The Risks of Wet Feeding Programmes

Liberia has been in the grip of a brutal civil war for the last six years and severe food insecurity exists in much of the country. In the autumn of 1996 Grand Cape Mount and the neighbouring Bomi counties were particularly badly affected as a result of intense conflict between two opposing groups of the United Liberian Independence Movement (ULIMO). Although figures are not available for the area around Vonzula village, epidemiologists from Epicentre and MSF estimated that the crude mortality rate in the adjacent Bomi county had been 8/10,000/day (over 30 times normal levels) for the period since June 1996. Despite the severity of the famine, food was present in the area and each day we witnessed long lines of villagers being forced by fighters to carry their produce to the road junction, where it was transported to Monrovia and sold. The villagers received nothing for this and instead were often beaten or killed by the fighters. Many had, as a result fled into the bush where they lived as best they could foraging for food. The population of Vonzula village was therefore very variable with villagers coming and going depending upon the levels of harassment. Many of the residents in October 1996 were villagers displaced from other areas of Liberia whilst many of the former residents had fled. The village had been cut-off from the outside since November 1995, its infrastructure was in tatters, the school and health centre no longer functioned and there was little central authority save for the occasional chief whose power had all but been removed by the fighters. When the Oxfam team arrived there were high levels of severe malnutrition with several children and adults dying a day.

The joint NGO approach in the area was to concentrate upon wet feeding and not give out a general ration as it was feared that dry ration distributions would attract fighters and put the villagers lives in danger. This did indeed appear to be the case in Sinje, the next village to Vonzula only 4 kilometres away, where 40 people were murdered by fighters the day after a WFP general ration distribution. However wet feeding presented many problems especially as gaining daily access to the village was almost impossible. Frequently our Monrovia based team was unable to go to the village because of the nearby fighting and massacres. This disrupted the operation of the feeding centre particularly as the frequent threats by the fighters and the fear that food stocks would make the centre staff targets, prevented any more than a day's food ration being left over.

The Oxfam Programme
Despite these difficulties the Oxfam feeding centres in Vonzula began operating on Saturday 5th of October within a week of the first UN exploratory mission to Vonzula. Although the centre attempted to offer 24 hour therapeutic care the fighting in the area prevented effective night time supervision and the majority of the food was given out during the day. By Tuesday 8th there were 150 severely malnourished patients registered, the majority admitted according to the criteria '<70% weight for height'¹ or oedema in children and obvious severe emaciation or grade 3 oedema in adults. In addition there were approximately 375 beneficiaries receiving supplementary feeding (criteria, being all under five year olds and adolescents / adults who appeared moderately malnourished). By Thursday 12th October numbers of beneficiaries had risen to approximately 220 for therapeutic and 480 for supplementary feeding, swelled by a large influx of extremely malnourished people fleeing massacres in the area four kilometres to the west. At this stage the severely malnourished in the Vonzula centre were receiving four meals of therapeutic milk, and BPS biscuits per day,- sufficient calories to commence their rehabilitation.

In the event of an attack causing a break in the feeding for even a few days many would have died. Therefore 40 of the most severely ill patients were transferred to 24 hour therapeutic centres in Monrovia two days later, when it appeared that the village was soon to be overrun by another ULIMO faction. In the event the expected attack did not come and our Monrovia based team gained access to the village for five more uninterrupted days. Despite the poor supervision at night there was only one death from malnutrition in the centre during the first week and a half. The patients responded extremely well to the diet, quickly loosing oedema, gaining appetite and commencing weight gain. A major factor in the rapid response to treatment appeared to be the relative absence of diarrhoeal diseases. The village water supply was from a capped bore well with a functioning hand pump producing reasonable quantities of clear water. The water was tested by a Del Agua testing kit the day after we arrived and found to be free of coliforms and the village elders and previous health care workers all reported that there was no diarrhoeal disease in the village. As a result we decided initially neither to chlorinate the water nor to boil the milk which mixes well and was preferred by the children when cold.

Cholera
The first cases of cholera in the village presented on the 13th October 1996. The infected were all members of the same family who had drunk from a creek half a mile downstream of the village. One was an adult patient in the therapeutic centre. The next day we called a village meeting where we and the village chief informed people about cholera, its causes and prevention and about the specific dangers of the creek water. A second follow up meeting with the village elders to reinforce this information was held on the 19th October. We also instituted an out reach programme using four well respected female health workers from the village; fixed the second of the three water pumps in the village; converted part of the feeding centre into a make shift cholera centre and commenced construction of a separate cholera centre. When the villagers were questioned further it became apparent that there had in fact been frequent and severe cholera epidemics in the area during the preceding few months and that the last cholera outbreak in Vonzula had in fact only been one month before our arrival when 25 people were reported to have died. According to the village elders there had been approximately 150 deaths associated with diarrhoea in Vonzula and the surrounding villages within four hours walk, since August 1996. A very different story to the one we had obtained upon our arrival. On the 15th October two more cases were identified, including the cook in the feeding centre. As a result we closed down the supplementary feeding programme and instituted chlorination of all pump water used in the therapeutic centre. There were no further cases until the morning of the 17th when approximately 20 severe cases were admitted to our makeshift cholera centre. At this point we decided to suspend the therapeutic feeding and focus all efforts on the prevention, early case finding and treatment of cholera cases. Sadly on the 18" fighting meant that our Monrovia team and a truck load of supplies donated by MSF were unable to reach the village and the local staff were forced to make do as best they could in the temporary shelter that we had constructed. From the cholera centre admission data, presented in figure 1 it can be seen that the majority of the cases and 85% of the deaths occurred on the 17th and 18th.

