FDA and the Media: Lessons from Tysabri about Communicating Uncertainty

 

Abstract

Despite the fact that they have gone through a rigorous approval process, uncertainty about newly approved drugs always exists. The uncertainty is amplified when drugs are approved on preliminary or interim evidence (i.e., accelerated approval) or on surrogate outcomes and when new serious harms emerge.

The case of Tysabri—a multiple sclerosis drug that received accelerated approval because of promising preliminary evidence, was quickly withdrawn from the market several months later after the emergence of an unexpected, typically fatal harm (progressive multifocal leukoencephalopathy), and was subsequently remarketed with a risk management plan—illustrates many of these uncertainties. We quantified the extent of the media coverage of Tysabri and compared Food and Drug Administration (FDA) press releases with the major media coverage in order to draw lessons on how to improve the communication of uncertainty about new drugs.

 

Take-Home Messages

• FDA could help the media by improving its press releases. Better press releases that routinely quantify benefits and harms with absolute numbers (e.g., with means or absolute risks) and that highlight uncertainties about a new drug would help journalists do a better job
• FDA and the media should be clear that accelerated approval does not only mean “extra promise” but also “extra uncertainty.” To create realistic expectations, FDA and drug company press releases and the media should explain the reasons for accelerated approval and highlight the inherent extra uncertainty.
• FDA should flag new drugs for their first few years on the market. FDA should routinely use graphics or text to help ensure that the public understands that a new drug’s limited track record means uncertainty about long-term benefits and harm.

 

Despite the fact that they have gone through a rigorous approval process, there is always uncertainty about newly approved drugs: Will the benefits observed in typically short and small approval studies hold up over time? Will important side effects emerge? The uncertainty is amplified when drugs are approved based on preliminary evidence (accelerated approval) or on surrogate outcomes. Communicating uncertainty about new drugs is an important challenge facing FDA in fulfilling its mission to promote the safe and effective use of prescription drugs.

To explore systematic and structured approaches to the characterization and communication of uncertainty, in February and May 2014 the Institute of Medicine (IOM) held two public workshops, “Characterizing and Communicating Uncertainty in the Assessment of Benefits and Risks of Pharmaceutical Products” at FDA Headquarters in White Oak, Maryland. ( The full workshop summary report, Characterizing and Communicating Uncertainty in the Assessment of Benefits and Risks of Pharmaceutical Products, is available for free download at: http://www.nap.edu/catalog/18870/characterizing-andcommunicating-uncertainty-in-the-assessment-of-benefits-and-risks-of pharmaceutical-products.)  The workshops focused on the drug Tysabri as one case study in which workshop participants could consider how uncertainty in the pharmaceutical regulatory environment is communicated to the media and the public. Tysabri is a multiple sclerosis (MS) drug with an eventful history: It received accelerated approval because of promising preliminary evidence, was quickly withdrawn from the market after the emergency of an unexpected, typically fatal harm (progressive multifocal leukoencephalopathy, or PML) and was subsequently reintroduced with a risk management plan (IOM, 2014). We quantified the extent of media coverage of Tysabri and compared FDA press releases with the corresponding coverage in major newspapers in order to draw lessons on how to improve the communication of uncertainty about new drugs.

 

Methods

Media Coverage Search

As shown in Figure 1, we used LexisNexis to identify U.S. media coverage (using the search terms “Tysabri” or “natalizumab”) in the 2 months following seven separate key events in Tysabri’s lifecycle: the approval of the drug, its withdrawal from the market, the FDA Advisory Committee meeting, FDA’s decision to approve remarketing, the occurrence of two PML cases after remarketing began, and two label changes that better quantified the risk of PML.

 

 

For the quantitative analysis assessing the extent of media coverage, we identified 806 U.S. stories in newspapers, wire services, national television, and radio. To characterize what the media reported, we conducted a detail content analysis of the coverage in major newspapers, specifically, the 20 U.S. newspapers with the greatest circulation and two major national wire services, the Associated Press and United Press International. Of the 192 major newspaper stories, 114 met our inclusion criteria that they were long enough (>100 words) and focused on one of the Tysabri events of interest (i.e., not just a passing mention or pure business or stock news.)

