U.S. insurers can unlock $7 billion in total value — 10-15% of operating expenses — in 18 months by using artificial intelligence to automate certain core administrative functions, according to a new study from Accenture.
The savings could stem from streamlining core functions for payers across the board, including customer service, billing, enrollment, claims and quality and compliance. Automating these functions for an individual health plan would bring in $1.5 million in operating income for every 100 full-time employee by the end of next calendar year for large and small payers alike, the report concludes.
In 2017, 72% of payer executives said that, within the year, AI would be among the top three strategic priorities for their organizations, according to a separate Accenture survey, adding to industry buzz around the technology’s potential. The top three current areas to target for near-term value were anticipating and resolving customer questions, improving the benefits loading and design process and accelerating prior authorization and clinical review of claims.
Although AI in healthcare is still going through growing pains, it holds big potential — especially for insurers under pressure to stay competitive in a continually shifting industry.
In addition to mounting consumer cost concerns and rising deductibles, payer loyalty is low. The Accenture study touted the benefits of using strategic AI initiatives to drive capital in the midst of this market disruption, changes that require payers to forgo traditional sources of unlocking value.
The consultants pointed to six areas of an insurer's operating model. The largest savings reside in managing customer interactions and applying AI technologies to anticipate, and respond to, customer demands.
If used adroitly, AI could unlock $2.1 billion for health insurers in this one domain alone, followed by managing membership and billing ($1.4 billion), managing and supporting reimbursement ($1.1 billion), managing network and providers ($1.0 billion), performing health management ($900 million) and managing quality improvement and compliance ($500 million).
"The premise of the piece is that there's both significant value in focusing health plans' operations first, and that's where the quantitative value, the monetized value, comes,” Richard Birhanzel, managing director of Accenture's payer business, told Healthcare Dive.
"But it also creates foundational value," he noted, stressing that to set up and implement AI systems, an organization must address underlying data architecture constraints and determine its data network needs.
Birhanzel spelled out AI's usefulness in each of the outlined six categories. In managing customer interactions, for example — the biggest slice of the pie — AI is equipped to predict what a customer will ask based on prior calls or other system data, and then help a call center representative to have a more efficient and targeted interaction to satisfy that customer's needs.
It's about a virtual agent, Birhanzel said, or "taking some of the simple things that people ask about and answering them without human intervention, being able to anticipate need" and "infusing the conversation with just-in-time information based on what the customer is asking of the call center."
Accenture's analysis highlights two other areas insurers should target for near-term value: improving the benefits loading and design process and accelerating prior authorization and clinical review of claims.
The study also provides targeted AI and machine learning proposals for each area that could demonstrate near-term operating income impact, such as implementing robotic process automation (RPA) to auto-approve prior authorization requests — something Express Scripts is already doing, aided by its $3.6 billion acquisition of eviCore in 2017.
The report defines AI to mean "collection of multiple technologies enabling machines to sense, comprehend, act and learn, so they can perform administrative and clinical healthcare functions." Birhanzel stressed that AI augments, as opposed to replaces, human activity and will not directly threaten employees' jobs.
According to a survey of over 12,000 participants conducted by consultancy PwC in 2016, lack of trust and a need for the human element were the biggest hurdles to using AI in healthcare.
Birhanzel argued that AI can allow employees to focus on decision-making and processes that require human intervention, and create institutional capacity as payers will no longer have to "scale future jobs in areas that can be automated" or tackled by AI.
Other obstacles include weak existing infrastructure, ill-defined governance structure and bad or ill-fitting vendors.
One real world example is the recent expansion of IBM and Anthem's digital infrastructure contract. IBM has implemented more than 30 bots and automates more than 70% of Anthem's monthly high-volume repetitive tasks.
Although this study focused on payers, hospital systems can also reap the benefits of AI automation — benefits that, according to the report, can win over disgruntled customers who have been asked to bear increasing out-of-pocket costs and who now demand personally-tailored service.