Sierra Leone: a review of surveys from 2000 to 2007

Transcription

1 234 Briefing note Sierra Leone: a review of surveys from to 2007 Ruwan Ratnayake Olivier Degomme September 2008 Complex Emergency Database

2 Sierra Leone: a review of surveys from 2000 to 2007 Ruwan Ratnayake 1, Olivier Degomme 2 September 2008 Abstract Sierra Leone is known for extremes in poverty and poor health and statistically, the country ranks low in the life expectancy, child mortality and human development league tables. From 1991 to 2002, Sierra Leone suffered through intense civil war between the government and rebel groups. The analysis of CE DAT data for the period 2000 to 2007 reveals steady or decreasing trends in CMR, U5MR and GAM in selected districts and camps, but since these indicators are still crossing the emergency thresholds as late as 2006, it is recommended that NGOs continue to monitor the population using nutrition and mortality surveys. Requested by: Action Contre la Faim, France (ACF) 1 ruwan.ratnayake@gmail.com 2 olivier.degomme@uclouvain.be 1

3 1. National level indicators From 1991 to 2002, Sierra Leone suffered through intense civil war between the government, the Revolutionary United Front (RUF) and other rebel groups. After the declaration of the end of the war in 2002, UN peacekeepers helped to restore order from 2002 to By 2006, the UN established the Integrated Office in Sierra Leone (UNIOSIL). Pre war health services were not adequate and the conflict decimated the existing health infrastructure. Sierra Leone is now known for extremes in poverty and poor health and statistically, the country ranks last or second to last in the life expectancy, child mortality and human development league tables (1). Sierra Leone s crude death rate (CDR) is 23 deaths per 1,000 persons per year which corresponds to a crude mortality rate (CMR) of 0.63 deaths per 10,000 persons per day (2). This is above the baseline mortality rate for Sub Saharan Africa (0.44) (3). The under 5 mortality rate (demographic and measured retrospectively) for 2002 was 270 deaths per 10,000 under five which is lower than the 286 deaths per 10,000 under five measured in 1997 (4). However, the U5MR remains exceptionally high, ranking Sierra Leone as having the highest U5MR out of 189 countries (5). These values mark the tail end of the conflict period and as such, are expected to have remained static. The Plasmodium falciparium form of malaria is a leading cause of morbidity in Sierra Leone (6) (7). Transmission is hyperendemic to holoendemic and the disease contributes to 48% of outpatient consultations (6) (7). The long standing conflict severely degenerated disease control programmes and infrastructure and hindered infection control practice leading to lapses in malaria control and the emergence of Lassa fever (6) (8). MSF, Merlin and IRC have dedicated much of their efforts to building back disease control programmes. 2

4 2. Global Acute Malnutrition Figure 1: GAM (children under 5), to GAM, MICS, 2000 VAM, 2003 MICS, 2005 VAM, East North South West Numerous country wide surveys have been carried out since the end of the conflict: UNICEF s Multiple Indicator Cluster Survey (MICS, 2000, 2005), WFP s Vulnerability and Assessment Mapping (VAM, 2003, 2005) and Sierra Leone Integrated Household Survey (SLIHS, 2003). The 2005 MICS and VAM surveyss show the most recent trends in child malnutrition among children under 5 (Figure 1) (4). Figure 2: GAM (children 6 59 months) ),

