Financial Model
Ramp Timeline
Buildings
Penetration
Visit Scaling
10% Loan
Bottom-Up Billing-Code Engine · NP-Capacity-Derived Ramp · COGS-Net Revenue
NE Georgia 10-Clinic Portfolio P&L
—
Y1 Net Revenue
3 clinics · launch M1 / M3 / M5
—
Y2 Net Revenue
6 clinics ramping
—
Y3 Net Revenue
10 clinics · sustaining
—
Y4 Net Revenue
Steady state · no expansion
—
Y5 Net Revenue
Steady state · no expansion
—
Y1 EBITDA
—
—
Y2 EBITDA
—
—
Y3 EBITDA
—
—
Y4 EBITDA
—
—
Y5 EBITDA
—
Monthly Net Revenue · 60-Month Portfolio Trajectory · Membership vs Ancillary Net
Annual Income Statement · Net Revenue = Membership + Ancillary (Gross − COGS) − Denial · All figures USD
Line Item Y1 $Y1 % Y2 $Y2 % Y3 $Y3 % Y4 $Y4 % Y5 $Y5 %

Sources: 2026 CMS MPFS Final Rule (NP billing factor ×0.85, GA Locality 03) · 2026 CLFS (lab and rapid-test codes, confirmed against the CMS CY2026 Q3 Clinical Diagnostic Laboratory Fee Schedule Public Use File) · NP-capacity-derived ramp schedule (see Assumptions). Every rate is listed with its sourcing note in the Assumptions tab's Billing Code Sheet. No HPSA bonus applied (deliberate simplification — see Assumptions). RPM device COGS $40/enrolled/mo (Tenovi, unverified vendor cost). POC chemistry + PT/INR (Piccolo, CoaguChek), rapid tests, and in-house lab devices (Afinion 2, Clinitek, Sysmex XW-100) COGS = 25%/50% margin on rate (team-set percentages, replacing the old flat $8.00/cartridge placeholder). Vaccine product codes (90662, 90677) COGS = ASP+6% back-out. MSO overhead includes $132K/yr ongoing compliance/legal/marketing. All figures subject to final billing validation. Consult CPA + CMS billing counsel before investor distribution.

Month-by-Month · 60 Months · 5 Years
Portfolio Monthly Detail
NP-capacity-derived ramp: Each clinic independently ramps from 20 members (M1 open) to its own NP-capacity-derived terminal panel (see Clinic Stages tab for each clinic's actual FT/PT staffing and terminal member count) over the selected Ramp Timeline. PMPM goes from — (M1) to — (ramp completion): highest in the opening months, when new-member intake visits dominate, then settling as CCM/RPM/APCM enrollment matures — computed live from the billing-code engine, not a hardcoded curve. No HPSA bonus applied. Portfolio revenue reflects staggered clinic openings — M1 (Habersham), M3 (Lumpkin), M5 (Stephens), M13 (White + Rabun), M15 (Union), M25 (Fannin + Gilmer), M27 (Gordon), M29 (Chattooga). Y4/Y5 (M37–M60): All 10 clinics at steady state — no new clinics, no new CapEx, MSO and ops held at Y3 levels.
Mo Period Avg Stage Members Net PMPM Total PMPM Total Anc PMPM Memb Rev Anc Net Denial Net Rev Clinic Staff Clinic Ops Tech Stack D&A MSO Total Exp EBITDA Net
All data drawn from the revenue engine · updates with scenario toggles
Stage Evolution · Model Revenue at Each Stage
Stage Evolution
Year Evolution · Model Revenue at Each Year End
Year Evolution

Terminal Status · Y5 Steady State
Clinic Status

24-Month Ramp · Model Revenue Month by Month
Clinic Ramp
Mo Members G0558 CCM G0557 G0556 RPM Avg Visit Freq E&M New E&M Est Total PMPM Total Anc PMPM Ancillary COGS Claim Denial Net PMPM Memb Rev Anc Net Net Rev Clinic Staff Clinic Ops Clinic Tech EBITDA Milestone

