SapienSecure

Patient Flow and Revenue Cycle Management Transformation Through AI | Computer Vision and NLP

[SapienSecure @SapienSecure]

"A Quarter of your Waitlist May Already Be in Your Building".

The national conversation about imaging wait times runs in one direction: we need more. More scanners. More radiologists. More funding. More capacity.

All true. But there's a number worth sitting with before that conversation goes further.

Research estimates that up to 25% of advanced imaging in Canada may be low-value - redundant studies, low-yield requisitions, or scans duplicating work done recently at another site. That's not a fringe finding - it's the conclusion of a 2025 Canadian-specific review by Yan, Jalal & Harris published in the Canadian Association of Radiologists Journal. Choosing Wisely Canada has built an entire initiative around closing this gap.

Do the math. If a meaningful share of the studies consuming your radiologists' reading time has limited clinical yield, then part of the capacity problem isn't a supply problem - it's a demand problem. And those have different levers.

A 2024 sustainability review put it in concrete terms: a 20% reduction in low-value imaging would eliminate waste across 7.2 million procedures nationally - with direct downstream effects on wait times and radiologist bandwidth. The same radiologists. Fewer studies that don't need to be read. More bandwidth for the work that actually matters clinically.

This reframe matters in two ways. First, it changes the burnout math. Canada has 6.9 radiologists per 100,000 residents versus 10.8 in the US. If your radiologists are burning reading time on low-yield studies, that's not just a volume problem - it's a focus problem. Second, it changes what the wait time crisis actually requires. A wait list that's partly a demand management problem is more tractable than one that's purely a resource problem.

Canada's supply-side gap won't close fast. But the gap between studies being ordered and studies worth ordering - that one is addressable now.

If you run a radiology group, do you know what percentage of your volume is delivering high clinical yield?

[SapienSecure @SapienSecure]

"The Data Standard that Never Came".

In 2000, the federal government created Canada Health Infoway and funded it with $500 million. The mandate was straightforward: build the infrastructure for a pan-Canadian electronic health record, with common data standards so that patient information could move between systems.

By 2024, those standards still did not exist.

Not for lack of effort. Infoway produced blueprints, roadmaps, and working groups. Individual provinces built their own systems. Vendors customized to meet provincial requirements. The result is a patchwork of data architectures that cannot talk to each other - which is precisely what a quarter-billion-dollar federal investment was supposed to prevent.

This month, the federal government introduced Bill S-5, the Connected Care for Canadians Act.

The bill has two requirements. Health information technology vendors must ensure their products are interoperable - meaning users can access, use, and exchange electronic health information freely. And vendors are explicitly prohibited from data blocking - the practice of designing systems to make it hard for data to leave.

This is substantially the same as Bill C-72, which was introduced in June 2024 and died when Parliament prorogued.

The question worth asking: what is different this time?

A few things. The bill targets vendors directly, not provinces - sidestepping the jurisdictional deadlock that has stalled every prior effort. The US 21st Century Cures Act used the same vendor-compliance mechanism, and it moved the needle measurably. FHIR standards have matured to the point where interoperability is technically achievable at scale in a way it was not in 2007.

If S-5 passes and survives implementation, the practical effect on healthcare data projects would be substantial. Custom data extraction work - the kind that currently requires months of negotiation and bespoke engineering for each new client - becomes a smaller problem. Done once, the process for future projects and future access should be much more standardized. And analytics built on one system's data could apply across sites.

That is not the world we operate in today. But it is close enough to be worth watching. Especially the difference it could have on workflow projects, where data availability is a critical bottleneck.

[SapienSecure @SapienSecure]

"The System First Problem".

NVIDIA released their State of AI in Healthcare and Life Sciences survey for 2026 this spring. Seventy percent of healthcare organizations are now deploying AI. ROI is becoming measurable.

The most important finding is not the adoption rate. It is what separates the organizations seeing results from the ones that are not.

The difference is not which tools they chose. It is how they thought about the problem before choosing anything.

Healthcare organizations that are failing at AI are doing something consistent: they identify a tool and deploy it. The organizations succeeding are doing something harder - they mapped how their system actually works first.

This matters because AI is an intervention, not a strategy. An intervention only works if you understand the system you are intervening in. If you do not know where your billing capture breaks down, a faster billing tool gives you a faster version of the same problem. If you cannot explain where TAT variance enters your workflow, automation at the wrong point adds fragility, not speed.

