Chronic Medication Use and Inflammatory Skin Diseases: The Power and Limitations of the Case-Control Study

By Robert S. Stern, M.D.
Published in J Invest Dermatol 2007,127;12:2709–2712

Introduction

This commentary focused on the difficulties associated with reliably determining if the chronic use of specific medications is associated with skin diseases. Since placebo controlled, randomized and blinded prospective trials are most often impractical for this type of a question, one must usually resort to other methods that are more subject to potential bias.

Of particular importance is the case-control study wherein prospectively defined cases with the disease of interest are compared to concurrently identified matched controls as similar to the disease cases as possible in all other characteristics. The author chose to look closely at such a study that found a relationship between the chronic use of calcium channel blockers (CCBs) and newly diagnosed eczema in the elderly.

Results and DiscussionThe study being examined used the most reliable of the three established approaches for this type of question, the case-control study. This study design had less potential for bias than spontaneous reports or retrospective analysis of clinical experience, the other two accepted methods.

Additionally, the authors brought forth substantial other information, all with more potential bias than the main study, that did, however, support the link between CCBs and the development of eczema.

Of particular importance was the use of a hypothesis-driven retrospective review of cases of eczema in the elderly wherein removal of CCBs resulted in improvement in their eczema that was more pronounced than when other of the patients’ drugs were withdrawn.

Data from spontaneous reports was used that showed an association between the use of CCBs and eczema, and improvement when the CCBs were withdrawn; in a small number of cases eczema returned on rechallenge with the drug. The author of the commentary noted that in the retrospective results the impact of CCBs on eczema was much more pronounced than in the case-control study, possibly due to the potential biases associated with the retrospective approach, including bias in the selection of patients for reporting, drug withdrawal and re-challenge.

In the case-control study there was evidence that the cases were sicker and on more drugs than the controls, allowing for speculation that cardiac disease or multiple drug use may have been responsible for some cases rather than CCBs per se.

The author also noted that the case control study would have been much improved by more detailed description of the characteristics of each patient’s eczema and that the referral/specialized practice setting from which the cases came might indicate that the problem is less frequent in the community.

Overall, the author found the evidence convincing that in about 50% of patients taking CCBs who develop otherwise unexplained extensive eczema, the eczema is likely due to the CCBs, and that the data suggest that about 15% of chronic and otherwise unexplained eczema in the elderly is due to CCBs.

Reprint requests to: Robert S. Stern, M.D., Department of Dermatology, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, GZ522, Boston, MA, 02215.