Letter on background to 1999 WHO guidelines on malnutrition, by Mike Golden

Dear ENN,

Further to your article on the technical debate regarding the management of severe malnutrition, I wish to offer some contextual information to the development of the World Health Organisation (WHO) guidelines, and agency field protocols.

Initial guidelines for the treatment of malnutrition, written in the 1970s in Jamaica, were published by the Pan American Health Organisation (PAHO) and subsequently by the World Health Organisation (WHO) in the 1980s. They were based upon practice at the tropical medicine research unit (TMRU) in Jamaica, treating small numbers of children (about 50 children a year). The researchers, clinicians and staff were all highly trained and had ample access to sophisticated instruments, literature and funds. They were 'overstaffed' to facilitate research measurements, which allowed labour intensive individual treatment to be given. The guidelines were not even implemented in the paediatric wards next to the unit where they were written.

Non-governmental organisations (NGOs) and agencies used the practical guidelines published by Medecins sans Frontieres (MSF) - in most hospitals no generally accepted guidelines were being used. In the early 1990s, when I drafted the current WHO guidelines , I simplified them as much as I thought safe. The big step was to formulate the diets, instead of giving individual ingredients by weight of child. Nevertheless, the basis of the management was the experience of a research ward.

The draft guidelines were given to the NGOs in 1994. Action Contre le Faim (ACF) took these draft guidelines, wrote practical protocols based upon the principles and provisions, and persuaded Nutriset to start to produce the diets commercially. Their protocols have since been adopted by most NGOs. These NGOs have now treated many hundreds of thousands of children using the guidelines and have amassed an enormous body of data, information and experience. For example, in Burundi from January 1999 to December 2001, 80,419 severely malnourished patients' characteristics and outcomes were entered into the national database, coordinated and maintained by UNICEF. The main expertise in applying the protocols and the effects of their variation now lies with the NGOs. However, in most places the data are not systematically collected or analysed. Yvonne Grellety made a very detailed analysis of over 10,000 patients from 13 countries in Africa. Interestingly, she confirmed, en masse, the importance of the old physiological data showing the sensitivity of these children to sodium and their propensity to develop heart failure in several different situations and analyses.

The original draft guidelines have been treated as a 'living document' that has gone through an evolutionary process - like computer software, we no longer use dos 3.1 (the operating system used for the original draft). The analyses, together with frequent field evaluations, in many different contexts and countries, shows where the critical points in the protocols lie and where difficulties of training, understanding, application and scale arise in practice, particularly in resource limited situations or emergencies. This has resulted in a number of changes in both detail and emphasis. The modified protocols are greatly simplified. They are relatively easy to apply in the field by nurses and nurse-assistants. The results being obtained are, in some places, as good or better than those obtained in the research ward in Jamaica. It is relatively easy to get good results from an adequately resourced, dedicated team in a research setting, it is quite a different matter to maintain good results in routine service. Nevertheless, at a national scale the results are a success story. The mortality rate (2000/2001) for Angola was 6% (3,976 deaths, 66,165 discharges) and 5% (3,552 deaths, 74,759 discharges) in Burundi. The current data for Ethiopia appear to have an even lower motality rate, with the best centres reporting around 1% mortality.

The main changes are of emphasis. Some aspects we thought were vital fifteen years ago are now known to be either minor or in some cases detrimental. Conversely, other aspects that were not emphasised were omitted, or compromises made in the original guidelines are now seen to be critical. The aspect of emphasis and approach is far from trivial. It determines where resources are directed when they are limited.

The challenge is to find a mechanism for both translating this very extensive body of knowledge into internationally endorsed guidelines within a reasonable time span, and to create a procedure for regularly updating the guidelines so that they are not outdated or used to produce derivative training material years later, when the State of the Art has moved on.

Yours
Professor Mike Golden
Email: mike@pollgorm.net