A nursing diagnosis is a clinical judgment about a person's response to a health condition or life process. It is the second step of the nursing process and the profession's most explicit claim that nurses diagnose something — just not the disease.
The distinction is the whole point. A physician diagnoses heart failure. The nurse caring for that patient may diagnose impaired gas exchange, activity intolerance, or insufficient knowledge of the medication regimen. Two patients with identical medical diagnoses can carry entirely different nursing diagnoses, and the same patient's nursing diagnoses change from shift to shift while the medical diagnosis does not.
Structure of a diagnosis
The conventional statement has three parts, an arrangement usually credited to Marjory Gordon and often taught by the initials of its components: the problem, the etiology, and the evidence. Written out, it reads as a diagnostic label, related to a cause, as evidenced by specific findings.
The middle element is what makes the diagnosis actionable. The label alone tells you what is wrong; the related factor tells you what to aim an intervention at. A diagnosis of impaired skin integrity related to immobility calls for different action than one related to moisture or to inadequate nutrition, even though the label is identical.
Four types are generally recognized: problem-focused diagnoses, describing an existing response; risk diagnoses, where the problem has not occurred but vulnerability exists and there are no signs and symptoms to cite, only risk factors; health promotion diagnoses, expressing a readiness to improve; and syndrome diagnoses, clustering several diagnoses that tend to occur together.
The taxonomy
The best-known classification is maintained by NANDA International, which grew out of the first national conference on the classification of nursing diagnoses in the early 1970s. Its current structure, adopted in 2002 and built on Marjory Gordon's functional health patterns, sorts diagnoses into thirteen domains — health promotion, nutrition, elimination and exchange, activity and rest, perception and cognition, self-perception, role relationships, sexuality, coping and stress tolerance, life principles, safety and protection, comfort, and growth and development — subdivided into classes.
Each diagnosis in the taxonomy carries a label, a definition, and diagnostic indicators — defining characteristics, related factors, or risk factors — which are what allow one diagnosis to be distinguished from a similar one. The total number of diagnoses changes with each edition as new ones are approved, existing ones revised, and some retired; published counts for the same edition frequently disagree, so a specific figure is best taken from the current edition itself rather than quoted second-hand.
The other recognized terminologies
The diagnosis taxonomy is one piece of a larger effort to give nursing a language that can be recorded, aggregated, and studied. The American Nurses Association recognizes seven standardized nursing terminologies, all of which are integrated into the National Library of Medicine's Unified Medical Language System.
Three of them divide the work between them: NANDA International covers diagnoses, the Nursing Interventions Classification covers what nurses do, and the Nursing Outcomes Classification covers the results, with the latter two developed at the University of Iowa. Used together they are meant to make a complete care plan machine-readable end to end.
Four others each cover diagnoses, interventions, and outcomes within a single system: the International Classification for Nursing Practice, maintained by the International Council of Nurses; the Omaha System, developed for community and home health; the Clinical Care Classification; and the Perioperative Nursing Data Set.
The argument for all of this is straightforward. If nursing care is recorded only as free text, it cannot be counted, compared across institutions, linked to outcomes, or costed. Nursing's contribution then becomes invisible in exactly the datasets that drive funding and staffing decisions. Standardized language is the profession's answer to that invisibility. See nurse staffing and patient outcomes.
The case against
The criticism is substantial and comes from working nurses more than from academics.
The most common complaint is that the taxonomy's phrasings are stilted — constructions no nurse would use aloud, adopted because the taxonomy requires them. Nurses report translating what they actually observed into approved language for the record, which is the opposite of what a standardized terminology is meant to achieve.
The second is that the artifact produced, the formal care plan, is generated for accreditation surveys and rarely consulted before care. Electronic records that build plans from dropdown menus have industrialized this: the plan is now produced faster and read no more often.
The third is empirical. Evidence that using standardized diagnostic language improves patient outcomes is thin, and the studies that exist are difficult to interpret because the language is almost never the only thing that changes.
The defence does not really dispute any of this. It argues that the alternative — nursing work recorded as unstructured narrative, invisible to every analysis anyone performs on health data — is worse, and that the failures are of implementation and of vendor design rather than of the underlying idea. Whether that is persuasive is one of the live arguments in the profession.
Related
Sources
- Nursing diagnosis taxonomy — NANDA International. Accessed September 8, 2026.
- Nursing Diagnoses: Definitions and Classification, 2024–2026, 13th edition — NANDA International. Accessed September 8, 2026.
- Inclusion of Recognized Terminologies within EHRs and Other Health Information Technology Solutions — American Nurses Association. Accessed September 8, 2026.
- Nursing Resources for Standards and Interoperability — U.S. National Library of Medicine. Accessed September 8, 2026.
- Nursing Process — Toney-Butler TJ, Thayer JM, StatPearls. Accessed September 8, 2026.