American Journal of Law & Medicine

Medical Advice from Lawyers: A Content Analysis of Advertising for Drug Injury Lawsuits

This study examined the medical information contained in a sample of television ads soliciting consumers for lawsuits against drug and medical device manufactures. Almost all such ads involved drugs or devices that have not been recalled and remain on the market. These ads raise important public health questions because they may influence the prospective medical decisions of viewers.

The ads contained extensive descriptions of serious adverse events associated with the drugs or devices but almost uniformly failed to disclose information relating to the likelihood of such events. They also failed to effectively advise viewers to consult a doctor.

Results also identified a subset of ads that mimicked public service announcements, claiming to be a "medical alert" "consumer alert" or "FDA warning" at the start of the ad. Most such ads did not disclose the attorney source of the advertising until the final few seconds.

I.   INTRODUCTION
II.  METHODS.
     A. Sample.
     B. Content Analysis and transcription
III. RESULTS
     A. Representation of Drug Injury Advertisements.
     B. Source and Subject Matter of Advertising.
     C. Medical Info in the Advertising.
     D. Advise to Consult
     E. Advertisements Frames as Public Service Announcements
     F. Failure to Identify Advertising Firm.
IV.  DISCUSSION
     A. Defining "False or Misleading" Under Applicable Law.
     B. Advertisements Framed as Public Service Announcements.
     C. Failure to Identify the Advertising Attorney
     D. Omission of Risk Information
V.   RECOMMENDATIONS
     A. State Bar Approaches.
     B. Providing Additional Information About Risk
     C. Other Disclosure-Based Approaches.
VI.  CONCLUSION

I. INTRODUCTION

This study focuses on television advertising for lawsuits relating to drugs and medical devices (hereinafter "drug injury advertising"). The purpose of the advertising is to recruit consumers for lawsuits against drug manufacturers. In particular, the ads seek to identify consumers that took a drug or used a medical device and subsequently experienced a particular medical problem that the drug manufacturer failed to disclose. (1) The ads warn of the dangers associated with a particular drug or device. Ultimately, advertising attorneys hope that injured consumers will contact them for legal representation.

Drug injury lawsuits are typically brought as mass torts, consisting of thousands of individual claims against the manufacturer of a defective product. (2) Unlike class actions, no single attorney or set of attorneys has a monopoly over representing plaintiffs in such claims. (3) As a result, there is a lively market of lawyers competing over the airways for the attention of a limited number of injured consumers.

The business models of advertising attorneys vary. About half of the most prolific advertisers in this study appear to actively litigate at least some of the cases they receive. (4) Recoveries in such cases can be quite large in the aggregate. (5) For example, claims involving the birth control pill, Yaz, caused the manufacturer to set aside 1.5 billion dollars in reserve to settle claims. (6)

The other prolific advertisers in this study rarely, if ever, file drug injury lawsuits. (7) It is possible that these advertisers are "settlement mills," meaning they settle large volumes of cases very quickly, without ever filing in court. (8) Alternatively, their business model may be based primarily upon referrals. (9) The disclaimers in some of these ads describe referring cases to associated attorneys. (10) Although attorney ethics rules heavily regulate referral fees, they generally permit advertisers to recoup the cost of advertising. (11) It is also possible that advertising attorneys base their business model on "maintain[ing] joint responsibility" for cases they refer. (12)

This article is not about whether drug injury lawsuits, attorney advertising, or referral mills are good or bad. It is about the complex policy questions raised by advertising that disseminates important medical information to the public. This lucrative market operates with little to no regulation or scrutiny. The impact of such advertising on consumer medical decisions is unknown and demands further study.

Drug injury advertising is, in a sense, the functional opposite of direct-to-consumer pharmaceutical advertising. Whereas pharmaceutical advertising seeks to encourage viewers to take a drug, drug injury advertising may deter viewers from taking a drug by highlighting serious risks. The risks described in the advertisements are often relatively rare, affecting fewer than 1 in 100 or even 1 in 1000 consumers taking the drug. (13) Consequently, the vast majority of interested viewers are not the injured consumers targeted by the advertisers, but uninjured consumers trying to decide whether to fill next month's prescription for the drug.

