Helping doctors with health care solution RITA has been saved
Helping doctors with health care solution RITA
Perspectives on AI’s impact on the world
Each day, AI helps address critical challenges in industries around the globe. According to a study from the British Medical Association, the United Kingdom is facing a shortage of 50,000 clinicians. In this AI for Good perspective, Deloitte UK’s Sunny Dosanjh discusses an AI and a health care solution known as Referral Intelligence and Triage Automation (RITA) that aims to reduce clinical admin tasks.
What is RITA, and how does it help address health care problems?
RITA is an AI tool intended to triage outpatient referrals from primary care into secondary care. It aims to reduce the clinical administrative burden and help clinicians spend more time on patient care. While the national and global shortages of clinicians continue to increase, studies show that up to 20% of clinicians’ time is still focused on administrative tasks. According to a recent study by the British Medical Association, the United Kingdom is 50,000 clinicians1 short, and it takes 15 years to get to a consultant clinician level (senior grade-level doctor). There is a considerable demand for clinicians that is only growing without enough supply. Technologies like RITA help eliminate the administrative burden and enable clinicians to do what they do the best—focus on their patients.
How does the AI solution, RITA, work?
RITA tests the potential of using the latest AI and robotics to automatically triage incoming patient referrals and assign patients to appropriate pathways, reducing the administrative burden of lower-risk referrals by about 40%; allowing more complex and higher-risk referrals to remain in the purview of hospital clinicians.
RITA is integrated with an e-referral system that the National Health Service (NHS) built to extract referrals directly from the technical communication system. This technology uses natural language processing to extract medical terminology in multiple ways to map patient symptoms to clinical models that can pinpoint what diagnosis the patient is likely to have. Automating the triage process is estimated to reduce the overall time to triage.
What is the benefit of the solution?
The primary benefit is that RITA helps release the clinician’s time to focus on their patients. Deloitte’s research indicates that a consultant clinician takes about five minutes to review a referral. In the United Kingdom alone, doctors spend nearly 1 million hours reviewing referrals, which equates to the time of 500 full-time doctors. Imagine what impact 500 full-time clinicians could have when enabled to focus on patient care instead of administrative tasks. RITA is one solution the NHS can utilize to make an impact on the 50,000-clinician shortage.
Additional benefits of RITA include the solution’s alignment to national guidelines for standardization to provide the best treatment. This alignment to the national guidelines allows patients to receive a standardized care assessment the moment the referral hits the clinician’s inbox. Once a patient has been triaged, they can be funneled through the hospital system to receive additional care if needed.
Finally, RITA provides transparency. The solution gives increased visibility into what is otherwise quite an opaque process for health authorities to track a patient’s referrals in the system.
What are some of the key challenges and learnings?
There are two significant challenges: clinical safety by design and explainability.
In health care, clinical safety is the first and foremost priority when designing any solution. RITA has required a continued investment of time to create safety assessments and ensure national standards are being met. To alleviate the tremendous operational pressure clinicians face, system architects of RITA are also required to stay abreast of the evolving best practices in the medical field to ensure compliance.
Explainability is the second challenge RITA has worked to address. In health care, it’s difficult to build explainability around machine learning-based models. To tackle this challenge, explainability was a key consideration when building the solution to aid doctors and clinicians when adopting the solution.