How many times have you read a research article that concluded with – the study findings are “statistically insignificant” or “not statistically significant”. What does this really mean from a clinicians perspective. Can research findings of statistical insignificance still be of clinical significance and or importance that can lead clinicians to change their management of patient care.
This is a difficult question and is quite complex with many layers however it does require answering. Nurses need to have a good understanding of what the term statistically significant actually means. Research and an understanding of research is very important if nurses are to ensure our individual and our collective nursing practise is evidenced based.
Primarily the purpose of research is to generate and answer questions and to provide a framework for researchers to work within in an effort to establish a level playing field. The research question (hypothesis) is identified and the researcher generates the evidence for and against a hypothesis.
Statistical significance basically means that from a statistical point of view, the study result was not due to chance. The two most common measures for statistical significance is probability value (p-value) and confidence interval (CI).
P-value is measured as a value from 0-1 with the lower the measure the lesser the result being attributed to chance. A p-value (p < 0.05) is often considered significant, but the lower this figure, the stronger the evidence.
CI estimates the range within which the real results would fall if the trial is conducted many times. Hence, a 95% CI would indicate the range between the two treatments would fall on 95% of the occasions, if the trial is carried out many times
The research paper provides a summary of the research method, results and provides information around the statistical significance. In making a decision as a clinician (individually or as a group) other aspects should be taken into consideration.
Clinical significance (clinical importance), can be viewed as a difference between two therapy results that is large enough to justify the changing of management of a patients care. Caution is required though because whilst the research outcome is statistically significant the real difference may be too small to support a move to change a current practice. It’s important for clinicians to understand that statistical significance doesn’t necessarily equate to clinical importance.
Ultimately, in order to choose among different
treatments, clinical physicians have to consider
not only the P-value of the latest published paper,
but also the magnitude of benefit of each
treatment, side-effect profiles, direct and possibly
indirect costs, patients’ preferences and even
their own comfort with prescribing a new therapy.
Regular reading of research papers and being involved in professional discussions regards the research outcomes facilitates greater understanding of the clinical importance and implications for management of care.
Journal Clubs are a well established strategy for this and are well entrenched in many professions especially within medicine.The nature of nursing does present barriers to nurse participation. More effort is required by nurse leaders in breaking down the barriers to optimise nurses regular participation in research reading and discussion.