The Importance of Clinical Oversight in Automated Prior Authorization Systems to Ensure Quality Patient Care

Prior authorizations are an important step in healthcare. They ask doctors and providers to get approval from insurance before giving certain treatments or services. This system aims to stop unnecessary or expensive care, but often it causes delays and extra paperwork. A Senate report reviewed over 280,000 documents from big Medicare Advantage insurers and found some worrying trends in denial rates.

For example, UnitedHealthcare’s denial rate for post-acute care prior authorizations rose from 10.9% in 2020 to 22.7% in 2022. Humana saw a 54% increase in denials related to long-term acute care hospital services during the same period. CVS (Aetna) had a 57.5% rise in prior authorization requests for post-acute services, although its denial rate stayed the same. These trends happened as insurers started using more automated decisions based on AI.

The growing number of denials and longer wait times affect patients. Delays can make health problems worse, frustrate patients, and reduce trust in healthcare providers. Clinic administrators and IT staff need to find ways to balance automation and clinical review to keep patient care on track.

The Role of AI in Prior Authorization

Artificial intelligence can help make prior authorization faster by cutting down manual work. AI platforms like Onpoint Healthcare Partners’ Iris use natural language processing to read clinical documents, improve communication between providers and insurers, and automate repeated steps in referrals and approvals.

The Iris platform connects directly to payer systems, cutting down delays caused by paperwork. It supports referral management, so once an approval happens, the referral keeps moving without interruption. It also lets healthcare workers check AI decisions and change them if needed.

Even with these benefits, AI cannot work alone without supervision. The Centers for Medicare and Medicaid Services (CMS) rules for 2024 say Medicare Advantage plans must not rely only on AI to decide medical necessity. Humans must be involved to make sure decisions are fair and fit each patient’s situation. This helps protect patients from wrong denials just based on an algorithm.

The Rising Need for Clinical Oversight

Clinical oversight means having qualified healthcare professionals review and confirm decisions about patient care, especially those made with help from AI systems. Using AI more in prior authorization shows both good parts and risks.

Automated systems can handle many requests faster than people. But without clinical review, they might approve or deny care wrongly, which can hurt patients. CMS stresses that providers must stay involved to stop decisions based only on AI, which may miss important details.

States are making laws to require oversight:

• California’s Assembly Bill 3030 requires providers to tell patients when AI is used and get their consent before using it.
• California’s Senate Bill 1120 says a qualified human must review utilization decisions so AI systems can’t decide coverage alone.
• Colorado’s Consumer Protections in Interactions with Artificial Intelligence Systems Act of 2023 requires AI developers to assess impacts, notify patients of AI use in decisions, and guarantee a right to appeal AI-influenced decisions.
• Illinois’ H2472 bill demands utilization management programs that use algorithms to base decisions on evidence and peer review.
• New York’s pending Assembly Bill A9149 asks insurers to publicly disclose AI use, have state certified algorithms, and let clinical experts review AI decisions.

These laws promote clear use of AI and human involvement to keep patients safe. Clinic administrators and IT managers must follow these rules to comply and keep care quality high.

Impact of PA Delays on Patient Care

Delays caused by prior authorizations affect patients in several ways:

• Treatment Postponement: Patients wait longer for important services like home health care or therapy when approvals take too much time.
• Increased Administrative Workload: Staff spend lots of time handling denials, refiling requests, and appealing decisions, which takes time away from patient care.
• Patient Dissatisfaction: Confusion and waiting for insurance approval can upset patients and lower their trust in healthcare.
• Care Fragmentation: Breaks in referrals and authorization processes hurt care continuity and may lead to worse health results.

For example, UnitedHealthcare’s denials for post-acute care doubled from 10.9% to 22.7% in two years. Humana’s denials for long-term acute care increased by 54%. These problems affect staff workflows and patient health, so practice managers must find a way to follow rules while reducing impact.

AI and Workflow Coordination in Prior Authorization

AI-driven workflows can cut down administrative hold-ups but need thoughtful use. Using tools like Onpoint’s Iris can help teams by:

• Automating routine tasks such as processing standard authorization forms and matching requests to insurer rules for faster initial reviews.
• Improving communication between providers, insurers, and referral networks to lower delays caused by back-and-forth messages.
• Supporting clinical judgment by summarizing notes and patient history to help providers make better decisions.
• Tracking approval times and denial rates to find problems and improve processes.

At the same time, clinical staff must stay involved to check AI suggestions, especially for complex cases. The system should flag tough requests for review instead of deciding automatically. This “human-in-the-loop” model helps catch mistakes or gaps that AI might miss.

IT managers play a key role in choosing AI tools that meet CMS rules and state laws. They must also make sure these tools work well with existing electronic health record systems and payer platforms. Training staff to use AI and handle exceptions is also important.

Regulatory Environment Surrounding AI in Prior Authorization

The rules around using AI in healthcare prior authorizations have become stricter in recent years. Important actions include:

• Federal Oversight: President Biden’s Executive Order on safe AI use in healthcare tasks the Department of Health and Human Services (HHS) with creating policies for responsible use of AI technologies.
• CMS Rules: CMS requires Medicare Advantage plans to base medical necessity decisions on individual patients, not just AI results. They also want payers to use Prior Authorization APIs by 2027 to speed up decisions while keeping provider input.
• State Legislation: States like California, Colorado, Illinois, and New York have laws about transparency, patient consent, impact checks, and human review of AI decisions.

Healthcare groups must keep watching these rules because following them affects both legal standing and patient care quality.

Considerations for Medical Practice Administrators and IT Managers

Practice administrators and IT managers need to think about many things to balance AI’s benefits and risks for prior authorization:

• System Selection: Pick AI systems with clear algorithms, strong clinical oversight, and smooth connections to payer portals.
• Training and Protocols: Train staff on how to use automation tools and create rules for human review of AI decisions.
• Patient Communication: Let patients know when AI is part of their care and get consent when needed.
• Compliance Monitoring: Keep up with federal and state policy changes to adjust AI use and follow rules.
• Performance Tracking: Use data to check approval rates, turnaround times, and denial reasons to improve the system and patient care.
• Collaboration: Work with insurers to improve workflows and speed up authorization for the benefit of providers and patients.

Final Thoughts on AI’s Role in Prior Authorization

AI can help manage prior authorization by reducing paperwork and speeding decisions. But it must have careful clinical review to keep patient care good. Automated systems cannot replace healthcare professionals’ careful judgment, especially since every patient is different and laws have many rules.

Combining AI with human review and following rules can make authorization faster while keeping care timely and appropriate for patients in the United States.