In drug or medical device lawsuits, plaintiffs’ attorneys will often try to gain an advantage in the litigation by looking at a manufacturer’s post-market reporting of adverse events. No matter how well-intentioned or diligent a manufacturer may be, plaintiffs’ counsel will inevitably argue that a manufacturer failed to timely report a particular report or that they overlooked a key safety signal. These arguments often gain traction given the tremendous number of adverse event reports any given manufacturer may receive and the allegation that many adverse events are under or partially reported. To compound the problem, following up on an initial report that may contain little information can sometimes be a fruitless venture.

This dynamic is only part of a broader problem. Every year the FDA receives more than 600,000 adverse event reports pertaining to marketed drugs and biological products and approximately 300,000 reports regarding medical devices.¹ The reporting of these events can be voluntary. Consequently, an expansive and efficient surveillance of adverse events is nearly impossible. Recognizing these problems, Congress passed the Food and Drug Administration Amendments Act (FDAAA) of 2007, which required the FDA “ to collaborate with public, academic and private entities to develop methods for obtaining access to disparate data sources and to validate means of linking and analyzing safety data from multiple sources.”²

This collaborative effort is now known as the Sentinel Initiative. Pro Te Solutio first considered the Sentinel Initiative in January 2010 – just as Sentinel was about to launch. At the time, there was a great deal of uncertainty about what this project would ultimately look like and how it would structure its vast data-collecting apparatus. Now, more than four years later, the Sentinel Initiative is making deliberate strides toward meeting its intended purpose.

In August 2011, the FDA issued a report to Congress on the status of the Sentinel Initiative. Here, the FDA set forth its vision for this program, promising that it “will serve as an active surveillance system to monitor the safety of marketed medical products.”³ The FDA boasted that “significant progress” had been made and that the Agency was “committed to continuing this ambitious pace during the coming years.” Specifically, the FDA claimed that they had exceeded Congress’s goal of being able to access data from 25 million patients by July 2010.⁴

In this report, the FDA also explained why they had opted to execute the Sentinel Initiative through a distributed system as opposed to a centralized approach to collecting data. In a distributed system, “personally identifiable information would remain with its data holders in its local environment, protected by existing firewalls and managed by those most familiar with the data.”⁵ This approach enhances patient privacy by keeping data localized within an already protected system and enables the owners of the data to perform their own analyses, which only enhances their interpretative value.⁶

Within this model, the FDA envisioned three categories of activities focused on safety signals: signal generation, signal refinement and signal evaluation.⁷ The FDA defines a safety signal as a “concern about an excess of adverse events compared to what is expected to be associated with a product’s use.”⁸ These signals may be derived from “the product’s clinical development program, postmarket studies of a product or postmarket adverse event reports submitted to FDA’s spontaneous reporting systems.”⁹ However, in order to effectively monitor the products it regulates, the FDA must be able to identify and filter out false positive safety signals. To accomplish this task, the Sentinel Initiative employs “signal refinement,” a process in which the FDA “further evaluates a safety signal for which there is already some evidence of a concern, either based on data from the clinical development program, postmarket studies or adverse event reports, or due to a theoretical safety concern related to the type of medical product or the class that the medical product is in.”¹⁰

A key component to the initial implementation of the Sentinel Initiative is Mini-Sentinel, a pilot project sponsored by the FDA. Mini-Sentinel is designed to provide the FDA with three core capabilities: “1) work through the ‘nuts and bolts’ of designing safety assessments using multiple existing electronic healthcare data systems; 2) develop and evaluate scientific methods to increase the precision of active safety surveillance efforts; and 3) identify and address barriers and challenges to building a practical, accurate and timely system for active safety surveillance.”¹¹

Mini-Sentinel operates through a distributed database and common data model. According to Mini-Sentinel’s website, “The Mini-Sentinel Distributed Data model gives Data Partners complete autonomy over access to and use of data in their possession. This distributed approach is achieved by using a standardized data structure referred to as the Mini-Sentinel Common Data Model. Data Partners transform their data locally according to the Common Data Model, which enables them to execute standardized computer programs that run identically at each Data Partner site. The combined collection of datasets across all Data Partner sites is known as the Mini-Sentinel Distributed Database.”¹²

Mini-Sentinel would not be possible without the more than 20 collaborating institutions that collectively are putting into practice the Sentinel Initiative.¹³ By July 2012, 18 of these partner organizations together had data on more than 160 million people, covering 3.5 billion dispensings from over 85 million unique patients.¹⁴ These figures far exceed Congress’s goal of the Sentinel Initiative being able to access data from 100 million patients by July 2012.¹⁵ Significantly, Mini-Sentinel contains data going back more than 10 years, which greatly enhances its ability to detect and evaluate safety signals that pertain to more chronic adverse events.¹⁶ This initial success garnered Mini-Sentinel a 35% increase in last year’s funding.¹⁷

One writer recently noted some of the practical applications that are emerging from the FDA’s new-found data gathering capacity: “Having such a large data set gives the FDA a new way to quickly detect, analyze and evaluate the risks posed by drugs. In presentations, Mini-Sentinel staffers have demonstrated how they can dig into the database to show hospital visits for acute myocardial infarction stratified by age and sex.” Given its unexpected growth and the vast amount of data under its domain, Mini-Sentinel will likely evolve into a vital tool to not only identify safety signals, but to evaluate national health issues. In fact, the FDA’s most recent Mini-Sentinel public workshop focused on the Initiative’s emergence as a key component to our national healthcare infrastructure.¹⁸