 

FDA Press Release Search

To identify FDA press releases about Tysabri, we searched FDA’s 2004-2012 web archive for the relevant dates (FDA, 2015b). In order to compare that with current FDA practice, we also identified all 2014 press releases about accelerated drug approvals (reviewing all FDA “press announcements” about new drug approvals) (FDA, 2014a).

 

Content Analysis

We developed a structured coding scheme and revised it based on a pilot test of five news stories which were not included in this study. The final two-page form (see appendix) was organized into four main domains: benefit (magnitude, uncertainty), harm (PML, other side effects, uncertainty), sources quoted or mentioned, and overall impression. Two reviewers independently coded each story. We calculated inter-rater reliability (kappa statistic) and excluded the two items (both of the overall impression codes) that did not meet our pre-specified threshold for reliability: kappa > 0.4. Kappa for all remaining items ranged from 0.53 to 1, with an average value of 0.7, which is generally considered to represent “substantial agreement” (Landis and Koch, 1977). Drs. Schwartz and Woloshin resolved disagreements to determine the final codes for the analysis.

 

Results

In 2004, FDA announced the approval of Tysabri in a press release which emphasized the innovative nature of the drug (FDA, 2004b). The press release, which was titled “First monoclonal antibody treatment for Multiple Sclerosis approved,” began by saying that “FDA today licensed a new biologic approach to treat patients with relapsing forms of multiples sclerosis” and then went on to quote the acting FDA commissioner explaining why it believed the drug was important:

 

This innovative treatment for multiple sclerosis represents a new approach to treating MS—exciting news for patients with this serious disease. . . . [W]hile we eagerly await long-term results from ongoing clinical trials, we have reason to believe that Tysabri will significantly reduce relapses in MS. [bolded emphasis added]

 

The FDA press release noted that Tysabri had received accelerated approval based on interim results seen after the first year of a 2-year randomized trial “because it appears to provide substantial benefit for patients with a serious disease.” The release continued by reporting that “as part of that approval, the manufacturer has committed to continuing its trials of this product for another year.”

While factually correct, this portrayal failed to communicate the uncertainties inherent in “accelerated approval.” The manufacturer’s commitment to continuing the trial was not a sign of altruism but a legal postmarketing requirement. FDA’s standard for approving all prior MS drugs had been a reduction in relapses over 2 years (FDA, 2004a, p.6). The accelerated approval of Tysabri was contingent on confirmation of the interim 1-year results at 2 years because of the uncertainty about whether the benefit would hold up over time. The press release also failed to acknowledge that Tysabri’s effect on disability progression—a known patient-important benefit of existing MS drugs—was unknown at the time. Demonstrating that the drug reduced disability was also a postmarketing commitment.

The 2004 press release generated modest media coverage (Figure 2). There were 29 news stories generated in response to the release, eight of which appeared in major newspapers. Of the stories, four portrayed accelerated approval as meaning extra promise, and only one portrayed it as extra uncertainty. No story mentioned that Tysabri’s effect on disability was unknown (Appendix A).

 

 

Fear and Uncertainty: Deadly New Side Effect

Shortly after approval, Tysabri was removed from the market after two patients taking the drug developed progressive multifocal leukoencephalopathy (PML), a rare untreatable and frequently fatal brain infection. FDA issued a public health advisory and a press release (FDA, 2005), which generated 292 news stories (including 11 front-page newspaper stories), with 44 of those stories in major newspapers. All 44 in the major newspapers mentioned the number of PML cases, and almost all (42) mentioned that one patient had died (Appendix B).

The FDA’s press release explained how an unexpected serious adverse event like PML might occur:

 

During the review of Tysabri for marketing approval, FDA conducted an intensive analysis of possible adverse events that might be related to effects of the drug on the immune system. No cases of PML were seen in the clinical trials. However, for any approved therapy, new and unexpected adverse events may occur that were not seen in clinical trials. [emphasis added]

 

Ideally, FDA would have included a similar statement explicitly acknowledging the uncertainty inherent with all new drugs in the press release that it had issued when the drug was approved.

