Introduction: While narrative data are increasingly being recognized as essential for clinical understanding, they are still not considered truly amenable to scientific use, unable to meet the traditional standards of replicability, universality, and objectivity. Narrative data on individuals however can aid in the professional and consumer use of trial evidence in clinical decision-making by offering ways to move from evidence of group effectiveness to individual applicability, and by offering consumers/patients ways to imagine themselves in the stories that different options represent. And further, narrative data can be constructed to serve as outcome data on the effectiveness of interventions for particular individuals in particular contexts. Here one captures the total state of the person through narrative terms or narrative-derived data. Based on Gordon's years of ethnographic research in Italy on patients, family members, and health professionals experiences around breast cancer, the end of life, and genetic risk, individual cases can be generalized along some of the following lines: 1) 'paradigm cases' : cases emblematic of a particular type of story or outcome; 2) 'lived narratives': the movement of the individual through and into particular lived stories, based on the well-documented notion that our experience, including our illness and treatment experience, is shaped not only by stories we tell and hear, but by stories we actually live in or live for. This will be illustrated with results of a study conducted on the end of life with the Service for Palliative Care, Merate, Italy. 3) 'plot lines': cases are compared according to the types of outstanding 'plots' that dominate: e.g., a plot of gradual and unmitigated, downward spiral; a plot characterized by many ups and downs ; a plot of steady-state, followed by a sudden, dramatic, and quick decline. Together these ways of generalizing about individual narratives may be used to contextualize population evidence as well as to generate a new type of data-base on individuals that itself constitutes another type of outcome data. While not scientific in the traditional sense, these studies can be extremely rigourous, explicit, consensually based, producing data that both objective and explicitly positioned.