The general working principle of the Atelier Paludisme is to provide the participants with broadband internet access and enable them to find out about various aspects of malaria under the close supervision of international experts. The first week is devoted to practical sessions on improving computing skills and oral communication. Each of the following four weeks is organized around a major theme, and at the beginning of the week each participant is given a topic to research and expected, by the end of the week, to deliver a 10-minute presentation with power-point slides. The topics are selected in advance by the tutors, based on the topics covered each week and randomly distributed to the participants. Table 1 shows the overall course structure, including a selection of topics dealt-with by participants; full workshop content and participant presentations for the past five years can be found on the course web site 1. The final week of the workshop is devoted to various forms of group work, including round table sessions and role-playing exercises. Practical sessions are organized at various times in the course of the workshop to illustrate technical issues (malaria diagnostic, geographical information systems, entomology).

In addition to course organizers, the teaching faculty consists of a group of 10–15 invited experts, who act as tutors to the participants. Each tutor stays with the course for one week (two or three tutors per week). The role of tutors is to guide the participants in their search for information on the internet, to help them to critically evaluate the information found and to make sure that their presentation is accurate (in order to avoid exposing the other participants to presentations containing erroneous information or misunderstandings).

Each tutor is asked to give a 50-minute lecture on a topic of their choice, usually their own research, and the lectures are followed by an open discussion with participants, course faculty and guests. These lectures by tutors represent the only formal teaching of the course and they represent only 10% of the overall contact time of about 256 hours.

At the end of each week, when participants give their 10-minute presentation, tutors and course organizers act together as a panel of examiners to evaluate the work and the presentation in a deliberately formal manner (Figure 1), using an evaluation grid including 20 criteria related to the form and content of the presentation, scored on a five-point scale (from "not acquired" to "completely mastered"). The scores awarded are used to construct a histogram providing a visual indication of the overall quality of the presentation and making it possible to identify strengths and weaknesses and to assess the progress made (Figure 2). The weekly cycle ends with an individual interview between each participant and the tutors for that week to identifying what the participant has learned and his/her weaknesses.

During the annual sessions from 2003 to 2007, 76 participants from 15 different countries were trained at the Atelier Paludisme. From the beginning, the intention was to achieve a good mix of participants: 1/3 clinicians, 1/3 public health administrators (i.e. scientists or health professionals involved in malaria control at national or regional level, usually employed by their Ministry of Health) and 1/3 research scientists (either at PhD student or post-doctoral level). Over the five-year period, this optimal ratio was more or less achieved. As the course was held in Madagascar, it is not surprising that 60% of its participants were from Madagascar for the first course in 2003, but this was reduced to 25% by 2007. As the course became increasingly well-known, the number of applicants gradually increased and so did the geographical diversity of participants with 11 different nationalities in 2007 (Figure 3).

Both tutors and participants were asked to evaluate the training, and their comments and suggestions were used to improve the workshop from one year to the next. An external audit of the five first sessions of the workshop was carried out in 2007, based on theoretical educational concepts. The design of this training was considered appropriate to its aims, particularly with respect to the acquisition of transverse skills, the use of appropriate learning models, the phases of knowledge acquisition, the identification of problems to be solved, material aspects, the availability of the tutors during the workshop, exchanges of experience between participants and tutors and the self-confidence acquired, especially when giving presentations. The preparation and delivery of presentations, and online bibliographical searches were considered important skills that the participants will be able to use in their professional lives, whatever their future post. The audit identified the following weak points, including a tendency to information overload, due to intensive weekly cycles preventing adequate consolidation of the knowledge acquired. This external viewpoint will be taken into account to improve future workshops.

In order to evaluate the long-term impact of the workshop, a follow-up of former participants was performed through annual questionnaires, to find out what impact the training had on their professional activities. In the 2007 survey, targeting participants having attended the first four sessions of the workshop, 68% (41/60) of participants responded. An overview of the responses of former participants and matched controls (who had attended different training courses in other subjects) is provided on Table 2. The subjective criteria recorded by former participants (improvements in their ability to assume their functions, strengthening of professional relationships) are positive, although one cannot exclude a bias in these responses.