While considerable empirical research has been conducted on direct-to-consumer pharmaceutical advertising, (14) almost none has examined drug injury advertising. (15) Two small studies surveyed female urology patients about their opinions regarding transvaginal mesh, which is commonly featured in drug injury lawsuits. (16) In one study, patients that relied on television for their medical information were more likely to mistakenly believe that the mesh had been recalled. (17) In the other study, patients that learned about mesh through attorney ads expressed greater uncertainty about the safety of the product. (18)

The other limited research to date has been sponsored by partisan groups. (19) A 2003 Harris poll sponsored by the U.S. Chamber of Commerce found that 86% of consumers surveyed had seen drug injury advertising, and 21% had seen such advertisements about a drug they were taking. (20) Twenty-five percent of respondents reported that they would "immediately stop taking" a drug they had been prescribed if they saw it in a drug injury ad. (21) Similarly, a 2007 study sponsored by Eli Lilly and the nonprofit National Council for Community Behavioral Healthcare found that half of psychiatrists surveyed reported that patients had discontinued taking medication because of drug injury advertising. (22)

Anecdotally, a law review article by Daniel Schaffzin describes reports of consumer harm from a doctor and a mass tort attorney. (23) A recent New York Times profile of a mother making medical decisions for her son noted in passing that a drug injury ad had influenced her decision. (24) Message boards and blogs on the internet include anecdotal self-reports of consumers responding in alarmist ways to mass tort advertising. (25) While these anecdotal accounts and partisan research suggest the need for rigorous empirical research on the question, they are no substitute for such evidence.

This study provides context for policy discussions regarding the regulation of drug injury advertising. It analyzes content of drug injury advertising broadcast in Boston and Atlanta in 2009. An empirical snapshot of their content can inform the respects in which drug injury ads might influence consumer decisions, and how their content might be improved. This study also illustrates gaps in existing research that demand further study.

Drug injury advertising presents both public health opportunities and risks. It informs injured consumers of legal remedies, offers wider access to attorneys, and holds manufacturers accountable. (26) The advertisements also disseminate new information about medical risks from the FDA and medical journals. (27) Although the ads are not necessarily intended for uninjured viewers taking the drug, (28) the ads may nevertheless catch their attention. (29) As proponents of pharmaceutical advertising have argued, this additional information can motivate viewers to seek medical care, usefully inform medical decision-making, and improve interactions with physicians. (30) Paradoxically, they might even remind consumers to take their medication. (31)

At the same time, such advertising could also distort consumer medical decisions by causing them to misperceive drug-related risks. (32) Unlike consumers misled by pharmaceutical advertising, (33) those misled by drug injury advertising need not consult a doctor to make a medical decision based on the ad. They can simply stop taking the drug featured in the ad of their own accord.

Despite this potential to influence patient medical decisions, drug injury ads are not subject to the complex FDA regulations governing pharmaceutical ads. (34) Attorney advertising is theoretically subject to the jurisdiction of the FTC and state consumer protection law, but neither the FTC nor state attorneys general have a record of regulating drug injury ads. (35)

Attorney advertising is most commonly regulated through state attorney ethics rules prohibiting false or misleading attorney advertising. (36) However, as Fred Zacharias observed in a 2002 study of attorney Yellow Pages advertising, state bars have been reluctant to enforce ethics rules in the advertising context. (37) This appears equally true in the context of drug injury advertising. This research identified no ethics cases involving drug injury advertisements since the 1985 Supreme Court case, Zauderer v. Office of Disciplinary Counsel. (38)

Results of this study suggest that some drug injury ads could be considered misleading under state ethics rules. In particular, a subset of ads appeared to be public service announcements, by opening with the words "consumer alert," "medical alert," or "FDA warning." Most such ads also failed to disclose that the advertisement originated from an attorney until the very end. Such ads are misleading in two respects: (1) they suggest that the source of the advertising has a government, nonprofit or public health affiliation or that the advertising attorney has authority to award compensation or administer claims; and (2) they suggest that the medical information presented was selected based on consumer needs, rather than a particular adverse medical event that forms the basis for a lawsuit.