Staffing & Membership Growth · E&M Visit-Based
Staffing & Membership Growth (E&M Visit-Based)
Each month shows the clinic's members, its E&M visit demand (New + Established visits, matching the Clinic Ramp table), and the NP team's visit capacity for that month. Utilization is demand divided by capacity. When utilization passes 85%, the clinic adds an NP: full-time if a second full-time NP would be needed within 6 months, otherwise part-time.
Mo Members Avg Visit Freq NP Demand (E&M visits/mo) NP Team NP Capacity (visits/mo) Utilization Staff on Duty Hire Event
Model Inputs · Sources · Methodology · Rebuilt Bottom-Up
Key Assumptions
Ramp Model · FFS Pen Based (Demand-Driven)
Replaces the old hand-built ramp table entirely. That table capped every clinic at a hand-picked 950/1,000 members — a round number with no connection to what an NP can physically staff. Every clinic's terminal panel is now the raw addressable market (ffsPool × Penetration), a straight-line ramp with NO NP-capacity cap — NP staffing reacts to membership instead of constraining it. See the NP Staffing Model card for the reactive hiring mechanism. (An earlier "NP Based" alternative DID cap membership at NP capacity — removed; see FLAGGED_ASSUMPTIONS.)
Schedule—
M1 members (every clinic)20 — carried over from the original ramp's seed value
Terminal membersClinic-specific — FFS × Penetration, uncapped. See Clinic Stages tab for each clinic's actual number and NP staffing.
PMPM range (computed live)—
Breakeven (single clinic)Month 8–9 · ~350 members
Member cap per clinicNone — demand-driven only (see NP Staffing Model for how staffing reacts)
Y4/Y5Steady state · no expansion
Penetration—
Panel Composition · NE Georgia Mix
APCM G0558 · QMB/Dual30% of panel · 60% enrollment (one-time decision at the first visit; billing starts the next month)
CCM 99490/99439/9948745% of panel · 60% enrollment (one-time decision at the first visit; billing starts the next month)
APCM G0557 · Std FFS15% of panel · 60% enrollment (one-time decision at the first visit; billing starts the next month)
APCM G0556 · Low acuity8% of panel · 60% enrollment (one-time decision at the first visit; billing starts the next month)
Panel mix total98% — pre-existing rounding gap, carried forward as-is (not corrected — see Flagged Assumptions)
Care-mgmt enrollment capUNCAPPED — pure demand, no clinical-staffing ceiling (team decision)
Care-mgmt billing start—
RPM enrolledFlat % of mem (see Flagged Assumptions) — the old RAMP.rpm column it came from no longer exists
NP Staffing Model · Reactive Hiring, Evaluated at Ramp Completion
Every clinic's member CEILING is the raw addressable market (ffsPool × Penetration) — NP staffing has no say over how big the panel gets (see Mem. Distribution card above). This card shows what the SAME reactive hiring algorithm below produces once evaluated all the way out to the Ramp Timeline (actualClinicStaffingAt(clinic, TIMELINE)) — not a separate one-time Day-1 sizing formula. It's shown here for reference (FT/PT capacity math, utilization) alongside the Clinic Stages tab's month-by-month view of the same algorithm actually ramping up from 1 FT NP + 1 LPN.
NP annual visit capacity14 visits/day × 230 working days = 3,220/yr
Part-time NP capacity50% of a full-time NP's capacity
Visit frequency ramp—
Ramp Timeline (evaluation point shown below)—
FT / PT capacity at this Ramp Timeline—
Hire-justification threshold85% — a hire only triggers once CURRENT staffing's utilization against real demand crosses this bar
AlgorithmReactive, moment-by-moment, not a one-time formula: the instant a month's real visit demand crosses 85% of current NP capacity, look ahead 6 months (PT_CHURN_LOOKAHEAD_MONTHS) — if growth already implies needing a 2nd FT NP within that window, hire FT now; otherwise hire PT. Headcount is a monotonic ratchet (never un-hired). No unserviced residual — membership isn't capacity-throttled, so hiring simply keeps pace with (or lags behind) whatever demand shows up.
Reference clinic (FFS 4,000)—
New-Member Cohort & AWV Eligibility Logic
Replaces the original model's broken mechanics: (1) a decaying "new patient %" that hit zero, (2) a calendar-based AWV switch (G0438 for everyone before M12, G0439 for everyone after), and (3) a two-way new-to-Medicare/transferring split whose 12-month ramp didn't fast-track members who already had AWV history. AWV eligibility is now a DISCRETE, per-cohort event calculation — see awvVolumesAt() and Flagged Assumptions.
New-patient E&M volume= that month's actual new-member count (first difference of the demand-driven member curve) — never decays to zero, since new members keep arriving for the life of the clinic
Established visit volume= real per-member recurring-visit count (see E&M New/Established logic below), minus the AWV-day visits already priced through the modifier-25 same-day E&M line (repeat AWVs only)
New-to-Medicare share of new cohort—
Never-had-AWV share of new cohort—
Already-had-AWV share of new cohort—
New-to-Medicare eligibility pathIPPE (G0402, new code) at their first visit → first AWV (G0438) at +12 months of their OWN tenure (real CMS Part-B waiting period) → G0439 every 12 months after that
Never-had-AWV eligibility pathG0438 immediately at their first visit (no 12-month wait — already Medicare-eligible, just no AWV history) → G0439 every 12 months
Already-had-AWV eligibility pathG0439 immediately at their first visit (same code from day one — no separate "initial" visit) → G0439 again every 12 months
Modifier-25 same-day E&M w/ AWV75% of REPEAT AWV visits (12+ months after a member's first AWV) also bill an accompanying established-patient E&M under modifier 25. An AWV at a member's first visit does not: it rides on the new-patient E&M, and two E&M codes can't bill on one visit — a real, CMS-confirmed billing pattern (clarified eff. Jan 1 2025) when the E&M reflects a genuinely separate documented problem. The 75% rate is a team-set working assumption, not CMS-sourced. That same visit is subtracted out of the general established-visit pool so it's billed exactly once.
AWV uptake—
Behavioral Health Integration (BHI) · New to This Model
Did not exist in the original model at all. Two codes, routed by which base care-management track the patient is already on, since a BHI add-on can only attach to whichever base code that practitioner bills that patient that month.
Total BHI-enrolled population—
G0570 (APCM-track BHI add-on)Attaches to G0558/G0557/G0556 patients, split proportional to each track's enrolled headcount
99484 (CCM-track standalone BHI)Attaches to CCM patients, same proportional split
Sourcing note25% of panel was set explicitly by the team, overriding an earlier, more conservative 15%-eligible × 70%-enrolled = 10.5% formula from initial research. The proportional-split methodology is the team's own construction, not independently sourced, and arguably underweights CCM (more medically complex patients may have higher behavioral-health comorbidity) — but the team went with the simpler split anyway.
Explicitly excludedCoCM codes (99492/99493/99494, G2214, G0568/G0569) — require a contracted psychiatric consultant the clinic doesn't have. A phase-2 lever, not an oversight.

