Version classiqueVersion mobile
OpenEdition Books

Lutte contre le trachome en Afrique subsaharienne

Anne-Marie Moulin
Jeanne Orfila
Doulaye Sacko
et al.

Part One. Synopsis and recommendations. English Version

Evaluation and long-term prospects

Texte intégral


1The question raises two problems: the choice of surveillance indicators, and the choice of population samples for a good monitoring system.

2There is no ideal indicator for evaluating effectiveness and monitoring a trachoma control campaign. However, it is essential for the various users to have the same indicators, to facilitate comparisons. The World Health Organization has recommended using several indicators of effectiveness and activity, some of which are already used in programs in West Africa. Definitions may still change, especially as regards the monitoring indicators.

Effectiveness indicators

  • Prevalence of follicular trachoma (FT) among children over one year of age and under 10. FT is preferable to inflammatory trachoma because it is easier to identify by workers in the field.
  • Prevalence of trichiasis (TT not operated on or returning after operation, among adults of both sexes and aged 40 and over).
  • Scars (CT) could be an interesting indicator, but this stage is the most difficult to define for an inexperienced examiner.
  • Corneal opacity (CO), easy to recognize, is not very specific since it can come from causes other than trachoma.
  • Percentage of children with a clean face, being aware of the difficulties of such an estimate (how can we observe children by surprise?).

Activity indicators

3It is more difficult to define the monitoring indicators of a program. The WHO definitions are still changing, like those of the NGOs working to combat trachoma. The indicators currently used are set for each element of the SAFE strategy.


4Three indicators may be adopted:

  • the number of interventions in the previous year,
  • the surgical cover of trichiasis: TT operations/TT operations + the number of cases of trichiasis waiting for intervention,
  • the percentage of recurrence over time, which is an indicator of the quality of the surgery.

Antibiotic therapy

5Two indicators may be adopted:

  • the number of people treated in the previous year,
  • the antibiotic cover: the number of people treated/number of people eligible and targeted for treatment.

Hygiene and facial cleanliness

6The indicator adopted is the percentage of villages where more than 80 % of children have clean faces.

Environmental improvement

7There is no consensus for these indicators, but two could be proposed:

  • the percentage of the population with access to water at less than thirty minutes walk, or less than one kilometer,
  • the percentage of premises equipped with functional latrines.

Sampling populations

8There are several sampling procedures. The choice depends on the objectives and resources of each program. It is possible to examine:

  • all individuals in a sample of villages picked randomly,
  • all individuals in a sample of plots picked randomly,
  • all children between one to ten years of age, and all adults over 40 from a sample of villages or plots picked randomly.

9At the start, the geographical unit that appears most adequate is the medical district (or circle in Mali). When the prevalence of follicular trachoma in children decreases and it is necessary to change the antibiotics distribution strategy, it is advisable to pass to a finer geographical level such as the health area.

The WHO quick trachoma evaluation method

10This consists of examining fifty children in the poorest families. The investigator also asks if there are any cases of trichiasis in the village. He also estimates the number of latrines and the presence of flies. This method is useful for identifying regions of trachomatous endemic disease. It was not prepared for establishing prevalence or for following a program.

Conclusions and recommendations

11There are no perfect indicators, but it is important for all countries involved in trachoma control to have a minimum number of common indicators, in order to compare the evolution of their programs. This common body of information is also useful for organizations working in several countries.

12In addition to these indicators, national programs and NGOs can develop others to suit their own program management needs. However, it is preferable not to multiply the indicators. During the next five years, from 2005 to 2009, the indicators will be re-examined annually, to adapt objectives during the following year, still bearing in mind the approaching target of the year 2020.



13Geographic information systems (GIS) are increasingly being used by health planners for visualizing the extent of a disease, and for better understanding of its epidemiology, particularly by highlighting the relationship between prevalence of a disease, populations, infrastructures and resources of a country. This instrument allows us to evaluate the epidemiological situation of an infection and monitor its evolution. It is a valuable tool for decision-making. There is certainly need to provide investment to bring the system up to date, but finance is increasingly available in this field, particularly for reducing the "numerical fracture" between north and south.

14From the moment of creation, GIS attracted and even fascinated researchers and policy makers. They use advanced technologies resulting from space research and make it possible to illustrate complex problems simply and intuitively.

