HealthITPrep
CPHIMS guide Concepts Features Pricing FAQ
Sign in Start free
العربية

Donabedian's model: structure, process and outcome

By Adel Moustafa, CPHIMS® · Updated 22 October 2026

The three kinds of quality measure, how to tell them apart, why each is useful and where each misleads, and how the Swiss cheese model explains the errors that measures try to prevent.

In short: Avedis Donabedian divided the measurement of health care quality into three parts. Structure is the setting in which care is given: staff, equipment, systems, policies. Process is what is actually done for and with the patient. Outcome is the effect on the patient's health. Good structure makes good process more likely, and good process makes good outcomes more likely.
The three parts An example Telling them apart Strengths and limits Balancing measures Swiss cheese Common mistakes Summary Sources

What are structure, process and outcome?

  • Structure: the conditions of care, before any patient arrives: the number and training of staff, equipment, buildings, information systems, policies. "Is there a cardiac catheterisation lab available around the clock?"
  • Process: what is done in giving and receiving care: tests, treatments, timing, documentation, communication. "Did the patient with a heart attack get aspirin on arrival?"
  • Outcome: the result for the patient's health: survival, complications, readmission, function, and also the patient's experience and satisfaction. "Was the patient alive 30 days later?"

Donabedian first set the model out in 1966 and it is still the most common way to sort quality measures.

What does it look like for heart attack care?

TypeExample measureWhat it tells
StructureA catheterisation lab and an interventional cardiologist available 24 hours a dayWhether the means for good care exist
ProcessAspirin given on arrival; time from arrival to opening the artery; beta-blocker prescribed at dischargeWhether the right things were done, and on time
OutcomeDeath within 30 days; readmission within 30 daysWhat happened to the patients' health

How are the three told apart in a question?

  • If it existed before the patient came (a resource, a qualification, a system, a policy), it is structure.
  • If it is an action or a time to act by the care team, usually written as "the percentage of patients who received…", it is process.
  • If it is a change in the patient's state (alive, healed, infected, readmitted, satisfied), it is outcome.

A test: an information system is structure; the share of orders entered through it is process; fewer adverse drug events as a result is outcome.

What are the strengths and limits of each?

  • Structure: easy to check and to report, but the weakest link to results: having a lab does not mean it is used well.
  • Process: quick to measure, directly under the team's control, and it shows what to fix. It is valid only when the process is proven to improve outcomes.
  • Outcome: what matters most to patients, but it needs time, enough cases, and risk adjustment: a hospital that treats sicker patients will have worse raw outcomes without giving worse care.

So a good set of measures uses all three. Information systems make this possible: much of the data comes from the EHR, and in the United States many quality measures are specified to be calculated electronically from it (electronic clinical quality measures, eCQMs).

What is a balancing measure?

The Institute for Healthcare Improvement adds a fourth kind for improvement projects: a balancing measure checks that a change does not cause a problem somewhere else. If a ward works to shorten length of stay by discharging patients earlier, readmissions are the balancing measure: they show whether the earlier discharges are harming patients.

How does the Swiss cheese model relate?

James Reason's Swiss cheese model explains how harm happens in complex systems. Each defence, such as the prescriber, the pharmacist's check, the alert, the nurse at the bedside, is a slice of cheese with holes in it. The holes are active failures (slips and mistakes at the sharp end) and latent conditions (weaknesses built into the system, such as a confusing screen, understaffing or a missing alert). Harm reaches the patient only when the holes in several slices line up.

The lesson is a systems approach rather than a person approach: instead of only blaming the last person, find and close the latent conditions, which is what a root cause analysis does. It also explains why good structure and process matter: they add slices and shrink the holes.

What do candidates get wrong?

  • Calling a timing measure an outcome. "Minutes from arrival to treatment" is a process measure: it describes what the team did.
  • Calling an information system an outcome or a process. The system is structure; how it is used is process.
  • Comparing raw outcome rates between hospitals. Without risk adjustment the comparison punishes those that take sicker patients.
  • Treating patient satisfaction as a process. In Donabedian's model the patient's experience and satisfaction are usually counted among outcomes.
  • Blaming the last person in the chain. The Swiss cheese model points to the latent conditions that let the error through.

Summary for revision

  • Structure: what exists before the patient arrives (staff, equipment, systems, policies).
  • Process: what is done, and when ("percentage of patients who received…").
  • Outcome: what happens to the patient's health (death, complications, readmission, satisfaction); needs risk adjustment.
  • Balancing measure: checks that an improvement does not harm something else.
  • Swiss cheese: harm when holes in several defences line up; fix latent conditions, not only people.

Sources

  • Donabedian A., "Evaluating the Quality of Medical Care", Milbank Memorial Fund Quarterly (1966); and "The Quality of Care: How Can It Be Assessed?", JAMA (1988).
  • Institute for Healthcare Improvement (IHI): the Model for Improvement and its outcome, process and balancing measures.
  • Reason J., "Human error: models and management", BMJ (2000): the Swiss cheese model.

Related free practice questions

  • Donabedian’s framework: structure, process and outcome
  • The Swiss cheese model

The site files this concept under the domain Clinical Informatics.

Try exam-style questions

A 20-question trial exam with an explanation for every option and a score per domain, drawn from 40 free questions. No payment method needed.

Start the trial exam

Questions about the exam: CPHIMS guide · What is CPHIMS? · Exam cost · Eligibility · Passing score · Renewal · Exam day · How long to study · Is it worth it? · What is CAHIMS?

Concepts explained: HL7 v2 vs FHIR · Capitation vs fee-for-service · Safe Harbor vs Expert Determination · Incident vs problem vs change · Backup types, RTO and RPO · Privacy Rule vs Security Rule · CDS five rights · Lewin vs Kotter · SDLC and go-live strategies · Testing types · Earned value (SPI, CPI) · HIE models · Six Sigma vs Lean · Levels of interoperability · Bandwidth vs latency · EMTALA · Risk responses · RACI matrix · Build vs buy · Sunk vs opportunity cost · Data governance

The four domains: Healthcare and Technology Environments · Clinical Informatics · Healthcare Information and Systems Management · Management and Leadership

Written by Adel Moustafa, About us

Written and checked against the sources above on 22 October 2026. An independent site, not affiliated with or endorsed by HIMSS. CPHIMS® is a registered trademark of HIMSS.

CPHIMS guide Compare certifications Free practice questions Concepts explained About us Terms of Use Privacy Policy Refund Policy Contact us

© 2026 HealthITPrep. This site is not affiliated with or endorsed by HIMSS. CPHIMS is a registered trademark of HIMSS.