 

Advisory Committee and the Decision to Return Tysabri to the Market

About a year after Tysabri had been withdrawn from the market, an FDA advisory committee met to decide whether it could be reintroduced with a restricted distribution program. The vote fo the committee for remarketing generated a lot of media coverage: 292 stories, far more than the 101 stories that would be written when the FDA officially authorized the return to market (Figure 2). The drama over the decision and the human interest angle likely explain the difference in coverage. Patients or family members – including many who felt that the drug had helped, as well as a relative of one of the patients who had died – testified at the meeting. Thirteen of the 44 major newspaper stories about the meeting included quotes from MS patients or family members, and 10 of the 13 stories the quotes supported remarketing (Appendix C). In contrast, the marketing authorization was simply a regulatory event, where FDA typically follows the advice of its advisory committees.

When FDA issued a press release announcing the remarketing (FDA, 2006), that release focused primarily on efforts to reduce the chance of PML:

 

The Food and Drug Administration (FDA) today approved an application for resumed marketing of Tysabri subject to a special restricted distribution program. Tysabri is a monoclonal antibody used to treat patients with relapsing forms of multiple sclerosis (MS) to reduce the frequency of exacerbations (flare-ups). Tysabri is indicated for use as monotherapy . . . meant for patients who have not responded adequately to, or cannot tolerate, other treatments for MS . . . [only] available under the TOUCH Program with the following main features [limited distribution only through registered providers, careful monitoring]”

 

In line with the contents of the press release, most of the major newspaper stories mentioned the restricted distribution requirement (Appendix B). Similarly, the press release did not quantify the risk of PML, and neither did most (11 of 14) stories in major newspapers. But FDA could have quantified that risk. The company’s briefing material that was presented to the advisory committee stated, “We therefore estimate the risk of PML in natalizumab-treated patients to be 1 per 1,000 (95% CI 0.2 to 2.8 per 1,000)” (Biogen Idec and Elan, 2006, p. 93). And an FDA official repeated the company’s figures at the time of remarketing (Pollack, 2006).

 

Two New PML Cases

In 2008, about 2 years after Tysabri’s remarketing, two new cases of PML were reported in patients on monotherapy (who had been believed to be low risk at the time of remarketing). FDA issued a public health advisory and a press release affirming its belief that “Tysabri monotherapy may confer a lower risk of PML than when Tysabri is used together with other immunomodulatory medications” (FDA, 2008). This event generated little media coverage (55 news stories, 6 major newspaper stories). The major newspapers that did cover the story focused not on the health risks facing patients, but on the financial risks facing the manufacturer. According to the Associated Press, “Biogen Idec Inc. and Elan Corp. defended their multiple sclerosis drug Tysabri . . . after the announcement of two new cases of the rare viral infection called progressive multifocal leukoencephalopathy, or PML, sent shares for both companies plummeting and reignited already simmering concerns about the drug’s sales potential (Associated Press, 2008).”

 

Better PML Risk Stratification

In 2011, FDA issued a safety alert to announce a change in Tysabri’s label (FDA, 2011). The new label included the first quantitative PML risk estimates (stratified by the number of infusions) and highlighted the fact that the risk increased with prior exposure to an immunosuppressant. The announcement generated little media coverage—just 16 stories, with none appearing in major newspapers.

Within 1 year, FDA issued a press release announcing the approval of a new test which allowed for even more refined PML risk estimates (FDA, 2012). The test measured the presence of an antibody to the John Cunningham Virus, a common and generally harmless virus in people with a normal immune system, but a marker of increased PML risk in patients taking Tysabri. In addition to the virus test, the press release noted two other factors that increased PML risk: exposure to Tysabri for longer than 2 years and prior treatment with immunosuppressant drugs. FDA also announced an additional change to the label: a new table quantifying PML risk according to the foregoing risk factors. For example, about 1.3 percent of MS patients who have the John Cunningham Virus antibody and prior immunosuppressant use will develop PML over 4 to 6 years, compared with 0.6 percent without the prior immunosuppressants, and less than 0.1 percent of patients without the antibody. Despite its clinical importance, the improved PML risk stratification generated little media coverage: 21 news stories with only one appearing in a major newspaper. While the limited media coverage may suggest that editors are less interested in stories about developments that reduce risk (e.g., the antibody test) than in stories about increased risk (e.g., the initial occurrence of PML), it may also reflect the fact that another MS drug (Gilenya [fingolimod])—which was without known PML risk at the time—was approved in 2010, since then cases of PML have been observed with Gilenya (FDA, 2015a). Regardless of the reason, this was a lost opportunity to communicate tradeoffs: the PML risk given the information from the virus test (and the duration of treatment) versus the benefit of taking Tysabri.

