An Enhanced View, Not an Automatic Judgment
An NFL referee has a difficult job. A critical call may turn on a split-second view of a play involving several moving players. Instant replay gives the official another look, sometimes from angles unavailable on the field. The technology can reveal details that matter, but it does not make the call. The referee still has to interpret what happened under the rules.
That distinction offers a useful way to think about artificial intelligence in liability claims, particularly when evaluating a plaintiff’s future medical needs. AI can help medical professionals see more of the record, test assumptions, and identify inconsistencies. The final judgment must remain with a qualified expert who understands the patient and can explain the conclusions.
Future medical projections carry consequences well beyond a spreadsheet. They influence reserves, settlement negotiations, litigation strategy, and, most importantly, whether a claimant’s legitimate care needs are recognized. A forecast built on incomplete records, unsupported treatment assumptions, or unreliable pricing can distort all four.
Clarifying the Play on the Field
Consider a litigated claim involving orthopedic injuries, chronic pain, and intermittent neurological complaints. Over several years, the plaintiff has seen multiple specialists, undergone imaging and therapy, received prescriptions, and perhaps been offered competing treatment recommendations. The resulting file may contain thousands of pages. Even an experienced reviewer can struggle to reconcile the chronology and distinguish current recommendations from earlier possibilities.
AI-assisted review can organize those records into a usable timeline, identify repeated or conflicting diagnoses, associate documented conditions with relevant codes, and flag gaps requiring closer examination. It can also compare the medical history with the items proposed in a life care plan. In that setting, the life care plan becomes another piece of evidence to test rather than an unquestioned forecast.
This work still requires verification. A missing reference may reflect a missing document, not an absent clinical need. A coded diagnosis may be historical or provisional. A treatment mentioned once may never have been recommended as ongoing care. AI can bring these questions to the surface faster, but the reviewer must resolve them against the actual record.
That is the value of replay. It allows the professional to spend less time reconstructing the sequence and more time examining the moments that could change the decision.
The White Lines Matter
Replay would be far less useful if football had no defined line to gain or standard for determining when a player is down. Medical forecasting faces a similar requirement. An organized record alone does not establish whether a proposed service is medically reasonable, how often it may be needed or what it should cost.
Credible analysis draws on appropriate clinical guidelines, published evidence, documented treatment history, and reliable local pricing. Resources such as Official Disability Guidelines can help inform the clinical review. Pharmacy pricing and available hospital price-transparency information may provide useful cost references, although quoted prices must be evaluated for relevance, accessibility, and what the patient or payer could actually obtain.
The distinction matters in negotiations. A projected procedure cost may look precise while resting on an outdated charge or a rate that is unavailable to the plaintiff. Likewise, a lifetime treatment schedule can appear authoritative even when the supporting medical record is thin. AI can expose these assumptions, provided the underlying data are trustworthy and the analysis shows its work.
Objective standards do not eliminate professional disagreement. They make the disagreement more specific. Instead of debating two experts’ credentials in the abstract, the parties can examine whether a particular procedure is supported, whether its frequency is reasonable, and whether its estimated cost reflects the relevant market.
The Referee Still Makes the Call
Experienced medical professionals bring something no automated chronology can supply on its own: clinical context. They assess competing explanations, recognize when guidelines do not fit an individual patient, and know when the right response is to seek additional records or an updated examination.
Their role also provides accountability. Every consequential conclusion should be traceable to the source record, the assumptions used, and the applicable clinical or pricing reference. The expert should be able to revise or reject an AI-generated finding and explain the final opinion without relying on a black box.
This approach is consistent with the broader direction of guidance from NIST, the National Association of Insurance Commissioners, and the American Bar Association: define the use, manage risks, protect sensitive information, and maintain meaningful human oversight. Those principles matter especially when protected medical information and high-stakes decisions are involved.
For claims organizations and counsel, the operating question is practical: can another qualified reviewer reproduce the reasoning? If not, a faster forecast may simply create a more efficient way to defend a weak assumption.
Better Evidence Changes the Conversation
Future medical expenses are often a substantial point of disagreement in serious injury litigation. When each side presents a different total, negotiations can become a contest between bottom-line figures. A transparent, item-by-item assessment creates a better starting point and an ongoing, objective reference.
Unsupported services can be challenged with specificity. Legitimate needs that were overlooked can be acknowledged earlier. Pricing disputes can be separated from disputes over medical necessity. That does not guarantee agreement, but it narrows the questions that genuinely require expert judgment and makes the negotiation more credible.
The real test of AI in this work is therefore not how quickly it produces a number. It is whether the process helps the medical professional produce a more complete, explainable, and defensible forecast. Like instant replay, the technology earns its place by improving what the decision maker can see. The whistle stays with the person responsible for the call.
About the Authors:
Ken Paradis is chief executive officer at Chronovo. KParadis@chronovo.com
Steve Ellis is vice president, liability practices at Sedgwick. steve.ellis@sedgwick.com