Every Rate, Every Code · Rendered Live From the Model's Own Data — Cannot Drift Out of Sync
The Billing Code Sheet
This table is generated directly from the BILLING_CODE_SHEET JavaScript object near the top of this page's script — it is not hand-typed HTML. Changing a rate in that one object is the only thing needed to update the entire model's output. All rates already include the ×0.85 NP (nurse practitioner) billing-factor discount and exclude any HPSA bonus (removed — see below), except CLFS lab codes (83036, 80061, 82043, 86140, 85025), which are paid under the Clinical Laboratory Fee Schedule and are not subject to either adjustment. "Confirmed" = verified 2026 CMS rate this session. "Unconfirmed" = physician-fee-schedule figure carried forward without independent verification this session. "Approximate" = a rough proxy rate, not a precise CMS lookup. "Team-set" = a figure the team chose directly rather than sourcing from a fee schedule.
Billing Option (fixed)—
Code Description Rate Source Sourcing Note

Revenue Model · Bottom-Up Billing Engine
Revenue structureSum of real billing-code lines × volume, no target multiplier
Membership revenue linesE&M · AWV bundle · APCM/CCM · BHI · TCM · ACP · Tobacco · G0443 · 99483 · H&B
Ancillary revenue linesRPM · POC chemistry + PT/INR (Piccolo BMP · CoaguChek) · rapid tests (QW) · vaccine products (buy-and-bill) · in-house lab devices (HbA1c · lipid · urine ACR · CBC)
CHANGED (Oct 2026 CapEx update)Lab lines follow the new equipment: Lactate (83605), hematocrit (85014) and CRP (86140) removed (no device runs them); BMP moves to the Piccolo Xpress, lipid to the Piccolo, PT/INR to the CoaguChek (volumes unchanged); urine ACR re-priced from the single-code 82043 proxy ($7.00) to 82044 + 82570 on the Clinitek ($6.23 + $5.18 = $11.41)
CCM billing success rate82% (unsourced legacy assumption)
APCM billing success ratesG0558 92% · G0557 92% · G0556 95%
RPM device COGS$40/enrolled/mo (Tenovi, unverified vendor cost — largest COGS line)
POC chemistry / lab-device COGS25% margin on rate (team-set percentage, replacing the old flat $8.00/cartridge placeholder)
Rapid test COGS50% margin on rate (team-set percentage)
Vaccine product COGSASP+6% back-out (rate ÷ 1.06) — the actual CMS Part B buy-and-bill reimbursement design, not a team-set margin
Claim denial write-off2.5% of (membership gross + ancillary/RPM gross) — unchanged from original model
NP billing factor85% of physician PFS on all professional-service codes
Remaining Scenario Toggles
Ramp Timeline12 / 18 / 24 months (18 = Spark default) — all three run the same NP-capacity-derived linear member ramp; what differs is WHICH POINT on the visit-frequency curve (see Visit Scaling below) gets used to size NP capacity. At 24 months, capacity is sized at the curve's true steady-state endpoint. At 18, at the curve's midpoint. At 12, at the curve's early/flat value — a materially higher per-NP ceiling, since staffing is a one-time decision that doesn't keep tightening after the ramp completes. E&M revenue itself always follows the real curve uncapped, regardless of which Ramp Timeline is selected.
Visit ScalingLive toggle (Low 2.0→4.0 default · 4.0→6.5 · No Scaling 4.0 flat · 2.0→6.5) — sets the two endpoints of the SAME linear curve driving both E&M revenue and NP capacity sizing. See E&M Visit Frequency card below.
BuildingsLease ($30K/clinic/yr) vs. Own/Mortgage ($350K building, 20% down + closing costs Day 1, 7%/20-yr amortization). Affects CapEx, OpEx, and the capital raise — not revenue.
PenetrationRe-added as a live toggle (25% / 30% / 35%) — no longer fixed. Directly drives each clinic's demand target, NP staffing tier, and terminal member count. See NP Staffing Model card.
Removed: Enrollment UpliftRemoved. Enrollment rates (80%/70%/75%/60%) are now the model's only enrollment assumption, applied uncapped — see New-Member Cohort card.
Removed: PMPM CeilingRemoved entirely — this was the mechanism under audit. It force-fit the computed ~$143 PMPM up to a hardcoded $239 target via a pmpmScale = PMPM_CEIL / BASE_PMPM multiplier. No such mechanism exists anywhere in this codebase.
E&M Visit Frequency · Now Unified With NP Capacity
Visits per enrolled member per year · ramped by clinic-relative month · unmodified from the original model. This is now the SAME curve that sets the NP-capacity ceiling (see NP Staffing Model card) — previously these were two independent, inconsistent curves (this revenue curve vs. no capacity concept at all). A member who needs more visits/year is exactly the member who tightens the NP's panel ceiling; one curve now drives both. Revenue uses this curve UNCAPPED (real calendar months, forever); NP capacity uses it CAPPED at the Ramp Timeline (see below) — staffing is sized once, at ramp completion, not re-tightened every month after.
Y1 (mo 1–12) · flat—
Y2 ramp (mo 13–24) · linear—
Y3+ (mo ≥ 24) · steady state—
NP-capacity evaluation point—
Ramp basisClinic-relative month · not portfolio month
RationaleEach county starts fresh — new community, new relationships, same trust-building curve regardless of when the portfolio opened
Industry benchmark (chronic FFS)10–13 visits/yr across all physicians
Rural NE GA current baseline1.4 visits/yr (problem we solve)
Concurrent billing noteE&M + CCM/APCM/BHI separately billable

