The Radiology Structure-Process-Outcome (SPO) Model
An interactive governance architecture that connects staffing, safety, clinical workflow, communication, and disease-specific outcomes, translating the Donabedian tradition into an actionable radiology performance system.
Executive Summary
Radiology operates at the intersection of diagnostic medicine, throughput, patient safety, digital communication, and therapeutic decision-making. Many performance systems still overweight volume, turnaround, and budget variance while undermeasuring the department’s contribution to accuracy, clinical action, patient comprehension, and pathway-level outcomes. The SPO model reorganizes performance around an evidence-linked causal chain that runs from capability, to execution, to patient value.
The most defensible radiology quality strategy is a layered scorecard: a small set of universal structure and process measures for the whole department, paired with pathway-specific outcomes where radiology has a plausible and measurable influence on treatment timing, diagnosis, safety, or patient experience.
Executive findings and management implications
Conceptual Foundation and Model Design
The SPO framework derives from Donabedian’s model of healthcare quality (Donabedian, 1988), which distinguishes structural capacity, care processes, and outcomes. It remains especially useful in radiology because imaging’s clinical impact is mediated through multiple handoffs: an order must be appropriate, the exam accessible and technically adequate, the interpretation accurate and timely, important findings must reach the responsible clinician or patient, and the resulting information must alter diagnosis, treatment, surveillance, or reassurance.
The structural-measurement imbalance
Why the imbalance is a management risk
Structural measures are accessible, auditable, and familiar to accreditation systems, which is why they dominate. The hazard is that a department can appear compliant and technically capable while still experiencing delays, communication failures, missed follow-up, avoidable variation, or poor patient understanding. The SPO correction is to measure whether capability is consistently translated into clinically useful imaging care and patient value.
Every KPI must answer three questions. (1) What element of care is being measured? (2) What decision or action will the measure inform? (3) What outcome or balancing measure will test whether the improvement helped, rather than merely shifting the problem elsewhere?
Methodologically, this is an evidence-informed framework rather than a systematic review. It synthesizes peer-reviewed radiology, emergency, stroke, communication, and patient-centered reporting literature with current guidance from the ACR, ESR, CAR, and The Joint Commission, prioritizing measures with clear definitions, feasible data sources, and plausible links to patient outcomes.
The Radiology SPO Value Chain
The model is best understood as a value chain rather than three disconnected columns. Structure shapes what the organization can reliably do, process reveals whether those capabilities are consistently translated into useful imaging care, and outcomes test whether the work changed something that matters.
The capacity to deliver care
Scanner capacity, staffing, accreditation, interoperability, governance, and subspecialty alignment. These are the prerequisites for reliable, safe, and clinically responsive imaging.
The execution of imaging work
Appropriate ordering, safe acquisition, timely and clear reporting, and reliable closed-loop communication. Process measures are the most directly redesignable.
The consequences for patients
Diagnostic accuracy, time to treatment, avoidable harm, patient understanding, and disease-specific clinical results within defined pathways.
The causal connection is not perfectly linear. A turnaround dashboard without discrepancy monitoring can incentivize speed over deliberation. A dose dashboard without image-quality review can reward underexposure. An appropriateness dashboard without access monitoring can reduce orders but worsen delays. A robust scorecard therefore pairs each primary measure with a balancing measure and, where possible, an outcome measure.
External factors (referring-clinician behavior, bed capacity, payer authorization, patient preference, downstream treatment availability) influence outcomes. The value chain does not claim perfect attribution, it provides a transparent hypothesis of how improvement should occur, which is then tested through governance.
Structure Domain: Capacity, Governance, and Digital Infrastructure
The structural domain captures whether the department possesses the prerequisites for reliable imaging: workforce composition, credentialing, equipment capability, accreditation, interoperability, service coverage, quality governance, and a culture that supports peer learning and corrective action. Capacity should be measured relative to demand by modality, care setting, daypart, and case complexity.
