For a copy of the complete research paper, find me on LinkedIn
Critical Integrative Evidence Synthesis · July 2026
The Diagnostic Access Gap
When delayed imaging becomes a population health problem: why geography, workforce shortages, administrative friction, and operational delays contribute to preventable health disparities.
Central proposition
Health systems cannot claim to manage population health without measuring how long people wait for a diagnosis
Diagnostic imaging is population-health infrastructure. Access should be measured across the full interval from clinical need to completed examination, finalized interpretation, diagnostic resolution, and treatment transition. Departmental productivity alone is not a sufficient measure of whether a population can obtain a diagnosis when it is clinically needed.
Performed more than 10 days after the expected date among 97,160 patients whose orders were expected within one day (Lacson et al., 2024).
In the same quaternary health system cohort, against an expectation of one day.
Across 2.87 million abnormal screens, 2008 to 2021, with lower rates for Black and American Indian women (Oluyemi et al., 2024).
Integrative review of 493 publications identifying failure to close the loop as a recurrent mechanism (Graber et al., 2024).
Executive summary
The access gap is multidimensional.
Geographic distance is visible, but the operational access gap also includes authorization, scheduling latency, limited appointment hours, equipment downtime, staffing shortages, transportation, language, affordability, report backlogs, and failures to complete recommended follow-up.
The burden is not distributed equally.
Recent U.S. studies associate delayed MRI or missed imaging with public insurance, neighborhood deprivation, rurality, race, financial hardship, and health-related social needs. Breast, lung, and prostate imaging provide particularly clear examples of unequal reach and unequal completion.
Capacity measures can conceal inequity.
A scanner may be heavily utilized while patients who are clinically urgent wait. A radiology group may achieve strong report turnaround while referrals remain unscheduled. Health systems must measure access before acquisition and closure after interpretation, not only what occurs inside the department.
The operating response is actionable.
Health systems can reduce diagnostic delay through centralized authorization, demand-capacity management, extended hours, regional scheduling, navigation, transportation support, same-day pathways, mobile services, teleradiology, automated tracking, and executive accountability for stratified access metrics.
Leadership conclusions
- Imaging access should be governed as a systemwide clinical capability rather than a collection of departmental assets.
- Order-to-performed time, diagnostic closure, and equity gaps should join volume, utilization, revenue, and report turnaround in the executive scorecard.
- Access metrics should be stratified by ZIP code, rurality, race and ethnicity, age, sex, disability, language, insurance, and neighborhood deprivation when data quality permits.
- Capital investment without workforce, authorization, navigation, and follow-up capacity can increase nominal supply without producing meaningful access.
- Population-health contracts and community benefit strategies should explicitly include diagnostic access, especially for cancer screening, chronic disease evaluation, and high-risk rural populations.
Report roadmap
| Section | Purpose |
|---|---|
| 1. Reframing the problem | Why imaging belongs within population-health strategy. |
| 2. Evidence of the gap | Geography, timeliness, inequity, administrative burden, workforce, and follow-up. |
| 3. Consequences | How delay changes the diagnostic and treatment pathway. |
| 4. Measurement model | A proposed Diagnostic Access Equity Index and executive scorecard. |
| 5. Operating response | Interventions, governance, and a 12-month implementation roadmap. |
| 6. Research agenda | Evidence gaps, validation needs, and policy priorities. |
Scope and method
This report uses a critical, integrative evidence synthesis approach. Targeted searches conducted in July 2026 prioritized PubMed-indexed peer-reviewed research published from 2023 through July 2026, supplemented by influential earlier studies where they remain methodologically or conceptually important. Official sources included the Centers for Medicare & Medicaid Services, the Health Resources and Services Administration, the U.S. Bureau of Labor Statistics, the American Society of Radiologic Technologists, and the American College of Radiology.
The review focused on MRI, CT, mammography, ultrasound, and related diagnostic pathways in the United States. Evidence was organized across five levels: patient, referral, imaging practice, health system, and community. Because the literature uses heterogeneous definitions of delay and is concentrated in breast imaging and selected academic systems, this report does not claim a single universal threshold for acceptable access. Instead, it develops a measurement architecture that health systems can calibrate by clinical urgency and modality.
Evidence standard
Empirical findings are identified as sourced evidence. The Diagnostic Access Equity Index, the maturity model, and the operating architecture are original proposed frameworks and have not yet undergone external validation.
Section 1
Diagnostic imaging is population-health infrastructure
Imaging is the mechanism by which many populations progress from risk factors or symptoms to a clinically actionable diagnosis. It is central to cancer screening and staging, stroke and cardiovascular evaluation, trauma, maternal care, musculoskeletal disability, neurological disease, and the surveillance of chronic conditions.
Operational definition
Meaningful diagnostic access exists when a clinically indicated examination is geographically reachable, financially and administratively attainable, scheduled and completed within an urgency-appropriate interval, interpreted promptly, communicated effectively, and connected to follow-up or treatment.
Figure 1 · Author-developed framework
The diagnostic access continuum
Failure can occur before scheduling, during administrative clearance, at acquisition, during interpretation, or after an abnormal result. The continuum is the unit of management, not any single stage within it.
