As INKA advances the review of levels of evidence (LOE) for nursing diagnoses, one of the most important changes is not simply numerical. It is a change in the logic used to determine what counts as evidence for a diagnosis.
The previous LOE structure and the new framework should therefore not be read as two versions of the same numerical scale. The earlier classification placed greater emphasis on the presence of diagnostic components, consistency, and supporting references. Under that approach, a diagnosis could sometimes reach a relatively high level even when much of the available literature supported the general clinical phenomenon rather than the diagnostic structure itself. The new framework asks a more demanding question: What aspect of the validity of this specific nursing diagnosis does each study actually support?
This distinction has important consequences. The new architecture organizes evidence according to stages and types of validity: diagnoses still under development are classified within Level 1, while Level 2 distinguishes conceptual validity (2.1), content validity (2.2), and clinical validity (2.3). Clinical evidence is further differentiated according to the kind of diagnostic question that a study can answer.
For this reason, a change from a number such as 3.2 or 3.3 in the previous system to 2.1 or 2.3 in the new one should not be interpreted automatically as a reduction in evidence. The scales were constructed differently and reflect different methodological assumptions. What matters is not whether the number became larger or smaller, but whether the available research provides direct, methodologically appropriate support for the diagnostic interpretation being made.
Having references is not the same as having diagnostic evidence.
One of the most challenging issues encountered during the review is surprisingly simple: the same reference list may be submitted to support more than one nursing diagnosis. This commonly occurs when investigators begin with a shared conceptual core and develop several related diagnostic proposals. For example, the same literature may be cited for a problem-focused diagnosis, a health-promotion diagnosis, and a risk diagnosis because all three concern the same broad human response.
At first glance, this may seem reasonable. The diagnoses share terminology, theoretical background, and often a similar clinical context. However, they do not make the same clinical claim.
A problem-focused diagnosis asserts that a human response is currently present and can be recognized through manifestations and related conditions. A health-promotion diagnosis concerns a person’s readiness or disposition to improve a health-related behavior or experience. A risk diagnosis states that a person is more susceptible to developing an undesirable response in the future.
These are fundamentally different diagnostic structures. Consequently, they require different forms of evidence. This is one of the central principles behind the new LOE framework: a publication about the same clinical context is not necessarily evidence for the validity of every diagnosis related to that context. The material supporting the current review explicitly emphasizes that the critical issue is not whether references exist, but whether those references support the diagnostic structure itself. Studies of interventions, clinical outcomes, or related medical conditions may provide useful background while offering little direct evidence of diagnostic validity.
Why the same study can have different value for different diagnoses.
Consider three hypothetical diagnoses derived from a common conceptual domain: one problem-focused, one health-promotion, and one risk diagnosis. Their developers might reasonably identify a common foundational literature. That literature may establish the importance of the phenomenon and explain its clinical relevance.
But after that common conceptual foundation is established, the evidentiary pathways must diverge. A cross-sectional clinical study identifying manifestations associated with the presence of a problem-focused diagnosis could provide strong evidence for its diagnostic structure. The same study may contribute very little to a health-promotion diagnosis if it does not examine readiness, motivation, subjective experience, or behaviors indicating a disposition for improvement. Likewise, it may provide insufficient support for a risk diagnosis if it does not establish whether particular factors prospectively increase the probability of developing the undesirable outcome.
The reference is the same. Its evidentiary meaning is not. This explains why diagnoses sharing the same conceptual core may appropriately receive different LOEs. The level should reflect the match between the diagnostic claimand the research design capable of testing that claim, rather than the number of publications cited.
Evidence should follow the structure of the diagnosis.
This shift has practical implications for researchers developing or revising nursing diagnoses. For problem-focused diagnoses, evidence should demonstrate that the proposed diagnosis represents a clinically identifiable construct. Studies examining defining characteristics are therefore especially important because they help establish whether the manifestations attributed to the diagnosis actually characterize the human response in clinical populations. Evidence concerning related factors is also essential because a problem-focused diagnosis involves not only recognizing the response but understanding clinically meaningful relationships associated with its occurrence. Clinical construct-validity studies, diagnostic accuracy studies, and appropriately designed investigations of causal relationships can therefore make particularly important contributions.
Syndrome diagnoses require an even more demanding evidentiary approach. A syndrome is not simply a collection of diagnoses that happen to occur in the same person. The evidence should demonstrate that the componentdiagnoses form a clinically coherent structure, potentially sharing an origin, context, or common dynamic. The current methodological discussion emphasizes that co-occurrence alone is insufficient and that syndrome validation may require evaluation of latent structure, local dependence, classes, etiological factors, and clinical utility.
For this reason, evidence appropriate for an individual problem-focused diagnosis may be necessary but not sufficient for a syndrome. In addition to evidence regarding manifestations and causal relationships, researchers should investigate whether the proposed syndrome behaves as a coherent construct. Evidence of unidimensionality or of another clearly justified common latent structure becomes particularly important. Without this step, a group of independently valid diagnoses could be incorrectly treated as a single syndrome merely because they frequently coexist.