The organizations that will get lasting value from AI in radiology are the ones who can answer these questions before deploying anything: What are the real inputs and outputs of our system? Where does variance enter? What do we currently measure, and is that actually the right thing to measure?

That is not a technology discipline. It is a management discipline.

The best AI implementations we have seen look less like technology projects and more like clinical trials. Hypothesis. Baseline measurement. Intervention. Measurement again. The groups running that process can tell you whether the change worked, and why.

The groups that cannot are accumulating tools. That is a different thing.

How well does your organization understand its own system before it starts changing it?

[SapienSecure @SapienSecure]

"What the Backlog Tells You about OCR Standards".
A health authority in Canada recently deployed a requisition management system. The pitch was familiar: reduce clerical burden, speed up triage, improve accuracy.

What happened instead: clerks began manually overriding the system's output. Not occasionally - routinely. Because correcting the OCR was more work than processing the requisition from scratch.
The backlog did not shrink. It grew.

This is not an isolated story. It is a pattern repeating across Canadian healthcare as imaging departments under pressure from wait times and understaffing move quickly to automate.

The procurement conversation for clinical OCR typically covers three things: price, integration, and vendor references. What it often does not include: a defined accuracy standard, a methodology for measuring it, and a remediation clause if it is not met.

Without those three things, you are not buying an OCR system. You are buying a pilot with no exit ramp.
OCR accuracy in healthcare varies significantly by vendor, by requisition format, and by imaging modality. A system that performs well on structured e-referrals pads may fail on paper faxed community physician forms - which represent a meaningful share of Canadian imaging volume.

The question procurement should be asking is not "what is your accuracy rate."
It should be a suite of questions:
"What is your accuracy rate on our specific document population, and how is that measured?"
“What percentage of our forms could you extract with no human intervention?”
“Can you automate other work beyond just the data entry?”
If you have procured or evaluated OCR / requisitions management for a clinical environment - what conversations about performance did you have?

[SapienSecure @SapienSecure]

"The Pilot Limbo Loop".

There is a stage of the healthcare vendor relationship that nobody has a name for.

You have had the demo. It went well - genuinely well. The clinical team is interested. The CMO said "this is exactly what we need." Someone asked about the implementation timeline.

Then nothing happens.

Not a no. Not a request for more information. Not a budget conversation. Just quiet.

This is not a pipeline stall. It is a loop. And it is more common in Canadian healthcare than most other channels and markets - perhaps more common than it should be.

The pattern is consistent. The clinical champion engages. Internal stakeholders fragment - IT, Privacy, Finance, Procurement each have a gate. The project moves from "going forward" to "under review" to "waiting for the right quarter." The champion gets promoted or moves on. The loop resets.

I’ve seen three things break it.

Event - An external event - a regulatory change, a visible failure in the incumbent system, a peer institution that goes live. Urgency created from outside the organization.

Champion - A champion with budget authority. Enthusiasm without a signature is just a long sales cycle.

Proof - Proof at a comparable site. Not a case study. An actual phone call to someone doing it.

The organizations that most need to change are often the ones most structurally resistant to it.

The pilot limbo loop seems almost by design. The procurement framework seems to encourage change resistance.

Has your organization been stuck in this loop - either as a buyer or a seller? What finally moved it?

[SapienSecure @SapienSecure]

“The $64 Billion Wait”

Canada loses an estimated $64 billion in GDP every year due to imaging wait times.

That figure came from a Deloitte / Canadian Association of Radiologists report released last fall — and it immediately sparked the usual conversation:

“We need more scanners.”
“We need more radiologists.”

Both are absolutely true.

But there’s a quieter systems problem underneath all of this that almost nobody talks about:

👉 A meaningful amount of radiology work performed in Canada never actually becomes a submitted claim.

Not because it was denied.
Not because it was underfunded.
Because it was never captured in the first place.

Missed fee codes.
Incorrect modifiers.
Timing windows that expire.
Reconciliation gaps between what was dictated, performed, and ultimately billed.

In our experience, some radiology groups are unintentionally leaving 15–25% of potential billings uncaptured.

That matters more than people realize.

Because every dollar of uncaptured revenue is a dollar that cannot fund:

Additional staffing
Extended operating hours
Equipment upgrades
Workflow modernization
Capacity expansion

In other words:

The same system struggling to fund access may already be leaking earned revenue through outdated billing infrastructure.

