Using Artifical Intelligence to improve health insurance
Niels Thone, Co-Founder & CEO of Sprout.ai, discusses applying AI to the health insurance value chain
Healthcare services around the world are facing unprecedented strain due to the pressure brought on by the Covid-19 pandemic. Despite increased digitisation around us, healthcare services still rely greatly on manual processes, including writing medical reports, which is still done by hand.
Process lag
This lack of digitisation in the healthcare industry has a spillover effect in slowing down other sectors, including insurance. For every insurance claim, handlers have to go through and analyse 20 to 30 pages of documents which, in addition to medical reports, can include various receipts and bank statements. Of all the documents, only a third of them will be typed, the rest will be photographs, scanned documents or handwritten pieces into templates or on blank sheets of paper.
Additionally, some of the handwritten documents will be difficult for claims handlers to decipher, which adds to the amount of time they spend processing and reviewing each case. The average insurance claim is resolved in around 30 days, but it can take up to six months to reach a decision in complex cases.
Manually sorting documents is a very time consuming and inefficient process that drastically reduces the amount of time claims handlers have to engage with customers. The less time available to claims handlers to engage with customers, the longer the latter have to wait before hearing back on a claim. The process of filing an insurance claim is stressful and often has to be done at a time when customers are financially and emotionally vulnerable. Time becomes extra precious when facing an urgent physical or mental healthcare need. This is a particularly big issue in countries like the US, where most citizens depend on private insurers to cover medical procedures.
To date, the insurance industry has been more resistant to innovation than the wider financial services sector. This is slowly changing, as insurers look to improve customer experience throughout the value chain.
AI character recognition potential
At Sprout.ai, we have built Optical Character Recognition technology (OCR), a subset of AI, to analyse and read any kind of text, both machine typed and handwritten. Natural Language Processing (NLP) then extracts and analyses the unstructured text and information read by the OCR, which allows for true automation in the claims process.
In addition to sorting and classifying information, AI technology can also identify potential errors and anomalies, which it brings to the attention of the claims handler for them to decide. Where names have been misspelt or handwriting is hardly legible by humans, AI can provide a number of probable matches to the words for the handler to decide which option is most likely. Thanks to AI having access to comprehensive medical databases, it can match the handwriting with the most likely word.
Removing workload from people
By introducing automation to the claims process, the road to survey is much more efficient and claims handlers are relieved from large workloads that involve deciphering documents for hours on end. They therefore have more time to engage with clients who need urgent attention. This is especially important for customers who require urgent medical care and are therefore, at their most vulnerable. Waiting for months for a medical claim to be processed is not an option for many patients, since with many illnesses – like cancer or serious infections – time is of essence if a cure is to be found. This technology can free up a lot of time for claims handlers, which they can dedicate to engaging with clients and working on the less straightforward/more complicated insurance claims.
AI and treatment identification
All the data analysed by the AI technology also provides unique insights into which treatments are most effective for patients with certain characteristics – age, gender, preexisting health conditions – and given the vast amounts of data currently available, the combination is endless. The AI looks at each patient like a black box and does not focus on personal information at any point.
These insights mean medical insurance companies are informed on which treatments are most effective for their clients and can recommend for this route to be chosen by the healthcare providers. This is especially important in countries where medical insurance is the most reliable access to healthcare. Insurers ‘holding the wallet’ gives them great power of decision and they can have a great influence on the treatments their customers receive.
The power of technology in health insurance
Automation can dramatically improve patient outcomes and in many cases, save lives. Technology is impacting and revolutionising every aspect of healthcare, from online GP consultations to complex surgeries. Incorporating automation to the insurance claims process will soon become essential for all companies in the insurance sector. Automation is a win-win for all parties involved in the insurance process: it dramatically speeds up the claims process making insurers more efficient and it can help patients get better treatment faster, regardless of the medical emergency. When it comes to choosing a remedy to a medical emergency, accuracy is key for success.