How Ordis is helping care teams reclaim time for person-centred care.
About 2 out of every 3 people who reach older age will need longer-term care and support with daily activities like eating, getting around or bathing, in addition to specific healthcare needs. Nurses and care workers are essential to helping these people, but will there be enough trained professionals to meet our future care needs? The WHO predicts that there will be a shortage of 4.5 million nurses by 2030, and the EU predicts the need for a further 3.9 million care staff by 2035.(1, 2) With staffing shortages a daily occurrence and increasing rates of work-related burnout, it is in every health and care institutions’ interest to support staff in their daily tasks, to reduce work-related exhaustion and bolster staff numbers.
Long-term residential facilities and care homes offer ongoing care and support where care providers not only have to provide safe, quality care, but also keep accurate records of their observations and care activities. For care providers, like nurses, care assistants and nurse managers, documentation increasingly takes up more time and mental energy, time and energy that could be spent on providing the compassionate, in-person care they are trained to provide. There are distinct benefits to keeping up-to-date, clear and accurate care records including continuity of care, improved quality of care and the ability to analyse and improve health services and safety. However, the burden of maintaining timely, accurate and quality notes, is a notable contributor to care provider burnout.(3)
At Ordis Healthcare, we have recognised the early potential to use ambient artificial intelligence (AI) technology to support nursing and care workers in providing person-centred care. Ambient artificial intelligence uses software that combines audio and other sensors. Speech-to-text tools capture and transcribe the interactions between the care provider and care receiver, to produce documents such as care summaries, plans, observations or incident reports into formats that can be uploaded to electronic health records.(4)
The use of ambient AI technology has already demonstrated benefits for medical healthcare providers including:(3, 5-7)
Interestingly, while ambient AI technology is being increasingly used in medical care settings, it has not yet been widely adopted in the residential care setting, where nurses and care assistants could experience similar benefits for their work satisfaction and quality of care.
At Ordis Healthcare, we started our user journey with frontline care staff who carry the burden of documentation, and the care facility managers who have the regulatory requirements to focus on safety and results. With the ongoing requirements to meet care and documentation standards for oversight organisations like the Health Information and Quality Authority (HIQA) in Ireland, and the Care Quality Commission (CQC) and the Regulation and Quality Improvement Authority (RQIA) in the United Kingdom, our goal has been to help care staff collect accurate, real-time data while ensuring trust and requiring minimal effort.
The implementation of the Patient Safety (Notifiable Incidents and Open Disclosure) Act 2023 in Ireland emphasised the need for transparency and open disclosure along with accurate and prompt reporting for incidents. In our Ordis pilot project, we developed an incident reporting system as our minimal viable product. The goal was to help staff create reports on incidents, such as falls or sudden illness, using speech-to-text technology to capture the details of the incident. The information could then be readily updated and entered into the various electronic records and forms. While, fortunately, there were few incidents involving residents during the initial pilot, feedback from care staff showed how our technology could be more widely applied to support their daily document tasks in care summaries, progress notes, and handover reports.
Using a continuous testing and refinement loop, with feedback from Ordis Healthcare’s users we were able to react and update the system to address user concerns and system efficiencies. We identified voice capture accuracy, technological improvements and system requirements for development within the unique use case of the nursing home environment.
The unique care environment of residential care facilities and nursing homes is key to how ambient AI scribes need to be developed for long-term care needs. Where other medical AI scribe technology used in primary care and hospital environments typically handles long healthcare conversations, nursing homes need short, real-time capture of fragmented information. The context for the information on a person receiving care is built from multiple small inputs, from different care providers, throughout the day. A system needs to be easily accessible and usable to capture and summarise these multiple data points.
Care home staff noted that progress notes were the most time-consuming documentation task. They accumulate information from multiple streams including medication notes, handover notes, vitals, doctor prescriptions and daily activities and observations. Automation and summarisation of these pieces of data were identified as a key areas of impact for the Ordis Healthcare system.
Voice capture accuracy is a key metric for speech-to-text technology. With local care staff shortages requiring recruitment of international staff, many of these staff may have English as a second or third language. In this context, our early Ordis Healthcare pilot showed poor accuracy due to challenges with different accents and pronunciations. Indian (Malayalam), Filipino accents, Irish names, and medication pronunciations were especially challenging. Accent adaptation and training were crucial to achieving acceptable accuracy.
