Abstract
BACKGROUND: Professional knowledge aims at improving practice. It reduces uncertainty in decision-making, improves effectiveness in action and relevance in evaluation, stimulates reflexivity, and subjects practice to ethical standards. Heuristics is an approach to problem-solving, learning, and discovery employing a practical methodology that, although not optimal, is sufficient for achieving immediate goals. This article identifies the desirable, heuristic particularities of research in professional, medical practice; and it identifies what distinguishes this research from scientific research. MAIN TEXT: We examine the limits of biomedical and sociological research to produce professional knowledge. Then, we derive the heuristic characteristics of professional research from a meta-analysis of two action-research projects aimed at securing access to essential generic drugs in Senegal and improving physicians' self-assessment and healthcare coordination in Belgium. To study healthcare, biomedical sciences ignore how clinical decisions are implemented. Decisions are built into an articulated knowledge system, such as (clinical) epidemiology, where those studied are standardisable - while taking care of patients is an idiosyncratic, value-based, person-to-person process that largely eludes probabilistic methodologies. Social sciences also reach their limits here because descriptive, interpretative methods cannot help with gesture and speech quality, while the management of the patient's suffering and risks makes each of them unique. Research into medical professionalism is normative as it is intended to formulate recommendations. Scientific data and descriptions are useful to the practitioner randomly, only from the similarities in the environment of the authors and their readers. Such recommendations can be conceived of as strategies, i.e., multi-resource and multi-stage action models to improve clinical and public health practice. Action learning and action-research are needed to design and implement these strategies, because their complexity implies trial and error. To validate a strategy, repeated experiences are needed. Its reproducibility assumes the description of the context. To participate in medical action-research, the investigator needs professional proficiency - a frequent difficulty in academic settings. CONCLUSION: Some criteria to assess the relevance of publicly funded clinical and public health research can be derived from the difference between scientific and professional knowledge, i.e. the knowledge gained with real-life experience in the field.