5 The 2005 VAM survey found very high prevalence (12 17%) that is not compatible with the trends produced by the other MICS and VAM surveys. The report suggests that differences in sampling methods and inaccessibility to poor, rural areas in 2003 skewed results for that survey. Likewise, the 2003 SLIHS found GAM prevalence among children under five to be two to three times the magnitude of that reported in the MICS and VAM surveys (not graphed, East 36.1%, North 34.3%, South 31.5%, West 15%) (9). It is unclear why these values are so high, but they are inconsistent with the results of other surveys. The 2003 VAM shows more detailed results at the provincial level for children 6 to 59 months of age (10). In 2003, values in the Northern districts of Bombali (10.8), Tonkolili (10.6) and Port Loko (8.3) were near or above the emergency threshold for serious nutritional emergencies. Unfortunately, anthropometric indicators for the 6 to 59 month age group from the 2005 VAM have not been made available for comparison (11). 2.1 Determinants of GAM Although these surveys results are considerably divergent and the difference in geographical coverage precludes a straightforward comparison, the bottom line is that GAM is at or near serious levels ( 10%) to critical levels ( 15%) in all provinces in the most recently measured period of 2005 (Figure 1) 3. In this context, global chronic malnutrition showed an overall reduction from 40.5% in 2003 to 29.9% in The 2005 VAM found critical GAM levels in the Northern province (Bombali, Kambia, Koinadugu, Porto) and districts of Western, Bonthe and Kenema (Appendix 1). It is notable that acceptable (<5%) and poor GAM prevalence (5 9.9%) is found in the districts currently served by NGO driven large scale health initiatives (Table 1). Table 1: GAM, children under 5, by district served by NGO District GAM (0 59) NGO Kono 5.87 IRC Bo 7.99 MSF Pujehun 6.84 MSF Childhood nutritional status reflects access to food, lack of illness and appropriate feeding practices, including timely initiation of breastfeeding. Determinants of high GAM in Sierra Leone are likely related to poor dietary diversity, food access and consumption patterns in 3 Acceptable, serious and critical GAM prevalence levels have been delineated for the 6 59 month age group. For the <6 month age group, it can be expected that breastfeeding delays the onset of acute malnutrition. Therefore, standardized GAM thresholds for the 6 59 month age group are used for the 0 59 months age group. In essence, it is anticipated that GAM prevalence for the 6 59 age group will be higher, as it excludes children who are breastfed. 4

6 the Northern province (specifically in Bombali and Koinadugu) and borderline food access and consumption patterns in the Southern province (particularly in Bonthe, Moyamba and Bo) (11). According to the 2005 VAM, high food prices in the hunger season (June to September) combined with reduced agricultural production likely contribute to poor access to food. Finally, MICS 2005 found that exclusive breastfeeding (8%, 0 5 months), timely complementary feeding (52%, 6 9 months), continued breastfeeding (87%/57%, months/20 23 months) are all below West African standards and are believed to contribute to poor nutritional status (4). 3. Measles vaccination coverage In 2005, measles vaccination coverage (MCV) was 76.6% among children months (33.5% with vaccination card and 43.1% declared by the mother); this falls short of UNICEF s target of 90% coverage by one year of age but is still above the regional average and the 2003 MCV coverage of 62% (4). This indicates that the Expanded Programme on Immunization was successful in reaching more children in the reporting period. This occurred after substantial delays and challenges to immunization during the 1991 to 2002 conflict period (12). 4. CE DAT data: review of NGO data sources The data on CMR, U5MR and GAM was compiled from the CE DAT database (13). As of August 27, 2008, CE DAT has received 41 surveys from Sierra Leone for the period 2000 to present. Surveys were conducted either jointly or independently by ACF, GOAL, IRC, Merlin, MSF, Sierra Leone Ministry of Health, UNHCR, WFP and World Vision. The findings describe the period which is covered by the surveys received by CE DAT although a large drop in the number of surveys conducted can be seen after This timing coincides with the formal end to conflict. From that point, data is sourced from MSF and IRC s humanitarian operations in Bo, Pujehun and Kono. 5

