The use of risk scores in primary care settings—A focus group study

Abstract

Introduction Individualized risk scores hold substantial promise by helping to identify high-risk populations and subsequently guide shared decisions about surveillance, treatments or further testing. The use of risk scores in clinical settings is rare. The aim of the study was to identify barriers to the use of risk scores in primary care settings in Germany. Methods A network of primary care physicians was asked to participate in a focus group to discuss the use of risk scores. A literature review aided the design of the questionnaire guide. The focus group was conducted by a trained moderator. A second member of the research team observed the discussion. The focus group was tape-recorded. The record was transcribed verbatim and entered into atlas.ti for coding and analysis according to grounded theory. Two members of the research team coded the materials. Differences in coding were resolved in discussion. Analysis was conducted collaboratively. Results Three important themes were identified: (1) primary care setting as defined by regulation, (2) patients as resistant to prevention information, (3) feasibility of risk scores. Discussion The implementation of scientifically validated instruments like epidemiological risk scores in health-care settings is not a simple matter. They have to be transformed to fit the clinical setting. This transformation has to be in accordance with regular clinical practice. The literature on risk scores, risk communication, and prevention stresses education. This study shows that the failure of attempts to implement risk score instruments for primary care physicians is not merely the result of an educational deficit. The question that has to be answered prior to the use of risk scores in health-care settings is “for whom shall the risk score be used and for what purposes shall it be used?” Only then can epidemiological risk scores be used successfully in real settings and facilitate shared decision-making.