High levels of mortality, malnutrition and measles amongst displaced Somali refugees in Dadaab, Kenya

Summary of published research¹

Location: Dadaab, Kenya

What we know already: Routine vaccination of children aged 6 months to 15 years, supplemented by mass vaccination campaigns, is one of the most important public health interventions among crisis-affected populations.

What this article adds: Other age groups may be affected by measles in different contexts. Early identification of unusual age-distribution of measles cases should guide vaccination policy and target groups in different settings. This will have resource implications and requires timely registration and data analysis.

Dadaab refugee camp is located in the North Eastern province of Kenya, approximately 100km from the border with Somalia, and was established in 1991. With an estimated 472,420 residents (8th July 2012), it is reported to be the world’s largest refugee camp complex, comprising three camps: Hagadera, Ifo and Dagahaley. As a result of the deteriorating humanitarian situation in 2010 (caused by the continued conflict in Somalia and by the failure of rains from October to November 2010), the number of Somalis seeking refuge in Kenya (and elsewhere in the region) steadily increased. Between January and November 2011, 154,450 individuals were newly registered in Dadaab, in addition to the estimated 63,000 refugees registered during 2010. An added complication was that in 2007, the Kenyan government closed the transit and registration facilities in the border town of Liboi due to security concerns and starting in August 2008, new arrivals were no longer allocated new plots of land. Without sufficient registration facilities to process them rapidly and without ready housing, the majority of new arrivals settled in the plains surrounding the main camps.

Médecins Sans Frontières (MSF) has been working in Dagahaley since 2008, providing medical care and psychological assistance to the population living within the camp and its environs, estimated at approximately 123,833 people (8th July 2012). In August 2011, MSF activities included nutritional interventions (in- and outpatient therapeutic feeding centres and supplementary feeding programmes (SFPs)). MSF operated programmes in five health posts, six ambulatory and six supplementary centres for nutrition programmes, and a 120-bed second level hospital with a further 200 beds serving as the nutritional stabilisation centre in the hospital. In addition, a network of community health workers actively sought malnourished children and other medical emergencies for referral for treatment.

In August 2011, with a view to understanding the health status and needs of the then newly arrived population, Epicentre and MSF conducted an exhaustive survey of all households in Bulo Bacte (BB), an area of ‘self-settlement’ outside the camp of Dagahaley, and analysed the data collected during a measles outbreak that affected this population at the time of the survey. A recently published article presents the survey estimates of death rates and malnutrition prevalence, and the age and sex breakdown of the suspected measles cases. The implications of these findings are then discussed.

Method

All inhabited structures were visited by the survey teams and when possible, a suitable respondent was identified with whom the interviews were conducted. Respondents were members of the household aged at least 18 years and were usually the person identified as the head of the household. MSF collected information relating to mortality during the recall period in all eligible households. For any death reported during the recall period, information was collected from the household respondent regarding suspected cause or symptoms associated with the death. MSF also attempted to identify the timing of the death with the aid of a ‘calendar of local events’.

Mid-upper arm circumference (MUAC) measurements were taken and the presence or absence of bipedal oedema recorded, for all children of height 67- <140 cm (proxy for children aged 6 months - 9 years) living in included households on the survey date. These data were used to calculate: a) the age-specific prevalence of severe acute malnutrition (SAM) and global acute malnutrition (GAM) for children of height 67- < 87 cm (proxy for children aged 6 - 23 months) and children of height 87- < 110 cm (proxy for children aged 2 - 4 years) and b) the proportion of children of height 110- <140 cm (proxy for children aged 5-9 years) meeting the admission criterion for entry into the nutritional programme (MUAC<140 mm). Cochran-Armitage tests-for-trends were performed to explore the effect of duration of stay in BB on prevalence of acute malnutrition. MSF also collected information on arrival dates of individuals in BB. The study population included all people living in BB. An exhaustive method was selected to allow stratification of the results by period of arrival, location, and age-group.

Findings

The surveyed population included 26,583 individuals, of whom 6,488 (24.4%) were children aged under 5 years. There were 360 deaths reported during the 177 days of the recall period, of which 186 (52%) were among children aged under 5 years. The crude death rate for the entire recall period was 0.8 per 10,000 persondays. The under-5 death rate was 1.8 per10,000 person-days. More than two-thirds of all deaths were reported to have been associated with diarrhoea (25%), cough or other breathing difficulties (24%), or with fever (19%). Measles accounted for a reported 17% of all deaths, due to a measles outbreak that occurred between June and October 2011. GAM was observed in 13.4% and SAM in 3.0% of children measuring 67- <110 cm. Among children measuring 110- <140 cm, 9.8% met the MUAC admission criterion for entry into the nutritional programme. A trend of decreasing malnutrition prevalence with length of stay in BB was observed.

These findings reflect an alarming situation. Both crude and under-5 death rates and the prevalence of SAM and GAM were at emergency levels in the outskirts of a refugee camp that was served by several international nongovernmental agencies (NGOs) and United Nations (UN) agencies. The recall period coincided with an outbreak of measles in Dagahaley that began in June and continued until October 2011, and which partially explains the increasing under-5 death rate observed; 17% of all deaths were reported to have been caused by measles. As a result of the two decades-long civil war in Somalia, vaccination coverage among all ages has declined to the point that outbreaks of infectious diseases such as measles are increasingly likely. Indeed, there was a simultaneous measles outbreak among newly-displaced Somalis in Kobe refugee camp in Ethiopia and in several places in Somalia, including the capital Mogadishu. Outbreaks of measles among populations in crisis are common and well documented. The routine vaccination of children aged 6 months to 15 years, supplemented by mass vaccination campaigns, is widely accepted as one of the most important public health interventions for averting preventable morbidity and mortality among crisis-affected populations.