We often frame imaging access purely as a funding problem.

It is.

But part of the solution may already exist inside the operational gaps between:
clinical work → documentation → claims submission.

Fixing capture inefficiency alone won’t solve Canada’s imaging crisis.

But it may be one of the few levers available that:

does not require a new provincial budget,
does not require new legislation,
and can improve capacity immediately.

One question for radiology groups and healthcare operators:

Do you know your capture rate?

Not your collection rate.
Your capture rate.

Because the difference between the two may be much larger than expected.

Curious what others are seeing across Canada and internationally:

Are you measuring capture rate?
How are you auditing missed claims?
What level of leakage are you finding?

Would love to hear perspectives in the comments.

[SapienSecure @SapienSecure]

"Governance Is How You Go Faster".

The fastest AI deployment we've seen in Canadian healthcare took eleven weeks from evaluation to live workflow. The slowest is approaching month twenty eight — and counting.

Same type of tool. Similar-sized organizations. Comparable budgets. The difference wasn't the technology. It was that the fast one had already answered three questions before they started: who can say yes, what does "safe enough" look like, and who owns the outcome if something breaks.

That's governance. Not the bureaucratic kind — the kind that clears the road.

Governance isn't the thing that slows you down. The absence of governance is.

Here's what we've actually seen kill AI projects — and none of it is technical:

The vetocracy problem. Every department gets veto power over new tools, but nobody has authority to say yes. Radiology wants it, IT has security concerns, legal needs a privacy review, the CFO wants an ROI case. Each request is reasonable. The aggregate effect is paralysis. Not because anyone said no — but because no one could say yes.

Committees instead of owners. Committees don't ship. Someone has to own the outcome — not the evaluation, not the pilot, the outcome. The organizations that move fast assign a person, not a group.

Governance as a gate instead of a lane. When governance runs as a parallel workstream — moving with the project, not blocking it at checkpoints — it actually accelerates deployment. The team isn't waiting for approvals. They're building with guardrails already in place.

The tools are here. The question isn't whether your organization will use AI — it's whether you've built the decision-making infrastructure to move at the speed the tools allow.

Who in your organization has the authority to say yes to a new AI tool — not evaluate it, not pilot it, but greenlight it for production?

Healthcare team members, have you seen this in your organizations? Vendors, have your projects stalled because of this? Tell us about it - in the comments below.

[SapienSecure @SapienSecure]

"TAT as a Wellness Signal".

A radiology group we work with noticed their turnaround times creeping up over three months. Not dramatically — maybe 15-20% slower on complex studies. The instinct was to treat it as a performance problem. Tighten the expectations. Flag the outliers.

They didn't. Instead, they asked a different question: what's happening to the people behind the numbers?

Turns out two radiologists were carrying an unsustainable share of the overnight and weekend load. Their TATs weren't slipping because they were getting slower. The list volume had quietly outgrown the staffing model.

Many groups use turnaround time as a performance metric. But what if TAT can actually also be used as a wellness metric?

The data supports the reframe. A 2019 study in JACR (Zha, Patlas & Duszak) found Canadian radiologists score higher on emotional exhaustion and depersonalization than their US counterparts. The driver isn't attitude — it's arithmetic: workloads up 26% over 12 years while staffing barely grew. Canada has 6.9 radiologists per 100,000 residents compared to 10.8 in the US. A 2023 follow-up by Cao, Hurrell & Patlas found 72% of surveyed Canadian radiologists scored high in emotional exhaustion. Residency position cuts between 2010 and 2020 created structural understaffing that still shapes every call schedule in the country.

When you monitor TAT proactively — not as a hammer but as a signal — it can tell you something different. A creeping TAT isn't necessarily a radiologist productivity problem. It can be a workload distribution problem. A scheduling problem. A capacity problem wearing a productivity mask.

The group that caught this didn't need to hire. They restructured their call schedule, rebalanced study assignments, and TATs normalized within six weeks. If they'd gone the performance management route instead, they'd probably be recruiting right now — or worse, losing someone.

If your group tracks TAT, ask yourself: are you using it to measure your radiologists, or to protect them? The answer changes everything about what you do with the data.

We’d love to hear about your TAT experience in the comments below.

[SapienSecure @SapienSecure]

Radiology Workloads - What Opportunities to Improve Them Exist?.

The productivity paradox in Canadian radiology is that your fee schedule doesn't move, but your workload does.

