgM positive compared to age < 5 years and testing measles IgM negative OR 6.3 p-value < 0.001. Conclusion Rubella is endemic in Uganda and although rubella IgM positivity is highest in the age 5-15 years even the younger, older and women of reprodutive age are affected. This means the risk of Congenital Rubella Syndrome is high hence the need to introduce the rubella vaccine for infants and pregnant mothers and continued surveillance to enhance its control.


Background
Rubella and measles infections are similar in presentation with fever and rash as the main Symptoms. rubella is caused by an RNA rubella virus and is usually a mild condition with measles causing a more severe clinical manifestation and common among children. If vertical transmission occurs during the first trimester of pregnancy the virus causes miscarriage, still birth or Congenital Rubella Syndrome (CRS) which has teratogenic effects that include deafness, blindness, congenital heart disease and mental retardation [1,2].
Although the exact prevalence of CRS is not know it is estimated that more than 100,000 cases were born especially in the developing countries in 2015 [3,4]. The prevalence of rubella among all people in Africa is estimated at 52.9 to 97.9% and that of pregnant women in Africa has been reported to range from 54.1 to 95.2% a public health importance despite the reported high immunity in the younger children [4][5][6][7][8]. The general rubella-specific IgG sero-prevalance in pre-vaccination era in Africa is comparable with other regions in Southern America, India and Europe before vaccination. However, the level of natural immunity in these studies is lower than the immunity currently reported in Europe; this might be due to on-going vaccination programmes in developed countries [6,9].
Rubella is a vaccine preventable disease and the rubella vaccine contains a live attenuated virus that is effective in combination as trivalent with measles and mumps or bivalent with measles. The WHO global measles rubella initiative and strategic plan targets to eliminate measles and rubella by 2020 however the WHO African region has not yet set a target. In the WHO American region due to intensified vaccination against measles and rubella there has been no indigenous measles case since 2002 and no rubella case since 2009. WHO declared the American region to have eliminated rubella and CRS in 2009 and 2015 respectively. In the United States of America evidence from data collected and analyzed on the epidemiology and geno types of rubella and CRS showed that rubella had been eliminated by 2004 due to an intensive rubella vaccination programmes which had been started in 1969 [3,[10][11][12]. Rubella vaccine is widely available and is part of the routine vaccination programs in the developed countries but this is not the case in many African counties including Uganda. In Uganda there is limited or no pro-active process of routinely screening pregnant mothers for rubella and their possible vaccination.
Although the Uganda National Expanded Programme of Immunization (UNEPI) intends to introduce the rubella containing vaccine by 2019 there is no systematic documentation for epidemiological information up to 2012 pre vaccine era apart from information on the genotypes [13,14]. In Uganda, the Uganda Virus Research Institute (UVRI) EPI laboratory is part of the WHO network of laboratories that has its own database while the EPI program also has a more detailed measles case based surveillance epidemiological database within the routine Integrated Disease Surveillance and Response (IDSR). At the UVRI, Expanded Programme of Immunization Laboratory the serum samples for the suspected measles cases are tested for measles IgM and rubella IgM in parallel [15] & [16]. We analyzed data from the EPI program case based measles suspected cases data from January 2007 through December 2016 (10 years) to describe the epidemiology and associated factors and to provide a baseline for the monitoring of the impact of rubella vaccine which is to be subsequently introduced into the routine immunization in 2019.

Methods
We did a retrospective analysis of data collected during routine case based measles surveillance which is part of the routine integrated disease surveillance and Response (IDSR) in Uganda. This was also used to test rubella IgM among all measles suspected case samples.

Study setting
The study was conducted in Uganda an East African country with a population of about 34.7 million people and consisted of 112 districts for our study. The data was for suspected measles cases from all the districts in Uganda that were detected and reported from January 2007 to December 2016.
The case definition of a suspected case of measles was any person with fever and maculopapular generalized rash and cough or coryza or conjunctivitis or any person in whom a clinician suspects measles. The community case definition was any person with fever and a rash. A confirmed case was a suspected case that tested positive for the rubella Immunoglobulin M [17].

Data collection and serological testing
A case investigation form was filled by health worker in triplicate the third copy was submitted to the laboratory at Uganda Virus Research Institute (UVRI) with the blood specimen (serum). All collected specimen were tested for measles IgM and for rubella IgM in parallel using a standard Enzyme Immno-Assay (EIA), Enzygnost Anti-Rubella virus IgM test kit (Siemens, Marburg, Germany) [15]. The variables were those that are updated in the UNEPI case based database and included the following: Age (age group), sex (M/F), settings (rural/urban), region of residence, year of reporting, outcome of illness and whether hospitalized or not for illness.