Figure 1 shows a timeline of various measles vaccination interventions conducted in BB. A mass measles vaccination campaign was organised at the end of April 2011, targeting children aged between 9 months and 15 years, with a follow-up campaign in July 2011 to vaccinate those children aged under 5 who did not receive measles vaccination in April.

When a registration centre was opened within Dagahaley in June 2011, all children aged 9 months to 15 years were routinely vaccinated against measles upon registration. In response to the outbreak, a reactive vaccination campaign (RVC) targeting children under 5 years was launched throughout Dagahaley camp in early August, and the target age group for vaccination at the registration centre was increased to 30 years. In September 2011, a RVC was organized targeting individuals aged 15 to 30 years. It is worth noting that the RVC launched in August 2011 ran concurrently with the survey described in the article. The decision was therefore taken not to include measles vaccination coverage in the survey, because the results would have had no influence over any decision to launch such a campaign and because the coverage at the end of the survey would have been different to the coverage at the start, thereby rendering the results immediately invalid. UNHCR and partner organisations assessed measles vaccination coverage in BB shortly after this campaign and reported a coverage of 83.9% (95%CI: 73.7 - 94.0%) among children aged 9-59 months. The authors argue that the measles outbreak was preventable; the essential lessons from past mass displacements should have been learned and a suitable aggressive vaccination strategy implemented at an earlier stage.

In mitigation, the population most affected were those who had recently arrived, containing a large proportion of families unregistered by camp management due to the overwhelming arrival rate of these refugees. The late establishment of the registration centre and consequently vaccination at arrival permitted the development of a pool of susceptible individuals in BB. In addition, the measles outbreak in Dagahaley was characterised by an unusual age distribution; the median age of patients recorded in the outbreak line list was 23 years, with 75% of patients aged 15 years or older, suggesting that a wider age group could have benefited from vaccination. However, the current ‘one-size-fitsall’ recommendations are to vaccinate all children aged 6 months to 15 years and do not take into account the context-specific epidemiology, which in this case included a highly immunologically-naïve population due to the breakdown of healthcare services arising from the ongoing political crisis in Somalia.

Early identification of the unusual age-distribution of measles cases would have helped guide vaccination policy in this setting. Indeed, the disaggregation of deaths attributed to measles by age and by month shows that age distribution of measles cases was detectable in June 2011, at an early stage of the epidemic. However, this would have required information that was not available at the time: low health facility utilisation rates and under resourced community-based surveillance of epidemicprone diseases meant that most measles cases occurring before July were not detected. In July 2011, by which time an outbreak had been declared and active community-based surveillance strengthened, more data were available, which led MSF to advocate for a wider target age group for the RVC planned for early August. However they were unsuccessful due to the limited resources available for that particular campaign. The target age group for vaccination at registration was, however, expanded to include all individuals aged 9 months to 30 years. Owing to the failure of the August RVC to halt the epidemic, which peaked in August and September, adults aged 15-30 years were the target age group for the subsequent RVC.

More recent arrivals were in a significantly worse state, which was reflected both in death rates and in nutritional status. The survey found trends of decreasing death rates with length of stay in BB, such that those residents who had arrived more than six months prior to the survey date had death rates well below the emergency thresholds, while those who had arrived within three months of the survey date had death rates which were above the emergency thresholds. Similarly, MSF found a higher prevalence of acute malnutrition and children meeting the admission criteria for entry into the nutritional programme among those children who arrived during the three months prior to the survey than among those children who arrived earlier. The same pattern was reported in a subsequent survey conducted in BB.

This apparent improvement in health and nutritional status over time may be due to the assistance gained after registration and the development of coping strategies, but may also be due in part to: a) a high concentration of deaths in the period immediately prior to the survey date (in particular, due to the measles outbreak), which may have resulted in artificially elevated death rates among recent arrivals due to the relatively low person-days contributed by these individuals (in other words, a low denominator rather than a high numerator used in the calculation of death rates), and b) higher mortality among those children with poor nutritional status on arrival since both recovery and death have the effect of decreasing the prevalence of malnutrition by removing these children from the numerator in the calculation of malnutrition prevalence. Those malnourished children who had recently arrived had had less time in which to experience either of these outcomes.

One limitation of this study is that it did not use a validated verbal autopsy technique when obtaining information on cause of death as this was not a principal objective, and therefore these results should be interpreted with caution. Furthermore, standard case definitions for measles deaths (any death within one month after rash onset) were not used; it is likely that some of the deaths associated with diarrhoea, cough or breathing difficulties and fever were in fact cases of measles.

The authors conclude that there was an unacceptably high death rate and prevalence of malnutrition in the outskirts of a long established refugee camp, albeit among a recently-arrived population. Although the levels of malnutrition may be partly explained by the poor health of new arrivals, the high mortality among refugees after their arrival in Dagahaley reflects a failure of the various humanitarian and governmental actors to safeguard adequately the welfare of this population. While the massive influx of refugees did pose enormous difficulties, outbreaks of measles and long delays before registration (which permits access to food distributions) should not have occurred. The recommendations for measles vaccination among crisis-affected populations should be revised to take into account the epidemiologic context.

¹Polonsky. J et al (2013). High levels of mortality, malnutrition and measles among recently displaced Somali refugees in Dagahaley camp, Dadaab refugee camp complex, Kenya. Conflict and Health 2013.7:1. Doi:10.1186/1752-1505-7-1 Open access at http://www.conflictandhealth.com/content/7/1/1/abstract