Today most hospitals and health networks are making decisions about drug choices without adequate resources to analyze the most current post-approval drug side effect data. In many cases, warning signs from emerging data regarding drug side effects may be overlooked, thereby risking patient safety and increasing overall cost of care. In this new era of value-based payment models, health care providers must discover and implement an analysis of real-world adverse events to ascertain the true costs and risks of drug choices.

The landscape is shifting among high-reliability health care organizations like Memorial Hermann in Houston, where reviewing and responding to adverse events data and analytics are believed to improve outcomes and decrease avoidable adverse events-related medical costs, such as readmissions. Managed care organizations, anxious to lower health care costs, also are adopting new thinking about converting certain unpreventable adverse events into preventable ones.

Not far into the future, both public and private payers, eager to enhance outcomes and reduce capitation rates, will demand that providers implement adverse events algorithms and their money-saving potential in several ways. From Medicare to large HMOs, these powerful payers likely will require that provider formularies demonstrate a systematic review of side effect case reports from a broad-based patient population, not just an individual organization's internal tracking. This review must be coupled with an organization's analysis of what the data mean in terms of costs, safety and outcomes. Whether a formulary serves a managed care organization or hospital, it must embrace the inevitable change and paradigm shift in drug safety.

Incentives for Industry Change

Historically, hospitals and health care professionals focused primarily on interventions to minimize adverse events caused by medication errors in the hospital setting. Through acknowledgment of the issue and organized industry efforts to improve, eliminating medication errors in hospitals became a top priority during the past decade. This movement is now broadening to include improving seemingly unpreventable side effects.  The rationale for further change includes both outcome and financial incentives.

Beyond medication errors, adverse side effects cost the health care system $4.7 billion in 2013. In the same year, hospitals reported 148,000 admissions or readmissions as a result of these adverse drug events. Nevertheless, this figure does not tell the whole story.  The Agency for Healthcare Research and Quality and the Food and Drug Administration's Adverse Events Reporting System estimate that only 10 percent of adverse drug events are reported, bringing the true cost to a staggering $25 billion a year.

Clearly, the current procedures are flawed when most hospital formulary review processes include only an initial review of adverse events from clinical trials and a cursory glance of FAERS data. Furthermore, many health systems only rely on FDA warnings about specific drugs that typically occur months if not years after a severe side effect issue arises.  

With 10,000 prescription medications, and nearly one-third of adults in the United States taking five or more drug treatments, remaining complacent is the easy thing to do, but not the right thing to do. Some hospitals and health networks are embracing big data, and the algorithms that make the data actionable, to chart a course for the future management of adverse drug events.

Our Adverse Events Data Plan

At Memorial Hermann, the change in our formulary review process unfolded as a three-step process:

We assessed our current formulary review process for incorporating adverse events side effect data beyond clinical trials and FDA alerts.  We asked ourselves: Is the process using broad-based, post-approval adverse events data and technical analysis to compare true drug costs per patient across an indication, class or mechanism of action? Does it dig deeper into the data provided by the FAERS database to analyze what the information means? If the analytics reveal compelling results, does the process lead to changes in formulary?

We ascertained strengths and weaknesses of today's system. We asked: Is the hospital leadership able to track and prevent side effects of the most commonly prescribed medications in the hospital setting? Is it coupling its internal review findings with a review of FAERS? Is it taking the time to clean up the FAERS data to remove input errors or errant information? 

We reconfigured the process to include sophisticated analytics of post-marketing side effects. This was accomplished by calculating adverse event risks and costs as part of the decision-making process. While off-the-shelf software solutions can be an affordable, fast and efficient solution, we wanted to make sure that the chosen system includes:

— The right data. Does the system include on- and off-label usage of sanitized data from not only FAERS, but also the most recently reported adverse events that have not yet been entered into FAERS? These data can be obtained by employing the Freedom of Information Act and requesting this information from the FDA. However, we were fully aware that the FDA is sometimes six months behind in inputting data about adverse events into FAERS.

— The right analytics. Does the system utilize algorithms and analytics that identify, assess and rank medicines, and that deliver final outputs that are timely and actionable? Analytics are the key to generating a return on investment that will result in better patient safety and reduced readmissions. We aimed to dissect the data critically and reveal the costs related to serious adverse events and outcomes per patient, as well as to identify the specific adverse drug reactions that are driving increased costs.

— The right measurement. Does the system measure progress and success in terms of cost savings per prescription and outcomes?

An Example of True Costs

Memorial Hermann began putting adverse event analytics to the test with drugs commonly prescribed in the hospital setting, such as anticoagulants. With the relatively recent introduction of newer anticoagulants Eliquis (apixaban) and Xarelto (rivaroxaban), which are both Factor Xa inhibitors, and Pradaxa (dabigatran), a direct thrombin inhibitor, we set to determine the true downstream costs of these drugs if post-marketing adverse events were considered. The following surprising yet insightful analysis was generated by our approach, causing our organization to reconsider our formulary choices. As the following chart depicts, Eliquis offers significant downstream cost savings over both Pradaxa and Xarelto, for our hospital annually, when adverse events are considered in the true cost of the anticoagulants. 

Analysis of true costs of anticoagulants commonly used in hospital setting



Rx cost

price per

costs per


Rx cost savings
(per Rx if
switched to Eliquis)


















(Source: AdverseEvents Explorer)


The future of health care can benefit from the systematic use of advanced data analytics with regard to adverse drug events. Memorial Hermann is proud to have embraced this methodology early on, and we plan to push for continued improvements in formulary analysis. As our experience indicates, hospitals and health networks can improve drug decision-making by incorporating a thorough analysis of the most current post-approval drug side effect data. The benefits will be realized in a reduction in adverse event costs, such as hospital readmissions or other serious consequences, as well as improved outcomes among patients.

Patti Romeril, PharmD, is system director of clinical pharmacy services at Memorial Hermann, an integrated health system in Southeast Texas.