| Structural element | Operational definition | Primary data sources | Why it matters |
|---|---|---|---|
| Workforce capacity and coverage | Radiologist and technologist FTE relative to demand, vacancy rate, after-hours attending coverage, subspecialty assignment for high-risk cases. | HRIS, scheduling, PACS/RIS volumes, worklists. | Excess workload and fatigue can degrade diagnostic performance. Extended attending coverage has been associated with materially shorter turnaround for emergency CT (Jalal et al., 2021; Stec et al., 2018). |
| Equipment and protocol capability | Modality uptime, preventive maintenance, protocol availability, image-quality QC pass rate, technologist competency. | Vendor service data, physics reports, PACS metadata. | Reliability reduces repeat exams, delays, and variation. Dose monitoring reveals outlier protocols and wrong-protocol use. |
| Accreditation and audit maturity | Current modality accreditation, regulatory compliance, peer learning, audit calendar, corrective-action closure. | Accreditation files, committee minutes, peer-learning platform. | Creates a formal mechanism for image quality, qualifications, safety, and continuous improvement. |
| Digital interoperability | Integration of EHR, CPOE, RIS, PACS, reporting, dose monitoring, and follow-up tools. | Interface logs, EHR audit trails, architecture inventory. | Closed-loop programs require auditable notification and follow-up capability (Schwartz et al., 2021). |
| Governance and leadership | Named quality owners, multidisciplinary council, executive sponsor, escalation pathways, resource allocation process. | Charters, dashboards, corrective-action logs. | Sustains improvement beyond one-time projects and makes ownership visible. |
Radiologist fatigue and burnout are not merely satisfaction concerns. Systematic reviews describe fatigue as present in radiology and potentially associated with diminished diagnostic performance (Stec et al., 2018). Coverage adequacy should not be reduced to vacancy counts alone: monitor worklist volatility, studies per staffed hour, overnight burden, high-complexity case mix, and delay accumulation by shift, while retaining human review of whether the workload design is clinically sustainable.
Process Domain: Appropriate, Safe, Timely, and Communicative Imaging
Process measures represent the operational behavior of the imaging system and are typically more actionable than outcomes because the department can redesign them directly. Measure with timestamp-based data wherever possible, reserving manual audit for validation and nuanced review. The three engines below model the highest-leverage process domains.
Turnaround & coverage model
Turnaround stratified by care setting
Closed-loop communication calculator
Closed-loop success
Appropriateness and decision-support impact
Balancing measures to monitor alongside this estimate: delayed needed imaging and repeat-exam rate, so appropriateness gains do not become access harms.
Outcome Domain: Diagnostic, Safety, Experience, and Clinical Measures
Outcomes are the ultimate test of whether a radiology system helps patients, and the hardest to define and attribute. A valid framework avoids relying on generic hospital mortality or satisfaction scores, focusing instead on outcomes proximate to imaging and linked to specific pathways.
| Outcome family | Operational definition | Evidence and management link |
|---|---|---|
| Diagnostic accuracy and concordance | Clinically significant discrepancies, pathology-radiology concordance, second-review changes, interval cancers where applicable. | Breast imaging and selected oncology pathways offer mature, benchmarkable approaches. Subspecialty review can alter interpretation and treatment planning in complex cases. |
| Time to treatment and follow-up | Door-to-needle, door-to-puncture, door-to-reperfusion, biopsy-to-treatment plan, follow-up completion after actionable findings. | In the ANGIOCAT trial, a direct-to-angiosuite workflow reduced median door-to-arterial-puncture time from 42 to 18 minutes and improved 90-day functional outcome in selected patients (Requena et al., 2021). |
| Patient safety | Contrast reaction and extravasation, wrong-patient or wrong-site imaging, avoidable repeats, dose outliers, missed follow-up. | Dykes et al. (2015) reported a median CT contrast extravasation rate of 0.21% in a national registry. Interpret in terms of severity and case mix. |
| Patient experience and understanding | Wait-time experience, comprehension of report, complaints, portal use, anxiety related to release of results. | Patient-oriented formats and lay explanations can improve understanding, but implementations require clear scope, clinician alignment, and careful validation. |
| Pathway-level clinical outcomes | Functional outcome, cancer detection, treatment-plan change, biopsy adequacy, technical success, complication, ED length of stay. | Use only where radiology has a credible causal role and data linkage exists. Risk-adjust when clinical severity or referral selection influences outcomes. |
Acute-stroke time-to-treatment simulator
Door-to-arterial-puncture comparison
A fast, accurate report may still be followed by delayed clinical action for reasons outside radiology, and a poor outcome can occur despite optimal imaging. The response is not to avoid outcomes, it is to select pathway-specific outcomes, apply risk adjustment where feasible, make confounders visible, and interpret results through a multidisciplinary governance forum rather than as an isolated radiology score.