For health-promotion diagnoses, the evidentiary question is different again. These diagnoses concern a person’s disposition to improve health, well-being, self-care, or another positive human response. Evidence limited to poor outcomes, clinical problems, or the effectiveness of an intervention does not necessarily demonstrate such readiness. Studies should explore the behaviors, meanings, experiences, and expressions that indicate a subjective disposition toward change. For this reason, qualitative validation can be particularly valuable, alongside concept analyses and subsequent content and construct-validation studies. The source material specifically identifies qualitative studies of readiness, conceptual analyses, and research conducted in relevant populations and contexts as important directions for health-promotion diagnoses.
Finally, risk diagnoses require evidence that addresses probability. Their defining question is not whether a factor is associated with an existing problem, but whether exposure to that factor is associated with an increased susceptibility to a future undesirable response. Studies should therefore quantify the magnitude of association between proposed risk factors and subsequent outcomes whenever possible. Cohort studies, case-control studies, prognostic research, predictive models, and systematic reviews of risk factors are particularly relevant. Evidence that an intervention improves an outcome, or that a medical condition commonly accompanies it, should not be treated automatically as evidence that a proposed factor predicts the nursing diagnosis.
A practical roadmap for researchers.
A useful way to approach future diagnostic research is to begin not with the question, “How many references support this diagnosis?” but with “What claims does this diagnosis make, and what evidence is needed to test each of them?”
- Problem-focused diagnosis: prioritize clinical construct validity, defining characteristics, diagnostic indicators, related factors, and evidence supporting relevant causal relationships.
- Syndrome diagnosis: include the evidence expected for problem-focused diagnoses, but also test whether the proposed components form a coherent common structure, including appropriate investigation of unidimensionality, dependence among components, and contextual or causal coherence.
- Health-promotion diagnosis: prioritize evidence concerning readiness, motivation, subjective experience, behaviors indicating a disposition to improve, and qualitative and conceptual validity before moving to broader clinical validation.
- Risk diagnosis: prioritize longitudinal, prognostic, or otherwise appropriate analytical evidence demonstrating the magnitude and direction of associations between proposed risk factors and subsequent undesirable outcomes.
The committee’s methodological approach follows the same underlying principle: studies should first be classified according to design and direct relevance to the diagnosis, their methodological quality should then be considered, and the resulting LOE should avoid overvaluing evidence that is only indirectly related to the diagnostic structure.
A lower number does not mean a weaker diagnosis.
This distinction is especially important during the transition to the new system. Examples examined during the development of the framework illustrate that a diagnosis previously classified at 3.2 or 3.3 may have its strongest directly applicable evidence identified at 2.1 or 2.3 under the new architecture. In another example, much of the literature supporting a risk diagnosis was classified at 1.3 because it did not directly validate the proposed risk structure. The appropriate interpretation is not that these diagnoses suddenly became less valid. Rather, the new framework makes more visible how directly the existing literature supports the particular diagnostic claim.
This may initially create the impression that the new system is more restrictive. In an important sense, it is. It asks researchers and reviewers to distinguish background knowledge from diagnostic evidence, indirect relevance from direct validity, and shared conceptual foundations from evidence specific to a particular diagnostic structure.
But this greater methodological specificity also creates an opportunity. It can help researchers see where evidence is actually missing. A diagnosis may have an extensive literature supporting its underlying phenomenon while lacking studies of its defining characteristics. A health-promotion diagnosis may have many intervention studies but very little research exploring readiness for change. A risk diagnosis may be supported by dozens of cross-sectional associations but lack longitudinal evidence showing that its proposed factors precede and predict the undesirable outcome. A syndrome may contain well-established individual diagnoses while still lacking evidence that they represent a coherent common structure.
In each case, the new LOE does more than classify the existing evidence. It points toward the next research question that needs to be answered.
From accumulating references to building diagnostic evidence.
Perhaps the most important consequence of the new LOE framework is therefore cultural. The objective is no longer to assemble the longest possible reference list. It is to build a body of evidence in which each study has a clear role in supporting the diagnostic interpretation.
A shared conceptual foundation will continue to be important, and related diagnoses will naturally continue to share references. But as diagnoses diverge into problem-focused, syndrome, health-promotion, and risk structures, their evidence must also become more specific.
The future strength of nursing diagnosis classification will depend not simply on how much research exists around a phenomenon, but on how clearly that research demonstrates that the diagnosis says what we believe it says, identifies what we believe it identifies, and is supported by evidence appropriate to the kind of clinical judgment it represents.
That is the central promise of the new levels of evidence: not merely a new set of numbers, but a clearer connection between concept, research design, evidence, and diagnostic judgment.
About the Author
Dr. Marcos Venícios de Oliveira Lopes, RN, PhD, FNI, is Director of Research for the International Nursing Knowledge Association (INKA) and a Full Professor at the Federal University of Ceará, Brazil. A NANDA-I (now INKA) Fellow and recipient of the Dr. Kristine Gebbie & Dr. Mary Ann Lavin Founders Award, he has more than two decades of experience in nursing diagnosis, theory development, and research methodologies, with particular expertise in the validation and analysis of nursing diagnoses. He is also a CNPq Level 1A Research Productivity Fellow and serves as a reviewer for the McMaster Online Rating of Evidence (MORE) system.
Comments (0)
Leave a comment