How'd you fix Harris Health's revenue issues in 2026?

What it is and why it matters

Harris Health System is the public hospital district serving Harris County, Texas — the Houston metro, one of the largest and least-insured urban populations in the United States. It operates Ben Taub Hospital (a Level I trauma center), Lyndon B. Johnson Hospital, and a network of community health centers and specialty clinics. Its revenue problem is structurally different from a for-profit health system's, and that difference is the whole story. A commercial system optimizes for payer mix, service-line margin, and market share. A safety-net district cannot optimize for payer mix at all, because its charter is to treat everyone who walks in regardless of ability to pay. The revenue levers that remain are operational, and they are unglamorous: how fast a claim gets out the door, how completely a service gets documented, how many patients move through a fixed bed footprint, and how much supplemental government funding the system successfully captures.

That reframing matters because the wrong diagnosis produces the wrong fix. Boards under pressure reach for headcount reduction or service-line closure, which in a safety-net context reduces revenue faster than it reduces cost — you shed the billable encounters and keep the fixed plant. The right diagnosis says: the encounters are already happening. The question is what fraction of them convert into collected dollars, and how much of the cost of the unfunded remainder gets offset by supplemental programs designed for exactly this purpose.

The payer mix is the constraint you design around, not the one you fix. A large share of Harris Health's volume is uninsured or Medicaid, and Texas has not expanded Medicaid under the Affordable Care Act — meaning a working adult below the poverty line in Harris County frequently has no coverage pathway at all. That population still gets sick, still comes to Ben Taub, and still consumes trauma bays, imaging, operating room time, and inpatient days. There is no billing sophistication that turns an uninsured trauma admission into commercial reimbursement. What exists instead is a stack of supplemental funding mechanisms — Medicaid disproportionate share hospital payments, uncompensated care pools, directed payment programs under the state's Section 1115 demonstration waiver, and the 340B drug pricing program — each of which requires documentation, cost reporting, and attestation work to capture fully. A district that runs those programs at 70% of achievable capture is leaving eight figures on the table without a single new patient.

There is a second reason this matters beyond Harris County. Every large public health system in a non-expansion state — Parkland in Dallas, Grady in Atlanta, Jackson Health in Miami — faces a structurally identical problem, and the operational playbook transfers almost cleanly. So does a surprising amount of it to rural critical access hospitals, tribal health systems, and federally qualified health centers, which share the same pattern of high unfunded demand against a fixed cost base with government supplemental programs as the balancing mechanism. If you understand the safety-net revenue cycle, you understand a large fraction of American healthcare finance that never appears in a commercial-payer case study.

The final piece of "why it matters" is timing risk. Supplemental funding is politically negotiated and renewed on multi-year cycles. Waiver programs get extended, restructured, or reduced depending on federal-state negotiation. A system that has built its operating budget on a supplemental payment stream without a parallel effort to improve core collection efficiency is one waiver renegotiation away from a crisis. The operational work is not just upside — it is insurance against a funding cliff you do not control.

The step-by-step process

The sequence below works because each stage funds the next. Cash freed in the first ninety days pays for the analytics and staffing needed for the harder structural work later. Do not start with value-based contracting; it takes eighteen months to produce a dollar and you need cash in ninety days.

Stage one: measure the actual leak, not the assumed one (weeks one through four). Before changing anything, pull a stratified sample of denied and zero-paid claims across the last twelve months — several hundred at minimum, sampled across payer, service line, and facility so the results are not dominated by one clinic's habits. Code each denial to root cause: eligibility not verified at registration, prior authorization missing or expired, medical necessity documentation insufficient, coding error, untimely filing, coordination-of-benefits failure, or true non-coverage. In most systems that have not done this recently, the distribution surprises leadership. Front-end registration errors — wrong plan, wrong subscriber ID, coverage terminated — typically account for a large plurality of denials and are the cheapest to fix, because they require training and a real-time eligibility check rather than clinical documentation change. Untimely filing denials are pure self-inflicted loss and should be near zero; if they are not, you have a workflow queue that nobody is working.