Staffing · Reactively Staffed by Actual Membership (Rebuilt Again)
The old member-count-banded STAGES role table (NP/LPN/Care Coordinator/Remote MD/Admin, escalated by benefits+merit+headcount multipliers) has been removed from the cost side entirely. A prior pass replaced it with just NP(s)+driver, priced against staffClinic()'s TERMINAL capacity determination from Day 1 — but that conflates "how big this clinic can eventually get" with "how many people it actually employs this month." A 20-member clinic doesn't pay for the 2 FT + 1 PT NP team its terminal panel will eventually justify. This rebuild adds LPN/MA, Care Coordinator, and Front Desk roles, all priced against ACTUAL current-month membership via actualClinicStaffingAt(clinic, mo), not the terminal ceiling.
Full-time NP salary$147,000/yr
Part-time NP salary$73,500/yr (50% of FT)
LPN salary$60,000/yr — staffed alongside the first FT NP, Day 1
Medical Assistant (MA) salary$50,000/yr — REPLACES the LPN once a 2nd FT NP is actually justified by real membership
Care Coordinator salary$55,000/yr — hired once the LPN/MA's TOTAL workload (clinical-assist + front-desk + coordination) crosses 85% of a full-time week
Front Desk salary$45,000/yr — hired once the Care Coordinator's combined front-desk+coordination load alone crosses 85% of a full-time week
Driver — 1 per active clinic$45,000/yr, flat
Mem. Distribution—
Actual NP count sourceactualClinicStaffingAt(clinic, mo) — moment-by-moment reactive hiring against REAL current membership/visit demand, floored at 1 FT NP, monotonically ratcheted (never un-hired). The OLDER rawActualNPStaffing(mem, visitFreqAt(mo)) formula still exists but is only reachable from the dead 'Average Capacity' billing path — see Flagged Assumptions.
Cost timingReactive, not flat Day-1 — headcount ramps up as real membership does, month by month, per clinic. There is no separate capacity-sizing step anymore — the member ceiling is pure demand (FFS × Penetration), and this same reactive algorithm is what's shown, evaluated at the Ramp Timeline, in the NP Staffing Model card above. No benefits, merit, or year-over-year escalation applied to any figure here — deliberately deferred, not decided.
⚠ Still not modeledThe Admin role from the original STAGES table (physician costs are now modeled: see the Contracted Physicians card). A 2nd Care Coordinator or Front Desk hire once one FTE's workload alone exceeds a full week — see Flagged Assumptions for the compliance-risk nuance this raises specifically for CCM/APCM coordination hours.
⚠ Still-stale tech-stack scalingEHR/AI-doc/telehealth costs (techStackMonthly) still infer "1 or 2 providers" from the old member-count STAGES band, not from actualClinicStaffingAt()'s real ftCount/ptCount — not fixed this pass.
Portfolio breakeven—
Clinic Staff · Live Per-Clinic Table
Shows each clinic's TERMINAL (ramp-complete) actual staffing — the steady-state team, generated from actualClinicStaffingAt(c, TIMELINE). This is what the clinic grows INTO, not what it pays on Day 1 — see the Clinic Stages tab for the month-by-month ramp-up of NP, LPN→MA, Care Coordinator, and Front Desk hires. Updates automatically with the Penetration and Ramp Timeline toggles.
Clinic FFS Pool NP Staffing Support Staff Clinical+Support/yr Driver/yr Physicians/yr Total/yr Total/mo
Contracted Physicians · Permit Health
The PC owner doctor and the supervising doctors are contracted through Permit Health, priced monthly and charged in the P&L every month a clinic is open. Pricing as provided by the team.
PC owner doctor · first clinic$2,950/mo
PC owner doctor · each of the next 6 clinics+$645/mo each (clinics 2–7, in opening order)
PC owner doctor · clinics 8–10No additional cost
PC owner doctor · maximum$2,950 + 6 × $645 = $6,820/mo once 7 clinics are open
Supervising doctor$1,050 per NP per month (every NP head, full- or part-time, from the month the NP starts)
Permit Health setup fee$9,000 one-time · Day 1 · CapEx tab, Legal & Formation (5-yr straight-line)
Total physician cost by year—
Expansion · SPARK Senior Health, Inc. Sequence
Y1: Habersham (5,066 FFS)Launch M1
Y1: Lumpkin (3,946 FFS)Launch M3
Y1: Stephens (3,554 FFS)Launch M5
Y2: White + RabunLaunch M13
Y2: Union (5,571 FFS)Launch M15
Y3: Fannin + GilmerLaunch M25
Y3: Gordon → ChattoogaM27 · M29
Y4: No expansionSteady state · 10 clinics
Y5: No expansionSteady state · 10 clinics
MSO Staff · By Year
—
Y1 MSO Staff Total$1,400,000
Y2–Y5 formula(6 founding execs + Org Structure tab's lean-build MSO headcount that year, incl. Chief HR Officer from Y2 and Chief Medical Officer from Y4) × $125,000/role
Y2 · ——
Y3 · ——
Y4 · ——
Y5 · ——
Headcount always uses the lean build (the adopted plan), independent of whatever Year/Build the Org Structure tab's own toggles currently show. See the Org Structure tab's Metrics sub-tab, "MSO (execs + staff)" row, and its Lean Build sub-tab for the role-by-role build-out.
Capital Structure · SPARK Senior Health, Inc.
Series Seed · M1 (single raise)$5M
Instrument100% equity — no debt, no interest expense on the raise itself
Series A / B$0 — Series Seed is the only round modeled, no follow-ons
Total raise—
Per-clinic ops—
MSO staff Y1—
MSO staff Y2—
MSO staff Y3—
MSO staff Y4—
MSO staff Y5—
CapEx · Per-Clinic Summary
Clinic 1 (Flagship) total CapEx—
Clinics 2–10 (Expansion) total CapEx—
Y1–Y3 portfolio total CapEx (Y4/Y5 = $0)—
Leasehold improvements$98,500 · 15-yr useful life
Clinical equipment total—
Vehicles (van + wrap + WAV conversion)$46,500 · 5-yr useful life
GAAP depreciation (expansion)—
Funded byRaise tranches
Oct 2026 CapEx update · per clinic
Clinical Equipment$51,274 (was $23,329, +$27,945)
Furniture & Fixtures$14,032 (was $12,500, +$1,532)
Removedi-STAT Alinity −$3,000 (CHEM8+ no longer CLIA-waived; optional $4,500, not counted) · Automated urinalysis reader −$600 (duplicate of the Clinitek)
AddedPiccolo Xpress chemistry analyzer $17,178 · Clinitek Status+ urine ACR $1,049 · Accessible wheelchair scale $3,669 · Stadiometer $221 · Diagnostic sets (2) $1,646
MovedCoaguChek XS Pro → Day 1 as the primary PT/INR device, $635 (was a Month 9 backup, $650)
Re-priced equipmentAfinion 2 $4,701 (was $3,000; HbA1c only) · Sysmex XW-100 $6,000 (was $2,500) · 12-lead ECG $2,477 (was $1,500) · Vaccine fridge $2,165 (was $2,800) · Nebulizer $300 (was $600) · Vitals monitor $2,500 (was $1,100 station; scale split out) · Exam lights $653 (was $230) · Butterfly iQ+ $2,699 (was $2,999) · Spirometer $2,731 (was $2,000) · ABI Doppler $1,500 (was $1,200)
Re-priced furnitureExam tables (2, height-adjustable) $7,032 (was $6,000) · Storage incl. locked medication cabinet $1,500 (was $1,000)
Useful lifeUnchanged: Clinical Equipment 5 yr, Furniture & Fixtures 7 yr (straight-line)
Own Buildings Toggle · Mortgage Model
Building price$350,000/clinic · ~$159/sqft × 2,200 sqft
Down payment (20%)$70,000/clinic
Mortgage terms7% · 20-yr amortization · commercial
Active scenario—
Tech Stack · Dynamic Scaling
Reference: Early Scale = 5 clinics · 5 providers · 300 CCM · 150 RPM = $18,370/mo
EHR (Athelas AIR / Athenahealth)$750/provider/mo
CCM/APCM platform (Chronic Care IQ)$6/enrolled member/mo
RPM devices (Tenovi)In COGS — not double-charged
AI documentation (Commure Ambient)$100/provider/mo
RCM (Commure RCM)4% of monthly net collections
Patient SMS (Klara)$300/clinic/mo flat
Telehealth (Doxy.me + Zoom)$35/provider/mo + $200/clinic flat
Care transitions ADT (Bamboo Health)$1,500/mo flat portfolio fee
Connectivity (fiber + cellular)$400/clinic/mo
Fleet tracking (Samsara)$40/vehicle/mo · 1 van/clinic
HR/payroll/MDM (Rippling)$12/employee/mo + $35 base
HIPAA compliance (Compliancy Group)$400/mo flat practice fee
Early Scale reference total~$18,370/mo (5 clinics)