15They have already proved useful for monitoring the elimination of diseases such as the Guinea worm, for recording clinical cases on maps according to water points, and for directing research into the latest infections. At present, they are a research tool for more complex diseases such as malaria.

16Until recent years, GIS had to endure well-founded reproaches. The GIS used in public health required complicated and expensive software, which could only be handled effectively by specialists, who frequently had to follow refresher courses. After a brilliant start, they were in general not regularly fed with data and the laboriously prepared maps quickly became obsolete. Moreover, very often there were not enough links between the periphery where the data was collected and the center that used it.

17In the last five years, there have been considerable changes towards a simplification of the whole process. A certain number of simplified software programs now have user-friendly interfaces. There are increasing numbers of geographical databases, often free, accessible and usable by the countries for their health programs. GIS has become less time consuming than before. Data can often be entered simply by means of a palmtop computer connected to the GPS. An analysis can sometimes be performed instantaneously thanks to the interface software.

18The availability of the WHO Health Map, a simple to use mapping software, adapted to local conditions in the most difficult countries, has allowed many ministries of health to adopt GIS and form the basis for an effective partnership between the States, the WHO and many organizations working in the field of health.

The value of GIS in trachoma control

19The managers of trachoma programs will be confronted with the question of their use in the near future.

20Trachoma is an ideal candidate for benefiting from GIS. This is a disease for which there is an elimination program, a gathering of data by the ministries and/or partners in most of the countries with endemic disease, and where the planning of action is essential for using limited resources as well as possible.

21Today, however, in the countries concerned, GIS has not yet become a basic tool in daily use because of the complexity of most of the systems and the special training required by program directors and users who have to feed the system with data.

22For trachoma, GIS could be useful in several ways:

  • GIS can be used to gather epidemiological data acquired during investigations, to standardize it and make it available to the ministries of health.
  • It is possible to produce maps by area, district and locality, and have information about the endemicity levels and performance levels of the various aspects of the campaign, progressively as the program unfolds.
  • This base, partly to be built, can be used to classify each district into different categories of prevalence, from very low to high. Classification could be based on random sampling.
  • When planning, it could illustrate correlations between the extent and condition of the water points, for which there are databases in Mali, and pockets of trachoma.

23A national organization on the level of the ministry of health could centralize data collected in the district and recorded on a palmtop computer. Compilation would provide a broad database useful for planning interventions.

24The four components of the SAFE strategy can benefit from such a tool.

  • A record of trichiasis interventions could be set up, fully listing the interventions and their results.
  • A GIS could control the distribution of antibiotics according to the prevalence of active trachoma among children in the districts, monitor the mass treatment campaigns in these districts and manage drug stocks.
  • The percentage of clean faces among children, an element for monitoring the progress of hygiene, could be easily calculated through such a system.
  • The position of water systems, sanitary facilities and workers on a map can help spatial and chronological coordination of activities.

25With such an instrument, it is also possible to visualize and continuously follow a certain number of sentinel sites where more complete information could be collected at regular intervals.

26The data collected on trachoma could be shared with other programs of international cooperation in the field of hygiene and sanitation. It is desirable to share a maximum of medical data and the means of collecting it.

Conclusions and recommendations

27Today, GIS are better adapted for use in public health. However, they are not yet used routinely in campaigns against blindness and trachoma. However, in the near future, they could be a valuable tool for policy-making and surveillance. Technical support is good and human resources available. Maps made will make cooperation easier between those working locally, nationally and internationally.

28Before deciding on using it, there needs to be training for the persons who will work with it. Although the software can be free for the countries, GIS sometimes remain a tool that is too expensive. At the start, there was no cooperation between the various organizations. It is important to prepare for cooperation before deciding to use GIS.


At what level of prevalence of active trachoma should an antibiotic therapy campaign be launched in a medical area?

29To answer this question, we have to know the prevalence of eye infections due to Chlamydia. The prevalence of follicular trachoma (FT) among children between 1 and 10 years of age seems to be a good indicator.

30Until 2003, the launching of a mass treatment program for the whole community was recommended by the WHO when the prevalence of inflammatory trachoma exceeds 20 %. Between 5 and 20 %, it would be enough just to treat the members of communities with a high risk of trachoma (all children or children with active trachoma and their families). Under 5 %, individual treatment was enough.

31Since 2003, mass treatment has been recommended starting from a 10 % prevalence of follicular trachoma in children.

