THE EMRICK [SPO – Structure + Process + Outcomes] MODEL

Executive Summary

The Emrick SPO Model is a board-level quality governance system that connects the organization’s capacity to deliver care, what care teams actually do, and the results patients experience. Outcomes are never read in isolation — they are interpreted through the upstream structure and process conditions that shape them.

The Governance Pathway

Structure enables process; process drives outcome; outcome trends feed back into strategic resource decisions. Each link is a tested hypothesis, not a guaranteed causal chain.

Structure

Capacity to deliver care — people, equipment, systems, accreditation, governance.

→

Process

What teams actually do — guideline adherence, safety practices, documentation.

→

Outcome

Results experienced — mortality, infections, readmissions, satisfaction, recovery.

Domain 1

Structure

76/100

Leading indicators of readiness. Capacity gaps raise the probability of downstream process failure.

20 recommended metrics

Domain 2

Process

82/100

The most actionable lever. Workflows, protocols, and compliance can be redesigned directly.

25 model metrics

Domain 3

Outcome

71/100

Lagging signals requiring risk adjustment, trend analysis, and root-cause review.

22 model metrics

Snapshot composites above are illustrative. Open the Composite Scoreboard tab to model your own values and watch domain scores recalculate live using the model’s scoring formulas.

Board Decision Brief

Board Question	Recommended Position
What is the model intended to do?	Create an integrated quality governance system that connects capacity, execution, and results.
What should the board approve?	Approval to pilot the Emrick SPO Model as the structured quality monitoring framework, beginning with a defined set of high-priority metrics.
What management infrastructure is required?	Metric owners, validated definitions, data-source mapping, a monthly executive quality review, and a quarterly board dashboard.
What should executives & directors monitor?	Not only whether outcomes moved, but whether the variation is explained by structure, process, case mix, data quality, or true care variation.
What is the expected benefit?	A clearer line of sight from resource decisions and process reliability to patient outcomes, safety, satisfaction, and accreditation readiness.

Board use note: Directors should treat the SPO report as a management accountability tool and a learning system — it should prompt explanation, intervention, and monitoring, not simple ranking.

Metrics across three domains

Analytic methods, descriptive to predictive

90–365

Day phased implementation horizon

The SPO Model Architecture

Adapted from Donabedian’s classic structure–process–outcome logic and aligned with AHRQ and CMS measurement guidance, the model organizes every measure into a disciplined causal sequence rather than a disconnected list of indicators.

Domain Explorer

Select a domain to see its governing question, primary use, and how the board should interpret it.

Domains at a Glance

Domain	Core Question	Primary Use	Board Interpretation
Structure	Do we have the capacity and infrastructure to deliver reliable care?	Resource readiness, accreditation readiness, staffing, systems, governance.	A leading signal that determines whether process reliability is feasible.
Process	Are we consistently doing the right things in the right way?	Clinical reliability, workflow standardization, safety practices, documentation.	The main operational lever for improvement.
Outcome	Are patients and the organization experiencing better results?	Safety, recovery, satisfaction, readmission, mortality, adverse events.	A lagging signal requiring risk-adjusted interpretation and root-cause review.

Scholarly Basis

Consistent with the Donabedian model (2005), AHRQ’s quality-measure classification (2015), and CMS quality-measure guidance (2024). The model treats the SPO sequence as a governance pathway and analytic hypothesis — structure should support process, and process should improve outcome — while each relationship is tested and reinterpreted as the organization learns.

A Complex-Systems Caution

Jazieh (2020) notes that structural, process, outcome, and balancing measures serve different purposes; Geary (2024) argues the Donabedian approach should be updated with a complex-systems perspective, since quality emerges from interactions across organizational levels. The model therefore avoids treating SPO as a simplistic straight line.

Why not outcomes alone? Outcome-only governance can be incomplete and unfair — patient severity, socioeconomic risk, case mix, documentation, coding, and random variation all influence results. Process-only governance can reward compliance activity without proving patient benefit. Structure-only governance overemphasizes resources without demonstrating reliable execution. The SPO architecture forces a connected interpretation that balances these risks.

Composite Scoreboard

Each metric is transformed to a 0–100 scale where 100 is target performance and 0 is the floor. Adjust the current values below to watch domain composites and the SPO profile recalculate live using the model’s scoring formulas.

Scoring Logic

Higher-is-better metric score = 100 × [(current − baseline) / (target − baseline)]

Lower-is-better metric score = 100 × [(baseline − current) / (baseline − target)]

Composite domain scores (weighted) S_t = Σ(w_i · s_i,t) P_t = Σ(w_j · p_j,t) O_t = Σ(w_k · o_k,t)

Sample Metric Inputs

A representative subset of metrics per domain. Each shows its baseline (floor) and target; drag to set the current value.