Availability and access are not synonyms
A health system can own advanced imaging equipment and still have an access deficit. Equipment may be unavailable during evenings or weekends, constrained by technologist vacancies, limited by anesthesia or contrast support, or inaccessible to patients who cannot navigate authorization, transportation, cost-sharing, language, or scheduling processes. Likewise, a county-level facility count may overstate access when the facility lacks specialty protocols, accreditation, biopsy capability, open-bore accommodation, or reliable report coverage.
| Concept | What it measures | What it can miss |
|---|---|---|
| Availability | Whether a modality or service exists | Wait times, affordability, eligibility, staffing, protocol capability |
| Capacity | Potential examinations per unit of time | Demand mix, downtime, no-shows, bottlenecks, urgency |
| Utilization | Examinations actually performed | Unmet need, denied orders, abandoned referrals, inequitable queues |
| Timeliness | Elapsed time between defined steps | Whether the patient completed the downstream follow-up |
| Meaningful access | Reach, timeliness, affordability, completion, and closure | Requires linked longitudinal data and stratification |
The measurement reframe
Each row replaces a production-system question with an access-system question.
Evidence of the gap
Geography creates modality-specific diagnostic deserts
A 2024 U.S. ZIP-code analysis found that the average distance to the nearest breast MRI facility was 23.2 miles, compared with 8.2 miles for mammography and 22.2 miles for ultrasound. Breast MRI was therefore 2.8 times farther than mammography on average.
Across 29,629 U.S. ZIP codes (Christensen et al., 2024).
Compared with metropolitan areas, for breast MRI.
Adjusted, among 90,908 prostate cancer patients (El Khoury et al., 2025).
Among 1,009,040 Medicare beneficiaries with elevated PSA; rural odds 33% lower (Hansen et al., 2026).
Figure 2 · Interactive rebuild
Mean distance to the nearest accredited breast imaging facility
Source: Christensen et al. (2024). Distance measures spatial access and does not capture appointment availability or affordability. Small and rural areas carried an additional 23.2 miles to breast MRI relative to metropolitan areas, and the least advantaged neighborhoods also experienced greater distance.
Figure 2 as published in the research paper.
This is not simply a breast-imaging issue
Geographic differences also appear in low-dose CT lung cancer screening and prostate MRI. A nationally representative analysis of 2022 Behavioral Risk Factor Surveillance System data found that only 17.24% of eligible adults reported lung cancer screening. Uptake ranged from 21.95% in the Northeast to 13.41% in the West, and rural residents had 18% lower adjusted odds of screening than urban residents.
Figure 3 · Interactive rebuild
Lung cancer screening uptake by U.S. Census region
Source: Gudina et al. (2025). Estimates are based on self-reported 2022 BRFSS data and 2021 USPSTF eligibility criteria.
Figure 3 as published in the research paper.
Leadership implication
These findings indicate that advanced imaging diffusion can increase nationwide while geographic inequity persists. A scanner placement strategy based only on market share or facility economics may deepen diagnostic deserts. Community benefit and capital planning should overlay disease burden, social vulnerability, travel time, referral leakage, and modality capacity before placing new equipment or closing sites.
Evidence of the gap
Timeliness is an access measure, not merely a scheduling measure
The most direct recent evidence comes from a large quaternary health system study of 97,160 unique patients with outpatient MRI orders expected within one day. Nearly 48% were performed more than 10 days after the expected date, and the mean order-to-performed interval was 18.5 days.
Figure 4 · Interactive rebuild
Factors associated with prolonged outpatient MRI access
Source: Lacson et al. (2024). Odds ratios are not causal effects and are derived from a single large health system. Delay was associated with public insurance, female sex, cardiac MRI, and residence in the most deprived neighborhood quintile.
Figure 4 as published in the research paper.
Why this cannot be fixed at the scheduling desk
Appointment delays are produced by interactions among patient, community, and practice factors. Health systems must examine referral completeness, authorization, protocol review, appointment supply, modality-specific staffing, transportation, communication, and the clinical logic used to prioritize demand.
Delay compounds across the pathway
The clinical effect of access delay depends on disease biology, urgency, and the position of the delay within the pathway. A three-day delay in elective musculoskeletal imaging is not equivalent to a three-day delay in suspected cord compression. Population-health measurement therefore requires urgency-adjusted standards rather than one universal target.
Nevertheless, delays accumulate. A patient may first wait for a primary-care visit, then for authorization, then for imaging, then for interpretation, then for specialist consultation, then for biopsy or treatment. Each interval may appear operationally modest, but their sum can materially change the diagnostic journey. The risk is greatest when ownership is fragmented and no team measures the total elapsed time.
Cumulative delay principle
The population-health burden is determined by the total time from clinically meaningful need to diagnostic resolution, not by any single department’s performance.
Interactive tool
Cumulative pathway delay calculator
Enter the median elapsed days your organization observes in each interval. The tool sums the pathway and identifies where the greatest share of time is lost.
43
Total elapsed days
6.1
Weeks to resolution
+24.5
Days beyond the 18.5-day MRI benchmark
The 18.5-day comparator is the published mean order-to-performed interval for outpatient MRI orders expected within one day (Lacson et al., 2024). It is a reference point for a single interval, not a standard for the whole pathway. Set urgency-adjusted targets locally.
Cancer illustrates the stakes
An integrative review of 493 publications concluded that approximately one in five patients with cancer experiences a delayed or missed diagnosis and identified failure to close the loop as a recurrent mechanism (Graber et al., 2024). Treatment delay evidence also demonstrates that additional time can increase mortality risk for multiple cancers, although the magnitude varies by cancer type, treatment modality, and stage (Hanna et al., 2020).
Caution is necessary. Earlier imaging does not automatically improve survival, and observational associations are vulnerable to confounding, lead-time bias, and disease-severity effects. The defensible claim is narrower: clinically inappropriate delay can postpone diagnosis and treatment, and these delays are more likely to affect populations facing structural access barriers.
