The Butterfly Runs Into Flak: Critically Analyzing Lunesta's Television Ad

National Public Radio's show "All Things Considered" ran a segment last week explaining how direct to consumer (DTC) advertising can make drugs appear better than they may really be.

The show focused on a single television advertisement (with the butterfly) for Lunesta (eszopiclone) marketed by Sepracor for insomnia. The show focused first on the psychology of advertising design, pointing out the following features:

Using Images to "Sell You a Feeling" - The advertisement shows a lovely woman waking up after an apparently restful sleep produced by Lunesta. A former TV producer commented, "and when she wakes up, she's just gorgeous. She sits right up and stretches and looks great. Who wouldn't want that?" Thus, "you remember what you feel longer than what you know." So the image of restful sleep swamps all those facts about adverse effects presented later.

Obscuring the Adverse Effects - A number of tactics are used to make it less likely that the viewer will remember the obligatory recitation of adverse effects. These include

• Putting the litany in the middle of the advertisement, since it is hardest to remember the middle of a presentation.
• Reciting facts quickly, and using a more advanced vocabulary than the rest of the advertisement.
• Presenting the visual images during the recitation that do not correspond to the words spoken, so "when the eye and the ear compete, the eye wins."

The show also featured an analysis by Dr Lisa Schwartz of the claims made in the ad versus a critical reading of the most relevant evidence from clinical research (that is, using the evidence-based medicine approach). Her main points were about how the advertisement:

• Enlarges the Eligible Population - The ad suggests that the drug is suitable for "millions of Americans who once had trouble turning off their restless minds," that is, a large population of people who have occasional trouble falling asleep. But clinical studies of the drug were limited to defined populations of people who fit the DSM-IV criteria for insomnia, and who reported sleeping less than 6.5 hours a night, and taking >30 minutes to fall asleep for at least a month. Thus the ad appears to be trying to expand the population of patients for whom the drug is marketed well beyond those patients who were in the clinical trials of the drug.
• Exaggerates Results About Amount of Sleep - The ad suggests that people can sleep all through the night on Lunesta, implying that it leads to eight hours of sleep. Schwartz noted that the largest trial of Lunesta [Krystal AD, Walsh JK, Laska E et al. Sustained efficacy of eszopiclone over 6 months of nightly treatment: results of a randomized, double-blind, placebo-controlled study in adults with chronic insomnia. Sleep 2003; 26: 793-799.] only showed that patients given Lunesta slept 37 minutes longer, for an average of 6 hours and 22 minutes, than those given placebo. Thus, Lunesta did not prolong sleep much longer than placebo, and did not on average result in anything close to 8 hours of sleep.
• Exaggerates Results About Falling Asleep- The ad suggests Lunesta "works quickly." Yet the study results suggest that those taking Lunesta fell asleep only 15 minutes earlier than those taking placebo.
• Exaggerates Long-Term Usability - The ad suggests that the drug "is approved for long-term use." Yet the longest study of the drug only lasted six months, not what many people would call long-term.

Note that Sepracor said it based the advertising claims on a different smaller, shorter study that assessed patients in a sleep lab. [McCall WV, Erman M, Krystal AD et al. A polysomnographyt study of eszopiclone in elderly patients with insomnia. Curr Med Research Opinion 2006; 22: 1633-1642.] Yet this study only lasted 2 weeks, and it excluded patients with many chronic problems such as sleep apnea, chronic obstructive pulmonary disease, "uncontrolled medical abnormalities or unstable chronic disease, medical or psychiatric disorders," etc. Even so, this study only showed that people feel asleep on average of 11 minutes faster on Lunesta (average sleep latency -40.8 vs -29.6), and slept only 16.2 minutes longer (average change of total sleep time 48.6 vs 32.4 minutes).

The main points are that DTC advertising, rather than being educational, may be psychologically designed to convey positive impressions about the product far beyond the actual words used. Furthermore, the words used are still likely to suggest that the product is much better than what the data from clinical research suggest. In this case, rather than being a wonder drug, Lunesta on average makes people fall asleep a few minutes earlier and sleep a few minutes longer than they would have without it. These small advantages should be weighed against its adverse effects, including strange sleep behaviors which the FDA wants to add to Lunesta's (and other "sleep aids'") labels but which were apparently not in the advertisement.

Patients should be extremely skeptical about DTC advertising, especially about whether the psychological good feelings conveyed by the ads obscure the cold facts about the drug, and whether the facts spoken about the drug exaggerate its benefits and obscure its harms.

Knowing about these issues, physicians need to be appropriately circumspect when patients ask for that wonder drug they saw on TV. Remember that DTC ads are particularly likely to amount to pseudoevidence-based medicine.