Data analysis
The analysis was done using STATA version 13 software. Two-sided chi-square tests for association were computed to detect differences between categorical variables such as sex, age group, setting, region, admitted or not and outcome of illness. The means of continuous variables were computed. In order to investigate the association between the outcome variable of rubella IgM serology and other variables, bivariate logistic regression was first done then multivariate logistic regression models were run for all significant factors at 95% confidence interval in the bivariate analysis. The model building strategy was not only limited to significant variables from the bivariate analysis, but also included independent variables that were considered to have a clinical and social significance for the outcome of rubella IgM positivity. The evaluation of the associations of rubella IgM sero-prevalance was based on all 15,296 cases reported during the ten year study period.

Results: demographic characteristics of tested cases
The suspected cases were 16,551 and those that were tested for rubella were 15,296. There were 50.9% male, the largest age group was < 5 yrs. 58.3% (n = 8928). Most cases 10.9% were in the year 2011 and the largest number, 23.4% (n = 3985) from north central region of Uganda. Most 97.5% (n = 14,921) were from rural areas, 90.1% of the cases were not hospitalized and only ten of the cases died of the illness (cases fatality of 0.9% over ten years). Mean age was = 5.3 sd = 5 range 0 -80 years. See further details in the Table 1.

Rubella IgM+ prevalence
Overall prevalence from all 15 regions among 15,296 cases was 27.8% (ci: 27.1-28.5). The age group with the highest  Table 2.
Ten year rubella IgM prevalence trend

Factors associated to rubella prevalence
The factors that are significantly associated Rubella IgM prevalence are: age group 5-15 years OR 2.5 p-value    [20]. For the cases that tested negative for the rubella IgM depending on the effectiveness of our surveillance system the samples collected within 3 days or after 28 days of onset of the measles rash may not have sufficient detectable antigen. The WHO recommendation is that samples are taken within 28 days of the onset of the rash and this is what is done in Uganda to increase the possibility of not missing a positive test [15] & [16]. Among the measles suspected cases that were in the reproductive age group for Uganda of 15-49 years, 88 of the 322 (27%) were positive for rubella IgM this is a large percentage and an indicator that the female reproductive age group is also at risk. Although Njeru et al. 2015 had a lower prevalence in Kenya of 6% both  [13]. These out breaks are attributed to failure to provide the rubella vaccine containing vaccine. In the WHO America region we have seen that indigenous cases of rubella were eliminated. In this region rubella transmission was interrupted and CRS eliminated by 2009 and 2015 respectively it was declared to have eliminated rubella and CRS by the World Health Organization, this is consistent with [3,11,21] this is further evidence that introduction of the rubella vaccine containing vaccine is crucial for the elimination of rubella infection and Congenital Rubella Syndrome.

Limitations
Since we used the standard case definition for measles and yet we know that rubella is usually more mild in presentation than measles there is a possibility that the number of actual rubella cases may have been underestimated. About 50% of rubella cases may not present with a fever so this estimate may be lower than the actual rubella cases. Since the tested cases were suspected measles cases this data may not be inferred to the general population of Uganda. There is also a possibility that due to the varying surveillance performance over the 10 years some cases could have been missed. And also there are some socio and demographic characteristics of the cases that are not collected by the measles case form eg socio-economic status and level of education this could have left out some important associated factors.

Conclusions
The findings demonstrate a high level of IgM prevalence and endemic distribution of rubella in all the regions of Uganda throughout the year and over all the ten years that were analyzed. These findings provide further evidence of the high prevalence of rubella among the women of reproductive age, the infants, children and adolescents. This therefore gives us the evidence that the introduction of the rubella containing vaccine is long overdue given the success it has made in the America's and Europe.
In the context of the ongoing routine immunization performance of the past ten years and the success so far with measles, introduction of the rubella containing vaccine will enhance the control and elimination of rubella infection and Congenital Rubella Syndrome. The Uganda programme for Immunization is planning to introduce the rubella vaccine among and into the infants routine programme as a combination with the measles antigen as the MR vaccine formulation. This will immediately be followed with a catch up immunization to target adolescent females aged 9 to 15 years.
As a recommendation there is a need to start testing for IgG routinely in the future to help in the better classification of cases.