The Living Laboratory

To achieve systemic change at scale, the most successful health systems will treat their employees as what they already are: subject matter experts, working in the living laboratory of the healthcare enterprise — identifying the problems worth solving, forming their hypotheses, validating their solutions, and implementing them. Structured community engagement, inside a well-designed operating strategy, is not only a morale program. It is a critical component of the system.

Healthcare is among the largest and fastest-growing sectors of the American economy — and its largest employer. The Centers for Medicare and Medicaid Services put total U.S. health expenditures at $5.3 trillion in 2024: 18 percent of GDP, growing 7.2 percent in a single year, and projected to pass a fifth of the economy by 2034.

Driving the trend: the high cost of technological and pharmaceutical treatment, a surge in chronic illness, an aging population, and the heavy operating drag of outdated, wasteful, and inefficient business processes across provider systems.

And the spending does not buy outcomes. In the Commonwealth Fund's most recent ten-country comparison of high-income health systems, the United States ranked last — last overall, and last on health outcomes, with life expectancy more than four years below the ten-country average and the highest rates of preventable and treatable deaths. On the risk factors feeding those outcomes — obesity, chronic disease — the U.S. leads as well. Vast spending, poor results.

To characterize healthcare as stagnant would be wrong. Medical care is developing at a remarkable pace. The crisis is not clinical — it is operational: cost, complexity, and the human experience of working inside the system. While patient care advances, administrative processes, staff retention, and supply chain logistics move at a fraction of the speed.

What the industry requires is crisis intervention — a mandate for significant improvement in every part of the healthcare system beyond the clinical treatment of medical conditions.

The intervention begins with a change of venue: from the controlled mindset of the medical laboratory to the unpredictable living laboratory of the hospital floor. The method doesn't change — healthcare already trusts the scientific method: a testable hypothesis, validated through experimentation and data. What changes is where the hypotheses come from. The people experiencing the friction are the ones holding them. Structured engagement is how a health system collects those hypotheses at scale — from the front line, continuously.

Hospitals cannot run engagement the way a software company runs a hackathon. Clinical settings impose real limits: staff utilization inside patient-care pathways leaves little non-clinical time; hours are dynamic and unpredictable; units run one, two, and three shifts; workgroups sit centralized in one building and dispersed across a region. Around the clinical core sit HR, IT, legal, and finance — functions whose alignment is its own challenge.

The formats below exist because of those constraints, not despite them. Each is a working answer to the same question: how does a system take in signal from people who have no time to give it?

In a multi-disciplinary hospital network, how does a functional fix from a night-shift nurse get the same weight as a concept from the Chief of Surgery — without the evaluation collapsing into clinical hierarchy or a popularity contest?

The answer is an evaluation architecture, not an evaluation method. No single scoring system survives contact with bias, so the modules are composed — sequentially, in parallel, or selectively — so the flaws of each cancel the flaws of the others. Community evaluation lets employees read, comment, and rate: raw organizational enthusiasm surfaces, and the front line refines concepts organically. Pairwise voting replaces hundred-item ballots with rapid A-versus-B decisions: thousands of micro-choices across the enterprise float the strongest concepts with high mathematical accuracy. Token voting hands each participant a finite budget of digital currency: scarcity converts votes from applause into investment. AI evaluation clusters duplicates, finds semantic patterns, and scores raw text against strategic rubrics for an instantaneous, objective baseline. And expert review — clinical leads, IT directors, compliance owners — tests hard feasibility, regulation, and budget.

A high-volume challenge might run AI and pairwise scoring together to find the top ten percent before the community weighs in. A technical challenge might lead with expert review before the crowd allocates its tokens. The architecture is designed per problem — which is the point.

What it buys: bias reduction, psychological safety, mathematical accuracy, and the thing hierarchies never grant on their own — equal consideration.

Complex, multi-departmental bottlenecks — discharge delays, OR turnover — attract solutions to the wrong problem. Leadership solves what it assumes; the root cause lives on the floor.

The format is a two-phase discovery challenge. Phase one is a friction audit: instead of asking for ideas, leadership asks an observational question — "What is the single biggest barrier you face when trying to discharge a patient before noon?" Employees submit observations, pain points, and process breakdowns; the crowd filters and ranks them by frequency and impact. Thousands of individual frustrations synthesize into a validated hierarchy of systemic problems. Phase two is targeted solutioning: challenges launched against the top validated root causes. Because the front line defined the problem, the ideation that follows is focused, relevant, and immediately actionable.

Nothing in the system is more expensive than a solution to the wrong problem. This format is how an organization stops buying them.

Episodic, campaign-style challenges capture strategic input and miss daily operational friction — and nobody on a clinical floor wants another administrative layer.

The format is the distributed micro-challenge, embedded in routines that already exist: shift huddles, M&M conferences, departmental meetings. Staff submit observations from a phone or workstation during or just after a shift. Filtering runs inside the cohort — a single nursing unit reviews and ranks its own week of submissions. The highest-ranked items skip executive review entirely and land on the unit's own huddle agenda for resolution.

That last step is the doctrine in miniature: act at the lowest competent level. The unit that found the problem fixes the problem — and the system's cadence, not a committee's calendar, decides when.

In-person collaboration is the scarcest resource in a shift-based clinical environment — and most design frameworks squander it on brainstorming.

This format separates intake from design. Before anyone books a room, concepts are gathered, sorted, and selected asynchronously — broad participation during standard shifts, no disruption to clinical operations. The in-person session then starts where most workshops end: the highest-ranked concepts arrive pre-filtered, and the group spends its scarce hours prototyping and testing instead of generating and sorting. Afterward, the platform captures the prototypes and gathers enterprise feedback while teams report progress.

Collaboration time is spent on the one thing that requires a room: building.

None of these formats is a program to run once. They are components of a system — intake, evaluation, governance, cadence — and the system only produces if it is operated: run continuously, measured honestly, adjusted as the organization learns. The living laboratory is already staffed. The experts are already on the floor, holding the hypotheses. What most health systems are missing is not talent, and it is not ideas. It is the structured and operated system of change that lets the laboratory produce.