15MWTH - myhealthspecialist.com
13: Doctors Recommend Other Doctors With Dr Kartik Modha
13th October 2017
15MWTD - Kafoodle
15: Optimising the Dining Industry with Tarryn from Kafoodle
22nd January 2018
Show all

14: Using Machine-Learning to Reduce No-Shows with Benjamin Fels

15MWTD -Benjamin Fels - Macro Eyes
Today, we have Benjamin, the Co-founder of Macro-Eyes, a machine-learning company that simplifies personalised patient care. One of their products is Sibyl, it’s a scheduling tool that can help predict when patients are most likely to attend their medical appointments, therefore maximising appointment use. Learn about Benjamin’s background in hedge-funds and finding meaningful patterns in data, learn how Sibyl works, and why he wanted to address the not-so-shiny problem of patient scheduling.

“Patient scheduling is one of those things that maybe is so ubiquitous that it gets brushed aside for the bigger, shinier, and perhaps more superficial problems.”

Sibyl predicts when each patient is most likely to show and then builds a strategic schedule to maximise utilisation and minimise gaps in the schedule.

Scheduling is the front-door of care. However, it is incredibly complex and often overlooked. The focus of the Sibyl is to maximise the probability that exactly one patient is going to show up for each slot of the day.

“Slot by slot, it recommends the best-fit time for the patient and the provider of care.”

In the US, overbooking is a common practice that results in longer wait times during some periods, and idle times when patients don’t show up to appointments. Although in the UK overbooking is not practised, it has been estimated that patient no-shows cost the NHS £1 billion per year.

The Sybil technology is 75-80% accurate in its predictions, which allows medical staff to focus their interventions on only the patients most likely to miss appointments. The intelligence inside the technology has been deployed and refined at leading academic medical institutions and in one of the largest health systems in the United States. Over 500 000 medical records and 2 million appointments were analysed to make these predictions, and the tech learns in each new institutions where it’s used.

Also covered in this episode:

How Benjamin’s background working in a hedge fund helped prepare him for this role
– The similarities between machine learning in finance and medicine
– How the Sibyl technology works
– The lessons we can learn from Amazon about scheduling efficiency
– The costs associated with using and developing Sibyl
– How efficient and effective scheduling processes can save money

Find out more:

You can find Sibyl at the website https://www.gosibyl.com/ or on Twitter @gosibyl
You can find the company Macro-Eyes at https://macro-eyes.com or on Twitter @macroeyeshealth
To connect with Benjamin, you can email him at Benjamin@macro-eyes.com

Leave a Reply

Your email address will not be published. Required fields are marked *