Every licensee wants to give good care. Quality improvement (QI) is the discipline of finding out, with data, how good the care actually is, and then changing the system so that it gets better and stays better. This lesson sets out the language US health care uses for quality and shows how QI differs from the activities it is often confused with.

In this lesson, you will be able to

Define health care quality using the National Academy of Medicine (formerly Institute of Medicine) definition and its six aims.
Use Donabedian's structure, process and outcome model to classify a quality problem.
Distinguish quality improvement from quality assurance, clinical audit, research and root cause analysis.

A working definition of quality

The Institute of Medicine (now the National Academy of Medicine) defined quality of care in 1990 as the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge. Two ideas sit inside that sentence. Quality is about likelihood: good care raises the odds of a good result even when an individual outcome is poor. And quality is tied to current knowledge: what counted as good care ten years ago may not count today.

In 2001 the Institute's report Crossing the Quality Chasm described six aims that US health care should pursue. They are still the most widely used checklist for asking whether care is good.

Aim	What it asks	Example of a gap
Safe	Does care avoid harming the people it is meant to help?	Falls with injury on a medical unit
Effective	Do patients receive care that evidence shows works, and avoid care that does not?	Antibiotics prescribed for viral bronchitis
Patient-centered	Is care respectful of and responsive to the patient's preferences, needs and values?	Discharge instructions given only in English to a patient who reads Spanish
Timely	Are harmful waits and delays reduced?	Abnormal mammogram results with no follow-up for six weeks
Efficient	Is waste of equipment, supplies, ideas and energy avoided?	Duplicate laboratory tests ordered on admission
Equitable	Does quality vary because of personal characteristics such as race, ethnicity, sex, income or where people live?	Lower blood pressure control rates for one patient group in the same clinic

The six aims are sometimes remembered by the acronym STEEEP. A useful habit is to ask which aim a problem threatens. A long wait for a dental appointment is a timeliness problem; if it affects only patients on Medicaid, it is also an equity problem.

Structure, process and outcome

Avedis Donabedian, a physician and health services researcher, proposed in the 1960s that quality can be judged through three linked lenses. Structure is the setting in which care happens: staffing, equipment, information systems, policies. Process is what is actually done for and with patients: assessments, prescriptions, education, follow-up. Outcome is the effect on the patient's health and experience. Good structure makes good process more likely, and good process makes good outcomes more likely, but none guarantees the next. Most improvement work changes processes and checks the effect on outcomes.

QI and its neighbors

Activity	Main question	Typical output
Quality assurance (QA)	Are we meeting the required standard?	Pass or fail against a threshold; corrective action for outliers
Clinical audit	How does our practice compare with an explicit standard, and did it change after we acted?	Compliance rates before and after change; re-audit
Quality improvement	How can we make this process better, and how will we know it improved?	Tested changes, run charts, a new standard way of working
Root cause analysis	Why did this event or close call happen, and what will stop it recurring?	Causal statements and an action plan for one event or cluster
Research	What is true in general, beyond this setting?	Generalizable knowledge, usually under IRB oversight

The boundaries are not rigid. A good audit leads into an improvement project; an RCA often recommends changes that are then tested as QI. The difference that matters most is in the mindset. QA asks whether people met the standard; QI assumes that most problems come from the way work is designed and asks how to redesign it. W. Edwards Deming, whose ideas underpin modern QI, made the point that most variation in performance belongs to the system rather than to the individual working in it.

Worked example: one problem, two mindsets

A family medicine practice in Wisconsin notices that only about half of its patients with diabetes have had a retinal eye exam in the past year. A QA approach would remind clinicians of the guideline. A QI approach asks how the work is organized: who identifies patients due for the exam, whether the referral is easy, whether results come back into the record, and whether patients face barriers such as transport.

Analysis: the problem is a process gap (Donabedian) that threatens effectiveness and, if some groups are affected more, equity. The improvement team will measure the eye exam rate (outcome of the process), test changes such as a standing order or an in-office retinal camera, and track the results over time.

Apply it

Pick one recurring frustration in your own work, such as a delay, a duplicated task or a result that gets lost. Which of the six aims does it threaten? Is it mainly a structure, process or outcome problem?

Lesson 1.1 · Key points

Quality is the degree to which care increases the likelihood of desired outcomes and is consistent with current knowledge.
The six aims are safe, effective, patient-centered, timely, efficient and equitable.
Donabedian's model links structure, process and outcome; most QI changes processes and measures outcomes.
QI differs from QA, audit, RCA and research mainly in its question and in its focus on redesigning systems rather than blaming individuals.