In the dynamic and demanding world of healthcare, delivering high-quality patient care while maintaining efficiency in administrative duties is a perpetual challenge. Few environments embody this tension more than the Musculoskeletal and Trauma and Orthopaedics departments at Guy's and St Thomas' NHS Foundation Trust (GSTT), one of the United Kingdom's foremost teaching and research hospitals. With approximately 23,000 employees, a sprawling portfolio of patient needs, and a reputation for clinical innovation, GSTT is emblematic of the operational pressures facing the modern NHS. Within this context, the trust’s strategic adoption of Dragon Medical One—an AI-driven speech recognition platform hosted on Microsoft Azure—represents a bold, technology-forward response to the documentation burden that often impedes clinical flow and efficiency.
October 2023 marked a watershed moment for digital healthcare transformation at GSTT. Together with King’s College Hospital NHS Foundation Trust (KCH), GSTT launched a major, unified electronic patient record (EPR) system powered by Epic, in what became one of the largest single-site Epic go-lives worldwide. This ambitious move consolidated multiple legacy systems into a single, cohesive digital record, providing a potential bedrock for seamless care but posing formidable challenges in change management. Hospitals often encounter resistance and workflow disruptions during EPR transitions, with concerns around the learning curve, integration issues, and the risk of increased administrative load during the bedding-in phase.
Recognising these challenges, GSTT and KCH made a deliberate, strategic decision: deploy Dragon Medical One before the full Epic rollout. This calculated move sought to familiarise clinicians with voice-enabled workflows, improve overall confidence in digital documentation, and proactively mitigate the risk of clinical productivity loss during the transition to Epic.
As the Epic EPR went live in October 2023, thousands of doctors, nurses, and allied healthcare professionals embraced Dragon as the normative standard across routine and advanced clinical settings. It is in this broad landscape that the experience of Ryan Strother, Musculoskeletal Physiotherapist and Fracture Clinic Physiotherapy Team Lead at St Thomas’ Hospital, highlights the compounded value of advanced speech recognition within daily clinical practice.
Strother first began using Dragon Medical One in May 2023, primarily as a means to relieve the typing burden associated with producing clinic notes and formal correspondence. However, following the Epic EPR’s implementation, he took further steps, dedicating 45 minutes to advanced Dragon training and becoming a “super-user” capable of optimising the platform’s full array of features—notably, step-by-step voice commands and integration with Epic’s SmartTools.
It is critical to note, as Strother himself emphasizes, that this does not account for the additional time saved through dictation in place of manual typing; those gains vary widely based on the complexity of each patient’s presentation and the length of the clinical note. Nevertheless, the ability to curtail repetitive administrative tasks at scale offers a significant improvement, especially when extrapolated across hundreds of appointments each month.
This insight supports a broader truism in the digitisation of healthcare: generic, top-down IT solutions can easily flounder if they fail to reflect the distinctive, sometimes idiosyncratic realities of frontline practice. Dragon’s real-world efficacy at GSTT owes much to its flexibility and the ability for super-users like Strother to co-create workflows that genuinely alleviate clinician pain points.
While challenges remain—around training, resilience, and long-term cost management—the early evidence is clear. In environments where every minute reclaimed can be redirected to patient care, the judicious application of speech technology is emerging as a critical enabler of both clinical efficiency and record quality.
As the NHS and global health systems continue their march toward more advanced digital workflows, the balance of automation, user empowerment, and data stewardship will likely define not only care standards but also the day-to-day lived experience of clinicians on the frontline. Dragon Medical One’s story at Guy’s and St Thomas’ suggests that, with the right strategies, technology can serve as a true ally in this journey for better, more human-centred healthcare.
Source: Microsoft How a Dragon Medical One super-user improved efficiency at Guy's and St Thomas' | Microsoft Customer Stories
The Setting: A New Era of Digital Health at GSTT
October 2023 marked a watershed moment for digital healthcare transformation at GSTT. Together with King’s College Hospital NHS Foundation Trust (KCH), GSTT launched a major, unified electronic patient record (EPR) system powered by Epic, in what became one of the largest single-site Epic go-lives worldwide. This ambitious move consolidated multiple legacy systems into a single, cohesive digital record, providing a potential bedrock for seamless care but posing formidable challenges in change management. Hospitals often encounter resistance and workflow disruptions during EPR transitions, with concerns around the learning curve, integration issues, and the risk of increased administrative load during the bedding-in phase.Recognising these challenges, GSTT and KCH made a deliberate, strategic decision: deploy Dragon Medical One before the full Epic rollout. This calculated move sought to familiarise clinicians with voice-enabled workflows, improve overall confidence in digital documentation, and proactively mitigate the risk of clinical productivity loss during the transition to Epic.
