Star ratings are hard—5 ways AI can help

Centers for Medicare & Medicaid Services (CMS) Star Ratings are a key component of the Medicare program, designed to evaluate Medicare health plans on the quality of care and services provided.

CMS assigns hospitals and health systems an overall star rating based on various quality measures, such as safety, readmission rates, patient experience, and mortality rates. The ratings allow patients to compare hospitals and assess healthcare quality based on the organization’s star rating. 

However, the calculation and interpretation of Medicare stars have become increasingly complex due to several interrelated factors, such as changes in scoring methodology, expanded scope of metrics, data complexity, regulatory issues, social determinants of health, and financial implications, making it even harder to achieve and maintain 4- and 5-Star ratings.

A quick look at the 2025 Medicare Advantage and Part D Star Ratings shows a significant drop in star ratings over the past three years. Although the number of contracts has increased since 2022, 4-Star contracts decreased from 322 to 209. Worse, the number of 5-Star contracts has decreased from 74 to 7.

As healthcare organizations look to buck the trend of decreasing star ratings, AI and automation represent a viable solution to help hospitals and health systems deliver the high-quality, affordable care demanded by CMS.

Here are five ways AI can help:

1. Improve workforce productivity

There are more and more eligible patients in Medicare contracts at a time when the workforce is contracting. This makes it extremely challenging for healthcare staff to execute the manual tasks necessary to engage patients to close gaps. 

With AI, hospitals and health systems can improve the productivity of their workforce by automating open care gap identification, identifying closed care gaps in charts and documentation, and engaging patients in preventative care without adding to the administrative burden on staff and caregivers. AI can decipher non-discrete data using LLMs to showcase preventative care that has occurred, sifting through 1,000s of charts and documents in ways that are not plausible for clinical support staff. AI Agents can then engage patients via voice, text, email, or portal messages, enhancing the workforce and providing greater capacity so staff can focus on patient care and more complex administrative tasks.

2. Schedule appointments 24/7 

Many health systems only staff call centers or clinics during traditional business hours when patients are working or otherwise unavailable. With AI Agents, call centers can operate 24/7, so patients can schedule or change appointments and follow up on care gap tasks at any time convenient. Moreover, using natural language processing, AI can engage in natural ways of speaking, using particular accents, tonality, or polite forms of speaking, and generally outperform the automated call centers of the past.

AI can also connect with the patient privately and asynchronously while integrating with systems of record to provide an easy experience to close care gaps.  AI can assist with securing transportation or completing financial assistance forms, which often deter patients from getting routine care. 

3. Capture completed care

Maybe the health system did all the work but didn’t turn in the homework based on payer rules for “attestation.” AI Agents can identify the appropriate clinical documents to demonstrate that payer gaps are complete so that the health system gets the dollars in the door that it deserves. 

And, when homework is completed at an outside system, it is often saved in PDFs or faxes within the EHR. AI can capture completed care by searching and reading documents, extracting information, and submitting to payers. It can also tap into health information exchanges (HIEs) to find relevant data on care gap closures

4. Understand health plan incentives

Health systems contract with many health plans, and it can be challenging to understand which health plan wants which measures completed and how much money they will pay for each measure to be closed. The differences between targets can be minute and truly cumbersome to streamline.

AI can sift through and read each contract, calculate values on each care gap, create prioritized lists, and track the system performance over time, surfacing the right patients at the right time for each system to optimize their payouts while ensuring patients receive the proper care.  

5. Measure what matters

CMS is re-evaluating which measures matter most for Star Ratings. For instance, patient experience, complaints, and access measures used to be quadruple-weighted but will decrease in importance significantly in SY2026 when ratings are assigned.

With AI, health systems can evaluate just how much each open gap is worth financially and how these gaps will be evaluated in terms of the quality of outcomes so clinicians can have clear pursuit lists when meeting with patients.