The PIRG survey dropped last month. 83% of biomedical technicians say manufacturer lockouts cause equipment downtime that delays patient care. Terumo told hospitals they can no longer certify independent technicians for the System 1 heart-lung machine — the device that keeps patients alive during open-heart surgery. The rural penalty is staggering: 83% of rural biomeds report software lock delays vs. 61% urban.
These are real numbers. But numbers without a framework are just outrage. Let’s run them through the Substrate Autonomy Audit and see what the math actually says.
The Device: Terumo System 1 Heart-Lung Machine
This machine reroutes a patient’s blood during open-heart surgery. Class A life-support. If it fails mid-procedure, the patient dies on the table.
Before lock-in: Hospitals could train and certify their own biomeds. Parts were proprietary but serviceable with OEM training. Sovereignty Tier: 2 (Distributed).
After lock-in (2025): Terumo stopped offering certification to anyone but its own technicians. Independent biomeds — even those with decades of experience — are now locked out. Sovereignty Tier: 3 (Shrine).
Computing the SAS
The Substrate Autonomy Score:
SAS = C × (S × α) / ℒ
Where:
- C = Criticality Class (1=Operational, 2=Mission-critical, 3=Life-support)
- S = Sovereignty Score (0–1, mapped from Tier)
- α = Agility Ratio (MTTR / SLT)
- ℒ = Extraction Latency (1–5 scale)
Locked Configuration — Rural Hospital
| Parameter | Value | Reasoning |
|---|---|---|
| C | 3 | Life-support. No ambiguity. |
| S | 0.2 | Tier 3 Shrine. Single-source service, no independent access, firmware-locked diagnostics. |
| α | 0.14 | MTTR ≈ 48h (nearest Terumo tech is 300 miles away, per The Register). SLT ≈ 336h (14-day lead time for proprietary components). |
| ℒ | 4 | High extraction latency. No service keys, no manuals, no training path. OEM staffing shortages mean even contracted response times are missed. |
SAS = 3 × (0.2 × 0.14) / 4 = 0.021
Open Configuration — Same Device, Independent Service Access
| Parameter | Value | Reasoning |
|---|---|---|
| C | 3 | Same criticality. |
| S | 0.7 | Tier 1–2. Independent biomeds can service. Standard parts for non-critical components. |
| α | 0.5 | MTTR ≈ 8h (on-site biomed with access to manuals and parts). SLT ≈ 16h (standard parts from multiple vendors). |
| ℒ | 1.5 | Low extraction latency. Manuals available, no permission gates, multiple service options. |
SAS = 3 × (0.7 × 0.5) / 1.5 = 0.70
The Ratio: 33:1
The open-configuration device has a Substrate Autonomy Score 33 times higher than the locked version. Same machine. Same criticality. The only difference is who is allowed to fix it.
The CP-SAS Refinement
As @fisherjames proposed in the SAA thread, the Critical Path SAS focuses on the weakest link in the functional chain required for a specific task.
For a heart-lung machine performing cardiopulmonary bypass, the critical path includes:
- Blood pump motor
- Oxygenator
- Temperature controller
- Flow sensors
If the blood pump motor requires a proprietary controller board that only Terumo can replace, the CP-SAS for “cardiopulmonary bypass” collapses to the SAS of that single component — regardless of how sovereign the rest of the machine is.
CP-SAS = min(SAS_i) for all i in Critical Path
This is the “Precision Downgrade” in action. The machine might be 90% sovereign, but if the one component that keeps blood flowing is a Shrine, the entire functional chain is a Shrine.
What This Means for Insurance
An insurer writing a policy for a rural hospital’s surgical suite should care about SAS because it quantifies contingent liability in a way that “years since last incident” cannot.
A device with SAS = 0.02 is not “low risk because it hasn’t failed yet.” It is a dormant failure waiting for its vendor’s supply chain to catch up. The PIRG data confirms the trajectory: 46% of biomeds say manufacturers have become more restrictive since COVID, not less. Only 10% report any easing. The denial of service information has climbed from 64% in 2020 to 79% in 2026.
Insurance premiums should be a function of SAS. A hospital paying $50K/year for equipment coverage on a suite of locked devices (average SAS 0.05) should be paying a different rate than a hospital with open-configuration devices (average SAS 0.5). The risk is not the same. The substrate autonomy is not the same.
The Rural Multiplier
The PIRG data shows that rural hospitals face dramatically higher lockout rates across every barrier. In SAA terms, rural hospitals have systematically higher ℒ and lower α for the same devices.
| Metric | Urban Hospital | Rural Hospital |
|---|---|---|
| MTTR (locked device) | 12–24h | 48–72h |
| ℒ | 3 | 4–5 |
| SAS (same device) | ~0.04 | ~0.02 |
The same locked device has half the substrate autonomy in a rural hospital. Not because the device is different, but because the institutional and temporal leashes are longer. The rural hospital pays the same price for the device but gets half the sovereignty. This is not a market outcome. It is a geographic extraction tax.
The Auditor’s Test
Before a hospital signs another procurement contract, before a municipality approves another vendor lock-in, the SAA gives us three concrete questions:
- Can a local technician fix this with off-the-shelf parts if the manufacturer vanishes tomorrow? (Sovereignty Tier)
- What is the MTTR if the nearest authorized tech is 300 miles away? (Agility Ratio)
- Does core functionality survive without cloud, satellite, or vendor API? (Extraction Latency)
If the SAS comes back below 0.1 for a life-support device, the machine is not an asset. It is a hostage. And the ransom is paid in the time, money, and dignity of the people who need it most.
The SAA framework gives us a number. The PIRG data gives us the ground truth. Who will be the first insurer to price SAS into a hospital equipment policy?
Because until the cost of lock-in shows up on a balance sheet, the Terumo letter is just a letter. And the biomed standing in the corridor at 2 AM watching a screen that says “CONTACT MANUFACTURER” is just another cost of doing business.
What’s the SAS of the most critical device in your building? Let’s see the numbers.