Xsolis

The Care Level Score is the product. Continuously scoring medical necessity and anticipated level of care for every patient in real time is a prediction task, and the platform's entire argument, that it removes subjectivity from status determination, only holds if a model rather than a reviewer is producing the score.

The oversight position is stated plainly and the design supports it: the company describes a human in the loop approach explicitly, and the platform produces a score and supporting evidence rather than a determination, leaving the medical necessity decision with clinical staff and physician advisors.

That is the correct ceiling for a system whose output influences whether a patient is admitted or observed, and it contrasts with the several vendors in this index that claim autonomous determination without publishing a review step.

Unusually well evidenced for a proprietary score. Peer reviewed research across three United States health systems has evaluated level of care accuracy, observation discharge rates and operational forecasting in live utilization management workflows, and a third party study measured determination speed against fax and electronic health record routes. External evaluation of a model's performance in production is rare in this index and it is what lifts this above the category norm.

Held at B rather than A because the model itself remains closed. The Care Level Score is described by what it does rather than how it works: no architecture, no input feature set, no confidence interval, and no statement of how the score maps to a status recommendation.

The missing piece that matters most is subgroup performance. Published accuracy across three health systems is an aggregate, and nothing reports whether accuracy holds across payer mix, admission source, diagnosis category or patient demographics. For a score that influences admission status, an aggregate figure can conceal exactly the variation a buyer needs to see.

The architecture changes a payer's access to clinical documentation from a request into a standing window, and that is the substantive finding rather than an enumeration gap. The stated design intent is that a health system and a health plan work from the same evidence, with shared clinical access through the platform so both parties read the same record view.

There is a real case for it: a payer reviewing the actual documentation rather than a summary should produce fewer disputes and fewer rounds of back and forth. It also means a plan sees live clinical documentation for admitted patients through a vendor's window rather than by asking for it, which is a different relationship to the record than the one a hospital's release process was built around.

Establish what the payer side can see, whether that view is scoped to the minimum necessary for the determination at hand, whether access is logged and reviewable by the provider, and how a patient would learn their record is visible this way. Externally certified controls sit behind the platform, which is a stronger foundation than most records here carry.

Nothing else is enumerated: no model or model family, no hosting arrangement, no sub processor list, no retention period and no model training statement. Ask for the payer view scope and its logging, retention, and the training position.

The only vendor in the utilization management cluster with peer reviewed research. Published studies across three US health systems evaluated level of care accuracy, observation discharge rates, and operational forecasting in live utilization management workflows.

That is corroborated by a named health system result, AnMed reporting 14.6x return on investment over twelve months and 1,221 hours of front end review time saved, and by third party analysis finding determinations rendered 38 percent faster than fax and 15 percent faster than through the EMR. Graded on the existence and venue of peer reviewed evidence; this index does not re verify the underlying studies.

Externally validated controls through HITRUST r2 certification, which is a stronger foundation than most records in this index carry, combined with a stated human in the loop design where the platform produces a score and supporting evidence rather than a determination.

Held at B rather than A because of a question the architecture raises and nothing published answers. The stated design intent is that a health system and a health plan work from the same evidence, with shared clinical access through the platform so both parties read the same record view. That is the product's central idea and there is a real case for it, since a payer reviewing the actual documentation rather than a summary should produce fewer disputes.

It also means a plan sees live clinical documentation for admitted patients through a vendor's window rather than through a request. Establish what the payer side can see, whether that view is scoped to the minimum necessary for the determination at hand, whether access is logged and reviewable by the provider, and how a patient would learn their record is visible this way. No retention period or model training statement was located either.

No business associate agreement terms and no standalone HIPAA posture document were located. What supports the grade is that the HITRUST r2 certification held by the platform incorporates the privacy and security rule requirements among the authoritative sources it harmonises, so the controls have been validated against them by an external assessor rather than merely asserted.

That is the standard middle rung: demonstrable controls, undisclosed contract.

The contracting question here is more interesting than usual because of where the company sits. Precision UM is a shared arrangement between a health system and a health plan working from common clinical data, so the vendor holds business associate relationships with two parties who are counterparties to each other in the determination it informs. Establish what each agreement permits the other side to see, and whether the terms are symmetrical.

HITRUST risk based two year certified status, announced by the company for the platform then known as CORTEX and now sold as Dragonfly. The r2 tier is the strongest healthcare security credential in common use: a tailored risk based assessment rather than a fixed control set, validated by an authorised external assessor with the certification decision made by HITRUST itself, and maintained through interim testing.

That places this record alongside Cohere Health at the top of this axis, and above the vendors here holding i1 or an unspecified HITRUST claim.