7 4.1 Crude Mortality Rates (CMR) and Under Five Mortality Rates (U5MR) 4.2 The end of the conflict period: high, sustained mortality rates across the country The conflict is Sierra Leone is well known for its brutal violence and gross human rights violations committed against civilians, particularly of the senseless amputations of civilians, systematic sexual violence and the forced recruitment of children (14). Much of the violence occurred in the South and East but spread quickly to the rest of the country and culminated in the siege of Freetown. As one of the most violent battles between the RUF and the government, the senior government pathologist registered burials for 0.7% of the population (7,335 persons) who were killed over a three week period in January 1999 alone (14) (15). Using multiple data sources, researchers have estimated that total excess mortality for the 1991 to 2002 conflict period was 460,000, only 6% of which was due to direct violence (16). The CE DAT data covers a period of increasing stabilization and quelled violence (2000 to present) which includes disarmament programs starting in September 2001 and the formal declaration of the end of the conflict on January 18, 2002 (see Figures 3 and 4). Despite a gradual return to peace, crude mortality rates (CMR) from surveyed areas (districts of Kenema, Kono and Bo and cities of Kenema and Porto Loko) taken during the conflict period to the end of conflict uniformly exceeded the emergency level threshold of 1 death per 10,000 per day. Similarly, U5MR in Kono, Bo and Porto Loko city exceeded the threshold of 2 under five deaths per 10,000 under fives per day. Violence was a very small contributor to deaths in this period and deaths were mainly disease related. However, with the exception of Porto Loko in the Western province and Kono in the Eastern province, the districts surveyed were relatively secure by the end of the conflict period. These districts were plagued by a decimated health infrastructure and extensive population influx, with displacement from the war affected Northern and Eastern provinces into camps and towns in Bo and Kenema. 6

8 2,5 End of conflict period 2? Bo Bo, Pujehun MSF catchment area 1,5 Kambia, Tonkolili, Bo, Bombali Kenema 1 Kono Porto Loko 0,5 Kenema City 0 déc. 99 avr. 01 sept. 02 janv. 04 mai 05 oct. 06 févr. 08 Figure 3: Crude mortality rates (deaths/10,000 persons/day), 2000 to End of conflict period 6 Bo 5 4? Bo, Pujehun MSF catchment area Kambia, Tonkilili, Bo, Bombali Kono 3 Porto Loko Kenema city 2 Kenema 1 0 déc. 99 avr. 01 sept. 02 janv. 04 mai 05 oct. 06 févr. 08 Figure 4: Under five mortality rates (deaths under five/10,000 under fives/day), Kenema District in the Eastern Province was the site of early violence but had been secure since its last period of rebel violence between RUF and CDF in 1999 (17). The survey period of the year 2000 was still characterized by poor humanitarian access and heavy displacement from other conflict affected Eastern Districts (18). CMR ranged from and U5MR rose to 1.8, all surpassing or nearing emergency thresholds. 40% of deaths were due to febrile illness (suspected malaria), with a striking 30% of deaths borne by infants under one 7

9 year of age producing an extremely high infant mortality rate of 303 per 1,000 live births. Violent deaths were rare and accounted for only 2.5% of total deaths. Malaria and malnutrition (for under fives) was repeatedly self reported as the main causes of death in Bo District in the Southern Province and in Porto Loko town in the Western Province. 4.3 The Post conflict situation in areas of high displacement: Kono District Surveys appear to be scarce after The exception is found in catchment areas receiving humanitarian assistance from MSF and IRC. Kono, which shares a border with Guinea, is an economically lucrative area due to the presence of diamond mines. As a result, the RUF invaded and controlled this region throughout the war, causing the displacement of residents to neighbouring districts. Sporadic fighting between the RUF and Guinean forces continued from 2000 to 2001, just before disarmament initiatives. By May 2000, UN peacekeepers embarked on a largely successful disarmament campaign of 25,000 of the 45,000 rebels only to begin battle with RUF troops in Kono. Kono has been the site of a comprehensive child mortality prevention program introduced by the IRC in October The aim of the program is to target three high burden childhood conditions, malaria, diarrhoea and pneumonia (19). Large decreases of almost CMR and U5MR can be seen between the July 2001 and May 2007 period (see GREEN ARROWS in Fig. 3 and 4). This occurred in the presence of the community treatment program for which community workers used evidence based interventions including Vitamin A and insecticide treated to combat malaria. Between 2005 and 2007, the symptoms of death in under fives decreased for presumptive malaria and presumptive pneumonia in the 2007 survey. Although U5MR was consistently lowered, there was an unexplained increased in over five mortality rate (the difference of CMR and U5MR) between the 2005 and 2007 survey. 8