Results also raise the question of whether the advertisements omit risk-related information that would be important to viewers currently taking the drug. Although the absence of such information does not necessarily qualify as misleading under state ethics rules, it may lead consumers to overestimate the likelihood of adverse events associated with a drug and ultimately distort their decision-making. This article argues that including some risk-related information would substantially enhance the positive public health impact of such ads, and mitigate potential harms.

Specifically, a basic understanding of a medical risk requires information about both the likelihood that harm will occur and the particular harm that will result. The ads in the sample devoted considerable time and text to describing the alarming and sometimes fatal adverse events associated with the drugs or devices at issue. However, none of the ads in the sample explained the likelihood of harm in quantitative terms. As a result, the repeated and emphasized presentation of adverse events could lead a reasonable consumer to assume that the adverse event is very likely or even inevitable.

Another critical component of medical risk information is what a patient can do to mitigate the risk of harm. (39) Given the complexity of medical decisions, it is not practicable for these advertisements to provide any concrete guidance in this regard, other than suggesting that viewers consult a doctor before making a medical decision. However, only 39% of advertisements ever advised viewers to consult a doctor before discontinuing their medication, and they did so via small written disclaimers competing with other content in much larger font sizes. The inconspicuous nature of these disclaimers almost certainly rendered them ineffective.

In some circumstances, relative risk information--meaning information about certain subpopulations disproportionately or exclusively affected by an adverse event--is central to understanding risk. For example, if a recall in a manufacturing defect case only applies to a single manufacturer, disclosing the name of the manufacturer would be important to viewers taking the drug. Likewise, if the risk of adverse events primarily affects viewers who have taken a drug on a long-term basis, disclosing that fact would be important to both viewers who have taken the drug on a long-term basis and those who have not. Information on affected populations can reduce the likelihood that viewers will rely on the otherwise incomplete medical information to their detriment.

Section II of this Article describes the methodology of the study, followed by the results in Section III. Section IV analyzes the extent to which the content observed in this study would be considered false and misleading under relevant law. Lastly, Section V provides some policy recommendations.

II. METHODS

A. Sample

The data used in this study originated from media aggregator Kantar Media ("Kantar"). Kantar records cable television broadcasts in each Nielsen designated market area. (40) The sample consisted of digitized video for all advertising coded by Kantar as "legal services" broadcast between June 15, 2009 and January 4, 2010 in the Boston and Atlanta media markets. (41) The sample was limited to advertising broadcast targeted to the Boston and Atlanta markets, as opposed to advertising broadcast nationally. (42) For that reason, few ads in the sample were broadcast in both Boston and Atlanta.

The dataset consisted of 234 unique attorney advertisements broadcast in Boston and 654 in Atlanta. Each of these ads was reviewed to determine whether it related to lawsuits involving drugs or medical devices. (43) Both over-the-counter and prescription drugs were included in the definition of "drug or medical device." The above-described review produced a subsample of 56 unique advertisements (hereinafter, the "subsample"), 10 of which were deemed duplicates of other advertisements within the subsample. (44)

The attorney ethics rules in Massachusetts and Georgia are typical of most states in that they do not require attorneys to file their advertising with the state bar for review. (45) Following the Model Rules of Professional Conduct, both Massachusetts and Georgia prohibit false or misleading advertising. (46)

B. CONTENT ANALYSIS AND TRANSCRIPTION

Pairs of trained law students coded the subsample of drug injury ads using a coding instrument. The coding occurred in three successive rounds, with the trained students using a different instrument to identify different content in each round. (47) Two law students applied each of the three coding instruments to the ads. (48)

The instruments instructed coders to identify content relating to (1) adverse medical evidence; (2) the frequency of adverse medical events; (3) the benefits or use of the drug or device; (4) advice to consult a doctor before making medical decisions; and (5) the attorney or law firm sponsoring the ad. Definitions from the instruments are included in Table 1, below. Coders identified the time stamp(s) at which specified content first appeared in the ad, and the duration of that content. They also transcribed the audio and text of specified content.

Coders were not informed of the purpose of the research. (51) They were instructed to code the ads independently. Each coder then met with his or her counterpart to resolve inter-rater disagreement by reviewing the disputed portion of the relevant ad together. Their joint work product formed the basis of the analysis herein.

