There is an awkward gap between deciding what a patient needs and getting permission to do it. In cancer care, that gap can contain a treatment plan, an insurance policy, supporting documents, and several people who would prefer to be doing something else. NantHealth has made a business of working inside it. The company’s most revealing recent number, however, concerns the people operating the machinery.
- Eviti checks specialty treatment plans against clinical evidence and health-plan rules.
- NaviNet brings insurance transactions into a shared portal for health plans and medical offices.
- The lesson: measure the experience of the person doing the work, as well as the customer paying for it.
The payer likes it. The provider has questions.
In December 2024, NantHealth described an Eviti improvement project prompted by weak provider sentiment. Users wanted easier navigation and better training and support. The company responded with interface changes, tutorials, more communication, and help tailored to individual workflows. It reports that provider Net Promoter Score rose from -25 in 2023 to +31 in 2024. Payers scored it 55 in 2024. These are company-reported survey results, rather than an independent assessment of care quality.
The distinction matters. An insurer evaluates a system partly through spending and consistency. A clinician or practice employee evaluates it while trying to complete a task. Their interests overlap; their days look different. A product can satisfy its purchaser and irritate its operator. NantHealth’s own account gives that familiar enterprise-software problem a particularly clear set of numbers.
An insurer buys a result. A medical office has to live with the workflow.The product tension
A treatment plan meets a coverage policy
Eviti Connect works at the point where a proposed treatment meets the evidence and the payer’s requirements. For oncology, it examines the regimen: the treatment plan as a whole. Plans that align with the applicable guidelines can receive automatic approval. When a plan needs closer attention, the clinical team can discuss it with the prescriber. The exception has a human route through the system.
The intellectual work behind that arrangement is substantial. NantHealth says its Eviti library contains more than 4,600 evidence-based regimens and more than 9,500 federally registered clinical trials. Oncology and informatics specialists maintain the material, including information about efficacy, toxicity, outcomes, and cost. The library is mapped to payer policies and formularies. A treatment’s clinical rationale and its reimbursement requirements can therefore be examined together.
Maryland Physicians Care offers a concrete example of the appeal. In a NantHealth-published case study, the insurer describes wanting visibility into regimen costs before treatment began. Its medical and quality executive, Mary Barone-Leitch, puts the principle neatly: “The Eviti approach is to consider the total plan.” The case study reports a 10:1 return on investment after workflow integration. That is a customer example, with its own conditions, rather than a promise to the next purchaser.
The approach also extends to autoimmune disease. NantHealth’s 2021 results describe going live with Maryland Physicians Care on digital drug authorizations for autoimmune conditions. The current offering includes evidence-based treatment validation and support for biosimilar choices. Separately, Eviti Advisor gives providers access to oncology treatment intelligence at no cost. Its job is decision support; the clinician still has a patient to assess.
One login, and the price of using it
NaviNet handles another stretch of the same journey. A medical office needs to establish whether coverage is active, find benefits information, follow a claim, and exchange documents with a health plan. Having those tasks in one portal reduces the number of places staff must visit. The customer here might be a payer sponsoring workflows, or an office subscribing to broader access.
NantHealth’s January 2025 AllPayer announcement added electronic claim submission to the package. The network figures describe different doors: eligibility and benefits for more than 1,270 plans, claim-status inquiry for more than 660 commercial plans, and claim submission to more than 8,100 organizations, including plans and third-party administrators. A large list is only useful if it includes the destinations an office actually needs.
Paid tiers include transaction allowances. Usage charges can increase the bill. Free, partner-sponsored NaviNet workflows remain available.
Current published AllPayer prices run from $75 a month for Basic to $200 for Ultimate, with intermediate tiers. Access is office-wide and month-to-month. Each subscription includes allowances, and the cost depends on transaction volume. Buyers should calculate the bill with their actual workload. A starting price has excellent manners; an overage charge is less sociable.
There is evidence of practical expansion beyond the sales page. In September 2025, AmeriHealth Caritas District of Columbia told providers they could submit claims disputes and supporting documents through NaviNet. Its announcement describes replacing calls and mailed or faxed material with a portal form and a document ID for follow-up. This is the kind of small procedural improvement an administrator can immediately understand.
The expensive route to a narrower job
NantHealth’s original ambition was considerably wider. Founder Patrick Soon-Shiong brought a physician’s interest in cancer and a businessman’s appetite for assembling technology. Historical filings describe GPS Cancer molecular profiling, combining DNA and RNA sequencing to inform personalized treatment. In January 2016, the company acquired NaviNet, bringing payer-provider collaboration into that broader project. The June IPO sold 6.5 million shares at $14 each: $91 million gross before fees.
The early retrenchment was public. In 2017, NantHealth announced a reduction of roughly 300 positions through layoffs and employee transfers, and sold provider and patient engagement assets to Allscripts. Its stated purpose in the subsequent results announcement was to concentrate on core competencies and accelerate the path to profitability. This was a portfolio decision with consequences for employees, not merely a change in website language.
Further pruning followed. The company stopped offering commercial Liquid GPS in 2019 and divested its home healthcare services business. In February 2020 it completed the $47.25 million sale of Connected Care. Later that year it acquired OpenNMS, an open-source enterprise network-monitoring company. The resulting portfolio contains an unusual neighbor: alongside approvals and insurance transactions sits software that watches IT networks, including those outside healthcare.
The finances kept their own score. NantHealth reported $67.0 million of revenue and a $67.8 million net loss attributable to the company for 2022. It then announced a $22.5 million senior secured loan from Highbridge and Nant Capital in March 2023. Nasdaq trading was suspended that May; the company moved to over-the-counter trading. These dated figures describe the cost and pressure of the journey, rather than today’s financial condition.
The product story and the financial story deserve separate attention. A useful authorization workflow does not establish that its supplier is profitable. Nor does a troubled balance sheet tell you whether a particular billing task became easier. Judging this business requires both questions. The current leadership page lists Haris Naseem as chief executive and Soon-Shiong as chairman.
What a medical office can borrow
NantHealth occupies the market between providers and payers, where clinical information has to become an authorization or a transaction. Availity offers an alternative multi-payer portal; Evolent supplies oncology management. NantHealth’s distinctive combination is regimen-level specialty-care validation, administrative collaboration through NaviNet, and a separate network-monitoring subsidiary. Which part matters depends on the customer’s problem.
The buying exercise should begin with a real task. An office can check whether its frequently used payers support the workflows it needs, estimate its monthly transaction volume, and compare the total subscription cost with its existing process. A health plan can examine how exception review works and how its coverage policies enter the system. Demonstrating a routine approval is useful; demonstrating the awkward case reveals more.
The savings case weakens when the needed payer or transaction is unavailable, when staff duplicate work elsewhere, or when training never reaches the people doing the job. Automatic validation also has boundaries: unusual treatment plans may need documentation and clinical discussion. A library can organize evidence without settling every individual case.
For a software builder, the transferable idea is equally concrete. Survey the purchaser and the operator separately. Improve the step that frustrates the operator. Check again. NantHealth’s reported satisfaction turnaround makes that discipline visible. Medicine contains enough unavoidable difficulty. The login, the navigation, and the route to a human reviewer are places a software company can do something about it.