 

How To Do Better

Our analysis of the Tysabri story suggests several ways to improve the communication of benefit, harms, and uncertainties to the public. An obvious place to start is with FDA communications. Better FDA press releases could help journalists write more balanced stories with the key facts consumers need.

 

FDA Could Help the Media by Improving its Press Releases

Press releases matter. Press releases strongly influence the quality of media reporting (Schwartz et al., 2012; Sumner et al., 2014; Yavchitz et al., 2012). For example, we conducted a systematic review comparing medical journal press releases about original research with subsequent coverage in the news (Schwartz et al., 2012). If the press release contained information about absolute risks, harms, or study limitations, news stories were more likely to report them. For example, if the press release did not report absolute risks, only 9 percent of news stories reported them, while 53 percent of news stories reported absolute risks if they were provided in the press release. A recent study also demonstrated the influence of academic press releases on exaggeration in subsequent news coverage (Sumner et al., 2014). Unfortunately, medical journals, academic medical centers, and industry press releases often fail to include this fundamental information (Woloshin and Schwarts, 2002; Woloshin S. et al., 2009; Kuriya B. et al., 2008).

In the present study we identified a number of instances where better FDA press releases might have resulted in higher quality coverage by the news media.

Whenever Possible, Quantify Effects

Surprisingly, the FDA press release at the time of the remarketing decision failed to quantify the PML risk, even though credible estimates (quoted by FDA during the advisory committee meeting) were available. Only three of the 14 major newspaper stories presented the PML risk. The press release also failed to mention that Tysabri was now proven to reduce disability, let alone quantify this effect—and none of the14 major newspaper stories provided that detail, either.

Similarly, at the time of initial approval none of the FDA press releases (and few stories in the media) quantified the benefit of Tysabri. While press releases dealing with PML appropriately focus on harm, benefit information is still highly relevant. How can patients decide whether the PML risk is “worth it” if they are not reminded of the likely benefit of the drug?

Quantify Benefit Using Absolute Numbers

When benefit is quantified, the most transparent approach is to present the absolute numbers (e.g., means or absolute risks) of the outcome for both the drug and the comparator. Unfortunately, in announcing Tysabri’s initial approval, the 2004 FDA press release (FDA, 2004) only quantified the drug’s benefit as a relative risk reduction without providing the base rate (the absolute risk in the placebo group), a format which is known to exaggerate perceptions of drug benefit (Forrow et al., 1992; Malenka et al., 1993; Naylor et al.; 1992). This was the wording from the press release:

 

Tysabri was evaluated for safety and efficacy in two ongoing randomized, double-blind, placebo-controlled trials in patients with relapsing forms of MS. In the first clinical trial of the product’s safety and efficacy, the drug reduced the frequency of relapses by 66 percent relative to placebo. . . .In [the second trial it] reduced the frequency of relapses by 54 percent relative to placebo. [bolded emphasis added]

 

As expected, six of the eight major newspaper stories quantifying the effect of Tysabri used exactly the same format as the FDA press release: the relative risk reduction without the base rate. Only two reported the relapse rates in the Tysabri groups (a change from 7 to 2 relapses per 100 people per year). CONSORT, the international organization promoting reporting standards for trials (endorsed by most major medical journals), recommends including absolute risks when reporting outcomes (Moher et al., 2010).