Mini-Sentinel is not the only new initiative for enhancing the FDA’s post-market surveillance capabilities. The FDA, in conjunction with Epidemico¹⁹ and Boston Children’s Hospital, has developed MedWatcher, a free app that allows anyone to report an adverse event from their smartphone or tablet. According to MedWatcher’s website (where you can also report an adverse event), “MedWatcher is a free mobile app and web app that allows users to learn about side effects of drugs, medical devices and vaccines, and to easily report adverse events to the FDA. MedWatcher strives to increase transparency around medical products and to improve the safety profile of drugs, devices and vaccines.”²⁰ The app is remarkably simple and easy to use. You must first register by providing your name and email address. Once logged in, the user can search for specific drugs, vaccines or medical devices, and then follow specific products to see other user reports.

Reporting an adverse event through the app is intuitive and can be done quickly. MedWatcher asks what other medications or medical products were in use at the time of the adverse event, why the patient was prescribed the drug and the date of the event. It then asks for a description of the event and for the user to select from a list the different outcomes they attribute to the event.²¹ The app also allows a user to submit a photo with their report. After a report is submitted, MedWatcher will then forward that report to the appropriate FDA reporting system once they have processed, formatted and approved the report. ²²

With any such reporting system, patient privacy is a concern. According to their website, all personally identifiable information is kept private and confidential. “When we publish a MedWatcher report, we remove the reporter and/or patient’s personally identifiable information. We also remove any information provided in the adverse event description that may further identify a person (e.g. name of hospital). Only age, gender and home state are included in published reports.”²³

MedWatcher touts the ability to share information across the community of its users as the key advantage to reporting events through their program as opposed to directly reporting it to the FDA. Additionally, they claim that “submitting the reports we receive via MedWatcher to the FDA enables the FDA to link our reported adverse events to the high volume of reports they have received directly, which in turn will raise red flags faster.”²⁴

While it is difficult to say how prevalent the use of MedWatcher is among consumers, there is no doubt that it is a powerful tool that enables users of medical products to almost effortlessly report an adverse event. By promoting its use, drug and medical device companies could easily demonstrate their commitment to maximizing the scope of post-market surveillance and the degree to which consumers can gain access to information covering a range of events that relate to specific products.

The ubiquitous use of apps on smartphones and tablets, coupled with the faster-than-expected progress of the Sentinel Initiative, suggests that regulatory oversight of medical products has entered a new era defined by an ever-increasing amount of data. This data deluge is both a blessing and a curse – a blessing in the sense that more events from more patients are catalogued, and a curse in that false positive safety signals are increasingly common and a challenge to efficiently identify and filter out. On balance, though, drug and medical device companies – not to mention consumers – will benefit in the long term from a more complete sight-picture of the overall safety and effectiveness of their products.

---

[1] U.S. Food and Drug Administration, “The Sentinel Initiative – A National Strategy for Monitoring Medical Product Safety,” Aug. 19, 2011, Available at http://www.fda.gov/downloads/Safety/FDAsSentinelInitiative/UCM274548.pdf. Last accessed May 6, 2014.

[2] Id. p. 1.

[3] Id. p. 4.

[4] Id.

[5] Id.

[6] Id. p. 5.

[7] Id.

[8] Id.

[9] Id.

[10] Id.

[11] See <http://www.mini-sentinel.org/>. Last accessed May 6, 2014.

[12] See <http://www.mini-sentinel.org/data_activities/distributed_db_and_data/default.aspx>. Last accessed May 6, 2014.

[13] A complete list of all collaborating institutions as of January 1, 2014, can be found at http://www.mini-sentinel.org/about_us/collaborators.aspx. Last accessed May 4, 2014.

[14] See <http://www.mini-sentinel.org/work_products/Data_Activities/Mini-Sentinel_Year-3-Distributed-Database-Summary-Report.pdf>. p. 1. Last accessed May 6, 2014.

[15] See <http://www.fda.gov/downloads/Safety/FDAsSentinelInitiative/UCM274548.pdf>, Aug 19, 2011, p. 1. Last accessed May 6, 2014.

[16] See <http://www.mini-sentinel.org/work_products/Data_Activities/Mini-Sentinel_Year-3-Distributed-Database-Summary-Report.pdf>. p. 1. Last accessed May 6, 2014.

[17] Taylor, Nick Paul, “The FDA’s Mini-Sentinel Healthcare Surveillance Database is Growing Quickly,” Fierce Biotech IT, Feb 3, 2014. Available at http://www.fi ercebiotechit.com/story/fdas-mini-sentinel-healthcare-surveillancedatabase-growing-quickly/2014-02-03#ixzz2sKGyCmC9. Last accessed May 6, 2014.

[18] See <http://www.brookings.edu/events/2014/01/14-sentinel-initiativepublic-workshop>. Last accessed May 6, 2014.

[19] Epidemico describes itself as a “health data collection and analytics company” that’s a “spinoff from Boston Children’s Hospital, Harvard Medical School, and MIT.” See <http://www.epidemico.com/>. Last accessed May 6, 2014.

[20] See <https://medwatcher.org/faqs>. Last accessed May 6, 2014.

[21] Users can select from the following outcomes: non-serious, other serious, hospitalization, disability/permanent damage, birth defect, life-threatening, death.

[22] See <https://medwatcher.org/faqs>. Last accessed May 6, 2014.

[23] Id.

[24] Id.

Finis