Stage two: fix the front end (weeks two through twelve). Real-time insurance eligibility verification at registration, run again at the point of service, catches coverage that lapsed between scheduling and arrival. Financial counseling embedded in the emergency department matters enormously in a safety-net setting, because a meaningful share of "uninsured" patients are actually eligible for Medicaid, CHIP, marketplace coverage, or the district's own financial assistance program and have never been enrolled. Presumptive eligibility screening and on-site Medicaid enrollment assistance convert unfunded encounters into funded ones retroactively — this is the single highest-return front-end intervention available to a public district, and it is a staffing decision more than a technology one. Every patient enrolled is not just this encounter's revenue; it is every subsequent encounter's revenue, plus a primary care relationship that keeps them out of the emergency department.

Stage three: close charge capture in the clinical workflow (weeks four through twenty). Charge capture leakage happens where clinical documentation and billing diverge: implants and high-cost supplies not scanned, observation hours not documented, infusion start and stop times missing, evaluation and management levels undercoded because the note does not support the acuity actually treated. The fix is not a nag campaign. It is charge reconciliation reports that compare clinical activity against posted charges by department daily, plus clinical documentation improvement specialists working concurrently — while the patient is still admitted — rather than retrospectively querying a physician three weeks after discharge, when the answer will be a shrug. Concurrent CDI review also improves case mix index accuracy, which matters for both reimbursement and publicly reported severity-adjusted quality metrics.

Stage four: unclog throughput (runs in parallel, weeks one through fifty-two). Emergency department boarding is a revenue problem disguised as a clinical one. When admitted patients occupy emergency department beds waiting for an inpatient bed, the department cannot accept new patients, ambulances divert, and low-acuity patients leave without being seen. Every one of those is a lost billable encounter against fixed cost. The interventions are well-documented in operations literature: discharge-before-noon targets, physician rounding sequenced so discharge orders are written early, environmental services turnaround time measured in minutes rather than shifts, a discharge lounge so a discharged patient stops occupying a bed while waiting for a ride, and a fast-track or split-flow model that routes low-acuity patients away from main department beds entirely.

Stage five: capture every supplemental dollar (weeks eight through fifty-two, then permanently). This is finance and compliance work, not clinical. It means accurate Medicare and Medicaid cost report preparation, complete DSH survey documentation, rigorous 340B program compliance and contract pharmacy management, and full participation in whatever directed payment and uncompensated care programs the state currently operates. Under-documentation here is invisible — nobody sends you a denial letter for supplemental funding you failed to substantiate.

Costs, timelines, and typical ranges

Be honest about magnitudes, because false precision is how these projects lose credibility with a board. The public data available on any given system — audited financial statements, cost reports, bond disclosures — supports directional estimates, not decimal-point forecasts. What follows is the shape of the investment, not a quote.

Timeline reality. Denial root-cause analysis produces findings in four to six weeks. Front-end eligibility fixes show measurable clean-claim-rate improvement within one quarter, because the effect is immediate — a claim submitted with correct coverage today pays in thirty days instead of denying in fifteen and re-billing in sixty. Charge capture and CDI improvements take two to three quarters to show in net revenue, because they flow through coding, billing, and payment lag before appearing. Throughput redesign is the slowest cultural change and the most durable, typically nine to eighteen months to embed. Supplemental program optimization follows the cost-reporting calendar, meaning improvements may not settle until a subsequent fiscal year.

Accounts receivable days as the headline metric. AR days is the number a board understands. Industry benchmark discussion generally treats the mid-forties as strong performance and anything above sixty as a signal of process failure. The cash value of improvement is arithmetic: one day of AR equals roughly one day of net patient revenue. For a system with net patient revenue in the low billions, that makes each day of AR reduction worth several million dollars in one-time cash release. Pulling AR down by ten to fifteen days is a realistic target for a system starting from the high sixties, and the resulting cash is real, non-recurring, and immediately usable — but say "non-recurring" out loud, because a board that mistakes a working-capital release for recurring margin will budget against it and get burned.