Clinical Equipment · Introduction Schedule · Phase-Gated by Member Count
Equipment Schedule by Phase
Equipment introduction is clinically sequenced — each item is added only when the member panel is large enough to sustain utilization and the NP has had time to develop clinical confidence. Hardware costs reflect current market pricing. Revenue codes are 2026 CMS MPFS rates × 0.85 NP factor (no HPSA bonus); lab codes are 2026 CLFS rates (no NP or HPSA adjustment).
Equipment · Introduction Phase · CLIA Requirement · Hardware Cost · Revenue / COGS Treatment
Equipment Description Introduce At Members CLIA Hardware Cost Ongoing Cost Revenue / COGS Treatment
Phase 1 · Day 1 (Mandatory at Launch)
Tenovi 4G Cellular RPM Hub BP cuff + glucose + weight + pulse ox. No WiFi needed. Day 10 None $0 upfront $28–$35/device/mo Bills 99453 (one-time setup) + 99454 + 99445 + 99457/99458 (85% of management-qualified pts) + 99470 (the other 15%, new-for-2026 shorter tier). Device cost ($40/enrolled/mo, unverified vendor cost) deducted as COGS against RPM gross revenue.
Abbott Piccolo Xpress NEW POC chemistry analyzer — BMP, CMP, electrolytes, lipid (CLIA-waived discs, lithium heparin whole blood) Day 10 CoW $17,178 one-time purchase (⚠ CHANGED — replaces the i-STAT Alinity, removed Oct 2026: CHEM8+ is no longer CLIA-waived) Disc cost Bills 80048QW BMP ($8.46: 100% of new patients + 10% of established/yr) and 80061QW lipid ($13.09). CMP (80053QW) is available on the device but not priced. COGS = 25% margin on rate.
CoaguChek XS Pro PT/INR meter Day 10 CoW $635 one-time purchase (⚠ CHANGED — moved from a Month 9 backup to Day 1 primary) Strip cost Bills 85610 ($4.29): 5% of established panel at 17x/yr. COGS = 25% margin on rate.
Afinion 2 HbA1c analyzer (CLIA-waived, FDA CW210007) Day 10 CoW $4,701 one-time purchase (was $3,000) 25% margin on rate Bills 83036 HbA1c only ($9.49, diabetic panel 4x/yr). ⚠ CHANGED: the US Afinion 2 does not run lipid or CRP and its ACR test is moderate complexity, so lipid moved to the Piccolo, ACR to the Clinitek, and CRP was removed.
Siemens Clinitek Status+ NEW Urine ACR (semi-quantitative) + routine dipstick Day 10 CoW $1,049 starter kit Strips (operating cost) Bills urine ACR as 82044QW ($6.23) + 82570QW ($5.18) = $11.41, diabetic panel 1x/yr — replaces the 82043 proxy ($7.00). Replaces the separate urinalysis reader.
Sysmex XW-100 POC hematology analyzer — CBC Day 10 CoW $6,000 one-time purchase (was $2,500) 25% margin on rate Bills 85025 ($7.59, 2026 CLFS), 100% of panel 1x/yr. Also reports hematocrit, so 85014 is not billed separately.
12-Lead ECG Machine Standard 12-lead for AWV cardiac screening and acute evaluation Day 10 None $2,477 (Embra E-30; was $1,500) Minimal (paper/gel) Not itemized anywhere in the revenue engine — never had its own CPT code line, before or after the ancillary rebuild. Genuinely unmodeled, not folded into any bucket.
Rapid Tests (Antigen, QW) Flu A/B, Strep A, RSV, COVID, combo, Mono — 6 CLIA-waived antigen tests Day 10 CoW (included) $0 (consumables only) 50% margin on rate Live in Ancillary revenue: 87804/87880/87807/87811/87428/86308, all 2026 CLFS-confirmed rates — see Billing Code Sheet. Replaces the old, unitemized "Other" ancillary bucket (that bucket no longer exists).
Phase 2 · Month 6 (275+ Members · Clinical Confidence Established)
Butterfly iQ+ POCUS Handheld ultrasound — DVT screening, cardiac, abdominal, pulmonary views Month 6275 None $2,699 (was $2,999) Gel ~$2/scan + ~$420/yr membership ⚠ NOT MODELED — the legacy ancillary bucket this used to be folded into (16% POCUS share) has been removed and NOT replaced with a real POCUS revenue line, pending real Butterfly IQ device costing and CPT code decisions (76705 abdominal, 93308 limited cardiac, 76604 pulmonary, 93970/93971 DVT duplex were researched but never wired in). This is a genuine revenue reduction versus the prior version, not a reclassification.