Within a week the epidemic had subsided and the feeding centre reopened, relocating to the new centre that was originally intended to be the cholera centre. The old feeding centre compound was left as a cholera centre treating the few cases that continued to present. The explosive nature of the epidemic gives the impression that the source of the cholera was a contaminated water supply. This was confirmed by the vast majority of cholera patients who reported drinking directly from the creek. This information is presented in figure 2. However in some cases the probable source of infection appeared to be the feeding centre and it seems likely that contamination did occur in the centre. Indeed if there was an attack rate of between 2 - 10 %, which is usual for cholera, the existence of 12 severe cases apparently contracting the disease in the centre indicates that most if not all of those being treated there were exposed.

Lessons Learnt
Although cholera has been a frequent and serious problem in Vonzula several features of the Oxfam feeding programme in Vonzula promoted both the occurrence and extent of the above epidemic. There are several lesson which should be learned from this.. Despite our investigation of the level of diarrhoeal disease in Vonzula prior to the establishment of the feeding programme the local people, including the village health care workers had not mentioned the recent high degrees of cholera in the areas. An increased awareness of the importance of cholera in the local area would have increased the efforts made to prevent the disease in the feeding centre.

More structured information gathering as to the prevalence of specific locally occurring infectious diseases, especially diarrhoeal diseases, should be performed before the commencement of wet feeding programmes, even if this means delaying the start of such programmes. Non specific general questioning of local people or local health workers will not necessarily provide sufficient information.

The establishment of wet feeding in Vonzula encouraged people to congregate in Vonzula. Many of these people were internally displaced and unfamiliar with the village and its water / sanitation arrangements. These extra people over stressed the existing water supply from the only village pump (out of the original three pumps) that remained in operation. The displaced also drank from areas of the creek known by the locals to be contaminated. The use of the pump for the feeding centre further reduced the water supply available to the villagers thereby increasing waiting times. This encouraged people to use the creek.

Construction of wet feeding centres that encourage people to congregate in villages must be accompanied by programmes aimed to improve the infrastructure, in particular the water and sanitation. The monopolisation of existing water sources should be avoided and if necessary wet feeding should be postponed until after the rehabilitation or construction of additional water sources.

The person to person spread that occurred in the centre was encouraged by the failure to chlorinate the water used in both the milk and the washing up water and the delays in instituting hand washing for all recipients. Although the water from the pump had been tested and was found to be uncontaminated prior to commencement of feeding, the chlorination of all water used in the milk and the disinfection of equipment with chlorine during washing up would have helped to prevent contamination of the food in the centre after its preparation. The institution of hand washing from day one of the wet feeding operation, despite adding considerably to the organisational difficulties involved in opening wet feeding programmes in areas with few trained staff and in difficult circumstances, should be seriously considered.

Acting upon the above lessons is not as straightforward as it might appear. Time taken to improve the infrastructure slows the response to the nutritional crisis and may, although decreasing the mortality and morbidity from communicable disease, ultimately result in more death from malnutrition. In Vonzula the rapid establishment of wet feeding undoubtedly saved many lives. Had more water pumps been rehabilitated sooner and stricter hygiene instituted in the centre before commencing operations, the time taken to set up the wet feeding, and therefore the death toll from malnutrition, would have been higher. It is difficult to know what effect such measures would have had. What is clear however, is that an adequate assessment of the local disease patterns would have allowed for a more rational prioritisation of tasks. The cursory questioning that we carried out prior to setting up in Vonzula was inadequate. When a relief agency moves into a new area potential employees may be unlikely to tell them of anything that might, in their eyes, discourage the agency from starting up. Had we taken more time to investigate the local epidemiology in the region instead of believing the first few stories that we heard we would almost certainly have gained an increased perception of the risks of cholera. This would have resulted in us raising the priority of certain preventative measures, some of which would have required little extra time to implement. In particular, relatively little time would have been required to rehabilitate a second water pump and employ village water workers to advise newcomers to the village about the risks of the creek. Such action could well have slowed the spread of the cholera, decreasing the case load during the first few days, giving us more time to organise our curative response. Other measures to increase hygiene inside the centre itself, although unlikely to have affected the bulk of the epidemic, might have decreased the person to person spread in the centre. In retrospect disinfecting cups and hand washing should have been given a higher priority.

Wet feeding in areas of endemic diarrhoeal disease will always be fraught with the dangers precipitating epidemics, both through the overloading of the local infrastructure and the promotion of person to person transmission in the centres themselves. There are no clear cut answers as to when such risks outweigh the benefits. What is certain however is that an appreciation of those risks to enable rational prioritisation of programmes is an essential basis for the establishment of wet feeding programmes in such areas.

Tips to remember if considering a wet feeding programme: try to find out as much as possible about locally occurring infectious diseases, especially diarrohoeal diseases gather information in a structured way from a variety of sources if the programme is likely to encourage congregation in villages or overcrowding - improve the infrastructure, in particular the water and sanitation do not monopolise water supply rehabilitate or construct additional water sources if necessary chlorinate the water used in both the milk and washing up water

¹Indicates severe wasting when compared with a reference population

See also the Post Script to this article.