Background Audit and feedback (A&F) often successfully enhances health professionals’ intentions to improve quality of care but does not consistently lead to practice changes. Recipients often cite data credibility and limited resources as barriers impeding their ability to act upon A&F, suggesting the intention-to-action gap manifests while recipients are interacting with their data. While attention has been paid to the role feedback and contextual variables play in contributing to (or impeding) success, we lack a nuanced understanding of how healthcare professionals interact with and process clinical performance data. Methods We used qualitative, semi-structured interviews guided by Normalization Process Theory (NPT). Questions explored the role of data in quality improvement, experiences with the A&F report, perceptions of the data, and interpretations and reflections. Interviews were audio-recorded and transcribed verbatim. Data were analyzed using a combination of inductive and deductive strategies using reflexive thematic analysis informed by a constructivist paradigm. Results Healthcare professional characteristics (individual quality improvement capabilities and beliefs about data) seem to influence engagement with A&F to a greater degree than feedback variables (i.e., delivered by peers) and observed contextual factors (i.e., strong quality improvement culture). Most participants lacked the capabilities to interpret practice-level data in an actionable way despite a motivation to engage meaningfully. Reasons for the intention-to-action gap included challenges interpreting longitudinal data, appreciating the nuances of common data sources, understanding how aggregate data provides insights into individualized care, and identifying practice-level actions to improve quality. These factors limited effective cognitive participation and collective action, as outlined in NPT. Conclusions A well-designed A&F intervention is necessary but not sufficient to inform practice changes. A&F initiatives must include co-interventions to address recipient characteristics (i.e., beliefs and capabilities) and context to optimize impact. Effective strategies to overcome the intention-to-action gap may include modelling how to use A&F to inform practice change, providing opportunities for social interaction relating to the A&F, and circulating examples of effective actions taken in response to A&F. More broadly, undergraduate medical education and post-graduate training must ensure physicians are equipped with QI capabilities, with an emphasis on the skills required to interpret and act on practice-level data.

H ealth organizations have begun to explore ways to address social determinants of health to improve care and reduce health inequities. 1-6 One barrier is insufficient individual-level data on social determinants of health. 7-10 Gender identity and sexual orientation are important social determinants, 11-13 but such data are rarely routinely collected by health care organizations. 14-16 Gender identity is a predictor of health outcomes and can affect access to health care. 17,18 Transgender and gender-diverse individuals 19-22 (Box 1) consistently report negative experiences in health care related to their gender identity. 23-25 Sexual orientation is also an important predictor of health status and access to health care. 15 For example, women who identify as lesbian have lower rates of cervical cancer screening than those who identify as heterosexual 26,27 and bisexual people report more unmet health needs than heterosexual people. 26 Information on gender identity is particularly crucial to the care of transgender and gender-diverse individuals, and whether to disclose is not always a choice. 15 Being comfortable with disclosing gender identity and sexual orientation can be an indicator of trust and can influence the quality and effectiveness of care. 15,28 Certain health interventions, such as human papillomavirus vaccination or HIV pre-exposure prophylaxis, are focused on specific groups based on sexual orientation. 29 Recently, the Canadian Institute for Health Information suggested that data on gender identity be collected as 1 of 5 domains to identify health inequalities. 30 It remains unclear how Canadian patients react to being asked routinely about sexual orientation and gender identity, or what responses such questions would generate. The objective of this study was to examine patients' reactions to routinely being asked about their sexual orientation and gender identity in a primary health care setting. We also compared answers to the gender identity question against other data on gender identity in the medical chart. Methods Setting and context In 2011, health organizations in Toronto developed a sociodemographic survey to identify inequities at a system level, by reviewing RESEARCH VULNERABLE POPULATIONS