Integrated KPI Framework and Data Architecture
The minimum viable scorecard should be limited enough to drive action. Excessive KPI proliferation produces reporting fatigue and obscures accountability. The recommended portfolio is six universal enterprise KPIs and four pathway-level KPIs. Build your year-one set below.
First-year portfolio builder
Matches the recommended first-year discipline.
Portfolio balance
Recommended KPI specifications
| KPI | Operational definition | Cadence | Owner | Balancing measure |
|---|---|---|---|---|
| P Closed-loop success | Percent of critical/actionable findings communicated, acknowledged, and documented within policy timeframe. | Monthly | Quality lead, informatics | Missed follow-up, response-time distribution |
| P Report TAT by setting | Median and p90 completion-to-final time for ED, inpatient, outpatient, stat. | Weekly / monthly | Section chiefs | Meaningful discrepancies, ED length of stay |
| S Coverage adequacy | Vacancy rate, uncovered shifts, volume per staffed shift, subspecialty routing compliance. | Monthly | Operations | Worklist carryover, overtime, fatigue indicators |
| P Appropriateness / CDS | Percent of advanced orders rated usually appropriate, override pattern by service line. | Monthly | Informatics, referrer leaders | Delayed needed imaging, repeat-exam rate |
| P Dose and safety control | Percent of CT protocols within local benchmark, outlier rate, extravasation, avoidable repeats. | Monthly | Physicist, modality lead | Image-quality adequacy, clinical repeat imaging |
| P Structured-report adoption | Percent of priority studies using approved templates and coded recommendation fields. | Monthly | Subspecialty chief | Report TAT, referrer satisfaction |
| O Stroke pathway outcome | Door-to-image, door-to-puncture, door-to-reperfusion, functional outcome where registry exists. | Monthly / quarterly | Stroke program | Hemorrhage, transfer delay, selection bias |
| O Breast / oncology outcome | Cancer detection, recall and PPV where relevant, pathology concordance, MDT influence. | Quarterly / annual | Breast / oncology leads | Case mix, follow-up completeness |
| O Incidental-finding follow-up | Percent of actionable nonemergent findings with documented intended follow-up and completion. | Monthly / quarterly | Navigator, quality lead | Patient access and contact failures |
| O Interventional outcome | Technical success, complications, specimen adequacy, symptom and function outcome, readmission. | Quarterly | IR service lead | Case complexity, patient acuity |
Display operational measures as run charts or statistical process control charts with explicit time windows. Stratify by modality, site, acuity, daypart, language preference, payer category, and disability accommodation when sample size permits. Avoid a single aggregate that conceals service-line variation.
High-Impact Clinical Pathways
Outcome attribution is strongest where radiology has a credible causal role and data linkage exists. These five pathways anchor the outcome layer of the scorecard.
Acute stroke
Imaging directly influences treatment eligibility and timing. Door-to-imaging and door-to-interpretation are central measures. ANGIOCAT showed direct transfer to the angiography suite reduced arrival-to-puncture time and improved 90-day functional outcome in selected large-vessel-occlusion patients (Requena et al., 2021). The generalizable levers: prebuilt protocols, rapid activation, clear roles, real-time interval review.
Breast imaging and oncology
The most mature radiology outcome infrastructure. Cancer detection rate, recall rate, PPV, sensitivity, specificity, and interval cancer rate link to registry and pathology data. MQSA requires an annual medical-outcomes audit. Extend this logic to other oncology services through structured staging reports, pathology correlation, MDT participation, and tracking of treatment-plan influence.
Emergency radiology
Focus on reliability of access, availability of final reports, critical-result communication, and downstream disposition. Lean redesign can reduce CT acquisition delays, and continuous attending coverage can reduce turnaround. The risk is over-focusing on speed: pair turnaround with discrepancy surveillance, repeat imaging, consultation impact, and clinician feedback.
