 |
 |
 |
 |
 |
 |
 |
|
|
|
|||||||||
|
|||||||||||
Research Fellow Division of Pharmacoepidemiology and Pharmacoeconomics Brigham and Women's Hospital 1620 Tremont Street, Suite 3030 Boston, Massachusetts 02120 fax 617/232-8602 jgagne{at}hsph.harvard.edu
Research Assistant Professor Department of Health Policy Jefferson Medical College Philadelphia, Pennsylvania
Published Online, March 3, 2009. www.theannals.com, DOI 10.1345/aph.1K635a
Before accepting the authors' conclusion indiscriminately, practitioners
should closely consider the results that gave rise to this interpretation. The
authors assert that self-efficacy increased significantly as a result of the
intervention. Indeed, a small increase in the self-efficacy measure of 3.75
points from baseline to follow-up at 4 months was observed among the
intervention group. However, a similar increase of 3.24 points was observed
among the control group. The authors conducted a Mann-Whitney U test
to compare these changes and reported a corresponding p value of 0.3332, a far
cry from traditional 
levels (eg, 0.05) used in hypothesis testing in
medical research. The correct interpretation of these results is that the care
program had little to no effect on self-efficacy compared with standard
pharmacy care.
A similar analysis was conducted for measures of quality of life pertaining to both physical and mental health. No difference was observed for the physical health measure in either group or between groups. As with the self-efficacy measure, a small improvement (2.73 points) in the mental health component of the Short-Form 36 was observed from baseline to end of follow-up among those in the intervention group. A smaller change of 0.95 points was observed in the control group. The between-group comparison of the difference between these observations (a difference of 1.78 points) yielded a p value of 0.027, indicating that the 2 observations are statistically different by conventional thresholds. While these values are statistically significant, practitioners should consider the clinical significance of a 1.78-point improvement in mental health score. However, the authors provide no contextual guidance to determine the relative clinical importance of such a seemingly small numeric improvement.
Practitioners should critically evaluate the small improvement in mental health score within the overall context of the pharmaceutical care program that failed to demonstrate marked benefit over standard pharmacy practice with respect to improvements in days per month with headache, headache attacks per month, intensity of untreated and treated pain, self-efficacy, and physical health quality of life. In light of the lack of evidence supporting the benefits of the pharmaceutical care intervention that this study produced, including the lack of benefit in self-efficacy improvements despite what the authors claim, the conclusions reached by the authors are unfounded. Future evaluations of pharmaceutical care plans must carefully consider not only ways to improve the programs themselves, but also ways to more appropriately measure, compare, and report the outcomes.
Footnotes
Comments on articles are submitted to the authors of those articles. When no reply is published, either the author chose not to respond or did not do so in a timely fashion. Comments and replies are not subject to peer review.–ED.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
 |
 |
 |
 |