10 4.4 Three years after conflict: high magnitude decline in mortality in Bo and Pujehun Districts MSF has operated in Sierra Leone since 1995, with programs in Bo and Pujehun and a particular focus on malaria control. Aided by other MSF sections, MSF Belgium has carried out three surveys of their catchment area of Bo and Pujehun in , 2006 and 2007 (20). Very high CMRs ( ) above the emergency levels and higher than conflict era CMR from Bo were recorded in 2005 and 2006 (MSF OCB, unpublished data). U5MR was similarly higher than emergency levels ( ) and at the same magnitude as conflict era U5MR. In both 2006 and 2007, malaria/fever was the most frequently reported cause of death for all age groups. In 2007 however, much lower CMR (0.7) and U5MR (1.3) placed the population below the emergency thresholds. The cause of such a large and rapid decline in mortality is difficult to specifically link to a specific intervention though the population certainly benefited from better access to free health care, malaria treatment and use of bednets. MSF plans to carry out another survey in October 2008 to assess this trend. 30 End of conflict period Bo Bo Pujehun MSF catchment area 25 Bombali SERIOUS CRITICAL Bonte Kailahun Kambia Kenema City Kenema Koinadugu Kono Moyamba Porto Loko town 5? Porto Loko Pujehun Tonkolili 0 avr. 00 août 01 déc. 02 mai 04 sept. 05 févr. 07 Freetown Western rural only Figure 5: Global acute malnutrition (%), The 2005 survey also included Bombali, Tonkolili and Kambia. This multi-district mortality rates and the Bospecific mortality rates are included in Figures 3 and 4. 9

11 5. CE DAT: Global acute malnutrition Global acute malnutrition (GAM) prevalence over time is reflective of trends in CMR and U5MR (see Figure 5). Unfortunately however, surveys for the post 2003 period are scarce and the results for the age group of 6 59 months of age from the 2005 WFP VAM survey have not been made available. The only recent GAM figures, from 2006 and 2007, in the MSF catchment area in the Bo Pujehun District, show a stable and acceptable nutritional situation (5%, 5.8%). This is generally in line with the poor to acceptable GAM prevalence among the 0 59 month age group in Bo (7.13%) and Pujehun (7.99%) found in the 2005 VAM survey. Paying closer attention to the conflict to post conflict transition, GAM were generally at, or close to, stable, acceptable levels ( 5%) in Bo and Kenema Districts from 2000 to the end of the conflict and in Freetown in mid 2004 (see the PURPLE BOX). GAM in the Northern districts of Tonkolili and Bombali nearly doubled from stable to serious levels. Kono District experienced a substantial drop in GAM from critical to stable levels; this could be due to the increased humanitarian access in an area previously a fiercely protected RUF stronghold for the diamond trade which experienced tremendous displacement. 6. Refugee camps: a typical relief scenario Liberian refugees of the second Liberian civil conflict between the government and LURD rebels, fled to Sierra Leone between 2001 and After large influxes which started in February 2002, UNHCR set up eight camps in the districts of Bo, Kenema, Pujehun and Moyamba to accommodate around 55,000 refugees by 2003 (see Figure 6). UNHCR repatriated 30,000 refugees, handed over management duties of camps to the National Commission for Social Action on June 30, 2007 and is now active in integrating the remaining refugees (21). Figure 6: Refugee camps in Sierra Leone, 2007 (21) Several humanitarian organizations were involved in the coordinated administration of activities and demographic and health surveillance in 10

