Try not to think of a robot from Star Wars, rather think of software that supports administrators and clinicians by removing ‘grunt work’ and, increasingly, by supporting decision making by leveraging Machine Learning. Josh Murray of psHEALTH looks at how new technologies can reduce cost and ensure that the right patient gets to the right place first time.
SPA/Referral Management Challenges
The use of a single-point of access or a triage hub is increasingly recognised as critical for getting the right patient to the right place. Whether your SPA is a physical or virtual hub and works for an acute trust, a community service or, provides referral management on behalf of a CCG or an ACO, the purpose and challenges remain the same. How do you safely and quickly make sure the patient is sent or referred to the right destination? Ensure that your policies are consistently implemented? Deliver timely and relevant management information? How do you minimise admin cost?
Removing repetitive, thankless admin tasks
Admin staff are often bogged down in checking incoming referrals for completeness, managing incoming emails, re-keying or copying & pasting information between systems (and a few excel sheets).
This is time consuming, sometimes complex, almost always repetitive – perfect for a ‘robot’. Software can now automatically capture referrals (from e-RS, email or even fax) and extract relevant information. With new intelligent form-reading capabilities, the software can automatically extract patient details, referral reasons, check for red flags and only escalate to a user if something is ‘not quite right’.
Studies have shown this type of software does the job 5 to 10 times faster than admin staff. It needs no break and, if instructed properly, will not make silly mistakes. What the software can’t do is have a friendly call with a patient, but it can free a person up from thankless tasks to have that call.
Helping make better, more consistent triage/referral decisions
A combination of Decision Rules and Machine Learning, a branch of Artificial Intelligence, has now become an important tool for helping triage of referrals. At a simple level, Machine Learning can identify and sort documents. Think about a referral with five attachments. How do you identify the referral letter and attached blood tests? By applying Rules, it is now possible to present the referral to the right clinician – maybe an Extended Scope Practitioner for MSK – or, send it to administrator because a template is missing or, even automatically send it back to the referrer with relevant advice.
Machine Learning can help us make better decisions. By learning from patterns and Algorithms, it can provide recommendations based on multiple data sources, including travel distance, outcomes, wait-times and even costs. Increasingly we must enable the patient to engage with the process and show them the basis for recommendations.
Where humans are happy and productive? Interacting with patients
Just a few weeks ago Jason Sims, psHEALTH’s Product Manager and a keen advocate of Automation and Machine Learning, spoke at a conference on the topic in San Francisco. “We are excited about the potential, but this is not about removing the human aspect from healthcare, it is really about maximising the time administrators and clinicians have to make a direct, positive impact on patients.”