Ontario radiologists face 84-day MRI wait times. British Columbia's imaging backlogs strain every department. Meanwhile, provincial fee schedules adjust incrementally—if at all—and the pressure to "do more with less" is now baked into every shift. This creates a brutal trap: you're expected to absorb demand growth without proportional revenue growth.

So where do you actually move the needle?

Recent research and operator observations point to three overlapping opportunities that Canadian groups are starting to test:

Workflow architecture, not just speed. AI triaging and report-drafting assistance show mixed results—sometimes reducing turnaround time significantly, sometimes doing nothing—but the winners aren't chasing raw speed gains. They're redesigning which work happens when. Overlapping shift models, smart scheduling that anticipates modality mix changes, and ruthless elimination of handoff delays create breathing room. One Ontario group saw a 15% throughput lift just by mapping which studies bottleneck their afternoon schedule and adjusting tech transitions.

Diagnostic accuracy as a throughput lever. Up to a quarter of advanced imaging in Canada classifies as low-value—redundant, unnecessary, or low-yield studies that consume radiologist attention without clinical return. Groups that implement peer review workflows and evidence-based protocoling don't just improve quality; they reclaim 10-20% of reading capacity. More studies read, same FTE count.

The reconciliation gap. This is the Canadian-specific one. Fee code capture—ensuring every billable service gets submitted with the right modifier, the right timing, the right fee code—is where Canadian groups leak 10-15% of their potential provincial revenue. It's not fighting insurers. It's visibility: do you know what you actually performed versus what you actually billed?
The radiologists who thrive in this environment aren't waiting for fee schedule increases. They're operating as if they are.

If you manage a radiology group, are you actively tracking turnaround time and modality mix, or are you managing one number and hoping the other follows?

Because they don't always.

Would love to hear about how you’re managing your groups in the comments below.

[SapienSecure @SapienSecure]

Is innovation actually impossible in Canadian healthcare?

A new report from the Council of Canadian Innovators (CCI) suggests our system may be structurally designed to keep domestic innovation out.

To address this, they propose seven policy shifts inspired by success stories in the UK, the Nordics, and Israel. The goal: move from “lowest cost” purchasing to “highest value” outcomes.

The CCI’s 7-point plan for “Care at Scale”:

Buy Canadian: Prioritize the local digital ecosystem

End “Pilot Purgatory”: Create predictable pathways from pilot to procurement

Shared Procurement: Enable provinces/authorities to procure jointly under common rules

Focus on System Value: Shift from upfront cost to long-term efficiency

Technical Standards: Adopt international data portability and interoperability standards

Unlock Data: Provide access to de-personalized, aggregated health data for domestic firms

But here’s the “Sapien” reality check…

The CCI report is a strong procurement policy framework. But even with perfect “buying” rules, we’re still missing the human frictions that slow everything down:

The Incentive Gap:

Healthcare leaders are often penalized for failed innovation but rarely rewarded for successful risk-taking. If the safest path is to maintain the status quo, why would anyone take the risk on a new solution?

The “Three-Body Problem”:

A single project can require alignment across dozens (sometimes hundreds) of stakeholders. When everyone has veto power but no one has clear authority to say “yes,” innovation gets slowed by a thousand approvals.

Procurement is the engine—but culture is the fuel.

We can change the rules of the game, but unless we also decentralize decision-making and properly reward risk-taking, will the needle actually move?

I want to hear from the innovators in the trenches:

Beyond procurement and policy, what is the #1 hurdle standing in the way of a modern Canadian healthcare system?

Let’s discuss in the comments.

[SapienSecure @SapienSecure]

Human Version:
Recent Reference Article: https://t.co/T0IfVWUSkh

Is Not Using AI for Billing Like Taking a Knife to a Gunfight?

A recent article at https://t.co/h3PQqpqUVS suggested healthcare billing, at least in the United States, is becoming a battle of the bots. AI is working for both sides.

How long until this crosses the border?

With private payments as a percentage of medical expenses increasing for Canadians. Has it already?