Evidence of the gap
Diagnostic follow-up is often incomplete or delayed
National Mammography Database research identified 2.87 million abnormal screening mammograms from 2008 through 2021. Documented follow-up was present for 66.4%. Follow-up rates were lower among Black women and American Indian women, demonstrating that screening participation alone does not guarantee diagnostic resolution.
For Black women compared with White women after an abnormal screening mammogram (Fayanju et al., 2024).
Across 13,670 recommendations; the program also generated attributable revenue (Jhala et al., 2024).
Even though 93.9% of incidental-finding alerts were acknowledged within 73 hours.
Among patients who missed a screening mammography appointment (Wang et al., 2024).
Figure 5 · Interactive rebuild
Breast imaging wait intervals by race after a same-day biopsy program
Source: Yoon et al. (2023). In a Duke cohort of 2,156 women, Black women waited an average of 18.0 days from screening to diagnostic imaging compared with 11.5 days for non-Hispanic White women, and 9.0 versus 4.4 days from diagnostic imaging to biopsy. Although the same-day biopsy pathway shortened the overall interval, the reduction was significant for White women but not Black women.
Figure 5 as published in the research paper.
Critical interpretation
An intervention can improve the mean while failing to close the equity gap. Every access initiative should therefore report both overall performance and stratified benefit distribution.
Closed-loop communication is necessary but not sufficient
Automated alerts, standardized reporting macros, patient notification, and tracking registries can improve acknowledgment and follow-up. In one large multi-hospital workflow, 93.9% of incidental-finding alerts were acknowledged within 73 hours, and follow-up orders were placed for 62.3% of patients. A randomized trial of direct patient notification found that telephone outreach increased completion compared with a tracking system alone, while portal-only notification performed worse.
The lesson is that technology must be paired with human navigation, workflow ownership, and modality-appropriate escalation (Loftus et al., 2024).
The economic burden extends beyond radiology
Delayed diagnosis can increase emergency utilization, repeat testing, treatment intensity, disability, travel, time away from work, caregiver burden, and patient anxiety. It can also increase malpractice exposure and create avoidable revenue leakage when recommended follow-up is not completed.
A radiology safety-net program found that 24% of eligible follow-up recommendations required active intervention. Completed examinations generated estimated revenue sufficient to support dedicated labor, suggesting that patient safety and financial stewardship can be mutually reinforcing (Jhala et al., 2024).
Evidence of the gap
Social and financial barriers are embedded in missed imaging
A 2024 study of adults who missed outpatient radiology appointments found that 44.4% reported their illness as a source of financial hardship, 28.8% identified imaging itself as a financial hardship, 35.2% reported material hardship, and 18.3% reported cost-related care nonadherence. Nearly one-third had at least one health-related social need, most commonly food insecurity. Only 12.5% had previously been screened for financial hardship or social needs.
Figure 6 · Interactive rebuild
Financial hardship and social needs among patients with missed imaging appointments
Source: Cuyegkeng et al. (2024), n = 282 adults with missed radiology appointments. The lowest bar is the screening rate, not a burden measure: the barriers were common, and they were rarely identified in advance.
Figure 6 as published in the research paper.
Missed appointments are not simply noncompliance
They can signal insurance instability, financial toxicity, transportation barriers, work and caregiving obligations, language mismatch, or an inability to understand preparation requirements. In screening mammography, patients with no-shows were more likely to have Medicaid or means-tested insurance, reside in high-poverty areas, and identify with racially or ethnically minoritized groups. 40% had still not completed the examination one year later (Wang et al., 2024).
Table 2. The operational anatomy of diagnostic delay
Seven levels at which the pathway fails, with the failure mode and the highest-value response at each level. Severity dots mark where the paper’s evidence base is strongest.
Prior authorization creates a hidden queue before the radiology schedule
Administrative clearance is part of the clinical access pathway. MRI prior-authorization research in orthopedic care found that nearly all denials advanced to peer-to-peer review were ultimately approved, suggesting avoidable delay and administrative waste. Other research has identified a higher denial risk for Medicaid patients. These findings do not imply that all utilization management is inappropriate; they demonstrate that authorization design can create unequal latency even when the requested examination is eventually approved.
CMS-0057-F requires impacted payers to provide decisions within 72 hours for expedited requests and seven calendar days for standard requests, with major API requirements primarily beginning in 2027. Beginning in 2026, covered payers must provide specific denial reasons and publicly report selected prior-authorization metrics. These requirements create an opportunity for imaging leaders to link payer response time, denial reason, appeal status, and order completion in a single dataset (CMS, 2024).
Use the transparency, do not just comply with it
Imaging organizations should create payer-level dashboards that distinguish request quality, medical-necessity disagreement, administrative error, and plan-specific delay. Public reporting will be most useful when it is connected to patient completion rather than treated as a compliance exercise.
Evidence of the gap
Workforce scarcity converts capital into idle or constrained capacity
Imaging capacity depends on coordinated professional labor: technologists, nurses, schedulers, radiologists, physicists, authorization staff, patient navigators, and subspecialty support. In the 2025 ASRT staffing survey, vacancy rates remained in double digits across every major imaging discipline reported.
The highest reported discipline in the 2025 ASRT survey of 475 department managers.
Improved modestly from 2023 but still severely constraining.
Radiologic and MRI technologists, 2024 to 2034, driven largely by replacement needs (BLS, 2025).
