Malnutrition Precipitated Measles Outbreak in Gewane District, Afar Regional State, Northeastern Ethiopia, 2016

Background: Measles is a contagious viral and vaccine-preventable disease that is continuing a public health problem. It is occurring as an outbreak not only in developing but also in developed countries. Methods: A 1:2 ratios, unmatched case-control study was conducted from September 19/2016- October 08/2016 in Undufo Kebele, Gewane district, Northeastern Ethiopia. Both conrmed by laboratory and epidemiologically linked measles cases were involved. Controls were those who had no clinical signs of measles during the data collection period and residents of the same community where cases were identied. An interviewer-administered questionnaire was used to collect data. Nutritional status was measured using MUAC and the result decided based on the world health organization's nutritional categorization of children by MUAC. Data were entered into Epi-info 7 and analyzed using SPSS-21. Binary logistic regression was done to identify risk factors for measles infection with 95% condent level of Odds ratio. Result: Fifty cases and 100 controls have participated in this investigation. Four out of ve samples collected were positive for measles IgM antibody. There were three deaths reported. There was no vaccination history of cases and controls. The majority of cases were female and between 6–15 years of age. In multivariable analysis, malnutrition (AOR = 3.21; 95%CI 1.871,6.334), and contact history (AOR = 12.24; 95%CI 6.992,28.121) increased the risk of contracting measles infection. Conclusion: children under 15-year age were more affected groups. The absence of vaccination in the area precipitated by a high rate of malnutrition was the main cause that aggravates the number of new infections. Contact history and malnutrition were the identied risk factors for measles infection. Therefore, increasing immunization more than 80%, securing food of

area precipitated by a high rate of malnutrition was the main cause that aggravates the number of new infections. Contact history and malnutrition were the identi ed risk factors for measles infection.
Therefore, increasing immunization coverage of more than 80%, and securing food security decrease the susceptibility of outbreak occurrence.

Background
Measles is a highly contagious, vaccine-preventable viral illness characterized by fever, malaise, rash, cough, Coryza, and conjunctivitis. Initial signs and symptoms typically include fever, often > 40 o C, cough, runny nose, and red eyes (1). Two or three days after the start of symptoms, small white spots may form inside the mouth known as kopliks spot. A red at rash which usually starts on the face and then spreads to the rest of the body typically begins three to ve days after the start of the symptoms. The complication occurs in about 30% and may include diarrhea, blindness, in ammation of the brain, and pneumonia among others. Infectivity is the greatest three days before rash onset, can also be transmitted four days after onset of rash. Secondary attack rates among susceptible household contacts have been reported to be 75-90% (2,3).
Eradication is possible because measles has only human reservoir; however, due to social and political factors and high transmissibility, elimination and eradication have been achieved in very few areas of the world(4). Worldwide, measles is a signi cant cause of morbidity and mortality. Before the introduction of the measles vaccine, over two million deaths occurred annually, the majority in children < 5 years of age (5). Nine out of ten people who were not immune and share living space with an infected person(6), children too young to be vaccinated or not vaccinated, older age, being female (7), not received a second dose, those for whom the vaccine failed to provoke a protective immune response, travel to areas where measles is endemic (1,8) having contact history(9) could contract measles infection. The risk of death among those infected is usually 0.2% but probably reached up to 10% in those malnourished peoples (5,10). In developing countries, the expected Case-Fatality Rate(CFR) is between 3% -5%, but it could beyond 30% in high-risk populations; like in infants aged less than one year of age (3,11). The attack rate of the measles outbreak recorded in Ethiopia as nationally was 4.1 per 100,000 populations, in 2008 (2). In 2013, 13.4% CFR of the outbreak from brief communication in South Gondar zone, Ethiopia, in2009 (12).

Study setting and period
Unmatched case-control study design was conducted from September 29/2016 to October 15/2016 in Undufo Kebele (Kebele: lowest admin level in Ethiopia), Gewane district, Afar Region, Ethiopia. The district is located 385 Km from Addis Ababa (capital city of Ethiopia) and 222 Km from Semera. The total population of the district is 38,991 as projected from the 2007 national census. There is no Hospital but has three health centers and nine health posts. During the outbreak period, the district was seriously affected by drought and there was a scarcity of safe water for drinking. Hence, the population enforced to use from unsafe open water sources (Awash River and lwuha unprotected spring).
Case de nition: A measles positive case was any person who developed any one of the following symptoms; fever, cough, Coryza, red eyes, maculopapular rash or tested IgM positive during the study period (13).

