Obtaining international funding in health care is becoming increasingly competitive. For example, to acquire resources needed to fight HIV/AIDS or malaria, African countries must now participate in Global Fund competitions. This situation presents health care managers with two new challenges. First, their requests and action plans increasingly need to be evidence-based. Managers therefore must be able to understand and assess the quality of data and of intervention evaluations. The second challenge is that, when assessing requests, funding agencies look at how well previously-funded programmes met their objectives. These programmes' effectiveness must therefore be demonstrated. Health care managers can no longer be just good planners. They also must be informed evaluators, or have at least the basic knowledge required to interact effectively with the evaluation experts whom they will recruit. Within the current trend of establishing New Public Management in health care in developing countries 1 and the Paris Declaration on Aid Effectiveness 2, programme evaluation will become a major sphere of activity for the coming decade. Yet programme evaluation is rarely addressed in training programmes for health planning 3 and, in Africa, evaluation processes are still too often imposed by external bodies 4.

A series of regional seminars on evaluation planned by the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD) was started in 1990 in Côte d'Ivoire 5. In 1999, the African Evaluation Association (AfrEA) was launched. Despite these efforts, training in programme evaluation remains a relative rarity on the African continent. There are some seminars and workshops, but few training programmes leading to degrees. This is particularly true in francophone Africa 67. The strengthening of evaluation capacity building (ECB) has thus become an urgent matter in Africa. Experts in this field are asking for more empirical case studies to document the range of practices in order to improve their knowledge 89, as ECB is "an emergent field of practice" 10. University training is one useful strategy for ECB 11. A review of articles published in this field between 1965 and 2003 reveals a lack of literature on practical evaluation training 12. This article presents the evaluation of a course on programme evaluation, a four-week (150-hour) course attended by health professionals from 11 francophone African countries.

The Université de Ouagadougou (Burkina Faso) and the Université de Montréal (Canada) have partnered to establish, in Burkina Faso, a master's-degree programme in population and health that includes a course in evaluation. This master's programme is part of a larger programme aimed at reinforcing human and institutional capacities in the analysis and evaluation of public policies and programmes. Its goal is to offer a credible alternative to training programmes offered in North America and Europe. The master's-level training programme includes 12 months of course work and a three-month internship (Table 1). The training is organized into modules of several consecutive days, to accommodate the teachers who are brought in from a number of African countries and from Canada.

The overall objective of the master's programme is to develop students' knowledge and aptitudes in analysis, formulation and implementation management, as well as in the evaluation of population and health programmes, including a specific course in programme evaluation (Table 1: 3.3). Before presenting the results achieved in the master's programme, we will describe how the course was implemented and its content related to evaluation.

A number of methodological limitations to the reported results should be mentioned. First, while our assessment was exhaustive, our sample sizes were small, and thus it is quite possible that the difference between behaviour and learning for Cohort 1 is not statistically significant (n = 8 or 7). Second, with respect to the tools, it is possible that a fatigue bias was introduced into the results of the evaluation of all the lessons and courses of Cohort 1. In the African context, where students are rarely asked to evaluate courses and teachers 27, a social desirability bias could also have been introduced. However, if this was the case, it would be true for all the courses and not only for the one described in this article. In addition, we believe we chose the proper instrument because "more than three decades of research on post + retrospective pretest method has unequivocally supported this approach" 25.

Our analysis of the teaching of programme evaluation using the process described above shows that not only was it much appreciated by the students but it also produced positive outcomes. The students gained much knowledge and the degree of mastery of competences was increased and maintained over time. The greatest progress was in competences that were very specific to programme evaluation, as opposed to those in which the students already had attained high levels (systematic inquiry and interpersonal competence). It should nevertheless be noted that the positive effects cannot be attributed solely to the evaluation course, since many other courses in the programme also reinforced certain competences that were on the list of 60. The effect, then, is that of the programme as a whole, which is not a master's degree in evaluation, but rather in population and health. The competences in which the students rated low at the end of the programme were in fact elements that were not addressed in the evaluation course or in the master's programme. That being said, students' low rating of the evaluation of data validity (2.11) should certainly be addressed rapidly by those responsible for the programme.

This double positive effect is definitely attributable in part to the skills-based teaching approach. The training in programme evaluation remained practical, dynamic and respectful of the students. This was not surprising, since most teachers in evaluation espouse this type of interactive teaching 3121821, which was also observed during an experience in Mali 28. Rapid integration of the concepts into concrete exercises was an effective strategy, as was the availability of the teaching staff during the lessons. The fact that the difference in knowledge acquisition after the course in the "systematic inquiry" cluster (Table 4) was not statistically significant for Cohort 1 can be explained by: (1) a very elevated pre-course self-evaluation (2.37); (2) a selection of students who had already acquired competences in their training prior to the master's programme; and (3) competences that were interdisciplinary.

With respect to level 3 (learning), the data show that it is more difficult to implement evaluation skills than to understand them. In addition, reflective practice remains the only cluster in which the reduction is statistically significant for levels 2 and 3 one year later, while improvements at the end of the master's programme were the highest (Table 4). Thus, the students learned from this perspective, but it is clear that for them, as for all health professionals 29, reflecting in action is not the easiest thing to do. Many skills cannot be sustainably acquired in a university programme; if evaluators' skills are to improve, they must be put into practice. Also, our results suggest the importance of organizing the field of practice in evaluation with the help, for example, of the AfrEA, which could propose continuing education programmes and support reflective practice.

With regard to modalities for evaluating the students' learning, this study shows the importance of integrating summative evaluation into the learning process. From the beginning of the course, students knew the course content, how they would be evaluated at the end, and on what criteria. Transparency was essential. However, the most helpful aspect was that the knowledge and skills considered indispensable for developing an evaluation plan (as an instrument for evaluating learning) were evaluated (through practical exercises) throughout the course.

The tool for assessing evaluation competences has rarely been used, except by its creator 24. In this case, we found it very useful for understanding the strengths and weaknesses of the teaching provided. It allowed us to measure the level of students' knowledge as well as those elements where there was still work to be done. However, this tool was developed in North America, and the question of whether African evaluators might not need other specific competences remains to be examined.

This study shows that skills-based teaching is feasible, much appreciated and well-adapted for a university-based evaluation training programme in a West African context. We highlight the importance of integrating summative evaluation into the learning process. Creating a master's-degree programme in population and health in Africa and providing training in evaluation to high-level health professionals from many countries augurs well for scaling up the practice of evaluation in African health systems. However, this cannot occur without significant investment being made across Africa to develop university-based and professional courses in programme evaluation.

The authors declare that they have no competing interests.

VR led the study, data collection and analysis, and wrote the first draft. All authors contributed to the study's conception and design and reviewed the final draft. CT did the statistical analysis. VR is a research fellow with the Fonds pour la Recherche en Santé du Québec (FRSQ).