Every Unsourced, Unconfirmed, or Team-Set Number In One Place — Rendered Live From the Model's Own Data
Flagged Assumptions Registry
This table is generated directly from the FLAGGED_ASSUMPTIONS array in this page's script. Every item that is NOT a confirmed 2026 CMS/CLFS rate — every code-mix split, enrollment rate, panel-composition percentage, and cost placeholder — is listed here with its flag type and a plain-language explanation of where it came from and why it was kept.
Assumption Flag Explanation

Documented, Not Fixed — Explicit Follow-Up Items
Known Limitations & Follow-Ups
Weakest-Sourced Part of the Model
Ancillary %-of-panel volumes (POC chemistry, rapid tests)The old $18→$42 PMPM legacy-ancillary ramp has been fully replaced with real, per-CPT-code line items (BMP, PT/INR, 6 rapid tests; Lactate removed Oct 2026) — a genuine improvement in structure. But their VOLUMES (30%/30%/3% for i-STAT labs; 25%/8%/12%/20%/15%/2% for rapid tests) are still team judgment calls, not independently sourced — same sourcing tier as the bucket they replaced, just itemized instead of blended. Also: the 87428 combo test and the standalone 87804/87811 antigen tests are modeled as independent volumes even though a real visit would only bill one or the other — summing them overstates combined revenue if read as additive. POCUS/Butterfly IQ revenue, previously folded into the old bucket's 16% POCUS split, has been dropped entirely pending real device costing (see the POCUS discussion in prior research) — a real revenue reduction, not a reclassification.
TCM hospitalization rate5%/yr is very likely too low for this population (rural, QMB-heavy, high chronic-disease burden). Published Medicare FFS TCM-eligible-discharge benchmarks run 10–15%/yr. The team explicitly decided NOT to change this number yet — deferred pending a follow-up decision.
Documented But Not Modeled
Venipuncture (36415)$9.34/encounter, 2026 CLFS confirmed. Would only apply to tests genuinely sent to a reference lab — the team has not yet built out the remaining send-out-only tests (TSH, B12/folate, Vitamin D, urine culture, PSA) that would trigger it. No revenue line built.
SBIRT (G0396/G0397)Ambiguous Medicare-vs-Medicaid coverage status — needs MAC-specific verification before use. Not implemented.
CoCM codes99492/99493/99494, G2214, G0568/G0569 — require a contracted psychiatric consultant the clinic doesn't have. Intentional phase-2 exclusion, not an oversight.
Group/family H&B codes96164/96165/96167/96168 intentionally excluded — Spark has no group/family visit structure, so including them would double-count the same population under a mutually-exclusive alternate billing pathway.
Per-Test Economics · Updated for the 25%/50% Margin COGS
Superseded: this card previously flagged ACR, CRP, and CBC as net losses under a flat $8.00/cartridge COGS assumption. That assumption has been replaced with a 25% margin on rate for i-STAT/Afinion/Sysmex codes (and 50% for rapid tests), which flips all three from a loss to a small profit — a genuine change in the model's economic claim, not just a relabeling.
Urine ACR (82044 + 82570)CHANGED Oct 2026: $11.41 revenue ($6.23 + $5.18, approx. 2026 CLFS), $8.56 COGS (25% margin) = $2.85 profit/unit. Was the single-code 82043 proxy at $7.00.
CRP (86140)REMOVED Oct 2026 — no device in the CapEx list runs CRP (the US Afinion 2 does not).
CBC (85025)$7.59 revenue, $5.69 COGS (25% margin) = $1.90 profit/unit. Still a conservative volume floor: CBC is the single most commonly ordered primary-care lab test, so 1×/year likely understates true frequency.
25%/50% margin COGS itselfTeam-set percentages, not sourced vendor/cartridge costs for any specific device or manufacturer — replacing the earlier flat-$8 team-set placeholder with a different team-set placeholder. Neither has been validated against a real vendor quote.
Depreciation (D&A)The monthly P&L now charges full GAAP straight-line depreciation from the CapEx schedule (replacing the original $500 + $650/clinic/mo proxies). Owned buildings depreciate on the full $350,000 value over 39 years. D&A is non-cash, so the Capital Raise cash waterfall adds it back. Confirm treatment with your CPA. See CapEx Plan tab.

Equipment costs sourced from Spark Senior Health clinical model. CMS billing rates: 2026 MPFS (CF $33.40) × 0.85 NP factor, no HPSA bonus. Lab codes: 2026 CLFS. Validate all independently before investor distribution.

Capital Expenditure · One-Time · Per-Clinic · Y1–Y3 Deployment · Y4/Y5 No New CapEx
Clinic CapEx Plan
—
Y1 CapEx
3 clinics (1 flagship + 2 expansion)
—
Y2 CapEx
3 expansion clinics
—
Y3 CapEx
4 expansion clinics
—
Y1–Y3 Total CapEx
Y4/Y5 = $0 · fully built out
—
Annual Depr/Clinic
Avg GAAP (expansion)
Annual CapEx by Category · Flagship = Clinic 1 · Expansion = Clinics 2–10
Category Useful Life Per Clinic (Flagship) Per Clinic (Expansion) Y1 Total Y2 Total Y3 Total Y1–Y3 Total

Line-Item Detail
Per-Clinic Item Schedule
Item Notes Phase Flagship Expansion

GAAP Depreciation Reference
Annual Depreciation per Clinic
Category Per-Clinic Cost Method Useful Life Annual Depr Funded by Raise
How the P&L uses this: The monthly P&L charges this straight-line schedule for each open clinic from its launch month, replacing the original $500 + $650/clinic/mo proxies. Owned buildings depreciate on the full $350,000 building value, not just the Day-1 cash. Depreciation is non-cash: the Capital Raise cash waterfall adds it back so CapEx is only counted once. Confirm treatment with your CPA.

