The influence of self-compassion on the mental health and well-being of women experiencing infertility: a systematic review.

The increasing burden of clinical documentation contributes to physician inefficiency and burnout, especially with the widespread adoption of electronic health records (EHRs). EHR implementation has significantly increased physicians' documentation time from approximately 16% to 28% of observed clinical time, adding strain to already busy schedules []. This trend is confirmed across studies, with evidence showing that after-hours documentation is perceived as one of the most burdensome aspects of physicians' work and is strongly associated with burnout symptoms []. The design and usability of current EHR systems exacerbate the problem, as physicians report that many documentation tasks are clerical in nature and do not require clinical training []. Furthermore, poorly designed interfaces and time-consuming navigation contribute to cognitive fatigue []. Ultimately, the longer work hours, heavy clerical demands, and interruptions to patient care that come with current documentation methods have created frustration among physicians []. This documentation burden has been associated with medical errors, risks to patient safety, poor documentation quality, job attrition, and burnout []. These findings highlight the need for targeted interventions and improvements to alleviate documentation burden and promote both efficiency and physician well-being in clinical practice. To address this, many practices have implemented human medical scribes, which have been shown to reduce documentation time and improve physician satisfaction [,]. In high-volume specialties like orthopaedics, scribes can have a large impact: one orthopaedic sports medicine clinic found that scribes reduced half-day documentation time from 87 to 26 minutes []. More recently, “ambient” artificial intelligence (AI) scribes have demonstrated similar benefits. These systems use advanced speech recognition and natural language processing to automate clinical documentation, easing administrative burden []. Ambient AI scribe systems, such as Microsoft DAX Copilot, integrate directly into the clinical workflow by passively capturing patient-clinician conversations through a secure audio interface and automatically generating draft visit notes within the EHR. These tools use speech recognition and natural language processing to identify relevant clinical concepts, structure the encounter narrative, and populate the documentation in real time. The clinician then reviews, edits, and signs the AI-generated note, offloading the majority of clerical work to the system while preserving physician oversight and documentation quality. Ambient AI scribes in outpatient clinics can reduce time spent on notes by 20.4% and lower after-hours EHR work by 30.0% []. Additionally, Rotenstein et al. [] showed that virtual scribe use reduced total EHR time per appointment by an average of 5.6 minutes. These findings suggest that AI-assisted documentation may improve efficiency and physician well-being. However, their impact in specialty practices, such as hip and knee arthroplasty, has not been well reported. This study presents a prospective evaluation of an AI scribe system in a high-volume hip and knee arthroplasty clinic, hypothesizing that integration of the technology would shorten encounter duration and lessen physician workload. Institutional review board exemption was obtained at our institution prior to data collection. This was a prospective quality-improvement cohort study at a single institution performed by a single surgeon. Hip and knee arthroplasty patients were divided into 2 cohorts based on the method of clinical documentation used during their arthroplasty clinic visits. The control cohort consisted of patients whose clinical encounters were documented using traditional methods by the physician and medical assistant (MA) within the electronic medical record (EMR) system (Cerner Oracle Health, Kansas City, MO). The AI intervention cohort included patients whose visits were documented with the use of an AI-powered clinical documentation assistant (Microsoft DAX Copilot, Redmond, WA), in collaboration with the orthopaedic surgeon. Patients were grouped according to clinic workflow, with the first series of encounters documented using the traditional physician-MA EMR workflow, followed by the next series of encounters documented using the AI-powered scribe system; no randomization or patient-level selection criteria were used, and assignment into each group was determined solely by the documentation method in use during that clinic session. All patients whose encounters were assigned to the AI intervention group were informed of the use of the AI scribe system prior to their visit and verbally consented to the recording of their encounter. No patients refused participation.