A 2024 pilot global survey also found radiologist vacancies in nearly half of participating practices (Omofoye et al., 2024).
Figure 7 · Interactive rebuild
Medical imaging vacancy rates, 2023 and 2025
Source: American Society of Radiologic Technologists (2025). Several modalities improved from 2023, but the workforce constraint remained severe, and mammography and nuclear medicine moved in the wrong direction. Survey results reflect respondents and may not represent every market or facility type.
Figure 7 as published in the research paper.
Reporting capacity is constrained as well
A 2024 pilot survey found unreported imaging examinations in up to 68% of participating practices and radiologist vacancies in nearly half of them. The survey was not nationally representative, but it underscores a structural problem: acquisition can outpace interpretation, and both can outpace diagnostic closure (Omofoye et al., 2024).
MRI technologist employment is projected to grow by 7%, but increased supply does not guarantee alignment with regional demand, specialty protocols, or rural facilities (U.S. Bureau of Labor Statistics, 2025). AI may improve selected workflow steps, but current evidence supports augmentation rather than treating AI as a substitute for comprehensive workforce and operating-model redesign (Jing et al., 2025).
Capital planning rule
Translate demand into staffed modality hours, not only installed scanners. Capital proposals should include a workforce and access-operating plan as a condition of approval. Where installed capacity exceeds staffed capacity, the scanner is an accounting asset rather than an access asset.
Table 1
Selected empirical evidence on diagnostic access
Nine studies spanning geography, timeliness, follow-up completion, patient burden, and workforce. Filter by domain to isolate the evidence relevant to a specific operating question.
| Domain | Population or unit | Key finding | Source |
|---|---|---|---|
| Breast MRI geography | 29,629 U.S. ZIP codes | MRI 23.2 miles versus mammography 8.2 miles; rural areas had substantial additional distance | Christensen et al., 2024 |
| Outpatient MRI timeliness | 97,160 unique patients | 47.9% delayed beyond 10 days; mean 18.5 days; deprivation OR 1.70 | Lacson et al., 2024 |
| Abnormal mammography follow-up | 2.87 million abnormal screens | Documented follow-up 66.4%; lower rates for Black and American Indian women | Oluyemi et al., 2024 |
| Same-day breast biopsy | 2,156 women | Black women had longer screening-to-diagnostic and diagnostic-to-biopsy intervals | Yoon et al., 2023 |
| Missed imaging appointments | 282 adults | Financial hardship and social needs were common and rarely screened | Cuyegkeng et al., 2024 |
| Prostate MRI utilization | 90,908 prostate cancer patients | Rural residents 35% less likely to undergo an MRI | El Khoury et al., 2025 |
| Prediagnostic prostate MRI | 1,009,040 Medicare beneficiaries | Only 4.87% received MRI; rural odds 33% lower than metro | Hansen et al., 2026 |
| Lung cancer screening | National 2022 BRFSS sample | 17.24% uptake; rural adjusted odds 18% lower | Gudina et al., 2025 |
| Imaging workforce | 475 U.S. department managers | CT vacancy 19.4%; MRI 17.4% in 2025 | ASRT, 2025 |
| Follow-up safety net | 13,670 recommendations | 24% required safety-net involvement; program generated attributable revenue | Jhala et al., 2024 |
Reading this table honestly
Much of the evidence is observational, single-system, or concentrated in breast imaging, which is overrepresented because it has mature registries and clearly defined screening-to-diagnosis pathways. Race and rurality are social and geographic markers, not biological causes, so analyses should focus on modifiable mechanisms and structural conditions.
Section 4 · Proposed framework
The Diagnostic Access Equity Index
Health systems need a composite measure that does not collapse access into a single wait-time average. The proposed Diagnostic Access Equity Index converts five domains into a 0 to 100 score. Higher values indicate stronger access. The model is intended as an executive management architecture, not a validated clinical instrument.
Figure 8 · Author-developed framework
Proposed Diagnostic Access Equity Index
Domain weights are provisional and should be tested against outcomes and local priorities.
Proposed formula
DAEI = 0.20G + 0.25T + 0.20A + 0.20C + 0.15E, where G is geographic reach, T is timeliness, A is administrative and financial access, C is completion and diagnostic closure, and E is equity performance. Each domain is standardized to a range of 0 to 100.
Interactive tool
DAEI calculator
Score each domain from 0 to 100. The index applies the published weights and flags when a strong composite is concealing a weak domain.
Formal remediation plan and monthly executive review.
Domain profile
Weighted contribution to the index
Domain definitions and candidate measures
| Domain | Candidate indicators | Principal data sources |
|---|---|---|
| Geographic reach, 20% | Travel time to appropriate modality; local protocol availability; regional capacity per eligible population | Geocoded orders, facility inventory, drive-time analysis |
| Timeliness, 25% | Order-to-contact; order-to-scheduled; order-to-performed; exam-to-final report; abnormal result to next step | EHR, RIS, PACS, scheduling, report timestamps |
| Administrative access, 20% | Authorization response, denial, and appeal; patient cost estimate; language and transportation support | Authorization platform, payer files, patient navigation |
| Completion and closure, 20% | Order completion; no-show recovery; follow-up recommendation completion; diagnostic resolution | EHR orders, registries, tracking systems |
| Equity performance, 15% | Largest and average subgroup gaps across core access measures | Demographics, insurance, ADI, rurality, language |
Scoring principles
- Use clinical urgency tiers. Emergency, urgent, expedited, routine, screening, and surveillance pathways require different thresholds.
