When Restricted Utilization Becomes the Baseline: Why Health Care Spending and Workforce Models Need Access-Constraint Review

Historical utilization and spending can understate true need when access is constrained. Hawaiʻi should test for restricted access before using current use as a baseline.

HHTF Staff · 2026-06-01

Health care planning often relies on historical utilization and spending to estimate future need, set payment targets, assess workforce demand, and evaluate system performance. These measures are useful, but they reflect care that was actually delivered and paid for within the limits of the existing system.

In an access-constrained system, historical utilization may understate true clinical need. Patients may need care but be unable to obtain it because appointments are unavailable, physician panels are closed, specialists are not accessible locally, travel barriers are substantial, or the provider network exists on paper but not in practice. In these circumstances, low utilization may represent unmet need rather than efficiency.

This distinction is especially important for Hawaiʻi. Statewide averages can obscure island-level shortages, and claims-based utilization may not capture care that patients could not obtain. If future spending targets or workforce projections are built from restricted utilization, the model may carry current access failures forward as the expected baseline.

The central finding of this report is that historical utilization and spending should not be used as stand-alone baselines for future health care planning unless the model first evaluates whether the baseline reflects adequate access. Hawaiʻi should require an access-constraint review before using historical utilization or spending to set future payment targets, workforce expectations, or network adequacy conclusions.

Figure 1. HRSA Health Workforce Projections Dashboard, All Physicians, Surplus/Shortage by State, 2038

The HRSA Health Workforce Projections Dashboard image reviewed for this report shows Hawaiʻi with projected physician supply of 4,770 FTEs, projected demand of 4,620 FTEs, a calculated surplus of 150 FTEs, and 103% adequacy for all physicians in 2038. As a practical access assessment for Hawaiʻi, that projection is likely wrong unless it is interpreted alongside island-level distribution, part-time clinical availability, closed panels, payer participation, travel burden, and unmet need.[1]

This image prompted the question examined in this report. A statewide projection can appear reassuring while local access remains fragile, especially when projected demand is based on expected utilization in a system where patients already face geographic, workforce, appointment, and cost barriers.

Purpose of This Report

This report explains a measurement problem that affects health care planning, payment reform, workforce projections, and network adequacy analysis. The problem occurs when models use current utilization or spending as a proxy for future need without first determining whether current access is adequate.

The report does not argue that all low utilization is harmful. Lower utilization may reflect better prevention, lower disease burden, reduced waste, fewer unnecessary procedures, or improved care coordination. In an undercapacity system, however, lower utilization may indicate restricted access, delayed care, or unmet need. A model that does not distinguish between those conditions can mistake scarcity for efficiency.

Methods and Sources

This report is a policy analysis rather than an independent quantitative workforce projection. It reviews federal workforce projection materials, Hawaiʻi physician workforce assessments, and selected literature on access measurement, network adequacy, and provider directory accuracy to identify a specific measurement risk: historical utilization and spending may understate future need when current access is constrained.

The analysis focuses on how demand, utilization, access, and spending are defined and interpreted. It gives particular attention to whether models based on current or expected use can distinguish efficient reductions in unnecessary care from restricted use caused by inadequate access. Hawaiʻi is used as the case example because statewide averages, island geography, provider maldistribution, cost of practice, and directory-based network measures can obscure the difference between nominal coverage and real access.

This report does not independently estimate Hawaiʻi’s total unmet need or calculate a revised spending target. It proposes a policy standard: before historical utilization or spending is used as a baseline for future planning, payment reform, workforce assessment, or network adequacy review, the state should evaluate whether the baseline reflects adequate access or restricted access.

Key Definitions

Clinical need refers to care that would be appropriate based on health status, disease burden, preventive standards, chronic disease management, age, risk, disability, and local health conditions. It is not limited to care that was billed, documented in claims, or successfully obtained.

Demand refers to care that patients seek or are expected to use under a given set of social, financial, geographic, cultural, and system conditions. HRSA’s Workforce Projections Dashboard defines demand as the number of workers required to provide services “that will be utilized given patient health-seeking behavior and the ability and willingness to pay for services.”[2]

Utilization refers to care that actually occurs. It is usually measured through claims, encounters, visits, procedures, prescriptions, admissions, referrals, or other recorded services.

Spending refers to what is paid for care that actually occurs. Spending data generally do not capture care that was needed but never delivered.

Access refers to the practical ability of patients to obtain appropriate care in a timely way. The National Academies definition of access emphasizes “the timely use of personal health services to achieve the best possible health outcomes,” which means access cannot be reduced to insurance coverage, provider directory entries, or historical utilization alone.[3]

Restricted utilization refers to utilization that is lower than clinical need because patients cannot realistically obtain care. Unmet need refers to care that would likely improve health or prevent deterioration but is not received.

The Measurement Problem

Historical utilization is often treated as a practical starting point for estimating future need because it is available, measurable, and tied to real spending. Claims data can show what services were billed, where care occurred, and how much was paid.

The problem is that claims data show delivered care, not failed access. They do not show the patient who called multiple offices and could not find a primary care physician. They do not show the referral that was never completed because the specialist was unavailable locally. They do not show delayed evaluation due to travel cost, appointment scarcity, or months-long waits.