 

The News Media and FDA Should Present Benefits and Harms of Drugs Symmetrically

Almost all of the major newspaper stories that quantified Tysabri’s benefit presented only relative change statistics (31 of 33 stories across all events). By contrast, none of the stories quantified the PML harm in this way, but presented absolute numbers instead. Because the relative changes are numerically much larger than the absolute numbers, this asymmetric approach has the effect of maximizing benefit and minimizing harm in the reader’s mind, a phenomenon that we previously described (Schwartz et al., 1999).

We observed a similar asymmetry in how major newspaper stories compared Tysabri to other MS drugs. While 40 of the 114 major newspaper stories mentioned that Tysabri’s benefit was bigger than that of other MS drugs, only five compared harms.

FDA press releases also used asymmetric presentations. For example, the press release announcing the initial approval of Tysabri presented drug benefit as a relative change without a base rate and presented harm as a long list of 15 possible and unquantified side effects. That asymmetric reporting, emphasizing benefit over harm, is an example of a general problem with the medical news (Cassels et al., 2003, Moynihan et al., 2000, Schwartz et al., 2002).

 

FDA and the Media Should Be Clear that Accelerated Approval Does Not Only Mean “Extra Promise” but also “Extra Uncertainty”

The initial press release about Tysabri did not communicate the extra uncertainty inherent in accelerated approval. Failing to do so created the sense that accelerated approval only meant extra promise, when in fact it also meant extra uncertainty.

Since 2004 accelerated approval has become an increasingly common approval pathway, with 20 percent (8 out of 40) of new drugs receiving accelerated approval in 2014 (FDA, 2014d). All of the drugs that received accelerated approval were approved in the context of substantial uncertainty: Every one of the eight was approved on the basis of a surrogate outcome, and seven were approved on the basis of an uncontrolled study. We examined all the approval press releases issued about these drugs to see if the characterization of uncertainty in the context of accelerated approval has changed.

In fact, the press releases for these eight used nearly the same language as the 2004 Tysabri press release. For example, the press release announcing approval of Opdivo (originally approved to treat advanced melanoma) stated:

 

Opdivo is being approved under the FDA’s accelerated approval program, which allows approval of a drug to treat a serious or life-threatening disease based on clinical data showing the drug has an effect on a surrogate endpoint reasonably likely to predict clinical benefit to patients. This program provides earlier patient access to promising new drugs while the company conducts additional clinical trials to confirm the drug’s benefit (FDA, 2014b).

 

None of the press releases highlighted the extra uncertainty inherent in an approval based on a surrogate. For example, Opdivo was approved based on the surrogate outcome of tumor shrinkage, and its effect on the clinical outcome – improved survival – was unknown. Only three of the eight press releases about drugs receiving accelerated approval in 2014 named the unproven clinical outcome.

None of the press releases about the seven drugs approved based on uncontrolled studies included a caveat explaining that the since everyone took the drug, there was substantial uncertainty whether the observed beenfits and side effects were clearly attributable to the drug.

At a minimum, FDA press releases could adapt language that now routinely appears in the official label. (i.e. the prescribing information) for drugs recieving accelerated approval. That language specifies the name of the surrogate outcome and explicitly states that “continued approval for this indication may be contingent upon verification and description of clinical benefit in the confirmatory trials” (FDA, 2014c). Ideally, the name of the clinical outcome would be included, as is sometimes done in the label.

 

FDA Should Flag New Drugs for the First Few Years on the Market

In a national survey of U.S. adults, many mistakenly believed that FDA only approves – and only permits advertising of – extremely effective drugs (39 percent) or drugs without serious side effects (25 percent) (Schwartz and Woloshin, 2011). Such beliefs are incorrect, however, since drug approval means only that FDA believes the drug’s benefit outweighs its harm. It does not mean the benefits are large or important or that the drug is very safe or even that all serious side effects are known.

