Author Affiliations: Regenstrief Institute and Indiana University School of Medicine, Indianapolis (Drs Duke and Friedlin); and UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill (Mr Ryan).
Product labeling is a primary source of drug safety information for physicians. However, the effectiveness of labeling in communicating adverse drug events (ADEs) may be diminished by the problem of “overwarning,” in which excessively long and complex lists of potential reactions can result in information overload.1,2 The Food and Drug Administration (FDA) highlighted this issue in 2006 as they unveiled new labeling guidelines, specifically discouraging the inclusion of “exhaustive lists of every reported adverse event, no matter how infrequent or minor.”3,4(p2) Yet, at present, there are no baseline data on overwarning, nor are there benchmarks against which the success of the FDA's interventions can be measured. The goal of our study was to address this gap by producing comprehensive quantitative data on ADE labeling patterns. We further sought to measure whether the 2006 guidelines were successful in reducing the burden of overwarning.
We retrieved all drug labels available as of December 17, 2009, on the federal Web site DailyMed.5 DailyMed provides drug data in Structured Product Label (SPL) format, an electronic labeling standard mandated by the FDA and available for more than 85% of prescription drugs.6 We developed a software tool, known as the Structured Product Label Information Coder and Extractor (SPLICER), that uses natural language processing to extract ADE data from SPLs. A previous study of its performance on 100 labels showed a recall of 92.8% and a precision of 95.1%.7
We processed 5602 SPLs using SPLICER and performed descriptive statistics on the extracted ADE counts. We then used Wilcoxon rank-sum testing to perform subset analyses along 4 parameters: prescribing frequency, therapeutic category, approval date, and labeling format. Prescribing frequency was determined based on 2008 national dispensing data, and SPLs were divided into the 200 most commonly dispensed medications and all other drugs. Therapeutic category was determined using the National Drug File Reference Terminology “mechanism of action” classifications8 (eg, angiotension-converting enzyme inhibitors), and SPLs were broadly grouped by clinical specialty (eg, cardiovascular). Approval date was based on the original FDA acceptance of the drug as a new molecular entity,9 and SPLs were grouped by decade of approval. Finally, to determine the impact of the new FDA regulations, SPLs were separated into those compliant with the 2006 formatting guidelines and those in any other format.
We extracted 534 125 ADEs from 5602 SPLs. As given in the Table, the number of unique ADEs per label ranged from 0 to 525, with a median of 49 and a mean of 69.8. At the upper extreme, we identified 588 labels having more than 150 ADEs and 84 labels with more than 300 ADEs. In terms of prescribing frequency, labels for the 200 most commonly dispensed medications contained significantly more ADEs than other labels (median, 79 vs 47; P < .001). Aggregating drugs by medical specialty, we found ADEs to be highest in medications associated with neurology (n = 168), psychiatry (n = 116), and rheumatology (n = 111). Looking at date of approval, we found that newer medications had significantly more labeled ADEs than older medications, with drugs approved during the 1980s and 1990s having the highest overall number of ADEs.
Structured Product Labels formatted in accordance with the 2006 labeling guidelines contained a greater number of ADEs than other SPLs (72 vs 47; P < .001). To control for the possibility that this differential was due simply to new format labels being associated with newer drugs, we repeated the comparison looking only at medications approved since 1980. Again, we found a significantly higher number of ADEs in new-format SPLs than in older label formats (113 vs 72; P < .001).
The goal of our research was to survey the current landscape of ADE labeling. We found the volume of ADEs to be remarkably high, particularly in newer and more commonly prescribed medications as well as in psychiatric and neurologic drugs. These patterns are not entirely unexpected. Newer drugs may face more rigorous clinical trials and postmarketing surveillance compared with older medications. Similarly, commonly prescribed drugs, by sheer volume of patient exposures, are likely to generate more ADE reports than less common drugs. The high volume of ADEs found in neuropsychiatric medications may relate as much to patient population as to the effects of the drugs themselves. Yet while a high number of labeled ADEs is not necessarily indicative of drug's true toxicity, the presence of such excess data still may induce information overload and reduce physician comprehension of important safety warnings.
Recent FDA guidelines do not appear to have reduced overwarning. Structured Product Labels formatted in compliance with the 2006 regulations actually contained more ADEs than other labels. This finding underscores the tremendous challenge faced by the FDA in reversing the long-standing trend toward overwarning. It is our hope that the baseline data provided by this study will inform the design and evaluation of future efforts to decrease the complexity of adverse event labeling.
Correspondence: Dr Duke, Regenstrief Institute, 410 W 10th St, Ste 2000, Indianapolis, IN 46202 (email@example.com).
Author Contributions:Study concept and design: Duke, Friedlin, and Ryan. Acquisition of data: Duke and Friedlin. Analysis and interpretation of data: Duke, Friedlin, and Ryan. Drafting of the manuscript: Duke, Friedlin, and Ryan. Critical revision of the manuscript for important intellectual content: Duke, Friedlin, and Ryan. Statistical analysis: Duke and Ryan. Administrative, technical, and material support: Duke, Friedlin, and Ryan.
Financial Disclosure: None reported.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Internal Medicine editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 4
Customize your page view by dragging & repositioning the boxes below.
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.