Presenting Outcomes Data in DNP Manuscripts

The outcomes section of a DNP project manuscript holds an unusual position in doctoral writing. It is a clinical document: a clinical leader should be able to read it and act on it, and practitioners, administrators, and policymakers will evaluate its findings in terms of their implications for care delivery. It is also a scholarly document, one a doctoral committee will evaluate as evidence of advanced clinical scholarship and that has to meet the methodological and presentational standards of a doctoral institution. The two sets of requirements are compatible. They are still in tension, and negotiating that tension is one of the distinctive challenges of DNP writing.

Statistical sophistication is where it shows. A PhD dissertation's results chapter demonstrates methodological rigor through the selection and correct application of inferential statistical tests, the reporting of effect sizes and confidence intervals, and the appropriate hedging of claims to match the design's capacity for inference. A DNP project's outcomes section is often working with smaller samples, shorter time frames, and simpler analytical approaches — pre-post comparisons, run charts, descriptive statistics.

Those approaches would be methodologically insufficient for a research dissertation. They are appropriate to the quality improvement context and its specific evaluative standards.

The appropriate analytical frame. The outcomes section of a DNP project should be explicit about what kind of evidence it is presenting and what kind of claims that evidence supports. Pre-post data from a QI project does not support causal inference in the way that a controlled experimental design does. It supports a more modest claim: that the outcome measure changed in the expected direction during the implementation period, in a context where the proposed intervention was being applied. Say so plainly. The discussion should be honest about what this shows and what it does not show, without either overclaiming causation or underselling the clinical significance of the observed change.

Run charts and statistical process control. Run charts are the standard tool for displaying QI outcome data over time, and their interpretation follows specific rules — rules about the number of consecutive data points above or below the median that signal a non-random shift, and about the number of ascending or descending points that signal a trend. Apply them, and report them. Run chart data displayed without these interpretation rules lacks the analytical framework that gives it meaning. Statistical process control charts — particularly P-charts for proportion data and X-bar charts for continuous data — provide a more sophisticated analysis and should be used when the data and the author's methodological preparation support them.

Connecting outcomes to the PICOT question. The outcomes section should return explicitly to the PICOT question established at the beginning of the project and answer it directly. A reader holding the PICOT question and the outcomes section side by side should be able to determine clearly whether the project achieved the outcome specified in the question, in the population specified, within the time frame specified. Outcome data presented without that reconnection leaves the answer to the project's central question implicit, in the one place in the document where explicitness is most essential.

Sustainability and spread. Many DNP programs require the outcomes section or the discussion to address the sustainability of the implemented change — whether the practice improvement can be maintained after the project period ends, what resources it requires, and whether it could be spread to other units or settings. The question is clinical as much as scholarly. It calls for the same honest analysis the rest of the document requires: what does the evidence from the implementation actually suggest about sustainability, and what conditions would need to be in place for the change to be maintained and extended?