Acknowledging the uncertainty that is inherent in new drugs might help counter the common misconception that with drugs new is always better. To do so, FDA could adopt the approach taken by the British drug regulatory and recently implemented by the European Union – requiring a warning on recently approved drugs to explicitly flag uncertainty related to newness (EMA, 2013; Medicines and Healthcare Products Regulatory Agency, 2009). For the first few years that the drug is on the market the warning, a black triangle symbol (Figure 3), appears on all prescribing and consumer information about the drug to remind prescribers and patients that despite a rigorous approval process much will be learned after marketing begins, and to encourage reporting of adverse events. An IOM committee recommended that Congress amend the Food, Drug, and Cosmetic Act to require that labels for new drugs carry a special symbol such as the black triangle or an equivalent symbol (IOM, 2007). The committee also recommended that FDA restrict direct-to-consumer advertising during the period of time the special symbol is in effect. As of today, these recommendations have not yet been taken up by Congress or the FDA. To our knowledge, while clinicians in the United Kingdom may recognize that the “black triangle” identifies a new drug (Belton et al., 1995; Belton and The European Pharmacovigilance Research Group, 1997), the effect of the black triangle on consumer perceptions of uncertainty associated with a new drug has never been tested.

 

 

To learn whether a text-based new drug explanation would lead patients to choose drugs with a longer track record, we conducted an Internet-based randomized controlled trial using a national sample of 2,944 U.S. adults. Participants were asked to choose between two equally effective heartburn drugs with the same side effects (Schwartz and Woloshin, 2011). The only difference was that one drug was “approved this year,” while the other had been approved 8 years earlier. The primary outcome measure was the percentage of people choosing the new drug. In the control group, participants were only told the year of drug approval, while in the intervention group participants were told:

 

PAXCID was approved by the FDA in 2009. As with all new drugs, rare but serious drug side effects may emerge after the drug is on the market—when larger numbers of people have used the drug.

 

The new drug explanation increased the proportion appropriately choosing the drug with the longer track record from 47 percent to 66 percent, absolute difference 19 percent (95% CI: 13% – 24%).

 

Limitations

Our assessment of FDA press releases and the subsequent media coverage concerning Tysabri should be interpreted with caution. We covered only selected milestone events in the history of Tysabri, and some events generated only a small number of stories. Furthermore, our detailed content analysis was limited to stories in major newspapers and wire services. Limiting the analysis to major newspapers, however, probably means our findings are a “best case scenario” since the major newspapers usually have the highest quality coverage.

 

Conclusions

Understanding how FDA and the news media responded to the many communication challenges raised by Tysabri provides important lessons for future efforts. Better FDA press releases would help the media do a better job communicating benefits, harms, and uncertainties to the public and would hopefully lead to the media routinely presenting absolute numbers for benefit and harm and clearly explaining the extra uncertainties inherent in new drugs receiving accelerated approval. By flagging all new drugs for the first few years on the market—as is done in Europe—FDA could clearly highlight the fact that a new drug’s limited track record means there is extra uncertainty. Finally, the media should be careful to communicate drug benefits and harms symmetrically. In the case of Tysabri, asymmetric presentations magnified perceived benefit and minimized perceived harm. Doctors and patients need clear, balanced communication about drugs to be able to make good decisions in the face of uncertainty.

 

 

 

 

 

 

 

