Inpatient Reimbursement without Traditional Inpatient Criteria

Abstract:

For decades, health systems have relied on tools like InterQual and MCG to determine whether a patient should be classified as an inpatient or an outpatient receiving observation services. These criteria sets have been widely accepted as the industry standard for guiding admission decisions and meeting payer criteria for reimbursement. However, in our own utilization review (UR) operations, we increasingly found that these "gold standard" tools, and the processes built around them, often failed to reflect the clinical realities of patients today, leading to inaccurate classifications of patient status and missed opportunities for reimbursement.

Long Tail conducted a retrospective validation study using our CareAssist tool, supported by physician advisors, standard industry methodologies, and actual payer reimbursement (outcomes). This study evaluates the relationship between the application of commercial inpatient criteria sets (e.g., InterQual and MCG) and payment status for inpatient versus observation hospital admissions, revealing a critical operational insight for health systems.

Results:

This analysis includes 8,766 hospital accounts (HARs) discharged over a six-month period from October 1, 2024, to March 31, 2025, across multiple acute care hospitals. Each account underwent concurrent utilization review using a combination of methods, including commercial criteria and holistic evaluations by physician advisors. Notably, more than 27% of accounts reimbursed as inpatient did not meet commercial criteria. Compared to a theoretical scenario in which only commercial criteria were used to determine medical necessity, this resulted in an estimated $41 million in additional insurance payments from a six-month data set. Annually, this particular situation suggests an $82 million reimbursement differential based on expanding beyond the use of commercial criteria.

Methodology:

The following patient scenarios were excluded as outliers that do not follow standard inpatient criteria:

• Obstetrics – due to the 48-hour/96-hour inpatient standard applied to normal and cesarean deliveries.

• Psychiatrics – due to standard inpatient criteria not being applicable.

• Elective Surgical – due to admissions being authorized as inpatient prior to surgery without application of standard inpatient criteria.

Additionally, accounts included in the study had to meet the following requirements:

• Account is fully paid by insurance: this includes accounts having received net insurance payments where either A) ≥80% of expected net reimbursement is received, or B) payments exceed >30% gross collection rate.

• Account requires inpatient authorization: this includes accounts with primary insurance of Commercial, Medicaid Advantage, or Medicare Advantage insurance product types.

  
  

Accounts were categorized by criteria review outcome based on the determination from InterQual and/or MCG. A review was classified as “Inpatient Met” if the patient had any documented criteria reviews where an inpatient criteria set was determined to be “met”. Accounts were assigned “No Inpatient Met” if the best available documented criteria review outcome was either an observation criteria set that was “met”, or the account failed to meet any criteria set.

Key findings include:

• Of the 6,241 inpatient-paid accounts, only 4,515 (72.34%) had supporting inpatient criteria documented by InterQual/MCG.

• Conversely, 1,726 accounts (27.66%) were paid as inpatient despite not meeting InterQual/MCG inpatient criteria.

• This pattern was consistent across payer types:

  • Commercial: 69.29% met inpatient criteria; 30.71% did not

  • Medicaid Replacement: 78.46% vs. 21.54%

  • Medicare Replacement: 72.15% vs. 27.85%

• In total, this sample resulted in over $48M of inpatient insurance payments over the 6-month period that were not supported by these traditional inpatient criteria tools.

  • The average observation payment during this time was $4,083 per account. If this was applied to the 1,726 accounts paid as inpatient, total insurance payments would have been just $7,047,258.

  • The net difference in reimbursement by securing inpatient payments on accounts not meeting inpatient criteria is therefore over $41M in net improvement.

  • Note: this net amount of $41M accounts for over 27.6% of the total hospital insurance payments received on this population ($149M).

Conclusion:

As health systems confront mounting financial pressures and increasingly complex payer dynamics, many are seeking AI solutions that can redefine long-standing operational assumptions and level the playing field in reimbursement strategy. This study reveals a striking disconnect between traditional status criteria tools like InterQual and MCG and the actual reimbursement outcomes achieved. A significant portion (over 27%) of inpatient payments were made without alignment to these criteria, amounting to more than $41M in recovered revenue from 6 months of admissions.

These findings highlight a critical opportunity: clinical justification for hospitalized care should not be reduced to a checklist of criteria but instead reflect the full clinical picture of the patient. AI-driven tools like Long Tail’s CareAssist have the potential to reshape this process by synthesizing real-time data to support decisions that mirror the complexity of actual patient care. By moving beyond rigid, rules-based paradigms, health systems can better articulate the medical necessity of inpatient care, recapture revenue that might otherwise be lost, and proactively align with payer strategies that increasingly favor nuanced, patient-centered documentation.

Supporting Data:

Criteria Outcome	Inpatient Met	No Inpatient Met
Commercial
Accounts Paid IP	1586	703
Total Accounts	1666	1451
% of IP Paid	69.29%	30.71%
Total Insurance Payments	($48,787,470.53)	($26,532,777.03)
Medicaid Advantage
Accounts Paid IP	965	265
Total Accounts	1029	769
% of IP Paid	78.46%	21.54%
Total Insurance Payments	($17,360,910.98)	($6,047,516.02)
Medicare Advantage
Accounts Paid IP	1964	758
Total Accounts	2110	1741
% of IP Paid	72.15%	27.85%
Total Insurance Payments	($34,779,710.75)	($15,532,857.85)
Total Accounts Paid IP	4515	1726
Total Accounts	4805	3961
Total % of IP Paid	72.34%	27.66%
Total Insurance Payments	($100,928,092.26)	($48,113,150.90)

Questions for the Future:

• What if traditional criteria tools are no longer sufficient for the complexities of today’s clinical environment?

• Do newer algorithmic tools that “segment and score” patient populations, without real clinical justification, offer any efficiency or workflow improvements?

• What payer strategy changes might health systems anticipate if AI adoption becomes the new standard?

Call to Action:

The implications of these findings are groundbreaking, but further validation is essential. More data is needed from different hospital systems, payer mixes, and patient populations to assess the broader applicability of this approach. This is just the beginning of a critical shift, and Long Tail is offering free assessments to help our providers validate the opportunity in their own health systems.