113 An exploration of physician perspectives on mammography screening for average-risk women

Abstract

Objectives Although the influence of practice guidelines on physicians’ ordering of mammography screening is well established, conflicts exist in the recommended mammography screening initiation ages and screening frequency among different mammography guidelines. Furthermore, growing evidence nuancing the benefits and harms of screening has put systematic screening into question. Understanding physician perspectives on the evidence that affects patient care will be important for informing future best practices as guidelines evolve. A large amount of data exists on these perspectives from thousands of physicians who read and react to clinical research synopses (Patient Oriented Evidence that Matter: POEMs) through an ongoing continuing medical education program. The purpose of this study is to explore physicians’ perspectives on clinical research regarding mammography screening for average-risk women and the extent to which they use this POEM information in their clinical practice. Method The Essential Evidence Plus database was searched from 2012 to 2017 with the term ‘breast neoplasm’ to identify relevant POEMs on mammography screening, screening decision-making, and overdiagnosis. Using the Information Assessment Method (IAM), physician ratings and comments about mammography evidence were extracted from reflections on clinical research summarized as POEMs. The items of interest in the IAM were those calling on physicians to reflect on the value of the information and its applicability. Quantitative data were assessed with descriptive statistics. Using an iterative approach, the qualitative data were subjected to both an inductive and deductive analysis. These data were coded thematically into sub-themes, which were grouped into major themes. Connections were sought between both quantitative and qualitative data. Results Four relevant POEMs were identified. The number of quantitative POEM ratings ranged from 1243 to 1351. Across all POEMs, among the physician ratings about using the information for a patient, over 50% were about using it in a discussion with a patient or other healthcare provider. Three major themes emerged from the analysis of 310 qualitative comments across all POEMs: 1) Perspectives on information presented in POEMs, 2) Applying this information in practice, and 3) Confronting clinical and cultural realities. Physicians held diverse perspectives on the value of the POEMs. Some physicians continued to support screening while others condemned harms such as overdiagnosis. Although physicians noted the potential of the POEM to improve patient counseling, access to this information did not necessarily diminish perceived challenges in screening discussions. Physicians advocated for the personalization of screening decision-making and patient-centered approaches to respect each patient’s values and preferences. Conclusions This study of POEMs data reveals important divergences in the ways physicians value clinical evidence on mammography screening and use it in practice. Physicians’ intent to use the POEMs to support balanced screening discussions and prevent unnecessary testing and treatment suggest the potential of this information to reduce overdiagnosis at the level of the patient-provider consultation. However, our results also revealed challenges experienced by physicians in understanding and explaining evidence about screening and overdiagnosis. This research acknowledges the constant evolution of evidence on mammography screening and therefore points to the difficulty in deciding what exact information should be shared with average-risk women considering screening. Despite continuing controversies in mammography screening, physicians expressed the importance of optimizing ethical screening decision-making and respecting women’s personal values and preferences. Further research should explore how primary care providers can implement shared decision-making on breast cancer screening with their patients.