In the wake of the Covid pandemic, the U.S. healthcare industry is straining under the weight of rising costs and complexity, worker shortages, and high demand for treatment of chronic conditions, undermining its ability to deliver better and more accessible care.

Generative AI (GenAI) represents a major opportunity to help mitigate these challenges and even transform parts of the healthcare system in coming years. By harnessing its core ability to process massive datasets and generate accurate responses to natural language queries, healthcare leaders see the potential to address the growing needs of populations through better productivity and workflows.

The following are four key areas where GenAI is well suited to benefit healthcare.

Reducing the admin burden

Both providers and patients suffer from administrative overload that significantly weighs on productivity and ease of access.

Doctors spend more than 15 hours a week on paperwork and administration, including writing up patient notes and filling out insurance forms. The heavy admin pressure on doctors and nurses distracts them from their core roles caring for patients, while tying clinical workers up on low-level tasks rather than more creative and managerial work.

Take the process for obtaining treatment preauthorization from insurers. It can take weeks, involving multiple back and forth communications that strain healthcare provider resources and delay vital care for patients. In 2022 alone, the U.S. healthcare industry spent about $1.3 billion on administrative tasks related to prior authorizations, up 30% from the previous year.

Now imagine how that might look with a GenAI solution. The technology could dramatically streamline and speed up the process by automating data extraction from patient records, verifying eligibility in real-time, and suggesting treatments aligned with insurer policies to minimize delays.

A Mckinsey survey last year of 100 U.S. healthcare leaders found three key ways in which GenAI could ease the burden on providers and patients.

1. Clinical / Clinician productivity. This includes automating and summarizing documents such as doctor-patient notes, as well as supporting clinical decision-making by retrieving medical histories and data.

2. Patient engagement and experience. The use of AI-powered assistants to provide more personalized patient service and automate communication before and after consultations.

3. Administrative operations. Increasing the speed and accuracy of back-office operations, such as insurance processes and revenue management.

With administrative spending amounting to 15-30% of U.S. healthcare spending, up to half of which is wasteful, according to one report, it’s an area that is ripe for impact through GenAI.

Applying preventative care

GenAI’s strengths align squarely with healthcare’s focus on preventative care. Detecting diseases and conditions in advance and ensuring patients stick with treatment plans leads to better long-term health outcomes and reduces the need for more costly, complex interventions later on.

Diabetes patients who miss their annual eye and foot examinations, for example, can end up needing far costlier treatments down the line. The cost of managing diabetic foot disease in the United States is estimated between $9 to $13 billion. Overall, the costs associated with medical non-adherence in the U.S. are a staggering $100-300 billion a year, due to avoidable hospitalizations and other consequences.

GenAI-enabled assistants' ability to check in with patients and nudge them on treatment plans and follow-up appointments at every stage of their care journey has the potential to unlock cost savings and improve health outcomes at scale.

A GenAI solution could integrate securely with electronic health records to check for risk factors across a population and then send custom notifications to patients. Patients with risk factors for diabetic foot disease, for example, could receive reminders through an AI assistant about the importance of daily foot checks, guidance on how to do them, and receive responses to any questions. The same would apply to other chronic diseases, such as heart disease and cancer.

Improving and scaling multilingual support

More broadly, the integration of large language models (LLMs) into patient care and support has the potential to narrow gaps in care and further the overall social goal of health equity – the principle that everyone has a just opportunity to achieve their highest level of health.

Large language models’ ability to provide a responsive, multilingual, personalized, 24/7 source of knowledge can help narrow many of the gaps in care caused by factors such as geography, education, and language.

Language barriers and lack of access to quality care are cited as major reasons why Hispanic Americans receive less preventative care and experience worse health outcomes than other ethnic groups. Patients with limited English language skills are more likely to fall through the healthcare net due to misunderstandings about diagnoses, appointments, drug instructions, or through being deterred from accessing care. While they are no replacement for doctors and clinics, AI assistants can help to bring a high level of support and personalized medical information to rural areas, where health outcomes are worse.

Specialist medical AI assistants also promise to close gaps in care created by cost barriers and professional shortages. The Woebot Health app, for example, trained on a specialized, closed database, has been used by 1.5 million people as a kind of pocket therapist for mental health issues.

Improving the claims reimbursement process

The complex and error-prone healthcare claims process is a prime target for GenAI support. The U.S. healthcare system loses about $125 billion annually due to billing mistakes, with medical coding errors playing a major role. In the U.S., there are over 200,000 unique medical codes for diseases, procedures, and services.

The complexity of choosing the correct code, inadequate training, and updates to the codes can delay treatment, disrupt reimbursement cycles, and cause incorrect claim denials. Those erroneous denials lead to delayed care, worse patient health outcomes, and cost healthcare providers billions of dollars a year in missed revenue and higher admin costs.

It’s here that AI solutions can score a win-win for patients and providers by analyzing large datasets to identify the correct codes, checking insurance details in real time, and decreasing the risk of denials. AI models can also be trained to predict trends in claims denials, helping healthcare providers get ahead of challenges before they affect revenue.

A crucial ally for healthcare

While GenAI is no panacea for the healthcare industry’s challenges, its ability to handle complex tasks, make sense of diverse data, and provide real-time, personalized insights holds immense promise for organizations struggling to meet rising demands with limited resources.

As healthcare organizations test and deploy AI solutions, the complex and sensitive nature of the industry will make it particularly important to consider safe and secure applications and to find partners that combine healthcare expertise with comprehensive support for their AI journey.

Building enterprise AI agents

A comprehensive blueprint for developing scalable AI agents in regulated industries.

Download the eBook

Jill Barrientos