- Measure the median and the upper tail. The 90th percentile often reveals access failure hidden by the mean.
- Report both performance and disparity. A system can improve overall access while widening subgroup gaps.
- Link pre-examination and post-report intervals. Order-to-performed and abnormality-to-closure are distinct but interdependent.
- Avoid rewarding inappropriate overuse. The index should apply to clinically indicated orders and be paired with appropriateness safeguards.
- Publish uncertainty and data-quality flags. Missing race, language, or payer data can make equity comparisons unreliable.
Illustrative performance bands
| Score | Interpretation | Executive implication |
|---|---|---|
| 85 to 100 | Population-optimized access | Sustain, benchmark, and test whether outcomes improve |
| 70 to 84 | Generally accessible with targeted gaps | Address specific modality, geography, or subgroup deficits |
| 55 to 69 | Material access constraints | Formal remediation plan and monthly executive review |
| 40 to 54 | Severe diagnostic-access gap | Capacity redesign, payer escalation, and community intervention |
| Under 40 | Critical access desert or pathway failure | Immediate systemwide response and board oversight |
Composite caution
Composite indices can conceal poor performance within a domain. The component measures should always accompany the DAEI, and improving access without appropriateness controls can increase low-value imaging, incidental findings, and downstream burden.
Section 5
What executives should measure
The measurement architecture should follow the patient rather than the departmental workflow. A useful scorecard begins with demand and ends with closure. It should distinguish controllable operating delays from clinical, patient-choice, and external-payer delays while preserving accountability for the total patient experience.
Core executive scorecard
| Metric domain | Measure | Cadence | Accountable owner |
|---|---|---|---|
| Access demand | Orders received per 1,000 attributed lives; demand by modality and urgency | Monthly | Population health and radiology |
| Scheduling latency | Median and 90th percentile order-to-scheduled days | Weekly | Access operations |
| Completion latency | Median and 90th percentile order-to-performed days | Weekly | Radiology operations |
| Administrative friction | Authorization days, denial rate, and appeal overturn rate | Weekly | Revenue cycle and authorization |
| Capacity reliability | Productive scanner hours; downtime; staffed hours; fill rate | Daily and weekly | Technical operations |
| Report timeliness | Exam-to-final report by urgency and modality | Daily | Radiology medical leadership |
| Diagnostic closure | Abnormal result to completed follow-up; unresolved backlog | Weekly and monthly | Quality and safety |
| Equity gap | Largest subgroup difference for each core measure | Monthly and quarterly | Population health and equity |
| Patient burden | Travel time, rescheduling, out-of-pocket estimate, navigation needs | Quarterly | Patient experience |
| Outcome linkage | Stage, avoidable ED use, treatment interval, disability, where feasible | Quarterly and annual | Clinical service lines |
Minimum viable equity cut
At minimum, stratify order-to-performed time and completion by ZIP code, rurality, race and ethnicity, age, sex, insurance, language, and neighborhood deprivation. Suppress unstable cells and report missing-data rates.
Metrics that should not stand alone
| Traditional metric | Why it can mislead | Required companion measure |
|---|---|---|
| Scanner utilization | High utilization may coexist with long queues | Demand-to-capacity ratio and wait-time distribution |
| Examination volume | Counts completed care, not unmet need | Order completion and abandoned-referral rate |
| Report turnaround | Starts after acquisition and ignores access before the exam | Order-to-performed and closure time |
| No-show rate | Can imply patient blame and obscure structural barriers | Reason-coded recovery and social-need assessment |
| Revenue per scanner | May reward profitable access rather than equitable access | Service-level equity and community need |
| Average wait time | Can hide extreme delays and subgroup differences | Median, 90th percentile, and stratified gaps |
Appendix A. Minimum viable diagnostic access dataset
| Data domain | Minimum fields |
|---|---|
| Patient and population | Patient ID; age; sex; race and ethnicity; language; ZIP code; insurance; ADI; rurality |
| Clinical need | Order date and time; ordering service; diagnosis; urgency; screening or diagnostic intent |
| Referral quality | Order completeness; protocol query; duplicate order; canceled or withdrawn order |
| Authorization | Submission; payer; status; decision time; denial reason; appeal; overturn |
| Scheduling | First contact; scheduled date; offered alternatives; patient decline; reschedule reason |
| Capacity | Facility; modality; protocol; duration; staffed hours; downtime; slot release |
| Acquisition | Arrival; exam start and completion; cancellation; no-show; preparation failure |
| Interpretation | Preliminary and final report times; critical result communication |
| Follow-up | Recommendation; due date; order; scheduled; completed; resolved; escalation |
| Outcome | Diagnosis date; stage where applicable; treatment start; ED use; patient-reported burden |
Appendix B. Questions the dashboard must answer
- Where are the longest order-to-performed intervals by modality, urgency, facility, and ZIP code?
- Which payer creates the greatest authorization delay and the highest appeal overturn rate?
- What proportion of abnormal screening examinations reach diagnostic resolution within the defined standard?
- Which patients repeatedly reschedule or miss imaging, and what barriers are documented?
- Where does installed capacity exceed staffed capacity?
- Which subgroup has the largest 90th-percentile wait-time gap?
- How many follow-up recommendations are overdue, and which service owns them?
- Did the most recent intervention improve access for the intended population without worsening another group?
Section 6
The operating model must close the full loop
Five linked capabilities convert measurement into access: demand sensing, capacity design, access operations, diagnostic closure, and an equity audit that drives executive action.
