Applied Clinical Trials
With cloud-based software dominating data systems, on-premise installations still exist and command the pharmacovigilance software market.
Salesforce officially launched in February 2000 at DEMO 2000, the then-premier launch venue for new technologies. The “End of Software” marketing promotion that accompanied the launch was nothing short of extraordinary, replete with a party where guests were required to bring a piece of physical software to throw in a gigantic trash bin and mock protests by hired actors at competitor client events declaring software obsolete.
Salesforce, for those readers that may not know, is considered the first cloud-based, software-as-a-service (SaaS) product to disrupt a large market- customer relationship management (CRM). Salesforce.com is now the bellwether in the CRM space, leading and continuing to define new markets for the Salesforce platform.
Prior to its launch in 2000, legacy vendors such as Seibel Systems (acquired by Oracle in 2006) dominated CRM with on-premise servers and software, long and complicated implementations, and equally slow and complicated user experiences. Salesforce came along with cloud-based storage, almost instantaneous implementation, a web-based user interface that could be rapidly iterated as web technology advanced, and killer marketing. The rest, as they say, is history. At the time of writing, Salesforce (NYSE: CRM) had a market cap of
around $100 billion.
So, what does the CRM market have to do with pharmacovigilance (PV)? A lot actually. The parallels are striking, with some of the same market players, the same story lines, and as argued ahead, the same results.
A brief history of PV software
Electronic databases have long existed for the storage and capture of individual case safety reports (ICSRs) that are generated both within clinical trials and from spontaneous reporting. In the late 1990s, this space was dominated by organizations such as Relsys, Phase Forward, Aris Global, and, to a lesser extent, Oracle. The companies provided software that paired with the database to capture ICSRs and to transmit the ICSRs to regulatory authorities such as FDA. Analysis of these data were more difficult.
Statistical methods for safety signaling were in their infancy and computer systems that could process the data efficiently were not resource efficient, both in cost and time.
Datamining started to become more mainstream in 1999 with a study that was funded by FDA and published in The American Statistician, entitled “Bayesian Data Mining in Large Frequency Tables, with an Application to the FDA Spontaneous Reporting System.”¹ Technology was developed by organizations like Lincoln Technologies to take advantage of advances in computing power to apply new statistical methods at scale. In the mid-late 2000s, consolidation in the space started to occur. In 2005, Phase Forward acquired Lincoln Technologies to advance the company’s data and analytics platform. In 2008, Oracle, in line with its sector-specific growth strategy, honed its sights on drug safety and bought Relsys, and then later in 2010, acquired Phase Forward.
The enterprise goliath
With its acquisitions, Oracle became the PV software market goliath, eventually consolidating all of the brands into its Oracle Health Sciences suite of applications, and specifically bringing the database and capture software under the ARGUS brand and the data, analytics, and workflow software under the organization’s Empirica brand.
In 2018, by all accounts, Oracle remains the market share leader. ARGUS and Empirica software have been synonymous with “drug safety database” and “PV software” for the past 10 years. Oracle has, unsurprisingly, taken several steps to engrain both ARGUS and Empirica into the PV ecosystem.
Being Oracle, the company certainly knows what enterprise IT wants and needs. On-premise servers (and more so now, Oracle’s cloud servers) and the software that runs off of these servers are highly customizable, configurable, and designed with the Fortune 500 in mind. This level of complexity means that there needs to be a small army of in-house database and software engineers that are dedicated to keeping ARGUS and Empirica up and running with maintenance, patching, updates, compliance, etc.
Even if you are one of the only 12 pharma companies in the Fortune 500, you most likely need to outsource some of these back-office tasks. Oracle knows this and welcomes it. By creating Oracle Gold Partners and other designations that license a third-party to resell Oracle products, Oracle has created a remarkably robust closed-loop sales channel. With hundreds of partners that depend on Oracle for their own business revenue, there is ample motivation to keep ARGUS and Empirica right where they are.
Sound familiar? This is exactly the challenge that Salesforce faced when coming to market, right down to the same exact Oracle (then Seibel) strategies. The pitch to those that were willing to listen back in 2000 was that spending 90% of your technology budget on maintaining systems of record was a bad idea and that firms need to focus on front-end services-based differentiators that are revenue builders, and not just back-end infrastructure. Salesforce hammered this point home by making its platform simple to implement without the help of IT and with a front-end UX/UI that their end-users would actually want to use.
Can the same approach that Salesforce took in 2000 work for PV software in 2018?
The paradigm shift in PV software
In the past 18 years since Salesforce began its battle to take down the CRM goliath, the market for software-as-a-service (SaaS) has changed dramatically, as have the products that companies have brought to market.
The SaaS model has become so commonplace that even Salesforce itself had to clarify back in 2015 that “No Software” really meant, “No legacy software, just cloud software.” SaaS and software are now synonymous, in all but a few cases. PV software is one of those few cases. While the software itself is no longer shipped via a CD-ROM, and it’s been well documented that Oracle is making its shift to the cloud, existing on-premise installations still dominate the market and even new cloud-based options are still just remote servers, with similar software installation.
Why has PV software lagged behind? First and foremost, the world of PV is one of extensive regulations that up until recently have seemed to be always changing. A significant repository of standard operating procedures and required process documentation paperwork is necessary to pass any regulatory inquiry or audit. Both IT and business have had to focus on making sure what was in place was compliant.