Acquisition Timing · Spark Equipment Introduction Schedule
CapEx Phasing

CapEx estimates based on Spark Senior Health clinical model for clinical equipment; published market rates for buildout, furniture, technology, and vehicles. Clinical equipment and furniture re-priced line by line in the Oct 2026 CapEx update (each line's status shown under its note); all lab devices are one-time purchases (no lease option). All figures are estimates — obtain 3 contractor bids for buildout and 2 dealer quotes for vehicles before finalizing. Depreciation: straight-line over each category's assumed useful life (5-yr for all lab devices — a judgment call, matching other Clinical Equipment; no explicit useful life was specified for these two new devices). Validate with CPA.

Capital Raise · Use of Proceeds · Cash Waterfall · 60-Month · 5-Year View
Capital Raise Plan
How to read this tab: A single Series Seed equity placement at M1 covers all 5 years — no Series A, no Series B, no debt. The raise size tracks the Building toggle: $4M in Lease mode, $5M in Own/Mortgage mode (which carries Day-1 building-acquisition CapEx the Lease mode doesn't). The raise must cover (1) one-time CapEx at each clinic opening, (2) operating shortfall during ramp months, and (3) maintain a positive cash buffer. Y4/Y5 are steady-state — no new raises, no new CapEx. Changing the Buildings toggle updates the raise size, operating cash flows, CapEx, and minimum cash balance dynamically.
—
Series Seed · M1
$4M–5M equity
—
Series A · M13
No additional raise
—
No Series B
Self-funded from M1 raise
—
Min Cash Balance
—
—
Y5 Ending Balance
After all ops + CapEx
Series Seed Equity Policy: One Series Seed placement at M1 covers all 5 years — no Series A, no Series B, no debt. Raise size is $4M in Lease mode and $5M in Own/Mortgage mode, since Own mode's Day-1 building acquisition CapEx (20% down + closing costs per clinic) requires more capital up front. Toggle between Lease and Own above to see both the raise size and how it's absorbed change dynamically.
Cumulative Cash Balance · 60 Months · 5 Years · Raise Events + CapEx Deployments + Operating Cash Flow
Use of Proceeds · By Year · Raise vs. Deployment
Category Year 1 · Series Seed Year 2 · Series A Year 3 · Series B / Self-Funded Year 4 · Steady State Year 5 · Steady State 5-Year Total

Month-by-Month Cash Position
Cash Flow Waterfall
Month Period Raise In CapEx Out Operating Cash Total Cash Out Cash Balance Notes

Operating cash = net profit from the monthly P&L engine plus depreciation added back (D&A is non-cash; CapEx is deducted separately in its own column) (no interest expense — the Series Seed raise is equity, not debt, whether $4M Lease or $5M Own/Mortgage). CapEx is front-loaded at each clinic's launch month for simplicity. Cash balance must remain positive at all times; if it goes negative, the raise is insufficient for that scenario.

Organization & Reporting Structure
Org Structure
Overall business · selected year

Business metrics

Five-year view

Adopted plan · executives work as line leaders

Lean build: hire as volume requires

The six Y1 executives run their functions hands-on (the Chief HR Officer joins in Y2 and the Chief Medical Officer in Y4). A role is added only when volume makes an executive the bottleneck, and specialist work is contracted until it is steady enough to bring in-house. The selected year is highlighted.

Hires by year

Watch Year 3. Four clinic openings and eight MSO hires land in the same year. Pulling the Controller and the West Regional Manager forward to late Year 2 would spread the load.

Contracted services

Who covers each function · lean build

Coverage map

MSO functions by year, plus the PC's clinical-authority line at the bottom. Summary counts cover MSO functions only.

EXECan executive covers itHIREin-house MSO hireCONTRACTcontracted serviceCLINICclinic staffPCPC physician authority (not MSO)coverage changes that year (bar in that year's color)
Reporting structure · two legal entities

Org chart

Solid grey: reports to, within the MSO Solid purple: reports to, within the PC (hire/fire, discipline, clinical guidance) Dotted gold: operational lead or setup/monitoring under the MSA. No hire/fire, no clinical authority
Board of Directors
SPARK Senior Health, Inc. · Delaware MSO
Chief Executive Officer
Founder
Clinical authority boundary · SPARK Senior Health, PC (Georgia)
Every hire event · MSO and PC

Positions by year and reason for hire

Each row is the year a position is first filled, gains seats, or is retitled. Clinic care-team rows show FTE added across all clinics.

New positionfirst hire of a roleAdded seatsmore seats in an existing roleRetitleno new seatIn placeexisting executive
PC · clinical delivery

Clinic and PC staffing

OPEN 1.0 clinic opens that year2.0 +0.5 NP/PA FTE added to an open clinic

Clinic team template

RoleEntityAt openingGrowth ruleNotes
Nurse Practitioner / Physician AssistantPC1 full-timeModel ladder, 85% of visit capacityBills under own NPI; covered by a collaborating-physician protocol.
Lead NP (designation)PC—Clinics with 2+ NP/PAsReduced panel plus on-site clinical leadership.
Licensed Practical Nurse (or MA)PC11 per ~500 patientsLPN preferred for CCM/APCM clinical-staff defensibility. With care coordinators, LPNs handle RPM monitoring and alerts in every year.
Care CoordinatorPC11 per ~500 patientsCCM/APCM/BHI/RPM time, care plans, RPM reading follow-up, TCM outreach, AWV scheduling.
Patient Services RepresentativeTBD12 above 750 patientsCheck-in, eligibility, desk collections. Phones centralize to Patient Access from Y3 (lean) or Y2 (full).
Community Transport DriverTBD11One van per clinic.