Enrolment of Cases and Controls
Case: Any person who ful lls the case de nition given above during the data collection period was included as measles cases.
Control: Any person who did not show any one of the signs from the case de nition stated above during the data collection period, and is residents of the district were taken as controls.
Since the vaccination status of the populations in which the outbreak occurred was zero percent, the presence of epidemiological linkages as well as clinical case de nitions were used to draw an indication of occurrences. Therefore, it was not compulsory to con rm all cases by laboratory tests. Thus ve blood samples were collected from patients and were sent to Ethiopian Public Health Institute for laboratory con rmation.

Data Collection Procedure
House-to-house searching was done to include and manage cases using line list recording. A structured questionnaire was used to collect data from cases and controls. A total of eight questions were asked to assess participants' knowledge about measles. Respondents who scored the mean and above were considered as having "good knowledge", whereas, those scored below mean had "poor knowledge" about measles. The vaccination coverage of the district and moreover Undufo Kebele was taken from the district's annual report.

Data Processing and Analysis
The data were entered using Epi-Info 7.1 and imported to SPSS version 21 for advanced analysis. Results were presented using graphs and tables. Attack rate and case fatality rates calculated. Odds ratio (OR) with a 95% con dence interval was computed to identify factors associated with cases and controls.
First, a bi-variable analysis was done for each independent and outcome variables. Here, those variables had a P-value of < 0.2 were tted to multivariable analysis. In the multivariable logistic regression model, a p-value < 0.05 to identify statistically signi cant variables and 95%CI of OR to indicate the strength and direction of association were computed.

Ethical Issues
Ethical approval was assured by the University of Gondar, Institute of Public Health, College of Medicine and Health Science Ethical review board. Permission was gained from Afar regional health Bureau to conduct the investigation. Verbal assent from children between 5-18 years of age and verbal consent from the parents were obtained and the information gained were kept con dentially.
Furthermore, each new measles cases were managed as per national guideline for measles surveillance and outbreak investigation, Ethiopia (13). Also, a vaccination campaign was conducted.

Epidemiological characteristics of participants
A total of 50 cases and 100 controls were included in the investigation. Measles cases were reported since September 19/2016 in Undufo health center, Gewane district. Those cases were originally from Ididi and Badawayne areas of Undufo Kebele and they arrived at the area of the outbreak before a month. There were three deaths (CFR = 6%) reported during the study period. Among 50 measles cases, 29(58%) were female. A majority, 27(54%) of 50 cases were in the age category of 5-14, followed, 19(40%) by 1-5 age group. All most all, 47 (94%)of the cases were from Undufo while the remaining 3 of the case were from Den hele area ( Table 1). All of the currently affected residents were from Badawayne and Ididi remote areas. These people had migrated to this area before two months because of the scarcity of water and food because of drought. It could be concluded that there was no vaccination activity in the area.
Children below ve years' age were more affected (Attack Rate = 43.5) followed by 6-15 age group (AR = 15.2) per10,000 populations ( Table 2). Overall the attack rate was 19.1per 10,000 population (108 total cases were recorded until the outbreak controlled over), Of which the attach rate was higher (32.2 per 10,000 populations) among females.

Clinical presentation of cases
Forty-seven (94%) of 50 cases had a fever, while conjunctivitis Coryza and rash were present in 89%, 75% and 66% of the study subjects, respectively. None of the respondents with the disease had blindness (Table 3).

