A MedPath service line
Visible Worker keeps the person behind the claim in view — from the first prescription to the final payment. Seen workers recover faster; unseen workers call lawyers.
The ledger
A claim arrives as a row. Jurisdiction, body part, first fill, days supply, morphine equivalents. Every field is accurate and none of them is the person.
The three marked rows open. We'll show you what each one left out.
| Claim no. | DOI | Juris. | Site | First fill | Days | MED | Case mgr |
|---|
Composite accounts drawn from the published literature below. Claim numbers, dates and figures are sample data — no real claimant is depicted.
Evidence
This is not a sentiment. Decades of occupational health research finds that the compensation process itself changes outcomes — and that the mechanism is depersonalization: stigma, powerlessness, and having to prove yourself over and over.
In a six-year prospective cohort, a third of claimants reported high stress from understanding what their claim required of them; 30.4% from delays and 26.9% from the number of medical assessments. Compensation-related stress correlated with poorer long-term recovery.
Grant, O'Donnell, Spittal, Creamer & Studdert. JAMA Psychiatry 2014;71(4):446–453. PMID 24522841Injured workers told a parliamentary roundtable they had cycled through a minimum of five case managers, and in one case more than sixteen, with no apparent handover — so they had to prove themselves and retell the story each time. For psychological injuries, that means recalling the trauma on demand.
Injured workers roundtable, summary report. NSW Standing Committee on Law and Justice. ReportBerkeley's Labor Occupational Health Program studied the California claims experience for the state commission. Workers reported frustration, anger and sadness at impersonal or condescending treatment — and concluded the person they most trusted for accurate information was themselves.
Sum & Stock, Navigating the California Workers' Compensation System. UC Berkeley LOHP for CHSWC. SummaryLippel named the mechanism: workers — including some whose claims were accepted without contest — described being treated like a criminal, or like David facing Goliath. Workers with social support got through intact; isolated workers did not. The damage sits inside provisions that look procedurally neutral.
Lippel, Preserving workers' dignity in workers' compensation systems. Am J Ind Med 2012. Full textCollie and colleagues surveyed nearly eleven thousand workers 6–24 months after claim acceptance, extending Kilgour's systematic review finding that claim delays and claims-management practices drive worse mental health, social and vocational outcomes.
Collie, Sheehan, Lane, Gray & Grant. BMC Public Health 2019;19:927. Full text · Kilgour et al., J Occup Rehabil 2015;25:160–181.ProPublica and NPR's Murrow-winning investigation documented that since 2003, legislators in 33 states passed laws reducing benefits or making qualification harder — while workers battled insurers for years to get the surgeries, prescriptions and help their own doctors recommended.
Grabell & Berkes, Insult to Injury. ProPublica / NPR, 2015. SeriesA claims benchmarking study found employee trust to be a key predictor of successful return to work, and that employees are more likely to retain counsel when communication or trust with the employer or administrator breaks down.
Rising Medical Solutions benchmarking study, reported in Claims Journal, 2016. CoverageThe clinical case and the financial case are the same case. That sentence, from the JAMA Psychiatry cohort study, is the business model of Visible Worker stated by somebody with no stake in it.
Grant et al., JAMA Psychiatry 2014, concluding remarks.First sighting
Long before a case manager is assigned or a treatment plan is written, something concrete happens: a prescriber sends a script. That is often the earliest structured signal in the whole claim — and it arrives with a name on it.
MedPath·ERAI — Enhanced Recovery After Injury — is the interception point. Every injury script flows through one screening point, in near-real time.
Injury e-scripts are e-prescribed with MedPath as the designated pharmacy and travel the standard e-prescribing path. No new workflow is imposed on the prescriber, and nothing depends on the claimant remembering to tell anyone.
Because the script is structured data on arrival, the claim has a clinical fingerprint before the first adjuster note exists.
Where scripts originate
Transport
Each inbound script is screened against the claim, the jurisdiction's treatment guidelines, the rest of the drug profile, and the patterns that historically precede a long claim. The AI decision-support layer drafts the read; a pharmacist or clinician decides.
Early opioid prescribing is treated as a triage risk marker — a reason to look more closely at a person, not a causal claim about their outcome, and never a judgement about the prescriber.
What screening looks for
Routing decides who reaches out, how fast, and with what. Most people need a clear explanation and a working pharmacy benefit. A minority need clinical intervention early enough to matter. The point of screening is to tell those two groups apart on day three instead of month nine.
Whoever picks up the phone starts with the record already assembled — because the most reliable way to make somebody feel like a number is to ask them to explain themselves to a fifth stranger.
Pathways
What we do
Interception is where a person becomes visible. Everything below is how they stay visible — through the pharmacy, the paperwork, the authorization, the payment, and the year after that.
One assembled record, one point of contact, one story. The research on case-manager churn is unambiguous, and it describes a solvable engineering problem rather than a personality problem: when the history travels with the person, nobody has to prove themselves again to be believed.
In practice — a maintained health and claim summary the worker can actually use, plain-language explanations of what happens next, and outreach that starts before somebody gets frightened enough to call a lawyer.
Screening every workplace-injury e-script at the point of arrival, routing into care pathways in near-real time, and flagging the claims quietly heading somewhere bad while there is still time to change the direction.
Network and mail-order fulfilment, first-fill programs, formulary and prior-authorization handling, and the unglamorous work of making sure a person in pain can actually collect the medication somebody already approved. Workers' compensation and auto lines.
Pharmacist-led regimen review, morphine-equivalent monitoring, tapering support, interaction and duplicate-therapy screening, and alternatives-to-opioids conversations with prescribers — collaborative rather than adversarial. Stewardship starting at fill one is a clinical program; stewardship starting at year two is a recoupment fight.
Structured pathways aligned to the guideline set that actually governs the claim. State-adopted occupational medicine guidelines, formularies and utilization review standards vary enormously, and a defensible decision in one state is a denial letter in another.
We maintain the mapping so authorizations stay consistent, explainable, and able to survive a dispute.
Medical records, IMEs, correspondence, remittance advice and scanned faxes turned into structured, searchable findings with the source page attached. This is the machinery behind continuity: the reason nobody has to retell their story is that the record is genuinely readable.
Multi-payer remittance and EOB processing, denial analytics, recoupment dispute support, reconciliation and receivables analysis. Denials are not only a revenue problem — a payment that stalls is a prescription somebody doesn't collect, which is why denial patterns are watched per payer and per program rather than in aggregate.
Cohort and outcome reporting, payer scorecards, savings and utilization analysis, and early detection of the shifts nobody announced — an eBill transition, a new documentation requirement, a denial rate that moved in April and went unnoticed until June.
The screening, summarization and drafting layer runs on enterprise infrastructure inside the tenant that already holds the PHI. Protected health information is not sent to third-party consumer AI endpoints.
And the model never acts alone: it proposes, a person decides, and the proposal is logged either way. A system that reduced people to numbers because a model said so would be the same failure with better latency.
The ask
The fastest way to test this is on claims you already believe went wrong. Send a closed cohort — the long tail, the litigated ones, the ones that surprised you — and we'll show you where the person stopped being visible, and how early the signal was there.
Then we do it on the open ones, while it still changes something.