In response to analysis of accuracy concerns in the Ordis Healthcare MVP, we changed the technology we were using for transcription and noted an immediate improvement in the quality of the captured information. Iteration of the Ordis Healthcare system requirements enabled us to identify and implement quick and accurate voice capture technology to adapt to diverse accents, especially for those increasing international staff numbers from South Asian and African countries like India, Philippines and Nigeria where different languages and dialects could affect English pronunciation. Pronunciation directly impacts off the shelf speech-to-text models like Whisper, Azure STT and Google.
In the early Ordis Healthcare incident reporting product, the speech-to-text dictation would appear in real time for users to see. We soon realised that this caused distraction and at times anxiety if an error was noted. We recognised that passive processing of the audio without the live dictation text display allowed the care providers to focus on the caregiving and treatment with the ability to review and update the captured audio at the end of the episode of care. This was further enhanced by the system capturing the speech unobtrusively, then presenting a structured summary of the assessment and care for detailed progress notes.
Fundamental to integrating any system into the care environment, we mapped existing workflows and data sources that would require deep integration by the Ordis Healthcare system as well as interoperability needed with other electronic health record systems.
With any AI system, particularly related to ambient technology, personal and data privacy are essential considerations in building secure, trustworthy systems. In the development of the Ordis Healthcare system, careful consideration has been given to the General Data Protection Regulation (GDPR) and its application to the processing of personal data in the EU, the obligations on data controllers and processors, and strong protections for the data subjects as stipulated in the Data Protection Act 2018.(3, 4) The Ordis Healthcare system team has focused on meeting the highest standards of GDPR compliance and ensuring patient privacy and security.
To address privacy concerns, the Ordis Healthcare system ensures that all processing occurs within a secure, locally hosted Ordis cloud environment. The foundational AI model is deployed in this controlled infrastructure, ensuring that no data leaves the organisation’s secure environment.
The audio technology is used solely for real-time transcription. Audio input is not stored or retained once converted to text, and there is no continuous or background recording. Recording must be manually initiated at the start of a care episode or incident, maintaining full control in the hands of care staff.
Any identifiable information captured during transcription is automatically stripped before being used in the system, ensuring that only abstracted, non-identifiable data is retained for documentation purposes.
Access to resident information is strictly governed through the facility’s existing care management software, which must explicitly authorise the AI Scribe to access only the data necessary for documentation.
Above all, Ordis upholds the principle of informed consent. Care givers (or their representatives) must be fully informed about the technology’s use in their care delivery and retain the right to withdraw consent at any time.
The workplace wellbeing of nurses and care assistants within the long-term care environment is essential to the health, wellbeing and safety of residents. Ambient AI technology has already been shown to help healthcare staff to feel less burdened by documentation and be able to focus more on meaningful, person-centred care. Along with the physical activities of compassionate care, documenting care and outcomes is an ever-increasing priority. Managers and directors of nursing carry a significant responsibility to ensure that residents in their care receive the right care, at the right time and with the best possible results. When things go wrong, (and accidents do happen), these incidents need to be quickly addressed, accurately documented and analysed to reduce future risk.
At Ordis Healthcare, we develop and implement ambient AI using proprietary speech-to-text models tailored for the frontline healthcare staff and their existing workflows. Plugged into their existing care management tools, our technology helps care providers to gather relevant information about people receiving care and how that care is provided. Designed by a multidisciplinary team of healthcare professionals and technologists for the complex, multilingual, and fast-moving world of care facilities, Ordis adapts to every environment. It understands varied accents, context, and terminology ensuring that every spoken interaction becomes part of a connected care record. Our system enables the care provider to focus on the person needing their attention while ensuring that correct and useful information is collected, summarised and saved to add to relevant records and documents after the moments of person-centred care have taken place. Our current product focuses on information transcription and summarisation for residential and nursing home facilities. Our goal is to offer ambient assisted living (AAL) solutions with voice technology to support people in their health and daily living and enhance the essential work of care providers.
Please feel free to contact us and schedule a demo of the system.
Discover how Ordis can automate compliance and documentation effortlessly, giving your staff the freedom to prioritize patient care.
Viewplus Ireland Limited © 2025 | All Rights Reserved | Privacy Policy
Viewplus Ireland Limited © 2025
All Rights Reserved
Privacy Policy
