Center-Led in a Six-Hospital System

Sequencing principle. Do not announce a structure in the first 30 days. The 2019 attempt failed on credibility and ownership, not on design. Rebuild the fact base and the coalition before the org chart.

Days 1-30: facts and stakeholders.

• Build a spend cube from AP and group purchasing data, then validate it against each hospital's own ledger so no CFO can say the numbers are wrong. Publish price variance by hospital; the seven glove prices are your most persuasive single artifact.
• Interview all six CFOs, the CMO, the heads of OR and nursing, and every materials manager. Ask what went wrong in 2019 and write it down verbatim.
• Diagnose 2019 explicitly: a deal overridden with no escalation path, staff told to comply without training, and no local benefit. Circulate the diagnosis so it becomes a shared constraint rather than a rumor.

Days 31-60: coalition and rules.

• Stand up a steering council chaired by the CEO with you as secretariat; recruit the CMO and one skeptical CFO as members. A converted skeptic is the most credible messenger you can have.
• Publish the savings-sharing rule, for example 70% of verified savings back to the hospital P&L and 30% to fund the corporate function. That changes the incentive from losing control to gaining margin.
• Define decision rights before naming a structure: corporate owns strategy, contracts, and supplier selection above a threshold; hospitals own requisitioning, local service, and clinical input.

Days 61-90: prove it small. Three low-clinical-risk categories (gloves, sutures, office supplies) with a six-month target of $3-4 million, each hospital nominating one. Commit to no involuntary reductions in materials management in year one; redeploy instead.

What I would not do: publish a new RACI and reorganization chart in week two. That is how 2019 started.

The model. Center-led, not centralized. Corporate sets category strategy, contracting, and supplier governance; hospitals execute and own clinical choice. Write it into a one-page decision-rights matrix with named roles and dollar thresholds: who decides, who consults, who is informed.

Mechanisms that make it stick.

• Category councils with clinical representation for physician preference items (31% of spend). No PPI contract without clinician sign-off, which removes the done-to-us objection.
• Savings shared to the hospital P&L (70/30), verified by finance rather than by supply chain.
• Local service commitments: requisition-to-delivery SLAs and a named account manager per hospital. Centralization fails when the OR cannot get gloves at 2 a.m.
• A documented exception process. Any hospital may request an exception against written criteria such as clinical necessity or an existing capital commitment. Exceptions are logged and published quarterly, which gives autonomy without giving a veto.
• Capability: a six-month category management program for the existing materials managers, because 2019 told people to do a job they had not been trained for.
• Communication cadence: monthly CEO message, biweekly CFO call, quarterly town hall with a published savings scoreboard.

The two opposing CFOs, specifically. Meet separately before any group session and assume the objection is rational: they lose budget control, headcount, and the ability to solve problems locally. Offer each a council leadership role with real scope (one chairs the PPI council, one chairs savings verification), show them their own hospital's off-contract leakage, and give written guarantees on local service levels and no year-one involuntary reductions. If one still refuses, keep them informed and proceed, but never let them learn about a change from a vendor.

Adoption, leading indicators, at 6 months.

• Contracted spend share moving from 59% to 75%, measured from AP data, not self-report.
• Price variance within category under 3% across the six hospitals.
• Active suppliers from 1,900 to 1,500.
• Council attendance, exception-request volume (a falling number is not automatically good, it can mean people stopped asking), training completion, and a short pulse survey on whether local staff believe the process is faster or slower.

Realization, lagging indicators, at 12 months.

• Verified savings run-rate of $10 million, validated by finance against baseline prices rather than contract signatures.
• Percentage of PPI spend under a clinician-approved contract.
• Local SLA performance: fill rate and order-to-delivery.
• Materials management turnover. A spike means the change is being absorbed through attrition, which is a failure mode, not a saving.

Sustainment at 18 months.

• $19 million run-rate in budgets, not just in contracts.
• 85% contracted spend and 1,100 suppliers.
• Savings embedded in each hospital's budget so next year's baseline starts lower.
• Internal audit of a contract sample confirming price paid equals price contracted.

Design point. Adoption and outcome metrics must be tracked separately, because a change can look successful in a savings report while local staff quietly revert to off-contract buying. The honest test is price paid, supplier count, and whether a hospital CFO defends the model when you are not in the room.