The data revolution has launched a disruption of giant proportions in all industries. Medicine and biopharma are finding themselves in the midst of it, and leaders are working hard to get their operations up to task for the avalanche of data that screams to be processed. The most effective data operations are those that account for the human element in all its facets and that systematically include qualitative information.
The potential of data, machine learning, and artificial intelligence is dominating much business talk, including in medicine and biopharma. Many novel therapies and diagnostic procedures came to fruition because we’re now able to efficiently process large quantities of data—and the future promises personalized medicine and overall better treatment decisions. Of course, these blessings are paired, in the minds of many people, with fears of job loss, of Big Brother-like data trawling, of machines becoming our boss, and of large-scale reality distortion through inaccurate data.
“I only believe in statistics that I doctored myself,” Winston Churchill once famously said. One doesn’t have to go as far as to suggest deliberate mishandling of data, of course, to know that data processing is a delicate task that requires clear thinking. As far as the reality on the ground is concerned, data truly does hold enormous promise, and it deserves deliberate handling to truly deliver desirable outcomes and guidance for decision-making.
In terms of pharma marketing, our experience shows that there’s a need to shift the emphasis away from sheer number crunching toward truly objective knowledge generation. It’s not enough to answer the same standard questions in every survey for every disease category; and it’s not helpful to think that we can just leave it to the machines to figure it all out.
On the consumer end, we’re dealing with humans, which means they have more to offer than objective data: They also have subjective insights, which will never be captured in a multiple-choice survey. It also means they:
- Care about their data.
- Need to be able to trust the integrity of their data.
- Care that their privacy is reliably protected.
- Expect empathy when they contact a brand by phone or another means by which one ordinarily expects to interact with a person.
Don’t Just Ask, Listen
Companies should capture qualitative information in order to further develop and evolve their quantitative questioning. A recent example of this is when we operated inbound and outbound calls for a client in the infectious disease space. And when we analyzed the standard quantitative data that was gathered during these calls, no particular pattern emerged.
However, once we took a closer look at the call summaries written by our associates, we were struck by the number of people who all asked the same question: “How do you contract this disease?” This highlighted that there is a major lack of awareness in the community about the mode of transmission of the disease, and we could also reasonably deduce that a large portion of the patient population must still be undiagnosed.
The questions you are asked are often a lot more revealing than the questions you ask of others. Listening to patients means listening to all of it, not just what we set out to hear. Unless we seek direct interaction with patients and listen to what they have to say, it is almost guaranteed that we’re not going to ask the right questions.
The standard questions, such as age, location, or years of diagnosis, are an important starting point, but the questions have to evolve as you gain insights. Listening is an important aspect of learning. As a matter of routine, any questionnaire should include a field for free comments next to the multiple-choice boxes—and those comments of course need to be evaluated. True commitment to qualitative analysis also means opening up opportunities for direct conversations with patients.
Deciphering Prescribing Patterns
If all we’re after is data that quantitatively answers the questions we defined a priori as being relevant, we are likely to spin our wheels for knowledge of limited value. Of course, even then you can still sometimes extract useful insights.
Three examples include:
- In the autoimmune category, we came across geographical prescribing patterns and could show objectively how patients in some states had better access to the medicine than their peers in other states, even if they were insured with the same payer. However, interpreting these patterns correctly still required qualitative analysis.
- How people who found us on TV versus on the web show different treatment and behavior patterns.
- We were able to show how prescribing practices varied from clinic to clinic, which highlighted the need for a customized approach for each health system.
Yet it took qualitative analysis of call logs to see that we probably needed to ask callers if they are being seen by a specialist. Because, as it turned out, those who are on treatment also tend to have a specialist. Access to specialists has an intervening variable—that’s actionable insight when you’re facing a concrete product for a concrete part of the population.
The Human Element
Some of us welcome the data revolution with open arms, while some of us dread what it will mean for jobs and society. Whichever way we may feel about it personally, competitive businesses must harness the insights data can generate. And ironically, in order to harvest the full potential of data, we must recognize how important the human element remains in generating and interpreting the data.
Machines can do many things, but for the time being, it remains our responsibility to gather the right data, to treat sensitive data safely and with respect, to interpret the data in useful ways, and to interact with patients like only humans can interact with one another.