Impact of cross-sectoral approach to addressing konzo in DRC

By Marie-Morgane Delhoume, Julie Mayans, Muriel Calo and Camille Guyot-Bender

Marie-Morgane Delhoume is an agricultural engineer specialising in agro-development of tropical regions. She led the 2011 impact study of ACF-USA’s Integrated Programme for the Eradication of Konzo in the Territory of Kwango in DRC.

Julie Mayans is an agricultural engineer, specialising in food security and rural development programme management. She was the ACF-USA Food Security and Livelihoods Coordinator for the ACF-USA DRC mission.

Muriel Calo is the Senior Food Security & Livelihoods Advisor for ACF-USA who provided technical support to the ACFUSA DRC mission.

Camille Guyot-Bender is the Technical Programmes Assistant for ACF-USA.

Special thanks go to the whole of the ACF project and study team and the communities who participated in the study. Thanks to those who provided peer review of the article, namely Nick Radin of ACF, Marie-Sophie Whitney of ACF and Dr. J. Howard Bradbury of the Australian National University. The funding support of the European Union (EU) Food Facility Programme is gratefully acknowledged.

From December 2009 to October 2011, Action Against Hunger (ACF-USA) implemented a 22-month long intervention in the Bandundu province of the Democratic Republic of Congo (DRC) addressing several factors underlying the konzo epidemic affecting the population of Kwango district (see Figure 1). The ‘Integrated Programme for the Eradication of Konzo in the Territory of Kwango in DRC’ project was financed by the European Union (EU) Food Facility. It aimed to eradicate the disease through a cross-sectoral approach that focused on nutrition education and training, dietary diversification, improved water access and agricultural processing. A total of 22,000 households are estimated to have benefited from these activities. The project was implemented in 396 villages in the Territory of Kwango.

Prior to implementing the project, ACF conducted a baseline study¹ in collaboration with the Minstry of Health (MoH)’s PRONANUT or Programme National de Nutrition in DRC in 113 villages, across 51 health areas and 11 health zones. A total 2,388 suspected konzo cases were screened and 2,218 were confirmed. The average incidence of konzo was 1.07%. Among confirmed cases, 83% were located in savannas, 1% in hillside areas and 4% in valleys. Kahemba health zone was found to harbour the highest number of confirmed cases (1,639), and placed among the top three zones for incidence (2.08%) largely due to its density of population and associated risk factors.

Processing mill

Local beliefs and traditional customs were found to influence strongly the incidence of konzo in the area. Local eating customs that favour the male head of household were noted as likely contributors to heightening the exposure of women and children to konzo. There was a widely held belief that the disease is caused by black magic, while knowledge of the food-related origins of the disease was low.

Households rely primarily on agriculture for food and have limited dietary diversity. Cassava is cultivated as a main crop, with maize, groundnuts and beans as secondary crops. Diverse environmental factors, such as soil fertility and soil water retention, affect the quantity and quality of harvests.

Water access is a critical factor in konzo incidence, with access limited by both distance to and seasonality of water points. Water coverage levels are very low, with 5% coverage in the Feshi health zone and 4.3% in Kajiji.² Due to these challenges, cultivators in rural areas most often prefer to soak the cassava directly on river banks, in ponds or in swampy areas in order to avoid carrying heavy quantities of water back to their homes. In semi-urban areas, people prefer to ret the cassava in their homes (in buckets or barrels) due to the likelihood of theft if the cassava is left overnight in a public area. This practice can be hazardous as the quantity of water available in urban areas is often insufficient, and this often leads to not changing the water on a daily basis during the retting process which greatly increases the risk of cyanide intoxication.

Implementation strategy

The programme strategy sought to address directly the range of critical factors related to konzo disease that were identified in the baseline survey. These included knowledge of and attitudes towards the disease, limited agricultural and dietary diversity, low water access and poor knowledge and practices around cassava processing. The strategy also aimed to address indirectly the high rates of malnutrition seen in konzo cases (25.8% global acute malnutrition (GAM) prevalence in konzo affected children less than 18 years old, 69.3% of GAM in konzo affected adults). Project design used a cross-sectoral approach to address underlying factors in a holistic and integrated manner.

Community outreach, mobilisation and education

ACF employed a community outreach and mobilisation approach through the creation of community cells as a forum for discussion on konzo and nutrition. These served as launch pads for a broadly based educational campaign on konzo, which also extended to churches, schools, training of local health professionals, community volunteers and leaders, traditional authorities, etc.

Information, Education & Communication (IEC) materials on food processing and preparation, nutrition and konzo were developed in collaboration with the PRONANUT, including posters, brochures, training modules and other material with graphic illustration and supporting text in Lingala, Kikongo and French languages. Posters were distributed for display in places such as public areas, religious sites, health centres, schools and administrative offices. Radio messages incorporating songs and stories were crafted for broadcasting on two local radio stations. To complement, ACF organised 154 mass sensitisation sessions in churches, mosques and schools.

