Measles outbreak in 2018-2019, Madagascar: epidemiology and public health implications

Introduction In October 4th, 2018, a measles outbreak was declared in Madagascar. This study describes the epidemiology of the outbreak and determines public health implications for measles elimination in Madagascar. Methods Data have been collected using line list developed for the outbreak. Serum samples were collected within 30 days of rash onset for laboratory testing; confirmation was made by detection of measles immunoglobulin M (IgM) antibody. Results A total of 2,930 samples were analysed in the laboratory among which 1,086 (37%) were laboratory confirmed. Measles cases age ranged from a minimum of 1 month to a maximum of 88 years. The median and the mean were 7 years and 9 years respectively. Children between 1 to 9 years accounted for 50.6% of measles cases. Attack rate (39,014 per 1,000,000 inhabitants) and case fatality rate (1.2%) were highest among children aged 9-11 months. A total of 67.2% cases were unvaccinated. As of March 14th, 2019, all the 22 regions and 105 (92%) health districts out of 114 were affected by the measles outbreak in Madagascar. Conclusion Measles outbreak in Madagascar showed that the country is not on the track to achieve the goal of measles elimination by 2020.


Introduction
Measles is one of the most contagious diseases of humans [1,2] with a basic reproductive number (the average number of secondary cases produced by a primary case in a completely susceptible population) of 12-18 [3]. Its causal agent is the measles virus. Measles occurs as a seasonal disease in endemic areas. Transmission is primarily person-to-person by airborne respiratory droplets that disperse within minutes, and transmission can also occur through direct contact with infected secretions [1,4]. Measles incubation period generally lasts 10-14 days (range, 7-23) from exposure to onset of fist symptoms, which usually consist of fever, malaise, cough, conjunctivitis, and coryza. The characteristic morbilliform rash appears 2-4 days after onset of the early symptom or prodrome. Patients are usually contagious from about 4 days before occurrence of the rash until 4 days after eruption, when the levels of measles virus in the respiratory tract are highest. Prior to the onset of rash, bluish-white Koplik's spots, which are pathognomonic for measles, may be seen in the oral mucosa. In uncomplicated measles cases, patients improve by the third day after rash onset, and have fully recovered 7-10 days after onset of disease. In tropical zones, most cases of measles occur during the dry season, whereas in temperate zones, incidence peaks during late winter and early spring [1]. Before the introduction of measles vaccine in 1963, major epidemics occurred approximately every 2 to 3 years and it is estimated that 30 million cases of measles and more than 2 million deaths occurred globally each year [1]. Measles Rubella elimination initiative launched at international level in 2012 was adopted by Madagascar [5][6][7]. The last measles outbreak occurred in the country in 2003 [8]. On October 4 th , 2018, a measles outbreak was confirmed by the National Reference Laboratory. This outbreak started in the capital city, Antananarivo, and extended to all the 22 regions of the Madagascar [9,10]. This study describes the epidemiology of the outbreak, and determines public health implications for measles elimination in Madagascar.

Methods
Data collection: data have been collected using line list developed for the outbreak. Line list was available in health facilities. Patient's information was recorded on the line list and a blood specimen is collected when he/she visited the health facility. Active search of cases was conducted by community workers for patient who did not visit the health facility. Information of patients was sent to the health facility and patients were asked to visit the health facility for management. list included the name of health district, year,   suspected disease, epidemiological week, patient's name, health   facility's name, residence place's name, sex, age, date of rash onset,   health facility visit date, symptoms (maculopapular eruption, fever, conjunctivitis, cough, coryza), immunization status (vaccinated against measles, date of vaccination, not vaccinated), lab test (blood sample, date of sampling, lab result), outcome (alive, dead, unknown), places visited 2 weeks prior the beginning of the illness, and comment (uncommon sign, hospital, community).

Variables in the line
Laboratory test: serum samples were collected within 30 days of rash onset for laboratory testing; confirmation was made by detection of measles immunoglobulin M (IgM) antibody at an accredited national measles laboratory using a standard commercial enzyme immunoassay indirect kit [11].
Case definition: a suspected measles case was defined as (i) any person with generalized maculo-papular rash and fever plus one of the following: cough or coryza (runny nose) or conjunctivitis (red eyes); (ii) any person in whom a clinician suspects measles. Measles suspected cases at community level was defined as any person with generalized rash and fever. Laboratory confirmed measles was defined as a suspected measles case that is investigated, including the collection of blood specimen, has serological confirmation of recent measles virus infection (measles IgM positive) and had not received measles vaccination in the 30 days preceding the specimen collection. Measles confirmed by epidemiological linkage was defined as a suspected measles case that has not had a specimen taken for serologic confirmation and is linked (in place, person and time) to lab confirmed cases; i.e. living in the same or in an adjacent district with a lab confirmed case where there is a likelihood of transmission; onset of rash of the two cases being within 30 days of each other. A confirmed outbreak of measles was defined as 3 or more measles IgM positive (laboratory confirmed) cases in a health facility or district in one month [12].