---

References

1. Associated Press. 2008. Biogen, Elan stand by MS drug after illnesses. Available at: www.highbeam.com/doc/1A1-D929IR7G0.html (accessed August 13, 2015).
2. Belton, K. J., and European Pharmacovigilance Research Group. 1997. Attitude survey of adverse drug reaction reporting by health care professionals across the European Union. European Journal of Clinical Pharmacology 52:423-427. https://doi.org/10.1007/s002280050314
3. Belton, K. J., S. C. Lewis, S. Payne, M. D. Rawlins, and S. M. Wood. 1995. Attitudinal survey of adverse drug reaction reporting by medical practitioners in the United Kingdom. British Journal of Clinical Pharmacology 39:223-226. https://doi.org/10.1111/j.1365-2125.1995.tb04440.x
4. Biogen Idec and Elan. 2006. Advisory committee briefing materials. Available at: www.fda.gov/ohrms/dockets/ac/06/briefing/2006-4208B1_02_01Biogenbriefingmaterial.pdf (accessed July 6, 2015).
5. Cassels, A., M. Hughes, C. Cole, B. Mintzes, J. Lexchin, and J. McCormack. 2003. Drugs in the news: An analysis of Canadian newspaper coverage of new prescription drugs. CMAJ 168(9):1133-1137. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC153682/ (accessed July 17, 2020).
6. EMA (European Medicines Agency). 2013. The European Union (EU) has introduced a new process to label medicines that are being monitored particularly closely by regulatory authorities. Available at: www.ema.europa.eu/ema/index.jsp?curl=pages/special_topics/document_listing/document_listing_000365.jsp (accessed August 9, 2015).
7. FDA (Food and Drug Administration). 2004a. Medical review. Submission number: BLA 125104/0. Tysabri (natalizumab). Available at: www.accessdata.fda.gov/drugsatfda_docs/nda/2004/125104s000_Natalizumab_Medr_P1.pdf (accessed July 6, 2015).
8. FDA. 2004b. First monoclonal antibody treatment for multiple sclerosis approved. FDA news release P04-107. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2004/ucm108380.htm (accessed July 6, 2015).
9. FDA. 2005. FDA issues public health advisory on Tysabri, a new drug for MS. FDA news release P05-07. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2005/ucm108413.htm (accessed July 6, 2015).
10. FDA. 2006. FDA approves resumed marketing of Tysabri under a special distribution program. FDA news release P06-75. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2006/ucm108662.htm (accessed July 6, 2015).
11. FDA. 2008. Safety alerts for human medical products. Tysabri (natalizumab). Available at: www.fda.gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm110599.htm (accessed July 6, 2015).
12. FDA. 2011. FDA’s MedWatch safety alerts: April 2011. Tysabri: New safety information. Available at: www.fda.gov/forconsumers/consumerupdates/ucm254898.htm – Tysabri (accessed July 6, 2015).
13. FDA. 2012. FDA permits marketing of first test for risk of rare brain infection in some people treated with Tysabri. FDA news release. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm288471.htm (accessed August 9, 2015).
14. FDA. 2014a. 2014 Press announcements. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2014/default.htm (accessed July 6, 2015).
15. FDA. 2014b. FDA approved Opdivo for advanced melanoma. FDA News Release. Available at: www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm427716.htm (accessed July 6, 2015).
16. FDA. 2014c. Guidance for industry. Labeling for human prescription drug and biological products approved under the accelerated approval regulatory pathway (draft guidance). Available at: www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm390058.pdf (accessed July 6, 2015).
17. FDA. 2014d. NDA and BLA calendar year approvals. CDER drug and biologics approval for calendar year 2014. Available at: www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/DrugandBiologicApprovalReports/NDAandBLAApprovalReports/UCM435753.pdf (accessed July 6, 2015).
18. FDA. 2015a. FDA drug safety communication: FDA warns about cases of rare brain infection with MS drug Gilenya (fingolimod) in two patients with no prior exposure to immunosuppressant drugs. Available at: https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-about-cases-rare-brain-infection-ms-drug-gilenya-fingolimod (accessed July 17, 2020).
19. FDA. 2015b. FDA archive. Available at: www.accessdata.fda.gov/scripts/search/index.cfm?action=archive.search (accessed July 6, 2015).