When should an antibiotic therapy campaign in a medical area end?

32The answer is both crucial and difficult, and depends on two unknown factors:

  • is there a threshold of prevalence of active trachoma, which, when reached, implies that the disease will never reappear again?
  • what is the long-term impact of the programs concerning hygiene and the environment? Will they suffice to prevent the return of the infection, after stopping the antibiotics?

33As criterion for eliminating trachoma as a public health problem, the WHO recommends a drop in prevalence of follicular trachoma (FT) below 5 % in children. At that point, it considers that transmission of trachoma has then virtually stopped.

34As criterion for eliminating blinding trachoma, it recommends a prevalence of trichiasis (TT) of less than one in a thousand in the general population (table 1).

Table 1 – Elimination criterion for active and blinding trachoma

Table 1 – Elimination criterion for active and blinding trachoma

35These recommendations rest more on the concordant opinion of the WHO specialists than on published research. Nevertheless, they contribute a certain amount of information. Antibiotic therapy acts more on the reduction in bacterial infection (disappearance of the germs during laboratory tests) than on clinical infection. For example, after repeated treatment, we can still find 5 % of clinical cases, but the laboratory no longer detects any infection. It is thus possible that Chlamydia is still there, but undetectable by even the most sensitive current techniques, and is probably not a significant risk of contagion. For this reason, it would be desirable to choose the PCR as reference technique. Nevertheless, it remains more realistic for the moment to rely on the clinical examination.

Ultimate intervention objectives ("UIO")

36At a scientific meeting of the WHO in Geneva in August 2003, and drawing from experience gained in combating onchocerciasis and Guinea worm, a group of scientists defined the "ultimate intervention objectives" ("UIO"), which indicate the objective to be achieved for each element of the SAFE strategy in order to eliminate blinding trachoma. The idea of these objectives was dictated by a preoccupation with a planning aid.

37These objectives are designed to be part of annual intervention objectives in a medium and long-term program. They are to be calculated by each country according to needs estimated from the epidemiological data, taking into consideration the available treatment. They are regularly reassessed and modified over time.

38As an example, for surgery, it is necessary to operate on a certain number of people to obtain less than one case of trichiasis per thousand in the population after the WHO program. To establish this figure, we have to know the starting prevalence in the basic geographical unit (the district). The objective is then reviewed in annual objectives. A certain percentage of the persons to be patients is proposed taking into account local conditions.

39Objectives must be realistic, neither too ambitious nor too modest, in order to have serious chances to be achieved by emulation.

40For antibiotic therapy, aiming at a long-term prevalence of 5 % of follicular trachoma among children (end of transmission), the number of people to be treated is annually estimated and projected, according to the prevalence of the infection in each district, specifying the distribution methods (mass, targeted, individual) and the amount of distribution necessary. These estimates obviously take into account the demographic data and the teams available in the field.

41These objectives, now being adopted, should give us a clear idea of progress achieved and tasks remaining, while keeping in mind the overriding objective.


42The mathematical modeling of trachoma is complex since it utilizes very many untested factors. The dynamics of infection, cure and immunity are extremely poorly known. In addition, the impact of inventions, especially in the field of hygiene and the environment, has been the subject of very little quantitative research.

Modeling antibiotic therapy

43The only component of the SAFE strategy, which has been the subject of mathematical modeling, is antibiotic therapy; the simplest to model.

44Can we calculate the frequency of distribution needed to eliminate infection?

45A relatively simple mathematical model built by Lietman shows that regular mass distribution campaigns of antibiotics could eliminate ocular Chlamydia in the long term. Such elimination can only occur if:

  • the level of cover of the population is high (higher than 80 %),
  • the frequency of distribution is adequate.

46The frequency of distribution depends on the dynamics of the increase in the prevalence of trachoma after treatment, which itself depends on the starting prevalence. The essential element of calculation is the time needed for prevalence to double. It has been estimated by comparing several studies. According to Lietman, the desirable frequency of distribution depends on the prevalence of active trachoma among children.

  • In regions where prevalence exceeds 50 %, as in Tanzania, distribution should take place twice a year, provided cover is at least 80 % (fig. 3 and 4).
  • In regions where prevalence of active trachoma is between 10 and 50 %, distribution every year can be enough.