Second, PV has historically been seen as a cost center within an organization, existing solely to fulfill the obligations that were required of them. Thus, a legacy system of record was all that was seen as necessary. Enough to capture ICSRs, do the minimally required aggregate signal detection to be included in periodic reports, and record actions taken in the process of managing that signal.
There are other companies that have worked to gain market share from Oracle. For data capture, companies such as ArisGlobal, largely seen as Oracle’s most significant competition, has begun to make technological advances to create efficiencies. Smaller companies as well, like AgilePV, AB-Cube, and My Meds and Me, are working to create differentiated platforms to focus on case intake and processing. Vendors such as RxLogix, CommonWealth Informatics (recently acquired by Genpact), and this author’s company, Advera Health Analytics, have built analytics and workflow platforms that have begun to make an impact on the market. However, as a result of both the obstacles that Oracle tatically put in place, as well as the market barriers discussed earlier, the impact has been minimal.
Very recently, however, a shift has begun to occur. The last of the good pharmacovigilance practices (GVP) requirements have gone into effect. Big data has gone mainstream. And forced by demands by their customers, pharmaceutical companies are being asked to generate insight as it pertains to safety.
The last point is, in my view, the most important. In order to generate insight, appropriate tools need to be in place that are accessible to more than just a select few power users, in a select few resource capable companies.
Remember the Salesforce pitch, that spending 90% of your technology budget on maintaining systems of record was bad and that firms need to focus on front end services-based differentiators that are revenue builders, and not just back end infrastructure.
Data-driven insights gained from Salesforce have changed the way both large and small sales and marketing teams drive revenue, shifting the CRM paradigm by creating an intelligence that was previously inaccessible using legacy systems. Salesforce did this by not only making the back-end infrastructure irrelevant, but through a great user experience, analytics, and customer service.
The need for accessible, actionable drug safety insight to drive departmental as well as overall commercial success in large and small biopharma companies shifts the PV software paradigm in the same way, albeit 18 years later.
What does the Salesforce of PV software look like?
Companies that specialize in back-end infrastructure like Oracle, ArisGlobal, and other cloud competitors will no doubt continue to iterate on the best ways to deliver the back-end infrastructure for PV. The emergence of artificial intelligence and advanced machine learning will continue to evolve how databases are constructed and maintained. The efficiencies created in case processing by these technologies provide an opportunity to have a long-term, trickle-down effect that will free up resources to accelerate aggregate insight generation from those cases.
The immediate opportunity for intelligence, however, is how data and analytics PV software can better interact not only with traditional sources like ICSR databases, FDA Adverse Event Reporting System (FAERS) data, VigiBase, and clinical trial data, but with emerging, disparate sources such as social media, claims, electronic health records (EHRs), and other unstructured data. Bringing these pools of information together creates an opportunity to enhance signaling algorithms, make validations and assessment more efficient, and ultimately get answers to drug safety questions faster.
However, in order to bring these data together, extensive ontologies need to be created to link all of the data. Drug name and active ingredient represented as national drug code (NDC), Rx concept unique identifier (RxCUI), or ICSR drugs need to be resolved to one record. Adverse event coding in MedDRA needs to be mapped back to verbatim labeling and ICD-10 codes. Drugs need to be characterized by anatomical therapeutic chemical (ATC) classifications, NDF-RT, label status, etc. This burden should not fall on the end user, the business, or IT within an organization. Next-generation PV software that will disrupt legacy vendors need to provide these ontologies and mappings off-the-shelf to be able to further drive immediate, actionable insight.
Arguably more important than how software interacts with the data, is how end users engage with software. Complicated, slow, and unintuitive software leads to a poor user experience. Legacy software and platforms that were built during web 1.0 will no longer be acceptable. And datamining in 2018 should not require an end user to be a data scientist. The platform that is able to shift the PV software paradigm will be one that reinvents how end users feel about the tools they use for their day-to-day jobs. Bottom line, a safety reviewer has to want to use and engage with the software, rather than see it as a burden. When this happens, PV software will shift from not only just a system of record, or just one of engagement, but truly a system of intelligence that was not previously capable with legacy platforms.
The market for PV software in 2018 and beyond is one that is attempting to catch up to other verticals that have long been disrupted by the “death of software.” The barriers that were in place such as regulations with a moving target, drug safety’s role as a cost center only, and the difficulty of accessing data are breaking down. The result is that there are now choices in the PV software market that weren’t available just a few years ago, with potential clients that are more willing than ever to look past the systems of record and focus resources on building a system of intelligence that will drive PV workflow for years to come.
Vendors focused on data and analytics can shift the paradigm of PV software and allow the science of PV to advance at a rapid pace. Innovation will come with end users empowered to take advantage of disparate data sources through a modern user experience. New signal detection algorithms and new, more efficient workflow built on the cloud and infinitely scalable, will allow drug safety departments to provide actionable contributions to all areas of an organization.
Although it is unlikely that mock demonstrators will be seen at any drug information or regulatory conferences, PV software is indeed having its “Salesforce” moment.
1. Dumouchel, W., 1999. https://amstat.tandfonline.com/doi/abs/10.1080/00031305.1999.10474456#.W_1jnZNKh0J
Jim Davis is Executive Vice President, Advera Health Analytics, Inc.