When access is constrained, the absence of a claim is not the same as the absence of need. A model that treats low utilization as low demand can convert restricted access into a planning assumption.

This matters when utilization becomes the basis for spending targets. If a population historically used fewer primary care visits because there were not enough clinicians, the low spending level may reflect undercapacity. A future payment model built from that baseline may underfund the visits, staffing, and infrastructure needed to restore access.

The result is a circular error. A system lacks enough care, so utilization stays low. Low utilization produces low spending. Low spending becomes the future budget. The budget then reinforces the lack of care.

Why This Matters for Hawaiʻi

Hawaiʻi has access challenges that may not be visible in statewide averages or broad national models. Patients may have insurance but still struggle to find an available primary care physician. Neighbor island communities may lack local access to key specialists. Provider directories may list clinicians who are not accepting new patients, are only available part time, are not available to all plans, or are geographically impractical for timely care.

The geography of Hawaiʻi changes the meaning of supply. A clinician located in urban Honolulu does not create practical access for a patient on Hawaiʻi Island, Maui, Kauaʻi, Molokaʻi, or Lānaʻi if the patient cannot realistically obtain care there. State-level provider counts may therefore overstate access for specific communities.

Cost of practice also matters. The Hawaiʻi Physician Workforce Assessment Project identifies workforce data collection, recruitment, retention, continuing education, loan repayment, prior authorization simplification, Medicare reimbursement advocacy, and practice transition as active parts of Hawaiʻi’s physician workforce strategy.[4] These activities reflect a practical reality: the presence of a medical license does not necessarily mean accessible clinical capacity.

The 2025 Physician Workforce Report for the 2026 Legislature has been publicly summarized as finding that Hawaiʻi had 12,688 licensed physicians, but only 3,647 were providing patient care to Hawaiʻi’s population, with full-time-equivalent clinical capacity lower than the patient-care headcount.[5] This illustrates why license counts and statewide physician totals can overstate real care availability.

For these reasons, Hawaiʻi should be cautious about treating current utilization or current spending as a sufficient measure of future need. A low level of billed care may reflect a healthy and efficient system. It may also reflect a system where patients could not obtain care early enough for the care to be billed.

Workforce Projection Relevance

Workforce projections are valuable planning tools, but their interpretation depends on how demand is defined. HRSA describes the Health Workforce Simulation Model as an integrated microsimulation model that estimates current and future supply of and demand for health care workers by occupation, geographic area, and year.[6] The Workforce Projections Dashboard states that the model incorporates factors such as changing population size, demographics, population location, new entrants and exiting providers, and differing levels of access to care.[7]

The dashboard guide also states that the status quo demand projections “extrapolate current national health care use and delivery patterns by personal characteristics to the projected populations in the future.”[8] This is a reasonable modeling approach for many planning purposes, but it creates a specific interpretive risk in access-constrained settings. A population that historically had less access may appear to require fewer clinicians than it would require if patients had timely access to appropriate care.

HRSA’s own scenario design recognizes that access barriers can affect measured demand. The dashboard guide describes improved-access scenarios that simulate demand if populations with historical barriers used care at rates equivalent to groups with fewer barriers.[9] HRSA also states in its public projection summary that behavioral health estimates based on current use do not address the large amount of unmet need, and that including unmet need produces higher projected shortages.[10]

That behavioral health example is important because it establishes a broader principle. Current use can understate need when people face barriers to care. This report applies that principle to Hawaiʻi physician workforce and spending analysis, while recognizing that the precise adjustment for primary care, specialty care, and island-level access would require additional local data.

Policymakers should not read a reassuring projection as proof of adequate access unless the model accounts for restricted utilization, local maldistribution, closed panels, wait times, payer participation, and unmet need. A useful workforce assessment should combine projection models with local access measures, including practicing clinician counts, full-time-equivalent clinical availability, geographic distribution, open panels, appointment wait times, referral completion, practice closures, physician age and retirement risk, and payer participation.

Spending Model Relevance

The same measurement problem applies to spending models, global budgets, capitation, accountable care arrangements, and total cost of care targets. These models often use historical spending as a baseline. That can work reasonably well when historical spending reflects real access and when savings are achieved by reducing waste.

It is more dangerous when historical spending reflects restricted care. In that setting, a low spending baseline may not represent efficiency. It may represent delayed diagnosis, untreated disease, missing primary care, unavailable specialty care, or care shifted to emergency departments and hospitals.

Payment reform should distinguish between reducing unnecessary care and suppressing necessary care. A spending target should not reward the appearance of savings created by lack of access. If primary care utilization rises because previously unserved patients finally obtain care, that should not automatically be treated as failure. It may be the predictable cost of repairing an undercapacity system.

This distinction is essential for Hawaiʻi. A model that uses historical spending without an access-constraint review may set a budget below the level needed to support primary care access, specialty access, rural services, and neighbor island capacity. A model that includes access safeguards can distinguish waste from unmet need and avoid treating restored access as overspending.

Network Adequacy Relevance

Network adequacy cannot be measured only by whether providers appear in a directory. A paper network is not the same as access. A provider may be listed but unavailable, retired, not accepting new patients, limited to certain plans, working part time, or located too far away to provide practical care.

Recent provider directory research describes