Interventional radiology
A hybrid approach, both diagnostic and therapeutic. Structure: operator credentialing, nursing and anesthesia support, equipment readiness, emergency coverage. Process: time-out reliability, anticoagulation review, prophylaxis, specimen handling, follow-up. Outcomes should be procedure-specific: technical success, diagnostic yield, complication rate, symptom relief, functional recovery, unplanned utilization.
Outpatient and routine imaging
Often where access and communication failures accumulate silently. Priority measures: order-to-schedule time, cancellations and no-shows, authorization delays, completion, report availability, patient-friendly explanations, and follow-up closure for actionable incidental findings. Stratify by travel distance, language need, insurance category, disability accommodation, and scheduling channel to reveal inequities that aggregate averages conceal.
AI, equity, and patient-centered reporting
Controlled, not presumed
Position AI as a structural and process intervention whose benefit must be shown locally. Specify use case, eligible population, comparator workflow, responsible reviewer, performance threshold, escalation, equity analysis, and stop criteria. Evaluate as part of an SPO chain, not as a standalone product metric.
A measurement discipline
Apply equity to structure, process, and outcomes, not as a separate add-on. Do not read group differences as patient noncompliance without examining the structural and process conditions (transportation, digital access, language, work schedule, financial exposure, authorization, contact-data quality) that precede them.
Understandable and actionable
A plain-language layer should not replace clinical counseling, it should make the report clearer: lay summary, defined terminology, explanation of the impression, explicit next steps, contact pathway. Pilot and measure comprehension, satisfaction, anxiety, message volume, and clinician concerns (van der Mee et al., 2025).
Evidence-Based Intervention Portfolio
Prioritize changes that are both measurable and causally plausible. The strongest initial interventions are not the most technologically sophisticated, they are the ones that close known failure points: delayed communication, fragmented follow-up, avoidable access delay, inappropriate ordering, unsafe protocol drift, and lack of learning from discrepancies.
Prioritization matrix: resource intensity versus expected value
| Intervention | SPO target | Core steps | Resource | Evidence |
|---|---|---|---|---|
| Closed-loop critical findings program | Process to outcome | Alert tiers, embedded notification, acknowledgment and escalation, navigator pathway for nonemergent findings. | Medium-high | High |
| Report TAT and coverage redesign | Structure plus process | Demand and capacity modeling, worklist routing, dedicated acute coverage, escalation plan, report-status visibility. | Medium-high | Medium-high |
| Appropriateness / CDS deployment | Structure plus process | Integrate ACR criteria at order entry, feedback to high-variance ordering groups, govern overrides. | Medium-high | Medium |
| Dose monitoring and protocol optimization | Structure and process to safety outcome | Automate dose capture, benchmark, identify outlier protocols, quarterly physics-modality review. | Medium | Medium-high |
| Structured reporting in high-impact pathways | Process to outcome | Limited template set, required fields, coded recommendation capture, co-design with referrers. | Medium | Medium |
| Peer learning and clinical audit | Governance plus process | Just-culture review, targeted audit, action closure, feedback loops. | Low-medium | Medium |
| Patient-friendly report layer | Process to experience outcome | Lay summary, glossary, visual aids, explicit next steps, pilot comprehension testing. | Low-medium | Medium |
| Pathway redesign: acute stroke | Process to clinical outcome | Prenotification, predefined imaging, direct routing, real-time interval review, multidisciplinary simulation. | High | High |
Governance, Implementation Roadmap, and Evaluation
A functional SPO program requires governance with decision rights, not merely a committee calendar. A radiology quality and outcomes council, led by a radiology executive or medical director, reviews performance, validates data, identifies variation, authorizes corrective action, and escalates resource constraints.
Eighteen-month implementation roadmap
Foundation
Pilots
Standardization
Outcome linkage
Evaluation
Evaluation gates
A change should not be declared successful simply because a dashboard metric improved. It should also demonstrate no material deterioration in its balancing measures.
Strategic recommendations
SPO Maturity Self-Assessment
Rate your department across twelve items, four per Donabedian domain, on a 1 (not in place) to 5 (optimized) scale. The radar profile and tier update live, showing where capability, execution, or outcome measurement is lagging.
Rate the items to generate your profile.