Build the monitoring product. Skip the eighteen months of plumbing.
Ingest across a fragmented device market, unit normalization, personal baselines, threshold routing with owners and clocks, an audit trail built for an OIG look, and code selection under rules that changed in January. That is eighteen months of work that has nothing to do with your therapeutic area. It already exists.
What daily-resolution measurement gives a sponsor
The gap between prescription and discontinuation is where your commercial outcome is decided, and claims data shows it to you three to six months late with no reason attached.
Persistence, with reasons
Not a Kaplan-Meier curve from refill gaps. Time-stamped discontinuation with a structured reason captured in the two weeks before the stop, when the patient was still deciding. Tolerability, cost, plateau, or a side effect nobody warned them about.
Titration latency
How long a patient sat at a sub-therapeutic dose, and what was happening while they did. In heart failure and hypertension this is the dominant gap between trial efficacy and real-world outcome.
Response trajectory
Function and symptom trajectories at daily or weekly resolution, on therapy, in the real world. For the categories where your differentiation is functional rather than event driven, this is the evidence you cannot buy from a claims vendor.
Verified adherence before escalation
In asthma and severe respiratory disease, objective controller adherence is now what payers want before authorizing a biologic. A program that produces it serves the patient, the prescriber, and your access team simultaneously.
Trial instrumentation
Remote, repeatable, objective endpoints are scarce. Frequently the cleanest commercial fit, with none of the fraud and abuse exposure of a commercial program.
Instruments that already exist
Fifteen condition programs with device stacks, thresholds, and escalation designs already built and already running. You are configuring, not commissioning.
AI-derived signals, labeled as such
Adherence phenotypes, discontinuation-risk cohorts, deterioration-pattern flags — model outputs your evidence team can actually use, because every derived feature is labeled as derived, versioned to the model that produced it, and traceable to the raw observations underneath. Derived signals never masquerade as measurements, which is exactly the property a regulator or a journal reviewer will ask about first. Where AI runs, and where it is kept out.
Three structures we will build
The Anti-Kickback Statute prohibits offering anything of value to induce or reward referrals for items reimbursable by a federal health care program. Intent governs, and fair market value alone is not a defense. OIG has published on remote monitoring specifically — a 2023 consumer alert, a 2024 oversight report, a 2025 billing analysis, and settlements including one involving alleged payments to practices for enrolling patients.
Sponsor licenses the platform for its own research
You license Gathermed as trial or registry infrastructure. Sites are compensated at fair market value for research activities under a written protocol. No clinical claims are billed for the monitoring. Cleanest structure, and the one we recommend by default for evidence generation.
Sponsor funds a patient support program with no prescriber benefit
Monitoring and coaching delivered to patients by an independent entity, not billed to any federal program, with no free service flowing to prescribers and no data flowing back that identifies prescribing behavior for targeting. Requires careful design and existing OIG guidance on patient assistance programs applies directly.
Sponsor buys de-identified or consented data at fair market value
Practices run and bill their own clinical programs, entirely independently of you. You separately purchase de-identified aggregate data, or consented identified data under a patient authorization, at a rate supported by a valuation. No subsidy, no linkage, no targeting.
How we work
Structural review before commercial terms. We will ask which structure you are proposing in the first conversation and we will ask you to bring counsel to the second. It slows the sales cycle and it is the only version of this business worth operating.
Multi-tenant, white-label, API first
Tenancy
Per-tenant isolation with separate encryption keys. Cross-tenant queries are structurally impossible, not policy-prohibited. Configurable data residency.
Your brand
Patient application and clinician interface under your name and your domain. Gathermed appears in the subprocessor list and nowhere on the screen.
API surface
REST and webhooks over enrollment, device binding, observations, thresholds, dispositions, and monthly accounting. Sandbox tenant with synthetic patients from day one.
Consent as a field-level object
Consent is scoped per data category per recipient, versioned, and revocable. Revocation propagates to downstream exports. Sensitive categories are excluded from sponsor flows by default and cannot be enabled by configuration alone.
Audit
Append-only event log across enrollment, device binding, every observation, every threshold firing, every disposition, and every export. Per-patient per-month defense packet on demand.
Regulatory posture
HIPAA business associate agreement. Where your use makes the software a medical device, you are the manufacturer of your product and we document our role as a supplier with the design controls and traceability that requires. We will not let that ambiguity sit in a contract.