47Thus, if the starting prevalence is very high, an annual treatment cannot eliminate infection arising in the intervals between treatment. On stopping distribution, prevalence returns to its first level (fig. 3). If the treatment is more frequent, for example twice a year, and if the total cover is sufficient, elimination can be achieved (fig. 4).

Inadequate distribution frequency of azithromycin

Inadequate distribution frequency of azithromycin

Figure 3. Impact of annual distribution in area with strong prevalence of trachoma

Adequate distribution frequency of azithromycin

Adequate distribution frequency of azithromycin

Figure 4. Impact of biannual distribution in area with strong prevalence of trachoma

How often should we treat?

48We can speculate on the possibility that a single mass campaign could definitively lower the rate of infection to a threshold so low that the disease never reoccurs. This is what we can call the "magic threshold", referring to the hope of "magic shots" in the history of the epidemics (fig. 5).

Threshold of elimination

Threshold of elimination

Figure 5. Theoretical impact of a "magic shot" (single administration miracle) on the elimination threshold

49However, it is more probable that, in the absence of hygiene and environmental measures, all things being equal, the infection will gradually return to its original level (fig. 6), as soon as the antibiotic is stopped.

50The experience gained in Morocco and Tanzania shows that it is not realistic to control the disease with only one distribution and that satisfactory results were only obtained after at least three passages.

51A model must also take into account the population targeted by the antibiotic therapy. If only children are treated, it is probable that the operation should be renewed more often than if it is applied to all the population.

Return of trachoma

Return of trachoma

Figure 6. Probable impact of a single injection

Difficulty in modeling the effects of the SAFE strategy

52Indeed, there Is only few data on the long-term Impact, prevalence of trachoma, mass antibiotic therapy, and especially on the long-term effects of the components "facial cleaning" and "sanitation" in the environment.

53It is very difficult to quantify clean faces and introduce it into a mathematical model.

54It is also difficult to quantify fly control.

55If hygiene and sanitation measures had easily quantifiable effects, we could use mathematical models to consider the reduction possible in mass antibiotic therapy. For example, if hygiene alone could change prevalence of active trachoma from 30 to 15 %, we could be satisfied with a distribution of antibiotics every two years instead of every year.

The real question is: could trachoma disappear without antibiotic therapy as in Western Europe?

56How can we introduce the other components of SAFE into the model to decide on the threshold for stopping antibiotic therapy?

57The general trend towards the elimination of trachoma continues in other regions of the world affected by this disease. Thus, between 1957 and 1987, in Gambia, Malawi and Nepal, the prevalence of active trachoma has regularly decreased, apart from any specific campaign. Researchers broadly attribute this decline to general socio-economic factors. However, do we have to await the results of a historical process?

Eliminating blinding trachoma by 2020

58Even if active trachoma is eliminated, cases of cicatricial trachoma and existing trichiasis could still lead to blindness for some time. At least a generation must pass before blinding trachoma is actually eliminated.

59There have been few attempts to model the evolution towards blindness of old infections due to Chlamydia, after interruption of the transmission by C. trachomatis. It is possible to estimate the time of progression from scars to trichiasis and trichiasis to blindness due to corneal opacity, but nobody knows how cicatricial trachoma will behave in the absence of reinfection. It is possible that its progression will be slower. Identical questions arise regarding the spontaneous evolution of trichiasis in the absence of reinfection. Two current clinical studies are examining this precise question.

60It should soon be possible to accurately estimate the speed of long-term progression from cicatricial trachoma as the surgery failure rate. Consequently, a mathematical model could be created, which attempts to take into account all factors including economic ones, with an estimate of the time needed to eliminate blinding trachoma.

Table des illustrations

Titre Table 1 – Elimination criterion for active and blinding trachoma
Fichier image/jpeg, 111k
Titre Inadequate distribution frequency of azithromycin
Légende Figure 3. Impact of annual distribution in area with strong prevalence of trachoma
Fichier image/jpeg, 82k
Titre Adequate distribution frequency of azithromycin
Légende Figure 4. Impact of biannual distribution in area with strong prevalence of trachoma
Fichier image/jpeg, 78k
Titre Threshold of elimination
Légende Figure 5. Theoretical impact of a "magic shot" (single administration miracle) on the elimination threshold
Fichier image/jpeg, 73k
Titre Return of trachoma
Légende Figure 6. Probable impact of a single injection
Fichier image/jpeg, 80k

© IRD Éditions, 2006

Conditions d’utilisation :