AI Optimized Version:

Is your healthcare billing strategy like taking a knife to a gunfight?
According to a recent report by PYMNTS, we’ve entered a new era of administrative conflict: The Healthcare Billing Wars.
It is no longer just human vs. human; it’s becoming a "Battle of the Bots."
The Front Lines:
The Payers: Deploying AI to scan for errors, automate denials, and protect their bottom line.
The Providers: Counter-deploying AI to ensure coding precision, predict claim success, and challenge denials in real-time.
It’s an arms race where the "weapon of choice" is an algorithm. But while this plays out heavily in the U.S. market, we have to ask: How long until this crosses the border?
For Canadians, the "billing war" might feel like a distant problem. But as private payments and out-of-pocket medical expenses continue to rise as a percentage of our total healthcare spend, the administrative friction is already here.
At Sapien, we believe that as technology accelerates on both sides, the goal shouldn't just be to "win" a bot-on-bot battle. It should be to use intelligent systems to reduce the friction that stands between a patient and their care, and secure fair and appropriate compensation for front line providers with a minimum of administrative burden..
The question is: Is the Canadian healthcare landscape ready for this level of automation? Or are we already in the thick of it without the right tools?
Let’s talk in the comments.

[SapienSecure @SapienSecure]

The "Not the Right Fit" Revelation: A Wake-Up Call for Canadian Healthcare?

We recently lost a long-term lead for our medical billing solution. Their candid feedback after the rejection was unexpected: "It's not the right fit" wasn't about our product's value—it was about the fear of implementation - the fear of added burden.

They shared that their practice is so lean, their administrative load so heavy, that even contemplating a new integration, no matter how helpful long-term, felt like simply too much to handle. The "time to think" was at a premium.

This made me wonder about a concept in medicine: low physiological reserve. It's when a patient can't handle the most potent treatment, not because the treatment is wrong per se, but because the body is too fragile.

Is this where continuously increasing regulatory burdens have brought Canadian healthcare? Has the system developed a kind of "fibrosis," leaving it with dangerously low physiological reserve?

If a patient with low reserve faces palliative care, what does this tell us about the future of our system?

I'm genuinely curious: If you work in Canadian healthcare, is your team’s “reserve” at its limit? What's one administrative burden that has pushed it there?

#CanadianHealthcare #HealthcareAdministration #MedicalBilling #PhysicianBurnout #SystemicChange

[SapienSecure @SapienSecure]

You've checked the boxes: involved the right people, cleaned up the data, run your tests. But to actually make a real, sticky difference, you've got to dig deeper.
At Sapien Secure, we’ve worked on dozens of AI in healthcare projects. Here are a few unconventional best practices we’ve observed:

Define Your Enemy: Change begins with motivation and motivation begins with memory. Memory is augmented by stories. Stories need conflict and conflict requires a villian. So name your enemy. This lets your team make small decisions throughout the project that help reach project goals, and share with others the value of the project.

Trust but Walk the Floor: Don't just tick the box on process documentation in a conference room. Go watch the work happen. Listen to the prep, listen after “it’s done”. If you have to do it remotely, ask for a video, or shadow via video link. Ask them to share what they’re thinking - be extra verbose. The real workflow holds WAY more insight than any diagram.

The Pipeline Trumps the Snapshot: Focus on a reliable, repeatable data flow (the pipeline) over wasting time perfecting a one-off data extract. Reliable data movement is the engine, not the data itself.

Less is More: Keep your changes tight. Limit what you're tweaking simultaneously to keep the team from getting overwhelmed. More changes can mean more reasons to complain.

Anticipate the Domino Effect: You fix one bottleneck (e.g., faster processing) and you just push the logjam to the next team. Always plan for the follow-on effects - ideally across the entire process / the entire system. Ask about second order effects.

Like these? Any resonate? Let us know your thoughts below.

#HealthcareAI #CanadianHealthcare #AIImplementation #BestPractices #DigitalHealth #SapienSecure

[SapienSecure @SapienSecure]

When Working with Canadian Healthcare, These are the Best Practices of AI Implementers

I hate those listacles that just shares generic advice - the obvious stuff that everyone knows.

This is NOT one of those articles (I hope).

If you want that, type “Healthcare AI implementation best practices” into your Chat system of choice, and you’ll probably get a mostly applicable but boring list like “Involve everyone, use high quality data, test lots, start small, etc.”