One currency check belongs in any diligence and is stated rather than glossed. The certification was announced in March 2022 and the r2 credential runs on a two year cycle with interim testing, so more than one recertification window has passed since. Ask for the current certification letter and its validity dates. Separately, third party software directories attribute a long list of ISO standards to this product, several of which are not certifiable standards at all; none of that is credited and a buyer should ignore it.

No FDA pathway applies. Utilization management software informing medical necessity and status determination sits outside Software as a Medical Device, because the clinical decision remains with the reviewing clinician and physician advisor.

Graded C because the regime that does govern this territory is accreditation, and none was located. URAC Health Utilization Management accreditation and NCQA Utilization Management accreditation are the clearance equivalents here, and URAC version 8.0 added standards for algorithmic transparency, bias testing and clinical decision support tool oversight, which reach this product directly.

Eligibility rather than exemption is the point. The company sells Physician Advisor Services alongside the platform, supplementing coverage or operating fully outsourced, which puts its own licensed clinicians into the review and places it on the accreditable side of the line rather than outside it. Same position as Alaffia Health, and the same absence. Cohere Health holds both accreditations and grades A on this axis, which shows the ceiling is reachable in this lane.

No AI governance framework, bias evaluation or subgroup performance disclosure was located.

This is among the most consequential governance gaps in the index, and the reason is the specific decision the score informs. A medical necessity score shapes whether a patient is admitted as an inpatient or placed in observation. That distinction determines what the hospital is paid, and for Medicare patients it also determines the patient's own cost sharing and whether subsequent skilled nursing care is covered at all. A systematic tilt in the score would therefore move money and coverage in the same direction repeatedly, for a population that never sees the score.

The company's own framing sharpens the obligation rather than reducing it. The product is sold as removing subjectivity from status determination, which is a claim about fairness. A vendor making that claim carries a stronger duty to publish evidence for it than one selling efficiency, and nothing published tests whether the score performs evenly across groups.

Peer reviewed research across three health systems has evaluated level of care accuracy, observation discharge rates and operational forecasting in live utilisation management workflows, and a third party study measured determination speed against the alternative routes.

External evaluation of a model's performance in production, rather than on a retrospective dataset, is rare in this index and it is what lifts this above the category norm: a model that works on held out data and a model that works while clinicians and reviewers are acting on it are different claims, and only the second is what a buyer is purchasing.

The design also constrains what the system does with its conclusions, since the platform produces a score and supporting evidence rather than a determination, so a human makes the status decision with the underlying material in front of them. Held below the top grade on the disclosure that matters most for this output.

Published accuracy across three health systems is an aggregate, and nothing reports whether it holds across payer mix, admission source, diagnosis category or patient demographics. For a score influencing admission status, where the consequence falls on both the patient's cost exposure and the hospital's revenue, an aggregate can conceal exactly the variation a buyer needs to see. The score itself is closed, with no architecture, feature set or mapping to a status recommendation published. Ask for subgroup performance and the score to recommendation mapping.

Bidirectional integration with the EMR is stated, with data flowing automatically in both directions to support real time scoring against live clinical documentation. Shared clinical access through Dragonfly between a health system and a national health plan is a genuinely unusual integration pattern, since it means both parties read the same record view. Held back from A because named EHR systems and certification detail were not retrieved.

No hosting provider, region, tenancy model or data residency commitment published by the company was located, and no subprocessor list was found. A third party directory attributes United States and European Union regions to the product, but aggregator listings are not treated as vendor disclosure in this index, and the same listing carries obviously unreliable certification claims.

The HITRUST r2 certification is scoped to the platform and implies the infrastructure was assessed, but a certification scope is not a residency disclosure and does not say where the environment runs.

The tenancy question deserves a direct answer here because of the shared access model. Where a health system and a health plan both work inside the same platform against common clinical data, a buyer should establish how those two customers are separated, what sits in a shared layer, and whether the historical determination patterns the company describes using are drawn from a single relationship or pooled across many.

No public pricing. Contact the vendor. Note the hybrid: alongside the platform the company sells Physician Advisor Services covering night and weekend gaps or operating fully outsourced, so the buy can be software, staffed service, or both. Establish which is being priced, since a clinical staffing component behaves very differently from a licence.

Precisely bounded to hospital utilization management and the workflows adjacent to it: admission and status determination, continued stay and concurrent authorization, discharge readiness and length of stay, physician advisor escalation, and denials and appeals. The position between payer and provider is stated as the design intent rather than a marketing frame, and Precision UM operationalizes it as a shared program.