Where the recurring money is. Recurring improvement comes from three places: a higher clean claim rate (fewer denials means less rework cost and less write-off), better charge capture and coding accuracy (the same service documented completely reimburses more), and higher throughput (more encounters against the same fixed cost base). Each of these is typically expressed in low single-digit percentages of net patient revenue — which sounds modest until you multiply by a multi-billion-dollar base.

What it costs to get there. The spend falls into four buckets. Technology: eligibility verification, claim scrubbing, denial analytics, and patient-flow or bed-management software are generally annual subscription costs measured in hundreds of thousands to low millions, depending on scope and existing EHR capability. Consulting or interim revenue cycle management: outsourced RCM engagements are typically priced as a percentage of collections or as a fixed fee for a defined scope, and are worth it only if internal capability genuinely does not exist — the risk is renting expertise permanently instead of building it. Staffing: certified coders, CDI specialists, financial counselors, and denial analysts are hires, not projects, and each additional financial counselor in the emergency department has a direct enrollment-conversion return. Change management: training, workflow redesign, and physician engagement are the line items most often underfunded and most often the reason a technically sound project fails.

Payback discipline. Front-end fixes pay back fastest, often within two to three quarters. Charge capture and CDI investments typically pay back within a year. Throughput work has the longest horizon but the highest ceiling, because it lifts the volume across which every fixed cost is spread. Value-based contracting and population health infrastructure are multi-year investments — right to start, wrong to count on inside a twelve-month turnaround plan.

The adjacent economics. Two things move alongside the revenue cycle and deserve budget attention. First, 340B: for a safety-net system, drug pricing program savings are a material margin contributor, and both compliance risk and contract pharmacy management directly affect the amount retained. Second, the county tax levy — a public district's ad valorem revenue is set through a political process on an annual calendar, and any credible financial plan has to model scenarios where the levy is flat or reduced while demand grows.

Where teams get it wrong

Mistake one: attacking the payer mix. The most common strategic error is a "commercial growth strategy" that tries to compete with the Texas Medical Center's private systems for insured patients. The TMC is the largest medical complex in the world, with academic brand equity, physician networks, and facilities a public district will not out-market. Chasing that volume spends money on marketing and service-line expansion for share you will not win, while the actual revenue leak — the money already earned but not collected — sits untouched. Modest commercial gains are possible through narrow-network or direct-to-employer arrangements built on genuine differentiators like trauma capability, but framing them as the centerpiece of a turnaround is a misallocation.

Mistake two: cutting revenue cycle headcount to hit a cost target. Billing and coding staff are the people who convert care into cash. Reducing them lowers salary expense on a spreadsheet while raising denial rates, aging AR, and write-offs — a net loss that shows up two quarters later, after whoever made the decision has already reported the savings. If revenue cycle staff are expensive, the fix is making the work easier through automation and front-end accuracy so the same team clears more volume, not making the team smaller and the queue longer.

Mistake three: treating emergency department crowding as purely a clinical or facilities issue. Boarding is caused downstream — by inpatient discharge timing, bed turnover, and post-acute placement delays — not by the emergency department itself. Systems that respond by adding emergency department beds discover the new beds fill with boarders too. The lever is inpatient discharge process and bed turnaround, which is why throughput initiatives owned solely by emergency department leadership almost always fail. It needs a hospital-wide owner with authority over inpatient units.

Mistake four: buying software instead of fixing process. A denial analytics platform tells you what is denying. It does not work the denial. If nobody owns the rework queue with a named accountable person and a service-level agreement on appeal turnaround, the dashboard becomes a very expensive report on a problem that continues. The same is true for bed-management software with no discharge process behind it.