Structure

Process

Outcome

–

Structure

–

Process

–

Outcome

SPO Profile

Reading the profile. A high outcome score sitting atop weak structure or process scores is a warning, not a victory — it suggests results may not be sustainable. A favorable outcome trend should prompt the question: which structure and process changes produced it?

Correlation & Predictive Monitoring

The model’s mathematical value lies in moving beyond descriptive reporting. The board should expect management to test relationships — not merely display colors. The illustrative 12-month series below shows how the analytics behave when structure improvements precede process gains, which in turn precede outcome improvement.

Domain Composites Over Time

Illustrative monthly composites. Note how process tends to follow structure, and outcome follows process, after a plausible lag.

Correlation Matrix

Pearson correlations among the domain composites in the series above.

Correlation is not causation. A strong process–outcome association may reflect a true pathway — or unmeasured culture, patient selection, documentation changes, or case mix.

Lag Sensitivity

Correlation of process at month t with outcome at month t + lag. Outcomes often lag operational change by months.

Lagged Regression & Mediation

Lagged process model P_t = α₀ + α₁ S_(t−lag1) + γ X_t + ε_t

Lagged outcome model O_t = β₀ + β₁ S_(t−lag2) + β₂ P_(t−lag3) + δ X_t + ε_t

Mediation: does process explain how structure reaches outcome?

Structure

→α₁

Process

→β₂

Outcome

Indirect effect (via process) α₁ × β₂

Total effect β₁ + (α₁ × β₂)

Control variables (X_t) should include patient age, severity, comorbidity burden, payer mix, service line, procedure type, admission source, discharge destination, census, seasonality, staffing volatility, and social-risk variables where available. Without these, the organization may mistake case-mix variation for quality variation.

Analytic Methods

Method	Purpose	Board-Level Question
Correlation matrix	Measures association among the S, P, and O domains.	Are the domains moving together in the expected direction?
Lagged regression	Tests whether earlier structure or process predicts later outcome movement.	Are improvements appearing after a plausible time delay?
Mediation analysis	Estimates whether process explains structure’s effect on outcome.	Did resources improve outcomes by improving process reliability?
Statistical process control	Distinguishes common-cause from special-cause variation.	Is the trend meaningfully different, or ordinary variation?
Risk adjustment	Accounts for patient and population differences.	Are comparisons fair and clinically interpretable?

Metric Library & Data Dictionary

The full 67-metric inventory across all three domains. Every measure should mature into a formal data dictionary entry — with numerator, denominator, source, owner, frequency, target, and escalation threshold — before the board accepts trend interpretation.

Data Dictionary Minimum Standard

Element	Requirement
Measure name	Plain-language title and technical title.
Domain	Structure, process, outcome, or balancing.
Definition	Precise numerator and denominator with inclusion and exclusion criteria.
Directionality	Higher is better, lower is better, or target range.
Data source	Primary source system and any secondary validation source.
Owner	Executive sponsor and operational owner.
Reporting frequency	Daily, weekly, monthly, quarterly, or annual.
Escalation rule	Threshold that requires management action or board notification.

Implementation Roadmap

Implement the SPO Model as a living governance system, not a static dashboard. Begin with disciplined metric selection — resist the temptation to include every available data element — then move through five phases and a defined board action timeline.

Five Implementation Phases

Define

Board deliverable: Approved SPO metric charter.

Management work: Metric definitions, owners, data-source mapping, baseline selection.

Success: measures are documented, comparable, and usable

Pilot

Board deliverable: Pilot scorecard in the selected service line.

Management work: Build dashboard, test data feeds, validate numerator and denominator logic.

Success: leaders can explain each metric and detect obvious data defects

Analyze

Board deliverable: Correlation and trend report.

Management work: Composite scores, correlation matrix, control charts, lag analysis.

Success: variation is interpreted rather than merely reported

Govern

Board deliverable: Quarterly board SPO dashboard.

Management work: Integrate with the quality committee agenda and management action plans.

Success: board receives structured explanations and documented follow-up

Improve

Board deliverable: Annual model recalibration.

Management work: Retire weak measures, adjust weights, add balancing measures, refine thresholds.

Success: dashboard stays relevant and avoids metric overload

Board Action Timeline

Timeframe	Board Request	Management Deliverable	Evidence of Progress
0–30 days	Approve pilot charter.	Metric inventory, definitions, owners, data-source map.	Governance structure established.
31–90 days	Review pilot build.	Prototype dashboard and validation report.	Metrics render correctly; leaders can interpret them.
91–180 days	Review first analytic report.	Trend, correlation, and early process-control report.	Management distinguishes variation from actionable signal.
181–365 days	Review model maturity.	Expanded scorecard, recalibrated weights, balancing measures.	Dashboard supports strategic decision-making.
Annual	Reauthorize model design.	Metric retirement, addition, and target-reset recommendations.	Board dashboard remains relevant and lean.