12 camps including MSF, ACF, Merlin, IMC, WFP and World Vision. Two joint assessments were carried out in 2002 and The results reveal that CMR and U5MR were clearly kept under the emergency thresholds in all camps and remarkably, lower than the baseline CMR for Sub Saharan Africa of 0.44 deaths per 10,000 per day (see Figures 7 and 8). It should be noted however, that in the case of CMR in Jimmi Bagbo camp (0.9, C.I ), the higher bound of the confidence interval crosses the threshold. Malaria among all ages and malaria and lower respiratory infections for children under than five were reported as the leading causes of death. Both measurements were taken within the rainy season (May to October) and therefore provide a useful comparison. The phenomenon seen here is a reflection of a typical trend in relief operations, where mortality rates in camps generally display decreases largely attributable to provision of services in a secure environment (22). After the end of conflict in 2002, GAM in the Liberian refugee camps was initially serious ( 10%) or critical ( 15%, Jimmi Bagbo camp only) but had almost all declined to acceptable or near acceptable levels in 2003 (see Figure 9). The serious nutritional situation measured in October 2002 may have been associated with a high influx of at risk, mobile refugees arriving in February, May and June of 2002 which exceeded the space limits of many camps. Among the aggregate number of children in all six camps assessed (N=3697), those who arrived at the camp less than one month before the assessment (n=288) were twice as likely to be malnourished as compared to children present for more than one month (n=3409) (RR=2.14, p<0.05) (23). In the case of Gondama camp, lowered GAM rates are accompanied by a doubled U5MR indicating that malnutrition or other morbidity may have contributed largely to child deaths. Figure 7: CMR in camps, ,6 End of conflict period Bo Jimmy Bagbo 1,4 Pujehun Bandajuma 1,2 Bo Gondama 1 Bo Jembe 0,8 Bo Gerihun Kenema Largo 0,6 Kenema Tobanda 0,4 Moyamba Taiama 0,2 Porto Loko camp 0 mars 00 oct. 00 avr. 01 nov. 01 mai 02 Figure 8: U5MR in camps, déc. 02 juin 03 janv

13 5 Endof conflict period 4,5 Bo Jimmi Bagbo camp 4 3,5 Pujehun Bandajuma camp Bo Gondama camp Bo Jembe camp 3 2,5 2 1,5 Bo Gerihun camp Moyamba Taiama camp Kenema Niawama camp Kenema Largo camp 1 0,5 Kenema Tobanda camp Porto Loko camp 0 Figure 9: GAM in camps, End of conflict period Bo Jimmy Bagbo Bo Gondama Bo Gerihun CRITICAL SERIOUS Pujehun Bandajuma Bo Jembe Kenema camp Kenema Niawama Kenema Gofer Kenema Lebanese Kenema Nandeyama Kenema Largo Kenema Tobanda Moyamba Taiama Porto Loko camp 0 Mar 00 Oct 00 Apr 01 Nov 01 May 02 Dec 02 Jun 03 Jan 04 déc. 99 juin 00 janv. 01 juil. 01 févr. 02 sept. 02 mars 03 oct. 03 avr Notable initiatives As of late 2007, the following NGOs are working in Sierra Leone: Action Aid, ACF (France), CARE, Caritas, CRS, Concern, IMC, IRC, Oxfam, MSF, Save the Children and World Vision (24). Of note, MSF is planning mortality, nutrition and malaria survey of the MSF catchment 12

14 area for October 2008, ACF will conduct a survey in October in either Freetown or the Western rural area and Concern are conducting a nutritional survey in the near future. 8. Conclusions and recommendations A brief analysis of CE DAT data for the period 2000 to present reveals steady or decreasing trends in CMR, U5MR and GAM in selected districts and camps, but a lack of surveys in the post 2003 period makes further analysis and commentary difficult. Furthermore, recent national level surveys have produced GAM results that are somewhat inconsistent with each other when taken at face value. Further analysis of the 2005 and 2003 VAM datasets and sampling methodologies would be useful to fully assess the comparability of these surveys. Malaria continues to be a major public health concern in rural Sierra Leone and presented as such during the transition and post conflict phases. Given that CMR, U5MR and MCV are still crossing the emergency thresholds as late as 2006, it is important that NGOs continue to monitor the population using nutrition and mortality surveys. Acknowledgements This report is based on the field data and information provided by the following NGOs and organizations, to which we are grateful: ACF IMC MSF Merlin GOAL Concern IRC UNHCR World Vision UNICEF Sierra Leone Ministry of Health Statistics Sierra Leone WFP 13