Across the intervention zone, 35 senior MoH staff, 429 community leaders and authorities, and 1052 community volunteers were trained on the tools and in turn, used their knowledge and skills to pass the message more broadly across the population. Each community volunteer presided over a community cell or served as secretary to the president. Skilled in community mobilisation and training, this individual would facilitate community dialogues on a weekly basis and ensure regular reporting to the local health centre. Each cell had use of an office space and comprised one committee president, one secretary and two advisors. The creation of 647 community cells across 395 villages offered a setting for demonstration, discussion and exchange. These cells gathered members from a local neighbourhood or village hamlet and numbered roughly 45 members each.

Cooking demonstrations were organised at the level of each cell to complement discussions around balanced diets and promotion of kitchen gardens. ACF organized 1,808 demonstrations and volunteers organized another 2,600 demonstrations for cell members, averaging four demonstrations per cell. Improved fufu recipes based on mixed cassava and maize flour were introduced in the cooking demonstrations. ACF also extended practical support to the groups through provision of tanks and mills for cassava retting and processing.

Sensitisation being carried out on konzo in a mosque in Kahemba

Agricultural and hydraulic infrastructure support

As part of crop and diet diversification activities, ACF introduced cultivation of improved varieties of two food crops over two successive agricultural seasons, niébé (cowpea) varieties Vita 7 and Muyaya, and sweet cassava varieties TME119, Mwuazi, Nsasi, Disanka and Butamu. Cultivated in other parts of the DRC, their introduction in Kwango was aligned with local agro-ecological conditions and intended to support increased consumption of sulphurbased amino acids contained in leguminous foods (to counteract high levels of cyanide in the diet) and to complement consumption of traditional cyanide-heavy bitter cassava varieties with varieties low in cyanide. ACF delivered technical training on agricultural techniques to the 12,500 beneficiary households of food crop support. The bulk of the sweet cassava cuttings distributed in the framework of the project were produced by local agromultiplier associations partnered with ACF. ACF also supported the installation of 13 village based mills to increase access to maize and cassava milling services and improve the quality of the flour.

In order to increase water access, ACF implemented a variety of hydraulic constructions: public retting tanks to process cassava, boreholes, springs, rainwater harvesting systems and piped distribution networks. To encourage the participation of the community, the tanks established were given a supervisor who assigned a management team. Although the management committee owned the tanks, anyone in the area could use them in exchange for a small fee.

Cooking demonstration

Impact study

The impact study was conducted in six of the eleven health zones targeted by the project along two main axes - the western axis (Kenge, Boko, Popokabaka, Wamba Luad) and the eastern axis (Kahemba and Kajiji) (see Figure 1).

A stratified sampling approach was used, with six of eleven intervention health zones selected purposively and 40 of 395 intervention villages selected randomly. In each selected village, six beneficiary households were randomly selected to participate in household surveys (234 in total). Of these, 76% had participated in community cells and 24% had not. Household surveys were supplemented with information from key informants and focus groups.

Findings

Knowledge and attitudes on konzo and nutrition

Changes in knowledge at endline compared to baseline suggest that community outreach and education activities were effective in challenging long held local beliefs on konzo and nutrition. At project baseline, 74% of sampled population thought that the disease had a metaphysical or black magic origin; at endline this proportion had dropped to 7%. Eight-eight per cent of the sampled population correctly noted the food-related causes of konzo at endline, while 3% indicated a viral cause and 8% reported they did not know the cause.³ This finding represents the strongest indicator of project impact.

Participation in a cell was found to be correlated with knowledge of the food-related cause of konzo. In addition there was a strong inverse correlation between both ‘participation’ and ‘lack of knowledge’, and ‘participation’ and ‘belief in a metaphysical origin’. These findings reveal the importance of outreach and education activities delivered both within the community cells and directly by ACF.

Similar results were found regarding knowledge, attitudes and practice on prevention strategies. In particular, messaging encouraged appropriate processing of cassava and inclusion of increased levels of protein in diets through incorporation of maize flour into fufu preparation and legumes (pulses) in the diet. At baseline, households reported utilizing maize flour in their fufu preparation in only a few cases where milling services were available, while millet flour was used in the north eastern Feshi territory. Knowledge of prevention strategies linked to food preparation and dietary diversity was limited. At endline, a majority (78%) of respondents indicated that a diversified diet + correct cassava processing would prevent konzo, while 7% believed that witchcraft was the cause of the problem and 8% did not know. At endline, 47% of the population reported preparing fufu with mixed cassava/ maize flour. Nearly half (45%) of respondents prepare their fufu exclusively with cassava flour while 8% combine cassava and millet flours.

Lack of access to maize milling services explains 77% of surveyed cases of non-incorporation of maize flour into fufu, which are concentrated on the western axis of the intervention zone (Kenge, Boko, Popokabaka and Wamba Luadi). Just 3% of cases justified the exclusive use of cassava flour on the basis of food habits, indicating that the messages around fufu preparation were well appropriated, but cannot be translated into practice largely due to practical constraints.