Seamless Integration: Dragon Medical One Meets Epic
Dragon Medical One is designed to offer high-accuracy, cloud-based speech recognition that integrates closely with leading EPR platforms—including Epic. At GSTT, Dragon’s deployment was not limited to desktop environments; its capabilities extend to mobile workflows through Epic Haiku, Canto, and Rover, allowing clinicians to dictate clinical notes, complete documents, and navigate records by voice from a range of devices.As the Epic EPR went live in October 2023, thousands of doctors, nurses, and allied healthcare professionals embraced Dragon as the normative standard across routine and advanced clinical settings. It is in this broad landscape that the experience of Ryan Strother, Musculoskeletal Physiotherapist and Fracture Clinic Physiotherapy Team Lead at St Thomas’ Hospital, highlights the compounded value of advanced speech recognition within daily clinical practice.
A Clinician’s Perspective: Ryan Strother’s “Super-User” Journey
Ryan Strother’s clinical environment is one of relentless activity and precision. His daily schedule encompasses the fast-paced, same-day Fracture Clinic—handling up to 20 walk-in patients each morning—and the more detailed, 30–40-minute appointments of the MSK clinic, with up to 17 patients seen daily. The documentation demands are therefore immense: from swiftly capturing clinically relevant details during a crowded fracture clinic session to recording nuanced patient narratives after in-depth musculoskeletal assessments.Strother first began using Dragon Medical One in May 2023, primarily as a means to relieve the typing burden associated with producing clinic notes and formal correspondence. However, following the Epic EPR’s implementation, he took further steps, dedicating 45 minutes to advanced Dragon training and becoming a “super-user” capable of optimising the platform’s full array of features—notably, step-by-step voice commands and integration with Epic’s SmartTools.
Unlocking Workflow Efficiency Through Voice
A major innovation came through the development of custom, voice-driven workflows tailored to his clinical routines. Traditionally, each MSK appointment required a sequence of about 30 mouse clicks just to navigate through the necessary screens and templates before notes could even begin. Strother’s advanced Dragon workflow condensed this process to just four well-configured voice commands, reducing the pre-documentation phase from approximately two minutes to just 20–25 seconds per patient—a dramatic time reduction.It is critical to note, as Strother himself emphasizes, that this does not account for the additional time saved through dictation in place of manual typing; those gains vary widely based on the complexity of each patient’s presentation and the length of the clinical note. Nevertheless, the ability to curtail repetitive administrative tasks at scale offers a significant improvement, especially when extrapolated across hundreds of appointments each month.
Co-Design: Clinician-Led Technology Adaptation
Strother’s journey is not just about technical proficiency—it is a case study in the power of clinician co-design. As he observes, “Understanding workflow is critical — any solutions developed for clinicians must be purpose-built to integrate with the way they work. That’s why co-design with clinicians is essential; involving them from the outset ensures the technology truly supports clinical practice.”This insight supports a broader truism in the digitisation of healthcare: generic, top-down IT solutions can easily flounder if they fail to reflect the distinctive, sometimes idiosyncratic realities of frontline practice. Dragon’s real-world efficacy at GSTT owes much to its flexibility and the ability for super-users like Strother to co-create workflows that genuinely alleviate clinician pain points.
Critical Analysis: Strengths, Pitfalls, and the Future of Speech in Healthcare
The GSTT case study offers a window into both the compelling strengths and real limitations of speech recognition in healthcare environments.Notable Strengths
- Substantial Time Savings: The reduction from 30 mouse clicks to four voice commands, and from two minutes to under half a minute per non-note task, is backed by internal departmental time trials. This scale of efficiency is particularly potent in high-volume outpatient or ambulatory care environments.
