AI in Health Insurance: Algorithms Deciding Patient Care and Coverage

The Rise of AI in Health Insurance Decision-Making

Health insurance companies have increasingly adopted artificial intelligence (AI) algorithms over the past decade to make crucial decisions about patient care and coverage 1

1

2

2

3

3

. Unlike AI used by healthcare providers for diagnosis and treatment, insurers employ these algorithms to determine whether to pay for treatments recommended by physicians and how much care a patient is entitled to receive.

How AI Algorithms Influence Healthcare Coverage

One of the most common applications of AI in health insurance is in the prior authorization process. In this scenario, an algorithm decides whether requested care is "medically necessary" and should be covered 1

1

2

2

3

3

. These AI systems also help insurers determine the extent of care a patient can receive, such as the number of hospital days allowed after surgery.

When insurers decline coverage for a recommended treatment, patients typically face three options:

1. Appeal the decision (only 1 in 500 claim denials are appealed)
2. Accept an alternative treatment covered by the insurer
3. Pay for the recommended treatment out-of-pocket

Concerns and Criticisms

While insurers argue that AI helps them make quick, safe decisions and avoid wasteful treatments, there is growing concern about the potential negative impacts of these algorithms 1

1

2

2

3

3

:

1. Delayed or Denied Care: Evidence suggests that these systems may be used to delay or deny care that should be covered, prioritizing cost savings over patient health.

2. Lack of Transparency: Insurers have refused to disclose how these algorithms operate, citing trade secrets, which makes it impossible to evaluate their fairness and effectiveness 1

   1

   2

   2

   3

   3

   .

3. Disproportionate Impact: Research indicates that patients with chronic illnesses, as well as Black, Hispanic, and LGBTQ+ individuals, are more likely to experience claim denials 1

   1

   2

   2

   3

   3

   .

4. Financial Incentives: The use of AI for coverage decisions creates a disturbing possibility that insurers might withhold care for expensive, long-term, or terminal health problems to save money 1

   1

   2

   2

   3

   3

   .

Regulatory Landscape and Challenges

Unlike medical algorithms, insurance AI tools are largely unregulated 1

1

2

2

3

3

:

• They don't undergo Food and Drug Administration (FDA) review
• No public information is available about their decision-making processes
• No outside testing is conducted to assess their safety, fairness, or effectiveness

Recent developments in regulation include:

1. The Centers for Medicare & Medicaid Services (CMS) announced that insurers in Medicare Advantage plans must base decisions on individual patient needs 1

   1

   2

   2

   3

   3

   .

2. Some states, including California, have passed laws requiring physician supervision for insurance coverage algorithms 1

   1

   2

   2

   3

   3

   .

However, these regulations have limitations:

• They often leave too much control in insurers' hands
• They don't require neutral expert review of algorithms
• Federal rules only apply to federal health programs, not private insurers

Calls for Stronger Regulation

Many health law experts argue that the gap between insurers' actions and patient needs necessitates stronger regulation of health care coverage algorithms 1

1

2

2

3

3

. Suggestions include:

1. FDA oversight to provide a uniform, national regulatory scheme
2. Requiring algorithms to be reviewed by neutral experts before use
3. Stricter definitions of "medical necessity" and contexts for algorithm use

As the debate continues, the impact of AI on health insurance decisions remains a critical issue at the intersection of technology, healthcare, and policy.