Figure 9 · Author-developed framework
Population-health operating model for diagnostic access
The loop only closes when the equity audit feeds back into demand sensing and capacity design rather than terminating in a report.
Demand sensing
Forecast demand using attributed population, disease burden, screening eligibility, referral patterns, seasonality, planned service-line growth, and historical unmet need. Demand should be segmented by urgency, protocol complexity, contrast, sedation, and patient accommodation needs.
Capacity design
Translate demand into staffed modality hours, not only installed scanners. Capacity models should incorporate technologist and radiologist availability, room turnover, protocol length, downtime, cancellations, sedation, nursing, and contingency coverage. Capital proposals should include a workforce and access-operating plan as a condition of approval.
Access operations
Centralize or tightly coordinate referral intake, order validation, authorization, scheduling, preparation, and financial communication. Standard work should reduce avoidable handoffs and provide escalation when urgency and available capacity conflict. Regional scheduling should offer the earliest clinically appropriate appointment across facilities rather than defaulting to the ordering location.
Diagnostic closure
Every actionable abnormality should have an accountable owner, recommended interval, patient communication status, and escalation pathway. Tracking systems should distinguish acknowledgment, order placement, scheduling, completion, and resolution. A closed loop ends with completion or a documented clinically valid reason for noncompletion.
Equity audit and executive action
Access interventions should undergo a distributional review. Did same-day access improve for all groups? Did extending evening hours reduce delays for Medicaid patients? Did centralized authorization reduce denial latency? Did mobile imaging increase screening but create downstream biopsy gaps? Executive review should focus on residual disparities and reinvest resources where the index remains low.
Table 3. Intervention portfolio
| Intervention | Primary problem | Outcome metrics | Expected value |
|---|---|---|---|
| Centralized authorization | Administrative delay and denials | Authorization time, denial, overturn, completion | High |
| Regional single-queue scheduling | Fragmented capacity and local bottlenecks | Earliest available appointment, travel burden | High |
| Extended evening and weekend hours | Work and caregiving conflict | Fill rate, no-show, subgroup wait times | Moderate |
| Patient navigation | Complex preparation, social needs, and abnormal follow-up | Completion and closure | High for vulnerable cohorts |
| Transportation support | Travel and parking barriers | Completion among transport-vulnerable patients | Targeted |
| Mobile imaging | Geographic screening access | Reach, uptake, downstream resolution | Context dependent |
| Teleradiology and load balancing | Reporting backlog and subspecialty scarcity | Report turnaround, backlog, quality | High with governance |
| AI-assisted workflow | Prioritization, protocoling, communication, tracking | Time saved, errors, subgroup performance | Emerging |
| Same-day diagnostic pathways | Multi-visit delay after abnormal screening | Diagnostic interval, equity gap | High in selected pathways |
| Safety-net registry | Missed follow-up recommendations | Resolution, patient harm, attributable revenue | High |
Policy and market implications
Population-health contracts should include diagnostic timeliness.
Value-based contracts frequently hold organizations accountable for screening rates, emergency utilization, chronic disease outcomes, and total cost. They rarely specify whether patients can obtain timely diagnostic confirmation after a positive screening test or a new symptom. This omission separates accountability for detection from accountability for access. Payers and providers should test shared measures of order completion, diagnostic resolution, and disparity reduction.
Community benefit and capital planning should be geospatial.
Health systems should overlay disease burden, social vulnerability, travel time, referral leakage, and modality capacity before placing new equipment or closing sites. Conversely, mobile or satellite access should not be expanded without downstream diagnostic, biopsy, and treatment capacity.
Rural strategy requires regional networks, not isolated assets.
The feasible response may be a regional portfolio: local radiography and ultrasound, reliable CT, scheduled MRI access, mobile screening, shared specialty protocols, transportation support, teleradiology, and explicit transfer pathways. The unit of design should be the regional population, not the individual facility.
Section 7 · Proposed framework
A five-level diagnostic access maturity model
Maturity is sequential. Equity stratification and population linkage are not reliable until the underlying access and integration measures exist, so a level is only attained when every capability beneath it is in place.
Figure 10 · Author-developed framework
Diagnostic access maturity model
Each level adds a capability that the level above it depends on.
| Maturity level | Defining capability | Leadership test |
|---|---|---|
| Level 1: Departmental | Measures volume, productivity, utilization, and report turnaround | No linked access or equity view |
| Level 2: Accessible | Measures waits, cancellations, and completion by modality | Limited upstream and downstream linkage |
| Level 3: Integrated | Links referral, authorization, scheduling, acquisition, and reporting | Operational ownership spans departments |
| Level 4: Equity-managed | Stratifies core metrics and targets high-risk populations | Equity gaps trigger resource decisions |
| Level 5: Population-optimized | Links access to disease burden, outcomes, contracts, and community planning | The system manages marketwide diagnostic capability |
Interactive self-assessment
Where does your organization sit today?
Check every capability that is genuinely in production, not planned or piloted. The attained level is capped at the highest level whose capabilities are all complete.
Level 1: Departmental
0 of 4 in placeMeasures volume, productivity, utilization, and report turnaround.
Level 2: Accessible
0 of 4 in placeMeasures waits, cancellations, and completion by modality.
Level 3: Integrated
0 of 4 in placeLinks referral, authorization, scheduling, acquisition, and reporting.
Level 4: Equity-managed
0 of 4 in placeStratifies core metrics and targets high-risk populations.
Level 5: Population-optimized
0 of 4 in placeLinks access to disease burden, outcomes, contracts, and community planning.
No level is fully in place yet.