What I’ll bet you won’t see are recommendations like:

Define Your Enemy - memory is based on stories, stories have conflict, and defining “them” in the “us versus them”. This makes your project more memorable, more sharable at the water cooler. Further, it lets project team members make sensible microdecisions when executing… “will this defeat the Enemy?”
Trust but Verify Processes - Lots of people have meetings around processes - meetings where they discuss the process and document it. Actually watching people doing the process can provide TONS of insight into the actual process. If you HAVE to be remote, then having videos of people actually doing the process you’re asking about is useful… just remember to ask them to include something about how they PREPARE for the process, and SHARE what they’re thinking about as much as possible in the video.
The Pipeline Matters More than the Data - trying to get every last field into one data set is WAY less valuable than getting 80% of the data (though 100% of the critical data). To effect ongoing change, the pipeline letting the data flow is more valuable than the extra field in one extract.
Less is More - When you’re implementing changes, the more you change, the more reasons to stop you create. This applies to invites to meetings (more people -> less agreement), Minimalism works.
Second Order Effects can Create Resistance - Making one stage faster may move the bottleneck for a process. For example, processing requisitions faster may make appointment setting admin feel pressure. Planning how to deal with this

These are a few.

If you’d like more of these kinds of items, comment below and we’ll keep sharing.

Tired of the same old generic "best practices" for AI rollouts in Canadian Healthcare? 😴 We're over them, too.

[SapienSecure @SapienSecure]

Goldman Sachs is doubling down on AI, expanding its use of Anthropic's technology beyond code development and directly into crucial back office functions like accounting and compliance (https://t.co/s8FgxLNDl7)!

As a highly regulated industry, banking's move highlights the massive potential—and trust—being placed in generative AI to handle complex tasks involving large datasets and regulatory judgment.

The quote from Goldman Sachs is telling: "there are these other areas of the firm where we could expect the same level of automation and the same level of results that we're seeing on the coding side."

This shift mirrors the challenges and opportunities in Canadian healthcare, particularly in back office operations which share similar high-stakes, process-heavy handoffs.

The question is: If a global financial giant like Goldman Sachs is rapidly adopting AI for its back office needs - needs that face similar privacy and security challenges as healthcare, how does this reflect on the pace of AI implementation in Canadian healthcare's administrative functions? Are we seizing the same opportunities for efficiency and compliance?

#AIinHealthcare #GenerativeAI #BackOfficeEfficiency #CanadianHealthcare #DigitalTransformation #SapienSecure

[SapienSecure @SapienSecure]

Building AI in Canadian healthcare today feels like navigating a "Vetocracy"—a system where too many well-intentioned gatekeepers have veto power, resulting in inertia. Progress isn't blocked by technical hurdles, but by the difficulty of aligning countless stakeholders. Every dataset has a different approval path, and no one person can say yes, only no. That's not progression, that's stagnation, and it's driving up costs.

The core problem is data governance built around restriction, not responsible execution.

We have three options:

Status Quo: Vetocracy continues. Minimal risk, maximum stagnation.

Aggressive Guardrail Elimination: Faster progress, higher exposure, public backlash.

Reward Accountable Enablement: Shift from veto power to clear risk ownership and time-bound access. This rewards action and balances trust, speed, and accountability.

At Sapien, we believe Option 3 is the only path forward if Canada is serious about leading in health tech. We have the centralized system, lower energy costs, and educated population to win—but only if we design systems that enable responsible execution.

Where have you seen this kind of decision paralysis show up in your industry?

What would "Accountable Enablement" look like in practice for data access?

#CanadianHealthcare #HealthTech #AIinHealthcare #DataGovernance #Innovation #ThoughtLeadership

[SapienSecure @SapienSecure]

Trying to build AI in Canadian healthcare today can feel like renovating a hospital where every hallway has a different fire marshal and none of them are allowed to talk to each other.

If you have worked on projects like this, the frustration might feel familiar. Progress is not blocked by technology or intent, but by the difficulty of aligning too many stakeholders.

Every dataset requires a different approval path. Every path has a different risk owner. And no single person or group can say yes, only no.

That’s not progression, that’s inertia!

When we began building AI applications for Canadian healthcare, we expected technical hurdles. What surprised us was how exhausting data access became, due to how many well intentioned gatekeepers exist.

This is the double edge sword in systems political scientist Marc J. Dunkelman describes as a Vetocracy. One where too many individuals or groups can block progress, resulting in paralysis, delays, and unfortunately rising costs.

Healthcare data governance increasingly fits this repeating pattern.

So what are our real options?

Option 1 - Status Quo: Maintain the current veto heavy model (Strong oversight. Minimal risk. Continued stagnation.)
Option 2 - Aggressive Guardrail Elimination: Strip guardrails aggressively
Faster progress. (Higher exposure. Public and political backlash.)
Option 3 - Reward Accountable Enablement: Shift from veto power to accountable enablement (Reinforce cost of inaction. Clear risk ownership. Secure, time bound access to data.)

