Reading and Understanding Study Results: Implications for Practice

AADE Program Session
Friday, August 13, 2004
10:00 AM - 11:30 AM

Clinical practice should be based on research. The conscientious use of current best evidence in making decisions about patient care improves outcomes by 28%. When reading study data, it helps to have a basic understanding of scientific research. Clinicians should keep in mind that they are looking to find the results, determine whether the results are valid, and assess whether the results will help in caring for patients.

Here are 5 steps that can be used in formulating an evidence-based practice:

1. Identify the clinical problem and ask the clinical question
2. Systematically collect the most relevant and best evidence by literature search
3. Critically appraise the research evidence
4. Integrate all the evidence with one's clinical expertise; incorporate patient beliefs, values, and preferences in making practice decisions
5. Evaluate the practice decision

Research reports are comprised of the following components: introduction (which defines the purpose/objectives/hypothesis), methods, results, and discussion. The most efficient way to review medical literature is to read the abstract to determine if the research questions address the clinical question, and, if so, to move directly to the discussion. If the information gleaned from this process seems helpful, the reader can look at the methods and results more carefully. The careful reader will ask whether the study can be replicated and whether the study measured what it was designed to measure (does it have validity).

It is important to determine the study design and apply that information to the clinical question. The experimental design using a randomized controlled trial and dependent and independent variables is the gold standard. Other experimental forms may serve the purpose, however, including quasi-experimental studies, which do not use randomization or control, non-experimental studies, which are mostly descriptive, and methodological studies, which help determine what screening or testing tools are most useful.

P values, used in experimental studies, are important because they indicate the probability that the results were obtained by chance. P values of 0.05 and less are generally considered significant. Descriptive studies will often try to identify a correlation. The Pearson r, used with continuous variables, indicates the strength of a correlation. 0 equals no relationship, +1 equals a perfect positive (same) relationship, and -1 equals a perfect negative (opposite) relationship. When r values approach either +1 or -1, the correlation becomes stronger.

Other helpful pointers to remember when reading study reports include:

• Start with a reliable, peer-reviewed source
• Use randomized, controlled trials if any are available
• Note whether the sample size is large or small and rely on studies that use larger samples
• Evaluate whether or not the findings are significant
• Be mindful of the limitations of the study and decide if the conclusions are accurate
• Determine whether the study results can be applied to your practice
• Remember that evidence is continuously evolving