Capabilities checked overall: 0%
Completion by level
Section 8
Twelve-month implementation roadmap
Five periods that move an organization from charter to a validated operating model, with a required output at every stage.
0 to 30 days
Name executive sponsor and clinical-operational dyad; define the access pathway; inventory data and current queues
Required output: charter, definitions, baseline inventory
31 to 90 days
Build a minimum viable dataset; establish urgency tiers; validate timestamps; identify high-risk modalities and geographies
Required output: baseline scorecard and data-quality report
Months 4 to 6
Pilot centralized authorization and regional scheduling; deploy no-show recovery and a follow-up registry in one pathway
Required output: pilot results with stratified metrics
Months 7 to 9
Expand to CT and MRI and a single screening pathway; integrate transport and navigation; establish a monthly executive review
Required output: DAEI pilot and intervention portfolio
Months 10 to 12
Link access with clinical outcomes and financial effects; establish targets; publish governance and the annual improvement plan
Required output: validated operating model and year-two plan
First 90-day decisions
- Select one high-consequence pathway, such as abnormal mammography, lung cancer screening, outpatient MRI, or actionable incidental findings.
- Define the start and end of the access interval and identify every timestamp needed to measure it.
- Choose two vulnerable populations for initial stratification and verify data completeness before publishing comparisons.
- Create a joint operating review involving radiology, population health, revenue cycle, scheduling, service-line leadership, and quality.
- Fund one intervention with a testable hypothesis, a defined counterfactual, and a pre-specified equity outcome.
Research agenda
The literature establishes that diagnostic access is unequal, but important questions remain unresolved. Much of the evidence is observational, single-system, or concentrated in breast imaging. A stronger research program should connect operational measures with clinical and population outcomes while testing interventions prospectively.
- What urgency-adjusted thresholds for MRI, CT, mammography, and ultrasound are associated with clinically meaningful outcomes?
- How much of observed diagnostic delay is explained by capacity, authorization, patient burden, referral quality, or clinician prioritization?
- Which interventions reduce both average wait time and subgroup disparities?
- Can the DAEI predict stage at diagnosis, avoidable emergency use, treatment intensity, disability, patient-reported burden, or total cost?
- How should health systems measure unmet demand, including orders never placed because clinicians anticipate poor access?
- When do mobile imaging and teleradiology reduce inequity, and when do they merely move the bottleneck downstream?
- How should AI tools be evaluated for access impact, safety, subgroup performance, and workflow displacement?
- What public reporting standards would improve accountability without encouraging inappropriate utilization or gaming?
Recommended study design. A multicenter prospective study should combine geocoded population data, EHR and scheduling timestamps, authorization records, RIS and PACS data, patient-reported barriers, and outcome registries. A stepped-wedge or cluster-randomized implementation design could compare centralized access interventions across sites. Analyses should pre-specify urgency tiers, subgroup definitions, missing-data rules, and mediation pathways. The DAEI should be validated against both clinical outcomes and patient-reported access burden.
Caveats and assumptions
- No single wait-time threshold is appropriate for every examination. Clinical urgency, disease biology, and patient preference must determine targets.
- Spatial distance is an incomplete proxy. It does not measure appointment supply, insurance acceptance, protocol capability, quality, or downstream services.
- Many cited studies are observational and cannot prove that the measured delay caused the subsequent clinical outcome.
- Breast imaging is overrepresented because it has mature registries and clearly defined screening-to-diagnosis pathways.
- Race and rurality are social and geographic markers, not biological causes. Analyses should focus on modifiable mechanisms and structural conditions.
- Composite indices can conceal poor performance within a domain. Its component measures should always accompany the DAEI.
- Improving access without appropriateness controls can increase low-value imaging, incidental findings, and downstream burden.
- Current workforce surveys reflect respondents and may not represent every market or facility type.
Final leadership statement
Health systems cannot credibly claim to manage population health when they do not measure how long different populations wait for diagnosis.
Evidence base
References
Twenty-two sources spanning peer-reviewed research, federal agencies, and professional societies. Filter by theme to locate the evidence behind a specific claim.
American College of Radiology. (2026). The radiologist shortage: A workforce update from the Harvey L. Neiman Health Policy Institute.
acr.orgAmerican Society of Radiologic Technologists. (2025). Radiologic Sciences Staffing and Workplace Survey: 2025 vacancy-rate findings.
asrt.orgCenters for Medicare & Medicaid Services. (2024). CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F).
cms.govChristensen, E. W., Rosenblatt, R. B., Patel, A. G., Rula, E. Y., Carlos, R. C., Narayan, A. K., & Patel, B. K. (2024). Differential access to breast magnetic resonance imaging compared with mammography and ultrasound. American Journal of Preventive Medicine, 67(6), 897 to 905.
doi.org/10.1016/j.amepre.2024.07.007Cuyegkeng, A., Hao, Z., Rashidi, A., Bansal, R., Dhillon, J., & Sadigh, G. (2024). Prevalence of financial hardship and health-related social needs among patients with missed radiology appointments. Clinical Imaging, 113, 110232.
doi.org/10.1016/j.clinimag.2024.110232El Khoury, C. J., Freedland, S. J., Gandhi, K., Keith, S. W., Nikita, N., Shaver, A., Sharma, S., Kelly, W. M. K., & Lu-Yao, G. (2025). Disparities in the utilization of magnetic resonance imaging for prostate cancer detection: A population-based study. Journal of the National Cancer Institute, 117(2), 270 to 278.
