Quality of care: structure, process and outcome
- Concept
Quality of care is measured in three layers: what care is given with, what is done, and what happens to the patient.
What it is
Avedis Donabedian's model assesses quality of care through three linked components: structure (the resources and organisation where care is given), process (what is done in giving and receiving care) and outcome (the change in the patient's health attributable to care). Each component influences the next. It underpins quality indicators and outcomes research.
Worked example
- Data (made up): two hospitals with 500 operations each; surgical site infection is measured.
- Process: North gives prophylaxis on time in 450 / 500 = 90 %; South in 300 / 500 = 60 %.
- Crude outcome: North 15 / 500 = 3 % infections; South 20 / 500 = 4 %; the gap looks small.
- Risk adjustment: given the patient mix, 18 infections were expected in North (sicker patients) and 12 in South.
- Observed / expected ratio: North 15 / 18 ≈ 0.83; South 20 / 12 ≈ 1.67.
- Reading: South's O/E ratio is double North's (1.67 / 0.83 ≈ 2), and its process indicator shows what to fix first.
The three components of the Donabedian model
| Component | Example indicator | Advantage | Limitation |
|---|---|---|---|
| Structure | Nurses per bed, electronic record system | Easy to verify | Its link to outcomes is often weak |
| Process | % of operations with prophylaxis on time | Shows what to fix and responds fast | Worth it only if the process improves outcomes |
| Outcome | Risk-adjusted surgical site infection | Measures what matters to patients | Needs risk adjustment and many cases |
Markers
- Structure
- Process
- Outcome
- Risk adjustment
- O/E
In practice
- An improvement programme usually starts with evidence-based process indicators and then checks the effect on outcomes.
- Outcomes research studies what happens to patients in real care, including outcomes they report themselves.
- Indicators are defined with a numerator, denominator, source and period, just like a rate.
Common pitfalls
- Comparing crude outcomes between centres that treat patients of different severity.
- Measuring a process with no proven link to the outcome.
- Judging a hospital on an outcome with very few cases: chance dominates.
- Assuming good structure guarantees good care.
Test yourself
Think of the answer, then tap the question to see it.
What are the three components of the Donabedian model?
Structure (resources and organisation), process (what is done in care) and outcome (the effect of care on the patient's health).
Why are outcome indicators risk-adjusted?
Because outcomes also depend on patients' severity and characteristics; without adjustment, a centre that takes sicker patients looks worse.
What advantage does a process indicator have over an outcome indicator?
It needs fewer cases, responds quickly and points to what to fix; but it is useful only if the process is linked to better outcomes.
What does an observed / expected ratio of 1.67 mean?
That the centre had 67 % more events than expected for the mix of patients it treated.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: Performance Measurement: Accelerating Improvement. Appendix E: Methodology and Analytic Frameworks (The National Academies Press, 2006); Quality of care (World Health Organization, 2026).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.