OR WAIT null SECS
With specialty companies getting smarter in applying their big data insights to product marketing, the true commercial potential of machine learning and predictive modeling may soon be within reach.
With specialty companies getting smarter in applying their big data insights to product marketing, the true commercial potential of machine learning and predictive modeling may soon be within reach
As we advance to an age where almost every new drug entering the market will be a specialty drug, the word “specialty” itself may eventually become redundant. The theory and practice of pharmaceutical marketing, however, is being strongly guided and shaped by the tactical and technological advancements brought about by the challenges of communicating and commercializing specialty treatments.
Connecting with healthcare professionals (HCPs), care providers, and experts in multiple therapeutic areas is facilitating a more delicate balance of technology and human insight, and how marketers think about “traditional” patient populations is changing dramatically. Companies are striving to speak more effectively to individual patients with messages that are more relevant and personalized. In the next few years, predicts Sarah Alwardt, vice president of health and informatics and health and economics outcomes research at McKesson, pharma will move to catch up with the activities of companies like Amazon by providing “ultra-targeted messaging” driven by predictive modeling on big data sets.”
The further advancement of specialty pharma marketing depends on the smart use of data. While data has been available and accessible to pharma for a long time, with data sets becoming far more sophisticated, “the hard part,” says Paul Shawah, Veeva’s senior vice president of commercial strategy, is piecing the data sets together and matching them around an HCP or care provider and, ultimately, a target patient population.
Specialty companies are using a wider volume and variety of data to optimize their route to market. In some specialty treatments, the number of patients that exist for certain diseases is very small, and even in broader patient populations a drug treatment or therapy may only be relevant to a subset of that population. “The trick is identifying where those patients are and intervening at the right point in time,” Shawah told Pharm Exec. “To do that, you have to stitch lots of different data sets together so that you can identify patterns and then quickly shift your resources based on those patterns.”
For orphan diseases with very limited patient and prescriber populations, “there isn’t really any room for error,” says Remy Sukhija, senior vice president of commercial operations, Otsuka America, Inc. “As an industry, we are not poor when it comes to the amount of data we have. What we need as a next step is to improve our ability to shift from descriptive insights-which tell us what happened yesterday, last week, last month in the market or with a customer-to a diagnostic approach.”
Sukhija believes that data needs to inform not only about “what happened?” but “why did it happen?” as well. “The highest value after that would be using the data at our disposal to arrive at a predictive state, which would allow us to make better judgments about our resources, and to deliver the right information to the right customers at the right time through their preferred channel of communication,” he says.
Bringing data together, making sense of it, and then acting on it quickly has been a process that the industry has struggled with for a long time. But Alwardt sees large pharma companies advancing with their efforts in this area. Many companies are establishing centers for data competence with their own internal groups to focus on the big data sets, she says, making them less dependent on external partners or marketing consultancies. Having the right people on hand to analyze the data is another longstanding challenge that Alwardt believes the industry is starting to overcome. “I tend to think that the people who generate the data know the data best,” she says. “Bringing in people who really understand particular data sets can accelerate the insight that comes from that.”
However, custom-built data warehouses can be costly and time-consuming and outside the reach of smaller companies. Resources such as Veeva Nitro, an industry-specific commercial data warehouse in the cloud, offer an alternative. Nitro is built on Amazon Redshift-a cloud-based, petabyte-scale data warehouse infrastructure-to “ensure the highest levels of scalability and fast-query performance, even on the largest data sets,”
according to Veeva’s publicity. “It makes it easier for the industry to bring together their data sources in a single place in a way that’s ready for analytics and for more sophisticated technology like artificial intelligence (AI),” says Shawah, allowing “what took companies one or two years to build previously to be done literally overnight.”
Regardless of the route a company takes to process and analyze its masses of data, one thing certain is that the data will continue to grow at an unprecedented scale. “We used to talk about data ponds and data lakes, now we have data oceans,” Alwardt told Pharm Exec. “We’re far beyond even big data. It’s now about transitioning from big data to ‘the right data’.”
Such a transition requires a more informed understanding of the strengths and weaknesses of the data in question. “The data itself used to be enough. You could put it in a PowerPoint slide and it would tell you everything about the market,” says Alwardt. “But now you have to think about whether the questions you’re
asking of the data can be answered in the data sets you’re looking at.”