Governance ownership is not blame. Frontline teams own process reliability; department directors own workflow redesign and local accountability; executives own cross-functional barriers, resource decisions, and risk adjustment; the board oversees trends, escalations, and strategic alignment.

Governance & Monitoring

Process measures change quickly and can be corrected quickly; outcome measures need longer observation windows; structure measures are reviewed whenever readiness materially shifts. The board receives a concise, interpreted report — never raw data without explanation.

Reporting Cadence

Review Level	Frequency	Primary Content	Expected Output
Frontline huddle	Daily to weekly	Process reliability, safety signals, workflow barriers.	Immediate countermeasures and escalation.
Department review	Monthly	Structure readiness, process trends, short-cycle outcomes.	Action-plan updates and owner accountability.
Executive Quality Council	Monthly to quarterly	Cross-functional barriers, risk-adjusted outcomes, resource needs.	Prioritized improvement agenda.
Board quality committee	Quarterly	High-priority SPO trends, control-chart signals, risks, response.	Oversight, strategic direction, follow-up expectations.

Balancing Measures

Balancing measures protect against improving one metric at the expense of another. The framework includes sentinel checks for unintended consequences.

Shorter length of stay

…may raise readmission risk if discharge planning and follow-up are unreliable.

Faster diagnostic turnaround

…may worsen staff workload if staffing and equipment capacity lag.

More documentation

…may improve coding completeness but burden clinicians.

Risk Controls & Safeguards

Risk	Safeguard
Correlation mistaken for causation	Require lagged models, clinical interpretation, and root-cause review before assigning causality.
Metric overload	Limit the board dashboard to high-value measures; maintain a broader management library.
Data quality defects	Require validation, source-system reconciliation, and periodic audit of numerator/denominator logic.
Outcome instability	Use appropriate denominators, confidence intervals, smoothed rates, and caution with low-volume events.
Checkbox compliance	Audit whether process completion reflects meaningful care behavior, not merely documentation.
Unintended consequences	Add balancing measures for workload, complaints, access delays, and safety-event displacement.
Equity blind spots	Stratify measures by relevant patient groups when legally, ethically, and statistically appropriate.
Delayed effects	Use time-lag monitoring to prevent premature judgment of improvement work.

Measurement alone does not improve care. The model creates visibility; leadership action creates improvement. Every unfavorable trend should have an assigned owner, a defined response plan, a timeline, and a follow-up date — and every favorable trend should be explained, so reliable practices can spread.

Board Scorecard

A working version of the board scorecard template. Enter current and target values; the signal computes automatically. A quarterly board dashboard should not simply show red, yellow, and green — it should tell a disciplined performance story.

Live Scorecard

Signals reflect how close current performance sits to target, accounting for each metric’s direction. Within 5% of target is green; within 15% is yellow; beyond that is red.

Domain	Metric	Current	Target	Signal	Management Response

Board Review Questions

Which structure measures explain the most important process failures?
Which process measures have the strongest observed association with outcome improvement?
Which outcome measures require risk adjustment before the board interprets them?
Where are unfavorable trends likely due to common-cause variation rather than true deterioration?
What balancing measures protect against unintended consequences?
Which metrics should be retired because they no longer support governance decisions?
What management actions are assigned, dated, and ready for follow-up at the next board meeting?

References

Agency for Healthcare Research and Quality. (2015). Types of health care quality measures.
Agency for Healthcare Research and Quality. (2025). AHRQ Quality Indicators: Empirical Methods, Version 2025.
Ameh, S., et al. (2017). Relationships between structure, process, and outcome… BMC Health Services Research, 17, 229.
Centers for Medicare & Medicaid Services. (2024). Quality measures.
Donabedian, A. (2005). Evaluating the quality of medical care. The Milbank Quarterly, 83(4), 691–729.
Geary, U. (2024). Healthcare quality improvement: update the Donabedian approach… Int. J. Health Planning & Management, 39(5).
Havranek, M. M., et al. (2023). Validity of 16 AHRQ Patient Safety Indicators… Int. J. for Quality in Health Care, 35(4).
Jazieh, A. R. (2020). Quality measures: types, selection, and application. Global J. on Quality & Safety in Healthcare, 3(4).
Oostendorp, R. A. B., et al. (2020). Relationships among context, process, and outcome indicators… Patient Preference & Adherence, 14.
Waqas, M., et al. (2024). Control charts in healthcare quality monitoring… Int. J. for Quality in Health Care, 36(3).

Structure

Process

Outcome

Structure

Process

Outcome

SPO Profile

Define

Pilot

Analyze

Govern

Improve

Kelly Emrick, DHSc, PhD, MBA

Kelly Emrick, DHSc, PhD, MBA