Using Patient Experience Surveys to Assess Diagnostic Safety in Urgent Care

Research Objective Diagnostic error is the leading patient safety concern in the U.S. One in twenty patients seeking care in the U.S. is estimated to be harmed by diagnostic errors annually. Identifying diagnostic pitfalls is a significant challenge. This study explored the utility of negative sentiment free‐text comments from patient satisfaction surveys to identify patient‐perceived failures in the diagnostic process within an urgent care context. Study Design We used an exploratory mixed‐methods study design to explore perceived breakdowns in diagnosis in urgent care. We systematically coded negative sentiment free‐text responses from de‐identified Clinician and Group Consumer Assessment of Healthcare Providers and Systems (CG‐CAHPS) surveys using content and semantic analyses. The coding schema was designed to discern concepts within the definition of diagnostic error and the pitfalls in the diagnostic process presented in the National Academy of Medicine's 2015 report on improving diagnosis. The process includes six domains: patient engagement with the health system, information gathering, information interpretation, working diagnosis, diagnosis communication, and treatment. Descriptive statistics are used to summarize the results. Population Studied Adult patients (≥18 years) presenting to urgent care centers between January and December 2019 who completed a CG‐CAHPS survey and included a response to one of two open ended questions on the survey. We selected urgent care for the study because patients do not have an existing relationship with the provider and access urgent care for primary and emergent care concerns. Principal Findings During the study period, 7245 CG‐CAHPS surveys were completed and returned. Of these, 4315 (60%) patients included a free‐text comment. Most of the comments were from women (n = 3072; 71.2%) with a mean age of 58 years (SD = 17). Of the total reports, 1630 (37.8%) were sentiment coded as negative. Negative reporters were primarily women (n = 1175; 72.1%) and younger than those reporting positive, neutral, or both positive and negative sentiments. Of the free‐text responses coded as negative, 328 (20%) reported at least one breakdown in the diagnostic process. The 328 reports translated to 548 breakdowns in the diagnostic process. Breakdowns were reported by patients across all six domains of the diagnostic process. Most breakdowns occurred during information gathering (n = 172;31.4%) and treatment (n = 148; 27.0%). Phrases used to describe a wrong diagnosis included “misdiagnosed,” “incorrectly diagnosed,” “my diagnosis was incorrect,” “get the diagnosis correct,” “proper diagnosis and care,” and “wrong diagnosis” (n = 91; 27.7%). We also observed 15 (4.5%) reports of “missed diagnosis”. Diagnostic accuracy and timeliness were also questioned in 117 (35.7%) and 139 (42.4%) of reports. Conclusions Patients have unique and powerful insights on diagnostic safety in urgent care settings. Patient experience surveys are an unlikely but rich source of patient‐perceived diagnostic safety issues, particularly identifying issues in information gathering (history and physical examination) and treatment. Implications for Policy or Practice Learning healthcare systems should promote the utilization of patient experience surveys to estimate the impact of diagnostic errors and to design systems‐based solutions for improving diagnostic safety. Primary Funding Source Agency for Healthcare Research and Quality.