Food stocks and dietary diversification

New varieties of niébé were largely accepted across the intervention zone and integrated into the diet, notably on the eastern axis (Kahemba, Kajiji) where populations were unfamiliar with niébé. Sweet cassava was readily integrated into both east and west Kwango, with results showing a general increase in the intercropping of both bitter and sweet varieties, as well as increased cultivation of sweet varieties on their own. In Kahemba, the bitter cassava variety Mwambo is widely cultivated and consumed to the exclusion of other varieties, whereas both bitter and sweet varieties are cultivated and consumed along the western axis. However, sweet varieties were well accepted in Kahemba as they offer shorter processing times and are immediately consumable.

Food stocks at baseline (May 2010) and endline (August 2011) were assessed (see Figure 2). The results reveal a notable improvement in both the overall stocks and the diversity of food items held by households, including pulses. The surveys were not conducted at the same time of year, which would have ensured the greatest comparability, however neither one was carried out in the post-harvest period when differences in food availability are significant. The positive trend in diversity and volume of household food stocks may be attributed to project impact, in particular the IEC activities around balanced diets and food processing and preparation, as well as external factors such as climate, crop disease and seasonal fluctuations.

Cassava retting techniques and water access

Water source in Kigwangala

Knowledge of community leaders and member households of community cells regarding cassava retting and drying techniques were assessed before and after training. On the recommended length of time to ret and dry cassava, the share of community leaders correctly reporting optimal length (4 days) increased from 60% at baseline to 99% at endline. Member households showed a similar level of knowledge at endline but stated constraints around access to processing sites and water quality in applying the practice. At endline, a majority of households (92%) indicated they were processing cassava in rivers or ponds, with a minority using cassava retting tanks (4%) or containers at home (4%). At baseline, utilization of home retting techniques - that rely on prolonged use of the same water, saturated in acid and less effective in cyanide detoxification - was relatively common in urbanized sites (9% in Kahemba, 10% in Popokabaka, 5% in Kenge). At project end, it was noted that these practices have been largely abandoned, partly due to ACF’s implementation of peri-urban water points.

At the end of the project, households reported soaking cassava an average of 3.4 days, a significant increase from the average of 2 days noted at baseline across the intervention area. Constraints to optimal practice include the risk of theft of tubers at open river and pond sites, as well as dietary and income pressures. Impacts on cassava retting practice from ACF’s establishment of communal retting tanks maintained and watched over by community groups are not yet known as the infrastructures were in process of installation at the time of survey.

Konzo incidence

A surveillance system for screening and identification of konzo cases in Kahemba health zone was established by the local health structure in 2009, with annual caseload an estimated 1,300 individuals in 2009. MoH educational activities and ACF integrated activities on konzo were launched in early to mid 2010, with a marked decrease in cases (fewer than 200) recorded that year. A further reduction in caseload between 2010 and 2011 was noted during the critical months of June, July and August (dry season) with 47 new cases recorded in 2011(see Figure 3). This represents an 84% reduction in incidence between 2010 and 2011. The greatest reduction in new cases was observed among the under 5 years age group. Note that the observed reduction in konzo incidence cannot be attributed solely to project activities as numerous external factors are likely to influence this outcome.

The results on reduced incidence are corroborated by ACF analytical findings of urine and cassava flour sample cyanide content, taken from 100 randomly selected beneficiary households at project endline (see Figures 4a and 4b). A 50% reduction in flour samples presenting medium to high cyanide levels (20 to 40 ppm) was observed compared with baseline, as well as a 16% reduction in thiocyanate levels in urine samples (>300µmol). These reductions translate into a slightly lower risk of developing konzo.

Observed reductions in cyanide content of baseline and endline samples are similarly attributed to numerous external factors such as seasonality, migration, agricultural production, health condition, diet composition, water availability, as well as project impact.

Cassava retting tank in Feshi

Conclusions and recommendations

ACF’s multi-tiered community outreach and education strategy proved effective in the diffusion of information on a large scale. The community cell approach allowed for a deep, sustained and broad based appropriation of messages and activities around nutrition education and konzo that would not have been possible if only traditional IEC methods and materials had been utilised. Placing community members in leadership positions to carry out sensitisation allowed local taboos to be effectively mitigated through open discussion. This approach also permitted the affected population to control the educational process, encouraging better appropriation of messages, knowledge transfer and behaviour change. Impacts achieved through the community outreach and education approach were reinforced by improved access to water, agricultural processing infrastructure and opportunities to diversify diets.

Based on these findings, ACF-USA issued the following key recommendations:

• Continued promotion of messaging by community cells and MoH staff
• Continued epidemiologic surveillance of incidence of konzo cases by MoH in collaboration with local partners with a focus on high concentration areas
• Expansion of access to village-based agro-processing infrastructure and associated hydraulic infrastructure including cereal mills and cassava retting tanks
• Socio-economic support of konzo victims and their families through targeted support to income generating activities
• Further study on traditional diets, food habits and beliefs, including the contribution of wild foods and game, for improved strategies around dietary diversification

For more information, contact: Muriel Calo, email: mcalo@actionagainsthunger.org

¹Field Exchange 41, August 2011. A cross-sectoral approach to addressing Konzo in DRC. p2-5

²Based on national standards of user numbers per water point

³Respondents could list more than one cause.