And there are concerns such as the velocity of the data. It’s not uncommon for a data set to be two or three months old when it’s purchased or received. That can make a big difference to the questions that can be asked of it and can have major implications if a product is new to market. “Some data partners, such as McKesson, can give faster data,” says Alwardt, “but it’s also about asking the right questions upfront and making sure that whoever you’re associating with for your data needs understands them.”
In primary care marketing, says Otsuka’s Sukhija, “a pharmaceutical product could be considered successful if 50–60% of physicians prescribe it.” The orphan disease space, however, requires a different approach. There are fewer prescribers, which makes each one extremely important. Pharma manufacturers in the orphan drug space need to develop a highly targeted approach to determine the unmet needs for each prescriber and their patients.
“Data allows us to understand the unmet medical needs a prescriber is facing beyond what sales reps understand,” says Sukhija. “Using data and analytics properly enables the potential to educate more effectively
and help them treat their patients with appropriate therapies.”
Smarter analysis of the data moves things beyond the sales reps’ understanding, but rather than dilute or even remove the human element in sales and marketing interactions, the advance of technology is likely to facilitate a further evolution of the rep’s role. “We’ve been saying for about 5–10 years that the sales rep model in pharma/biotech is going to change, and to some extent it has,” says Sukhija.
Data is empowering reps with the understanding of what prescribers need “so that they’re more effective and more valued and more relevant,” notes Sukhija. As Shawah explains, if a company cannot present the most relevant information or offering for a customer in a way they care about, “it’s meaningless.” He adds: “Patients are humans, doctors are humans, providers are humans. They like to talk. Customers may have known reps for a long time, particularly in the specialty markets, where they offer both professional and personal value.”
The thinking that “all you need to do is to take the data, bring it together, and run a machine learning or intelligence program on it” is misguided, says Shawah. “Data still needs a lot of tuning,” he contends. “It will still need a human to help educate others about how it should be used.” And there remains a need for human beings to think about things that are not incorporated into the data, “such as a potential extension to a label or a new indication, or a competitive launch that’s about to hit the market place, or a shift in the payer environment that the data isn’t smart enough to know even exists yet.”
But change is happening in the rep space. Shawah notes that, on the specialist side, reps are increasingly required to display a new set of more specialized and targeted skills. What customers are demanding from a scientific information standpoint is becoming more advanced. “Your data may help you discern what is important to this or that customer, but you need someone to deliver the message who is credible and trusted,” says Shawah. “As such, we’re certainly seeing a shift toward more scientific engagement.”
On the orphan drugs side, for Sukhija, the field roles in pharma/biotech need to continue to evolve with prescribers’ needs. “Our customers’ needs go beyond just the efficacy and safety messaging,” he says. “For example, they are more interested in making sure their patients can access the therapies they’ve prescribed without unnecessary obstacles.” To address these unmet needs, pharma and biotech companies have a variety of field-based roles that interact with the same prescribers, with the goal of answering their questions and delivering a seamless prescribing experience for both the prescriber and the patient.
“As this trend continues, it is difficult to see the traditional ‘reach & frequency’ sales model remaining the primary way pharmaceutical companies interact with prescribers in the field,” says Sukhija.
Where the traditional goals of pharma were around patient adherence and drug utilization, the complexity and cost of specialized drugs has required more of a value- and outcomes-based approach. The shift, says Shawah, is toward “How do I get the patient better?” rather than “How do I get the patient to use my drug?” Again, technology is helping pharma become more precise about how and when a drug should be used, and which
candidate patients are likely to have the best outcome. “By going back to the data and understanding the most appropriate and meaningful use of your drug, you can start to achieve value,” he says. An organization’s data sets were once very siloed-for example, serving either the marketing/commercial or the health economics and outcomes research (HEOR) departments-but they are becoming more connected, and the insights being generated can be compared to effectiveness, efficacy, cost economics, and real-world outcomes.
For Shawah, this cross-functional approach-across sales, medical, marketing, market access, etc.-presents “the new frontier” in the commercial application of AI and big data. We will start to see more leading companies apply this approach in the next three to five years, he says. “While AI has been around for some 60 years, we’re still in the early stages of its impact on the commercial process.”
Julian Upton is Pharm Exec’s European and Online Editor. He can be reached at email@example.com