Table 2 summarizes the levels of inter-rater agreement for coding associated with the content defined above. Each type of content was coded for the timestamp at which the content was first discussed in the text or audio ("start time") and for the duration of such discussion, in seconds. The columns for observed agreement refer to the proportion of advertisements upon which the coders agreed. Coding differences of two seconds or less were deemed "agreement" for purposes of Table 2. The number of advertisements used to calculate inter-rater reliability exceeds the number of unique ads because the original sample included duplicates.

As Table 2 illustrates, the coders identified no advertisements disclosing the frequency with which adverse events occur. Because they were in 100% agreement as to that variable, no Kappa statistic could be calculated. All variables except the duration of adverse events resulted in a reasonable level of observed agreement.

Inter-rater reliability in Table 2 was generally higher for the start time of content than for the duration because the start time only required coders to identify a single point in time. Conversely, estimates of duration were less reliable because they required the coders to estimate the start and end points of an audio and/or text discussion. Coding the duration of an advertisement's discussion of adverse events proved persistently complex for the coders, and ultimately unreliable. (53)

Kantar identified the advertising firm for almost all ads. Where Kantar could not identify the source of the ad, the advertising firm could be ascertained through a very close review of the advertisement. Nevertheless, two unique ads, which jointly represented 7% of the ad volume, did not disclose the name of the advertising firm or attorneys.

III. RESULTS

A. REPRESENTATION OF DRUG INJURY ADVERTISEMENTS

Tables 3 and 4, below, summarize the representation of drug injury ads among all attorney ads. Because national advertising was excluded from the sample, the tables do not accurately represent the likely proportion of drug injury advertising among all ads. A separate dataset from Kantar suggests that the bulk of drug injury advertising volume is broadcast nationally. (54)

As Table 3 illustrates, the subsample of drug and medical device ads is somewhat small. Because the dataset consisted of a six-month sample for only two local media markets, results are not necessarily generalizable to other media markets or for advertising more generally. The sample is also somewhat dated and therefore may not reflect recent developments in advertising. (58)

B. SOURCE AND SUBJECT MATTER OF ADVERTISING

Drug injury advertising originated from a relatively small number of firms. Sixteen firms sponsored advertising in the subsample. The ten firms listed in Table 5 were responsible for more than 98% of the advertising volume. The majority of these advertisers were headquartered in states other than the state in which the advertising was broadcast. The three most prolific advertisers generated 70% of the advertising volume.

Figure 1 illustrates the volume of advertising associated with different drugs. The subsample included advertisements regarding 11 different types of drugs. (59) However, almost all of the advertising volume (87%) was focused on three drug types: Yaz/Yasmin/Ocella, Reglan, and Fentanyl. The combined advertising volume for all other drugs was substantially lower than the ad volume for each of these three drug types.

[FIGURE 1 OMITTED]

No single firm had a monopoly on advertising for the high volume drugs. Indeed, eight different firms advertised for lawsuits involving Yaz. (60) As previously discussed, the multiplicity of firms advertising for litigation over each drug type reflects the procedural and substantive structure of drug injury claims. (61)

C. MEDICAL INFORMATION IN THE ADVERTISING

All ads in the study sample discussed the adverse events associated with the drug/device in both the audio and on-screen text. By contrast, only about half (52%) of unique ads included some reference to the drug/device's benefits in either the audio or text. Only 39% of unique ads advised viewers to consult a doctor, and they did so only through on-screen text. None of the ads advised viewers to consult a doctor via the audio track.

Figure 2 illustrates the average number of words each unique ad devoted to three types of content: adverse events; the benefits or use of the drug/device; and advice to consult a doctor. (62)

[FIGURE 2 OMITTED]

These advertisements also spent much more time discussing adverse events than they spent describing benefits or advising viewers to consult a doctor, as illustrated in Figure 3. On average, the ads in the subsample lasted 28 seconds. Of these, the median ad devoted more than 20 seconds to discussing adverse events and 2 seconds to discussing benefits. …

Log in to your account to read this article – and millions more.