Misuse of the P Value: Using Quality Improvement Analyses to Identify Clinically Significant Improvements.

Quantitative research and quality improvement (QI) both seek to improve care provided to patients. However, clinicians often blur the lines between how to appropriately analyze data from these methodologies. Clinicians may inappropriately use statistical analyses for QI initiatives, rather than using run and statistical process control (SPC) charts to analyze improvements in outcomes.The purpose of this article was to address the analytic methods used for QI initiatives in the clinical setting in an effort to show clinicians how to identify meaningful improvements in clinical practice.In this article, we provide an example comparing the same evidence-based practice/QI initiative (chlorhexidine gluconate bathing in a medical intensive care unit) using a quasi-experimental pretest/posttest research design with statistical analyses completed with t tests with analyses using run and SPC charts to show the data trended over time. Using a pretest/posttest design, chlorhexidine gluconate bathing compliance improved from 63% to 65%, a nonsignificant change, P = .075. These same data plotted on run and SPC charts, however, show a shift and a trend, indicating clinically significant improvements per QI methodologies.The example in this article highlights the pitfall of relying only on statistical analyses and P values to determine the importance of a clinical project, and provides a practical example for how run or SPC charts can be used to identify improvements over time.

Duke Scholars

Author Staci Reynolds School of Nursing