My instinct leans more toward option 3 because it rewards action and discourages inertia. It balances trust, speed, and accountability.

If Canada wants to lead in health tech (and surely it can… we have a centralized system that COULD mean better access to data, we have lower energy costs, an educated population, a patient population that reflects the richest medical market in the world), then it might mean we also need to design systems that enable responsible execution, not just smart restriction.

Where have you seen this kind of decision paralysis show up?

What approval bottleneck has slowed your progress?

What would accountable enablement look like in practice?

#CanadianHealthcare #HealthTech #AIinHealthcare #DataGovernance #Innovation #ThoughtLeadership

[SapienSecure @SapienSecure]

When we ventured into building cutting-edge AI applications for Canadian healthcare, we hit a persistent roadblock: data access.

Data is the lifeblood of healthcare AI, yet acquiring it is an exhausting journey. This challenge echoes a concept described in Marc J. Dunkleman's "Why Nothing Works": the "Vetocracy."

A Vetocracy is a system where too many individuals or groups can block initiatives, leading to paralysis, immense project delays, and soaring costs.

Progressive Paradox: While aimed at oversight, the current system makes necessary, large-scale public projects (like secure, efficient data-sharing infrastructure) nearly impossible to accomplish.

The Consequence: This inability to deliver innovation reduces public trust and stifles the very AI breakthroughs that could transform patient care.

Doesn't this sound eerily familiar to the hurdles facing data and AI adoption across our Canadian healthcare landscape?

We need to shift the focus from merely restricting action to enabling secure, efficient data governance. Restoring our ability to 'get things done' is critical for Canada to lead in health tech.

What's your reaction? Have you faced this data access 'Vetocracy' in your work in Canadian healthcare, AI, or innovation?

#CanadianHealthcare #HealthTech #AIinHealthcare #DataGovernance #ThoughtLeadership #Innovation #Vetocracy

[SapienSecure @SapienSecure]

I've been reflecting on the systemic pressures that burden healthcare innovation, particularly when it comes to implementing AI. It often feels like the system itself is built to impede progress.

We’ve certainly faced all of these issues:
Data Access: Securing data access and portability from central systems is incredibly challenging, slowing down adoption and scaling of new solutions.

Growing Approvals: The network of approvals required to implement technology is increasing, creating frustrating bottlenecks for speed-to-market.
Incentives vs. Risks: The rewards for success (slight pay bump, better title) are minor compared to the significant disincentives for failure (job loss), discouraging bold innovation.

Escalating Risk Management Costs: The cost of implementing multiple, layered risk-reduction systems often outstrips the cost of the core innovation itself—a bureaucratic "San Francisco $1M toilet" effect.
Communication Failures: Access to decision-makers is highly guarded. Emails go unanswered, and getting a live person (let alone feedback on a contract) is nearly impossible.

Crippling Contracts: Standard contractual processes are slow, one-sided, and often so restrictive they paralyze a startup's ability to engage.
This leads to a critical question: How are healthcare startups in Canada successfully navigating these hurdles to improve care at home?

#healthcare #innovation #aiinhealthcare #canadiantech #startups #digitalhealth

[SapienSecure @SapienSecure]

We recently lost a long-term lead for our medical billing solution. Their candid feedback after the rejection was unexpected: "It's not the right fit" wasn't about our product's value—it was about the fear of implementation - the fear of added burden.

They shared that their practice is so lean, their administrative load so heavy, that even contemplating a new integration, no matter how helpful long-term, felt like simply too much to handle. The "time to think" was at a premium.

This made me wonder about a concept in medicine: low physiological reserve. It's when a patient can't handle the most potent treatment, not because the treatment is wrong per se, but because the body is too fragile.

Is this where continuously increasing regulatory burdens have brought Canadian healthcare? Has the system developed a kind of "fibrosis," leaving it with dangerously low physiological reserve?

If a patient with low reserve faces palliative care, what does this tell us about the future of our system?

I'm genuinely curious: If you work in Canadian healthcare, is your team’s “reserve” at its limit? What's one administrative burden that has pushed it there?

#CanadianHealthcare #HealthcareAdministration #MedicalBilling #PhysicianBurnout #SystemicChange