Reporting lines for clinic roles

Every clinician is a PC employee who reports up the PC. The COO chain is their operational lead, shown dotted: it directs how the clinic runs, not who is hired or fired and not any medical decision.

RoleEmployerSolid line: reports toDotted line: operational lead / setupWho decides what
NP / PAPCSupervising physician (Lead NP at clinics with 2+ NP/PAs) → Physician OwnerDirector of Clinic Ops / Regional Manager → COOHiring, firing, discipline, protocols, delegation, medical guidance: supervising physician. Schedules, workflow, supplies, onboarding logistics: COO chain.
LPN / MAPCClinic NP / Lead NP → supervising physicianCOO chainSame split. MAs work under physician delegation, so they sit with the PC. LPNs and care coordinators monitor RPM readings and alerts.
Care CoordinatorPCClinic NP / Lead NP → supervising physicianCOO chain; program workflows from the Care Management Program Manager (MSO)Clinical supervision stays with the PC. The MSO administers CCM/APCM/RPM program mechanics.
Lead NPPCChief NP Officer (from Y3) or supervising physicianCOO chainHandles day-to-day performance issues; escalates hire/fire decisions to the physician.
Behavioral Health care managers (hub, optional)PCPhysician OwnerCare Management Program Manager (program administration)Clinical review and escalation: PC. Platform and workflow administration: MSO.
Front desk (Patient Services Rep)TBDDirector of Clinic Ops / Regional Manager (COO chain)—Non-clinical; direct operational reports. Employer entity is an open decision.
Community Transport DriverTBDCGOCOO chain: vehicle setup, dispatch, Samsara monitoring, complianceGuardrail on the CGO line: rides not advertised, no marketing during rides.

PC-level roles

The PC is owned by a contracted Georgia-licensed Physician Owner who covers all clinics. Collaborating physicians are also contracted: one from Year 1, then one per 8 NP/PA heads under Georgia's combined APRN + PA limit (O.C.G.A. § 43-34-25). The owner is not counted as a collaborator. Contracted physicians are excluded from FTE totals; cost is pending the Permit Health bid.

Reference

Lean build vs full design

Seats by year for every role in both builds (Y1·Y2·Y3·Y4·Y5). Where the lean build has no seat, the last column shows who covers the work instead.

Rationale

Principles and open decisions

Executives are line leaders

Each executive runs a function directly in Y1. The first hire under each is the person who removes that executive as a bottleneck: Billing & Coding Lead (CFO), Health IT Implementation Specialist (CDIO), Community Outreach (CGO), then the Director of Clinic Operations in Y2 (the COO covers Medicare enrollment through a contracted consultant).

Contract first, hire when steady

Specialist or uneven work (bookkeeping, marketing, IT help desk, coding audit, MIPS, security assessment, Medicare enrollment, and contingency recruiting) is contracted until volume is steady enough to justify a seat.

Revenue integrity stays in-house

AI scribing and the RCM vendor do the mechanical work. The PC carries billing liability, care-management rules are strict, and a percentage-of-collections vendor has no incentive to chase small denials, so one in-house lead owns it from day one.

Two lines for every clinician

Clinical staff are PC employees. They report up the PC to the supervising physician, who alone decides hiring, firing, discipline, protocols, and medical guidance. The COO chain is their operational lead (dotted): scheduling, workflow, facilities, supplies. The Physician Owner holds clinical authority and is not the Medical Director. A Medical Director is hired in the MSO later (Y4) for policy guidance only, with no clinical authority.

HR: Chief HR Officer under the COO

People (HR), led by the Chief HR Officer, reports to the COO. Y1 uses a contracted HR consultant under the COO; the Chief HR Officer and an HR Generalist start in Y2.

Drivers: CGO line, COO setup

Drivers report to the CGO. The COO chain sets up vehicles and dispatch and monitors compliance through Samsara. The CGO line carries a guardrail: rides are not advertised and drivers do not market during rides.

Open decisions

DecisionOwnerWhy it matters
Employer entity for front desk and driversCCO/CLOAffects the MSA fee structure and Stark/AKS review of the related-party transportation arrangement. MAs sit with the PC (delegated clinical work). Drivers report to the CGO, so the transportation safe-harbor guardrail applies to that line.
Pull Controller and West Regional Manager into late Y2CEO / CFO / COOSpreads the Y3 load of four openings and eight MSO hires.
MSO Chief Medical Officer scope (Y4)CCO/CLOPolicy guidance only. Have counsel confirm the role description and reporting so MSO advice cannot read as direction of the PC's clinical decisions. The PC adopts all clinical policy.
Building ownership structureCFO / CCO/CLOThe model assumes the company owns its clinic buildings (mortgage option). Decide which entity holds title and how the PC occupies the space; have counsel confirm rent terms (fair market value, set in advance) against Stark/AKS.
Site-level day-to-day leadCOO / Physician OwnerA Lead NP in a dual role (operational direction from the COO chain, clinical direction from the physician) or a Clinic Manager. At single-NP clinics the NP would carry both.
Physician hire/fire authority in the Permit Health contractCCO/CLOSupervising physicians hold hiring, firing, and discipline for up to 8 NPs each. The contract must give that authority in practice, with the Lead NP or Chief NP Officer handling day-to-day performance issues.
Permit Health physician arrangementCCO/CLOConfirm the bid structure against the May 2026 GCMB position statement on paid physician-matching services, the PC ownership requirements, and fair-market-value terms before signing.
CFO full-time conversion dateCEO / CFOThe CFO's executive tranche is contingent on full-time hire.
Behavioral Health hub and CoCM in Y3Physician Owner / COOOptional in the lean build; requires a contracted psychiatric consultant.
Expansion-readiness rolesBoardPayer Contracting, Market Development, and AI & Automation only if expansion is funded.

This organization design is independent of the financial model's staffing costs and salaries. To price it, the MSO staff line and clinic staffing engine would need to be rebuilt against these roles.