doi.org/10.1093/jnci/djae234Fayanju, O. M., et al. (2024). Racial disparities and strategies for improving equity in diagnostic follow-up for abnormal screening mammograms. JCO Oncology Practice.
doi.org/10.1200/OP.23.00782Graber, M. L., Winters, B. D., Matin, R., et al. (2024). Interventions to improve timely cancer diagnosis: An integrative review. Diagnosis.
doi.org/10.1515/dx-2024-0113Gudina, A. T., Fitzgibbon, M. L., Peterson, C. E., Byrne, C., Das, A., & Hirko, K. A. (2025). Geographic disparities in lung cancer screening uptake in the United States using the 2021 United States Preventive Services Task Force Guidelines. Lung Cancer, 205, 108615.
doi.org/10.1016/j.lungcan.2025.108615Hanna, T. P., King, W. D., Thibodeau, S., et al. (2020). Mortality due to cancer treatment delay: Systematic review and meta-analysis. BMJ, 371, m4087.
doi.org/10.1136/bmj.m4087Hansen, N. F., Zurl, H., Korn, S. M., Zhang, J., Tan, H. J., Nielsen, M. E., Moore, C. M., Trinh, Q. D., Kibel, A. S., & Cole, A. P. (2026). Utilization of prediagnostic prostate magnetic resonance imaging among rural Americans: An analysis of Medicare claims for elevated prostate-specific antigen. Urology Practice, 13(3), 242 to 249.
doi.org/10.1097/UPJ.0000000000000956Health Resources and Services Administration. (2026). Health Workforce Shortage Areas and Area Health Resources Files.
data.hrsa.govJhala, K., Lynch, E. A., Eappen, S., Curley, P., Desai, S. P., Brink, J., Khorasani, R., & Kapoor, N. (2024). Financial impact of a radiology safety net program for resolution of clinically necessary follow-up imaging recommendations. Journal of the American College of Radiology, 21(8), 1258 to 1268.
doi.org/10.1016/j.jacr.2023.12.016Jing, A. B., Garg, N., Zhang, J., & Brown, J. J. (2025). AI solutions to the radiology workforce shortage. npj Health Systems, 2, 20.
doi.org/10.1038/s44401-025-00023-6Lacson, R., Pianykh, O., Hartmann, S., Johnston, H., Daye, D., Flores, E., Kapoor, N., & Khorasani, R. (2024). Factors associated with timeliness and equity of access to outpatient MRI examinations. Journal of the American College of Radiology, 21(7), 1049 to 1057.
doi.org/10.1016/j.jacr.2023.12.028Lawson, M. B., et al. (2025). Disparities in standard-of-care, advanced, and same-day diagnostic services among patients with abnormal screening mammography. Radiology, 314(2).
doi.org/10.1148/radiol.241673Loftus, J. R., Kadom, N., Baran, T. M., Hans, K., Waldman, D., & Wandtke, B. (2024). Impact of early direct patient notification on follow-up completion for nonurgent actionable incidental radiologic findings. Journal of the American College of Radiology, 21(4), 558 to 566.
doi.org/10.1016/j.jacr.2023.07.026National Academies of Sciences, Engineering, and Medicine. (2015). Improving diagnosis in health care. The National Academies Press.
doi.org/10.17226/21794Oluyemi, E. T., Grimm, L. J., Goldman, L., Burleson, J., Simanowith, M., Yao, K., & Rosenberg, R. D. (2024). Rate and timeliness of diagnostic evaluation and biopsy after recall from screening mammography in the National Mammography Database. Journal of the American College of Radiology, 21(3), 427 to 438.
doi.org/10.1016/j.jacr.2023.09.002Omofoye, T. S., Vlahos, I., Marom, E. M., Bassett, R., Blasinska, K., Ye, X., Tan, B. S., & Yang, W. T. (2024). Backlogs in formal interpretation of radiology examinations: A pilot global survey. Clinical Imaging, 106, 110049.
doi.org/10.1016/j.clinimag.2023.110049Sosa, E., D’Souza, G., Akhtar, A., Sur, M., Love, K., Duffels, J., et al. (2021). Racial and socioeconomic disparities in lung cancer screening in the United States: A systematic review. CA: A Cancer Journal for Clinicians, 71(4), 299 to 314.
doi.org/10.3322/caac.21671U.S. Bureau of Labor Statistics. (2025). Occupational Outlook Handbook: Radiologic and MRI technologists.
bls.govWang, G. X., Mercaldo, S. F., Cahill, J. E., Flanagan, J. M., Lehman, C. D., & Park, E. R. (2024). Missed screening mammography appointments: Patient sociodemographic characteristics and mammography completion after 1 year. Journal of the American College of Radiology.
doi.org/10.1016/j.jacr.2024.03.017Yoon, S. C., Taylor-Cho, M. W., Charles, M. G., & Grimm, L. J. (2023). Racial disparities in breast imaging wait times before and after the implementation of a same-day biopsy program. Journal of Breast Imaging, 5(2), 159 to 166.
doi.org/10.1093/jbi/wbad003The Diagnostic Access Gap
Kelly Emrick, DHSc, PhD, MBA, BSRT(ARRT)R · Critical Integrative Evidence Synthesis · July 2026
Empirical findings in this dashboard are drawn from the cited peer-reviewed and agency sources. The Diagnostic Access Equity Index, the maturity model, the cumulative delay calculator, and the operating architecture are original proposed frameworks and have not yet undergone external validation. Figures rebuilt interactively reproduce the values published in the source report; the published figure is available beneath each rebuild.