Mistake five: retrospective clinical documentation improvement. Querying a physician weeks after discharge yields low response rates and poor-quality answers. Concurrent review — while the patient is admitted and the clinical picture is fresh — produces materially better capture and does not feel like an audit to the physician.

Mistake six: ignoring the patient-experience side of collections. Aggressive collection practices against a low-income population generate political blowback for a public district, damage community trust, and produce very little cash. The higher-yield path is enrollment assistance and clear financial assistance policy — converting a bad debt write-off into a Medicaid payment beats pursuing an uncollectible balance.

Mistake seven: no single accountable owner. Revenue cycle work spans registration, clinical documentation, coding, billing, follow-up, and finance — five departments with five reporting lines. Without an executive owner who can compel cooperation across all of them, each fixes its own metric and the end-to-end result does not move. This is the RevOps discipline applied to healthcare: one owner, one instrumented pipeline, one definition of the numbers, and shared accountability for the outcome rather than departmental optimization that cancels out at the seams.

Mistake eight: measuring the wrong thing. Gross charges are meaningless — they reflect a chargemaster nobody pays. Track net patient revenue, cash collections as a percentage of net revenue, clean claim rate, denial rate by root cause, AR days, and cost to collect. If leadership reviews gross charges, the reporting itself is part of the problem.

Decision framework: when to choose what

Sequencing is the whole discipline. The framework below routes on two questions: how urgent is the cash need, and does internal capability exist to execute?

If cash is needed within ninety days, the only real options are working-capital levers: aggressive denial follow-up on aged AR, appeal of recoverable denials before timely-filing windows close, front-end eligibility fixes, and acceleration of any supplemental payment already earned but not yet claimed. Do not launch a value-based care initiative to solve a ninety-day cash problem.

If the horizon is six to twelve months, charge capture and CDI become the highest-return work, alongside throughput redesign. Both take a quarter or more to show in reported revenue but produce recurring, not one-time, improvement.

If the horizon is eighteen months or more, structural work becomes viable: value-based and managed-care contracting, population health infrastructure for high-utilizer cohorts, ambulatory capacity expansion that diverts non-emergent volume out of the emergency department, and community health worker programs that reduce avoidable readmissions.

On build versus buy: outsource when the capability gap is genuine and the timeline is urgent — an experienced RCM partner brings process and staffing faster than hiring can. Build when the function is core and recurring, which revenue cycle is. The healthy pattern is time-boxed outsourcing with explicit knowledge transfer, so you are renting expertise while building it, not permanently. Watch the contract structure: percentage-of-collections pricing aligns incentives on collection rate but not on cost to collect.

On payer strategy: for a safety-net system, supplemental government funding capture is nearly always higher return than commercial contract negotiation, because leverage in commercial negotiation comes from network essentiality, which a district competing against the TMC largely lacks in elective service lines. It does have leverage in trauma, burn, and other essential services no commercial competitor wants to staff — negotiate hard there and accept that elective orthopedics is not your fight.

Related questions

Does Medicaid expansion in Texas fix this automatically?

It would materially reduce the uninsured share and convert unfunded encounters into reimbursed ones, but it is a state legislative decision outside the system's control. It also would not fix denial rates, charge capture gaps, or throughput. Expansion changes the payer mix; operational discipline still determines how much of that mix gets collected.

How is a public hospital district's revenue different from a nonprofit health system's?

A district receives ad valorem property tax revenue set through a county political process, and it has a statutory obligation to treat indigent residents. That creates a funding floor a private nonprofit lacks, and a demand obligation it also lacks. The result is more predictable base funding but far less ability to shape volume or payer mix.

What single metric best signals revenue cycle health?

Cash collections as a percentage of net patient revenue, tracked monthly against a rolling trend. AR days is the more common headline, but it can be improved temporarily by writing off aged balances. Collections against net revenue is harder to manipulate and reflects what actually landed in the bank.