15 Bibliography 1. Rushton, Neil. Health and Peacebuilding: Resuscitating the Failed State in Sierra Leone. 4, 2005, International Relations, Vol. 19, pp CIA. Sierra Leone. World Fact Book. [Online] August 32, [Cited: August 21, 2008.] world factbook/geos/sl.html. 3. Sphere Initiative. Humanitarian Charter and Minimum Standards in Disaster Response. Oxford : Oxfam, Statistics Sierra Leone and UNICEF Sierra Leone. Sierra Leone Multiple Indicator Cluster Survey 2005, Final Report. Freetown : Statistics Sierra Leone and UNICEF Sierra Leone, UNICEF. State of the Worldʹs Children New York : UNICEF, Checchi, Francesco, et al. Evidence basis for antimalarial policy change in Sierra Leone:five in vivo efficacy studies of chloroquine,sulphadoxine pyrimethamine and amodiaquine. 2, 2005, Tropical Medicine and International Health, Vol. 10, pp Ministry of Health and Sanitation and Partners. National Strategic Plan for Roll Back Malaria Implementation in Sierra Leone: Freetown : MoHS, Gayer, Michelle, et al. Conflict and emerging infectious diseases. 11, 2007, Emergining Infectious Diseases, Vol. 13, pp Statistics Sierra Leone. Sierra Leone Integrated Household Survey (SLIHS) 2003/04: Final Statistical Report. Freetown : Government of Sierra Leone, World Food Programme. Vulnerability and Assessment Mapping (VAM). Freetown : World Food Programme, Food Security and Nutrition Survey conducted in March 2005, Sierra Leone. Freetown : World Food Programme, Senessie, C, Gage, G N and von Elm, E.Delays in childhood immunization in a conflict area: a study from Sierra Leone during civil war. 14, 2007, Conflict and Health, Vol Centre for Research on the Epidemiology of Disasters. CE DAT. Complex Emergency Database (CE DAT). [Online] [Cited: September 2, 2008.] Human Rights Watch. Sierra Leone: getting away with murder, mutilation, and rape: new report from the field.. New York : Human Rights Watch, Vol 11, No 3(A). 15. Salama, Peter, Laurence, Bruce and Nolan, Monica. Health and human rights in contemporary humanitarian crises: is Kosovo more important than Sierra Leone? 1999, British Medical Journal, Vol. 319, pp Bijleveld, Catrien and Hoex, Lotte. Direct and indirect mortality in Sierra Leone, [book auth.] Geneva Declaration Secretariat. The Global Burden of Armed Violence. Geneva : Geneva Declaration Secretariat, 2008, pp Smith, L Allison, Gambette, Catherine and Longley, Thomas. Conflict mapping in Sierra Leone: Violations of International Humanitarian Law from 1991 to Rome : No Peace Without Justice,

16 18. Fornah, J S, et al. Mortality In Kenema District,Sierra Leone. A Survey Covering January 2000 January Freetown : IRC, Amendola, Paul R. Kono Child Survival Program Mortality Survey Report. s.l. : IRC, MSF OCB. Unpublished data. 21. UNHCR. UNHCR Global Appeal Geneva : UNHCR, pp Salama, Peter, et al. Lessons learned from complex emergencies over the past decade. 9447, 2004, Lancet, Vol. 364, pp ACF. Nutritional anthropometrical survey in refugee camps in Bo, Pujehun and Moyamba districts (Sierra Leone). s.l. : ACF, OCHA. West Africa: Who is Where, October s.l. : OCHA,

17 Appendix 1: GAM among children under 5, by district (11) GAM, 2005, children under Eastern Northern Southern Western 16