- Improved Documentation Quality: By easing the process of recording clinical notes, Dragon enables clinicians to capture richer narrative detail—potentially improving continuity of care, medico-legal protection, and patient understanding.
- High Adoption and Familiarity Rates: The deliberate pre-Epic deployment strategy ensured that Dragon was familiar and trusted before the inevitable stresses of major system migration, minimising resistance and workflow disruption.
- Platform Agnosticism: Hosting Dragon on Azure and integrating it with Epic (desktop and mobile) means clinicians can access the system from almost any location, using hardware already in their hands.
- Clinician Co-Design: Involving end-users in workflow mapping and optimization led to more targeted, context-aware deployments, reducing the likelihood of “alert fatigue” or irrelevant automation.
Potential Risks and Unresolved Challenges
- Reliance on Internet Connectivity and Cloud Performance: Cloud-based speech recognition depends on reliable connectivity and adequate backend compute resources. While Microsoft Azure’s uptime is strong, any regional outage or network glitch could disrupt documentation at critical times.
- Initial Learning Curve and Ongoing Training Needs: While super-users like Strother adapted quickly, some clinicians may struggle to rewire established habits or fully exploit Dragon’s most powerful features without extended support and periodic retraining.
- Data Security and Privacy: Transmitting sensitive patient voice data across the cloud, even in encrypted form, raises perennial questions about compliance with GDPR, NHS Digital guidelines, and wider public expectations of health data privacy. No breaches have been reported in connection to the project, but continuous vigilance is required.
- Speech Accuracy and Accent Recognition: Context-sensitive, medical-grade speech recognition has advanced significantly, but challenges remain in environments with high background noise, non-standard dialects, or rare technical terminology. Periodic updates and customizable medical vocabularies partially mitigate these risks.
- Vendor Lock-In and Cost: As voice recognition platforms become embedded within broader EPR systems, organisations can become dependent on a small pool of providers (e.g., Microsoft/Nuance, Epic). While this ensures integration, it potentially limits price competitiveness and flexibility.
Quantifiable Impact: What the Data (and the Sources) Say
According to public case studies released by Microsoft and corroborated by GSTT internal progress reports, the pre-Epic Dragon rollout was a critical factor in the smooth transition. Review of user feedback and adoption statistics revealed a strong correlation between Dragon training and confidence in EPR workflows, with positive downstream impacts on documentation completeness and overall clinician satisfaction[/url]. However, these gains are not universally guaranteed, depending on how deftly technology deployment is aligned with existing practice.Lessons for the Broader NHS and Beyond
Guy’s and St Thomas’ experience with Dragon Medical One offers actionable lessons for other NHS Trusts and international health systems considering similar digital upgrades:- Phase Major System Changes: Deploy new workflow-enhancing tools before major EPR changes to build user trust and reduce cognitive overload.
- Champion Super-User Networks: Invest in super-user “ambassadors” who can train peers and act as in-house support, accelerating change adoption.
- Prioritise Interoperability: Ensure speech platforms not only integrate with the primary EPR but also work across mobile devices and different clinical contexts.
- Commit to Ongoing Co-Design: Treat technology adaptation as an iterative, adaptive process—not a one-off rollout.
Conclusion: Speech Technology’s Place in Clinical Care
The GSTT case illuminates the real-world benefits that cloud-based, AI-enabled speech recognition can deliver when deployed thoughtfully, in tandem with major digital health investments like Epic. By blending robust technical integration, clinician-led workflow design, and a patient-centred ethos, Dragon Medical One has shifted clinical documentation from a burdensome administrative afterthought to a streamlined, supportive facet of care.While challenges remain—around training, resilience, and long-term cost management—the early evidence is clear. In environments where every minute reclaimed can be redirected to patient care, the judicious application of speech technology is emerging as a critical enabler of both clinical efficiency and record quality.
As the NHS and global health systems continue their march toward more advanced digital workflows, the balance of automation, user empowerment, and data stewardship will likely define not only care standards but also the day-to-day lived experience of clinicians on the frontline. Dragon Medical One’s story at Guy’s and St Thomas’ suggests that, with the right strategies, technology can serve as a true ally in this journey for better, more human-centred healthcare.
Source: Microsoft How a Dragon Medical One super-user improved efficiency at Guy's and St Thomas' | Microsoft Customer Stories
