PFA expands efforts to combat long-term illness
PFA is significantly increasing its use of artificial intelligence in efforts to prevent long-term sick leave.
Since 2020, PFA has contacted nearly 5,000 customers identified by artificial intelligence as being at risk of long-term illness. Going forward, the initiative will be expanded so that 5,000 customers each year will receive a call offering assistance, if needed.
The initiative has already demonstrated remarkable results. Of the first 5,000 calls, 1,500 customers received additional support, and for this group, PFA has managed to reduce the risk of long-term illness by an impressive 70 per cent.
“We had hoped for a positive impact, but the results have exceeded all expectations. This is making a real difference for people before they fall ill,” says Ole Krogh, CEO of PFA.
Artificial intelligence identifies patterns that humans cannot see
PFA’s model analyses a range of data points after a customer has used their health insurance. This enables the system to identify patterns indicating an increased risk of long-term sick leave. PFA’s healthcare professionals then assess whether the customer should be contacted.
The technology serves as a supplement to human judgement, emphasises Ole Krogh:
“Artificial intelligence highlights patterns, but it is always humans who make the decision to call and offer assistance. This is absolutely crucial for us.”
An initiative designed to benefit more people
PFA’s ambition is for its experiences to inspire broader applications. According to PFA’s calculations, a national model based on PFA’s experiences could potentially prevent 2,800 new long-term sick leave cases annually while also delivering significant economic benefits to society.
“It’s about helping people in time. We are only at the beginning of what artificial intelligence can achieve in prevention and healthcare, and we hope our results can benefit even more people,” says Ole Krogh.
PFA’s efforts against long-term sick leave
- Since 2020, 5,000 customers have been contacted with a care call after artificial intelligence identified an increased risk of long-term sick leave.
- 1,500 customers have received additional assistance following the calls.
- The risk of long-term illness has been reduced by 70 per cent for customers who received help.
- PFA is now increasing its efforts and will contact 5,000 customers annually.
- The number of reported long-term sick leave cases among PFA’s customers has increased by 21 per cent since 2020.
Data ethics
- The model is entirely separate from pricing and other commercial decisions.
- PFA only uses artificial intelligence when it is in the customer’s interest.
- Customers decide whether to accept the assistance offered. There are no consequences if customers decline treatment.
- The calls offer voluntary support as a follow-up to previous care programmes the customer has been through with PFA in the past 12 months.
- PFA complies with legislation and has its own data ethics policies, focusing on data minimisation, continuous testing of the algorithm’s functionality, and close collaboration between healthcare professionals and AI developers.
- The model only uses information PFA already has as part of the customer’s care programme, and access to data is limited to a small group of employees who only see the necessary information.