Can AI meaningfully reduce denials?

Automation genuinely helps with eligibility checking, claim scrubbing, denial pattern detection, and prioritizing which appeals are worth working. It does not replace the clinical documentation judgment behind medical necessity denials, and deploying it over a broken process just produces faster failure. Fix the workflow, then automate it.

Does this playbook apply outside Harris County?

Largely yes. Parkland, Grady, Jackson Health, and rural critical access hospitals face the same structure: high unfunded demand, fixed cost base, supplemental government programs as the balancing mechanism. The specific waiver programs and state Medicaid rules differ; the operational sequence does not.

FAQ

Why start with denials instead of contract negotiation?

Denials are entirely within your control and produce cash in the current quarter. Contract negotiation depends on a counterparty, follows a renewal calendar you do not set, and requires leverage you may not have. In a safety-net system where most revenue comes from government payers with administratively set rates, negotiation upside is limited anyway. Fix what you own first.

How much of the revenue issue is genuinely structural versus operational?

Roughly speaking, the payer mix and the uninsured share are structural — no operational excellence converts an uninsured trauma case into commercial reimbursement. But the fraction of billable, funded care that fails to convert into collected cash is entirely operational, and in most systems that have not recently invested in revenue cycle, that fraction is larger than leadership assumes. The honest framing is that structure sets the ceiling and operations determine how close you get to it.

What does RevOps mean in a hospital context?

The same thing it means anywhere: one accountable owner for the end-to-end pipeline from first touch to collected cash, one instrumented set of metrics everyone agrees on, and process design that crosses departmental boundaries. In healthcare the pipeline runs registration to eligibility to clinical documentation to coding to claim to payment to appeal. Every handoff is a leak point, and departmental optimization at each stage reliably produces a worse end-to-end result.

Is outsourcing the revenue cycle a good idea for a public district?

It can be, under two conditions: the capability gap is real rather than a budget-cycle artifact, and the contract includes explicit knowledge transfer with a defined end state. Outsourcing to avoid building internal capability creates permanent dependency on a vendor whose incentives are not identical to yours. Time-boxed, scoped engagements with performance measures on clean claim rate and cost to collect work better than open-ended management agreements.

How do you protect frontline staff during a revenue turnaround?

Frame it accurately, because it is accurate: the goal is fewer denials, less rework, and less appeal fatigue, not fewer people. Billing staff working a large denial queue are being asked to absorb the cost of upstream errors, and fixing registration accuracy makes their work materially better. Redeployment into financial counseling and enrollment assistance both improves the employee's job and adds revenue, which is a rare alignment worth using.

What happens if supplemental funding gets cut mid-turnaround?

Model it before it happens. Any credible plan should include a scenario with reduced waiver or uncompensated care funding and a flat tax levy, and should identify which operational improvements are far enough along to absorb part of the gap. This is the strongest argument for doing the operational work even when supplemental funding is currently adequate: it is the only part of the equation you control when the political part moves.

Sources

• https://www.cms.gov/medicare/payment/prospective-payment-systems
• https://www.medicaid.gov/medicaid/section-1115-demonstrations/index.html
• https://www.hfma.org/revenue-cycle/
• https://www.aha.org/costsofcaring
• https://www.hhs.texas.gov/services/health/medicaid-chip
• https://www.kff.org/status-of-state-medicaid-expansion-decisions/
• https://www.hrsa.gov/opa
• https://www.macpac.gov/subtopic/disproportionate-share-hospital-payments/
• https://www.ahrq.gov/patient-safety/settings/emergency-dept/index.html
• https://www.harrishealth.org/

Related on PULSE

• How'd you fix Henry Ford Health's revenue issues in 2026?
• How'd you fix Forward Health's revenue issues in 2026?
• How'd you fix Blackbird Health's revenue issues in 2026?
• How'd you fix M Booth Health's revenue issues in 2026?
• How'd you fix Babylon Health's revenue issues in 2026?