What are some of the key performance indicators for measuring Utilization Management effectiveness?

HEDIS Measures from NCQA are performance indicators that one can compare their plan's performance to national standards. For example, in 2022 the national average for follow up after hospitalization for mental illness within 7 days post discharge was 46.5% in commercial PPO but only 27% in Medicare PPO. Therefore, there is a higher utilization rate in the commercial PPOs versus the Medicare PPOs for this HEDIS Measure.

It is of great importance that you understand the guidelines of the payer that you are requesting authorization for. Guidelines today for managed care will not be the same 6 months from now. It is important that your team has awareness of any changes in payer guidelines. Education in regard to medical necessity is also key.

Sadly with AI or automated online utilization methods, the office (usually a staff person) will quickly learn which boxes to check to get an approval. This results in the need for retrospective audits. Charts have to be collected or audited in the office, information there compared to the information entered and a evaluation of appropriate vs inappropriate requests. I am certain this is just as costly at upfront UM and not as effective.

It is time to flip the old model of mother-may-I, to point of care decision support. This simple change enabled by AI could instantly approve up to 80% of orders.

In my experience, a high denial rate is often impacted by a myriad of factors. It may indicate that a patient was admitted to the appropriate level of care but there is a need for more robust or specific documentation to support medical necessity (i.e. Two Midnight Rule). It may indicate a need for continuing education for UR staff to increase proficiency and appropriateness when applying payer guidelines (MCG, InterQual, LOCUS, ASAM). It could indicate a need for improvement in the organization's authorization process itself. Also, there are those cyclical ebbs and flows where there is an unexplained increase in denials by one specific payer indicating an opportunity for us as providers to partner with the payer.

In the behavioral health space, length of stay can be a challenge because there is the delicate balance of having enough authorized days to actually attain treatment goals but avoid unnecessarily prolonged treatment at a higher level of care. In these instances, low length of stay does not necessarily indicate a smooth recovery but rather may point to an increase in patients leaving AMA or treatment being cut short due to insurance denials.

Appropriate and comprehensive discharge planning, beginning at admission, including patient and caregiver education and assessment of SDOH is key to preventing readmissions.

Due to the complexity of coding, the clinical approval processes within the UM/PA need to constantly be updated. With the lag time between the practice of today and the most rigorous clinical studies taking years to update the overall process. This update is difficult on both the provider and payer side. Pricing in the 'cost savings' is very tricky. If AI optimizes both the clinical accuracy and the clinical workflow then if should benefit both parties.

This calculation can be tricky. What is the cost of the current manual review system? How much time is wasted in manual review? How much wasted with no-shows driven by denials? What is the cost to patients and the system in delayed care? What is the cost in poor customer experience? All this should be factored in. Is it time for a change? Can AI be the driver?

Risk mitigation and claims cost. Enhancement of continual survey readiness (not just lip service, but actual continuous PI and accreditation readiness), these activities while associated with some upfront cost will increase profitability, revenue, growth, patient and staff satisfaction while organically mitigating risk.