20. Forrow, L., W. C. Taylor, and R. M. Arnold. 1992. Absolutely relative: How research results are summarized can affect treatment decisions. American Journal of Medicine 92:121–124. https://doi.org/10.1016/0002-9343(92)90100-p
21. Institute of Medicine. 2007. The Future of Drug Safety: Promoting and Protecting the Health of the Public. Washington, DC: The National Academies Press. https://doi.org/10.17226/11750
22. Institute of Medicine. 2014. Characterizing and Communicating Uncertainty in the Assessment of Benefits and Risks of Pharmaceutical Products: Workshop Summary. Washington, DC: The National Academies Press. https://doi.org/10.17226/18870
23. Kuriya B., E. C. Schneid, and C. M. Bell. 2008 Quality of Pharmaceutical Industry Press Releases Based on Original Research. PLoS ONE 3(7): e2828. https://doi.org/10.1371/journal.pone.0002828
24. Landis, J., and G. Koch G. 1977. The measurement of observer agreement for categorical data. Biometrics 33:159–174. https://doi.org/10.2307/2529310
25. Malenka, D. J., J. Baron, S. Johansen, J. W. Wahrenberger, and J. M. Ross. 1993. The framing effect of relative and absolute risk. Journal of General Internal Medicine 8:543–548. https://doi.org/10.1007/BF02599636
26. Medicines and Healthcare Products Regulatory Agency. 2009. Drug safety update: The black triangle scheme. Available at: www.gov.uk/drug-safety-update/the-black-triangle-scheme-or (accessed July 6, 2015).
27. Moher, D., S. Hopewell, K. Schulz, V. Montori, P. C. Gøtzsche, P. J. Devereaux, D. Elbourne, M. Egger, and D. G. Altman. 2010. CONSORT 2010 explanation and elaboration: Updated guidelines for reporting parallel group randomised trials. BMJ 340:c869. https://doi.org/10.1136/bmj.c869.
28. Moynihan, R., L. Bero, D. Ross–Degnan, D. Henry, K. Lee, J. Watkins, C. Mah, and S. B. Soumerai. 2000. Coverage by the news media of the benefits and risks of medications. New England Journal of Medicine 342(22):1645–1650. https://doi.org/10.1056/NEJM200006013422206
29. Naylor, C. D., E. Chen, and B. Strauss. 1992. Measured enthusiasm: Does the method of reporting trial results alter perceptions of therapeutic effectiveness? Annals of Internal Medicine 117:916–921. https://doi.org/10.7326/0003-4819-117-11-916
30. Pollack, A. 2007. F.D.A. allows some patients to resume M.S. drug. Available at: www.nytimes.com/2006/02/16/business/16drug.html?_r=0 (accessed July 6, 2007).
31. Sumner, P., S. Vivian–Griffiths, J. Boivin, A. Williams, C. A. Venetis, A. Davies, J. Ogden, L. Whelan, B. Hughes, B. Dalton, F. Boy, and C. D. Chambers. 2014. The association between exaggeration in health related science news and academic press releases: Retrospective observational study. BMJ 349:g7015. http://dx.doi.org/10.1136/bmj.g7015
32. Schwartz, L., and S. Woloshin. 2002. News media coverage of screening mammography for women in their 40s and tamoxifen for primary prevention of breast cancer. JAMA 287(23):3136–3142. https://doi.org/10.1001/jama.287.23.3136
33. Schwartz, L., and S. Woloshin. 2011. Communicating uncertainties about prescription drugs to the public: A national randomized trial. Archives of Internal Medicine 171(16):1463–1468. https://doi.org/10.1001/archinternmed.2011.396.
34. Schwartz, L. M., S. Woloshin, and H.G. Welch. 1999. Risk communication in clinical practice: Putting cancer in context. Monographs of the National Cancer Institute 25:124–133. https://doi.org/10.1093/oxfordjournals.jncimonographs.a024187
35. Schwartz, L., S. Woloshin, A. Andrews, and T. Stukel. 2012. Influence of medical journal press releases on the quality of associated newspaper coverage: Retrospective cohort study. BMJ 344:d8164. http://dx.doi.org/10.1136/bmj.d8164
36. Woloshin S., L.M. Schwartz, S. Cassela, A. Kennedy, and R.J. Larson. 2009. Press releases by academic medical centers: Not so academic? Annals of Internal Medicine 150: 613-618. https://doi.org/10.7326/0003-4819-150-9-200905050-00007
37. Woloshin, S. and L.M. Schwartz. 2002. Press Releases: Translating Research into News. JAMA 287: 2856-8. https://doi.org/10.1001/jama.287.21.2856
38. Yavchitz, A., I. Boutron, A. Bafeta, I. Marroun, P. Charles, J. Mantz, and P. Ravaud. 2012. Misrepresentation of randomized controlled trials in press releases and news coverage: A cohort study. PLoS Medicine 9(9):e1001308. https://doi.org/10.1371/journal.pmed.1001308

Tweets by @theNAMedicine