Factors associated with measles infection
Since the variables were small in number, we analyzed all variables in multivariable logistic regression. Therefore, after adjusting for confounding effects of variables, in the multivariable model, malnutrition (AOR = 3.21; 95%CI 1.871,6.334), and contact history (AOR = 12.24; 95%CI (6.992,28.121) increased the risk of contracting measles infection by 3.21 and 12.24 times more likely than their counterparts respectively. The other independent variables like; vaccination history, age, educational status, occupation, and sex did not have an association with the dependent variable at α = 0.05 signi cance level (Table 4). This study was interested to identify risk factors and give an appropriate intervention in Undufo Kebele, Gewane district Afar regional state. The outbreak was declared after four from ve serum samples were positive for measles IgM antibodies. In the Ethiopian context, the measles outbreak is declared if ve measles suspected by a physician or three con rmed by the laboratory at the district level (13). Since the outbreak was local, all other cases were epidemiologically linked to laboratory-con rmed cases. Measles is highly contagious and one index case can produce up to 20 new cases in a community in which the whole is susceptible, and the secondary attack rate can be 75-90% (14). In this investigation, three measles deaths (CFR = 6%; 95%CI 5.934,6.065%) were recorded, which was higher compared to the expected case fatality rate (3-5%) in developing countries including Ethiopia (13,15). This might be due to sine this study area was affected by drought, where if the area is deprived by nutrition, the CFR will have reached up to 10% (13). This fatality rate was less than the outbreak investigation report from the South Gondar zone (CFR; 13.4%) (12). This might be due to the regional state, emergency management team deployed by the Federal Ministry of Health and MSF Spain collaborative case management effort in this study. In this study, the attack rate was 191 per 100,000 populations, which was higher compared to study in Tigray regional state in which less than 80 cases per 100,000 populations reported (2). The majority of the cases 32(64%) were in the age category of 5-14, under one population are not affected, which is incomparable with the observations made by other outbreak investigation studies (10,12). In this study females (58%) were more affected which also incomparable to the Somali refugee camp outbreak response, in which there is no sex difference, but similar to report in the South Gondar zone (2). One of the strategies achieving and maintaining high coverage ( > = 90% nationally and > = 80% in district) are to prevent outbreaks (13). But measles vaccination coverage in Undufo Kebele was zero, even the district's coverage is 70% (Gewane district 2015 annual health report). According to this study nding, there was no association between vaccination and the chance of acquiring measles virus. But Vaccination is known to be the main protective factor against measles (8,16). This might be no vaccination history of both the case and control groups.

Discussion
This nding revealed marital status, educational status, religion, occupation, ethnicity, area of living room, knowledge about transmission have no variability among the case and control and no association at α = .05, this was incomparable to other studies (8,11,16).
Malnourished children were 3.21 times more likely to be affected by measles than children who were not malnourished (AOR = 3.21; 95%CI 1.871,6.334). This was similar to different studies and guidelines, in which malnutrition aggravates acquisition of measles infections (1,13).
Children who had contact history with other measles cases were 12.24 times more likely infected with measles virus than those children did not have contact history (AOR = 12.24; 95%CI (6.992,28.121), which was similar to other studies (9,10,16,17).

Conclusion
Page 10/12 The main potential factors that led to a measles outbreak were the absence of vaccination at the area precipitated by a high rate of malnutrition. The leading causes of the spread of the disease were contact history at the household level, and malnutrition aggravated by the presence of drought in the area. There was no vaccination in the area, therefore, increasing the coverage to more than 80%, and securing the nutritional status of residents will decrease the numbers of susceptible individuals and risk of outbreak occurrence.
Abbreviations AOR: Adjusted Odds Ratio; COR: Crude Odds Ratio

Declarations
Ethics approval and consent to participate Permission was obtained from university of Gondar, College of Medicine and Health Sciences institute of public health ethical Review board, Afar regional state health bureau and Gewane district health o ce to investigate the outbreak. We have also obtained verbal consent from each study participants who are 18 and above years age, and for those between 5 to 16 years we taken verbal consent from them and written consent from their parents/guardians. But for the children under 5 years we obtained assent from their parents/guardians according to Ethiopian research ethics guideline.

Consent for publication
Not applicable

Availability of data and materials
Questionnaire used for data collection is available from the corresponding author on reasonable request. All other necessary data are included in the manuscript.

Competing interests
The author declared that there is no competing interest Funding This investigation received no speci c grant from funding agency in public commercial, or for pro t sectors.