A Phase 1 respiratory module for asthma and COPD: it builds a longitudinal picture from retrieved history, everyday vitals, FDA-cleared home spirometry, and local air quality — and it powers the reimbursement lane for pulmonary function and sleep testing. Reimbursement-safe today; anything that interprets lung data to flag disease stays behind FDA clearance.
Elderly and chronic-respiratory patients decline gradually — the value is in the trend. PulmoIQ fuses four honest data sources into one picture a clinician can act on.
The same discipline the rest of Ardia uses: the reimbursement/administrative work is being built first; anything device-like waits for the proper FDA pathway. We never blur that line.
PulmoIQ's tracking feeds Ardia's elder-first Companion — so an older adult with COPD gets gentle daily check-ins, medication and inhaler reminders, air-quality heads-ups, and an emergency alert to a trusted contact if their oxygen drops or they fall.
A Phase-1 respiratory decision-support module for asthma and COPD: an in-build reimbursement lane for pulmonary-function and sleep-test coding, plus a separate, FDA-gated prediction lane — non-diagnostic until cleared.
PulmoIQ assembles a read-only, time-ordered record: history via HL7 FHIR R4 per USCDI (labs/observations coded to LOINC, medications to RxNorm), everyday SpO2/HR from Apple HealthKit and Google Health Connect, readings from FDA-cleared home spirometers (FEV1, FVC, FEV1/FVC), and EPA AirNow AQI + NOAA/NWS exposure. Non-diagnostic summary.
A deterministic, rules-based engine maps documentation to CPT 94010 (spirometry), 94060 (post-bronchodilator), 94726–94729 (plethysmography/lung volumes/oscillometry/DLCO), 95810/95811 (attended PSG / PAP titration) and checks NCCI PTP edits; PulmoIQ drafts necessity notes aligned to GOLD/GINA staging (respiratory guideline retrieval in development). Runs as billing/admin support under FD&C Act §520(o).
When claims deny, the engine parses X12 835 CARC/RARC remittance codes to find root cause (missing spirometry values, frequency/NCCI edits), and PulmoIQ drafts appeal letters referencing GOLD/GINA thresholds (guideline corpus in development) plus the patient's own FEV1/DLCO trend and air-quality context. Human-in-the-loop — a clinician or biller reviews and signs before anything is sent.
Predicting exacerbations and interpreting PFT/imaging is a separate, regulated lane: a scikit-learn classifier over longitudinal vitals, spirometry-trend and AQI features. As Software as a Medical Device it is gated behind an FDA 510(k) with prospective validation — disabled and not shipped until cleared; decision support only, never autonomous diagnosis.
Assembles the longitudinal record and drafts necessity notes and appeal letters aligned to GOLD/GINA staging; guideline retrieval in development; non-diagnostic.
Maps documentation to PFT/sleep CPT codes, validates NCCI edits, and parses X12 835 CARC/RARC denials — rules-based, auditable, no ML.
Forecasts COPD/asthma exacerbation from vitals, spirometry and air-quality trends; regulated SaMD, deferred behind FDA 510(k).
Reimbursement lane: build the necessity/coding/appeal engine for PFT & sleep tests, plus FHIR, HealthKit/Health Connect and AirNow/NOAA connectors grounded in GOLD/GINA. Internal, de-identified data only.
Operate the reimbursement lane with a design-partner clinic under a BAA; clinician-in-the-loop review of coding packages and appeals. Begin assembling 510(k) evidence for the prediction lane.
Prospective clinical validation of the exacerbation-prediction model and PFT/scan interpretation; prepare and submit an FDA 510(k). Prediction lane stays disabled in production.
Expand across payers and sites; enable the prediction lane only after 510(k) clearance, always as clinician decision support.
Explore all seven modules, run the interactive tools, or book a working session with our team.