Sociodemographic characteristics of measles cases
From September 2018 to March 2019, a total of 112,693 measles cases were registered in Madagascar. The outbreak affected equally males and females. Overall, 56,356 cases (50.0%) were females and 56,061 cases (49.8%) males. The sex was not available for 276 (0.2%) cases. Measles cases age ranged from a minimum of 1 month to a maximum of 88 years. The median and the mean were 7 years and 9 years respectively, and the mode was 1 year. The most affected age group was 1-9 years. Children between 1 to 9 years accounted for 50.6% of measles cases. Also, 78.4% were below 14 years and 89.4% below 20 years old. Case fatality rate and attack rate were high among children less than 5 years (Table 1). Highest attack rates were among children below one year and specifically those between 9-11 months where this was almost 9 times the global attack rate. A total of 75,721 (67.2%) cases were unvaccinated. were also reported. The national attack rate was high with 4269 cases per 1,000,000 inhabitants. Overall, 52 (45.6%) districts had an attack rate greater than 2000 cases per 1,000,000 inhabitants (the vast majority of these districts are seaboard districts which are difficult to access), 24 (21.1%) districts had an attack rate of 1000-2000 cases per 1,000,000 inhabitants, 11 (9.6%) districts had an attack rate of 500-999 cases per 1,000,000 inhabitants, and 27 (23.7) districts had an attack rate of 1-499 cases per 1,000,000 inhabitants ( Figure 2).

Discussion
Children aged 1-9 years represented half of measles cases during 2018-19 outbreak in Madagascar. While children above one were more represented, measles seems to spread 2 to 9 times more rapidly than the average attack rate in children below one year. The highest attack rate among children between 9-11 months may be due to reduced maternal antibodies acquired at birth in this age group. In Madagascar, cases were mainly children aged less than 5 years.
Similarly, most of cases occurred among children under 5 years during a measles outbreak in Nigeria in 2013 [13], Cameroon in 2011 [14], in other African countries other the period 2002-2009 [15], and in Saudi Arabia in 2007. However, in Latin America countries such as Peru and Brazil people aged more than 15 years were most affected [16]. The average age for acquiring measles depends on biological and epidemiological factors, mainly population immunity and birth rate. In certain settings like low income countries or refugee camps, low population immunity, high birth rates and high population density, lead to increased transmission in younger age groups including infants and pre-school children. As vaccination coverage increases, the average age of measles infection can shift to adolescents and young adults [1,15]. The mean (9 years) and the median age (7 years) were comparable to those reported by Getahun Ghana, and Ethiopia [15]. However, in other countries such as Nigeria, Equatorial Guinea, Namibia, Rwanda, Tanzania, and Togo, males were more affected compared to females [13,15] while in Botswana, Zimbabwe, Mali, Swaziland [15] and Cameroon the majority of cases were females [14]. The variation of sex distribution of cases could be explained by the demographic features in the most affected age groups. In Peru outbreak, Sniadack et al. mentioned that being male and aged 16-20 years were risk factors for measles. According to the authors, one explanation for the predominance of young adult males among cases in this community may be that members of this group are more mobile or socially active than their female counterparts or elders [16]. In Madagascar, more than 60% of measles cases were not vaccinated. Measles occurred mainly among unvaccinated children as shown in previous outbreaks [13,18]. The last measles outbreak Global case fatality rate was 0.7 while age specific case fatality rate was greater or equal to 1% among children aged less than 5 years.
Most deaths occurred in health facilities at the beginning of the outbreak indicating gaps in the case management despite availability of treatment guidelines. These case fatality rate could be lower if free medicines for case management were available earlier and in adequate quantities in health facilities. However, this case fatality rate was lower than that reported in previous studies. This should be taken cautiously by taking in consideration poor reporting of deaths which occurred in the community due to weak community-based surveillance system. Generally, in Africa, the measles case fatality rate ranges from 3 to 5%, reaching up to 30% during severe outbreaks and outbreaks in closed communities such as refugee camps [4,20]. The severity of

Public health implications
This study provides four main lessons with potential implications for public health interest for policy makers, national EPI managers, and partners. Firstly, routine immunization programme has been performing poorly for more than a decade despite very high administrative coverage reported over the years and data quality issues that should be addressed. National forum on routine immunization could be an opportunity for the country and partners to rethink future orientations of the programme. Conducting a national EPI review after that of 2012 and an immunization coverage survey will provide baselines for this routine emergency programme. The place of the private sector in routine immunization implementation should also be discussed as an important part of the population seeks care in private health facilities. Secondly, introducing a second dose of measles vaccines in the national EPI programme is needed to provide a second opportunity to children missed during the first year on age to be vaccinated and reduce susceptible over the years.

Conclusion
Measles outbreak in Madagascar showed that the country is not on the track to achieve the goal of measles elimination by 2020. Much needs to be done to improve immunization performance in order to prevent other large-scale measles epidemics in Madagascar. We recommend a thorough review on the national EPI programme after the last one done in 2012. This will enable in identifying gaps and provide new orientations for better programme management.
Secondly, introducing the second dose of measles vaccine in the national programme will provide an additional opportunity for children • Failure to maintain high coverage of childhood immunization in all districts has resulted in a resurgence of the disease, in countries where vaccination has substantially reduced the incidence of measles.

What this study adds
• Children aged 1-9 years represented half of measles cases during 2018-19 outbreak in Madagascar; • The early mass immunization campaign against measles could have prevented many cases and the extension of the disease in all the health districts of the country; • There is a need of having sufficient measles vaccine stockpiles at global level.
We thank all people (community workers, health workers, health district management team, region management team, laboratory workers) and institutions (Madagascar Ministry of Health, and technical and financial partners) involved in the measles outbreak response in Madagascar.