Have researchers become too good at finding 
rare-disease patients to interview?

No more haystacks

Bj Kirschner is director, business development at Just Worldwide. He can be reached at bj@just-worldwide.com.

How many traffic lights does your great-uncle pass if he becomes a tuna? Absurd question, right? But you are curious, I bet.

Actually, in the context of market research, I think we can make sense of it. Let’s do what we do best, look more closely and analyze it.

We have a guy (great-uncle) with a startling change to his life (I would call becoming a fish pretty startling, no?) and a question only he can answer (he’s the only one passing traffic lights, only he can count them).

Sound familiar? If you have ever done any health care research, it should:

• The great-uncle is a patient.
• Turning into a tuna is the patient’s rare condition.
• The number of traffic lights is a question to which only this rare-disease patient could actually know the answer.

In other words, “rare patient” research. See? It’s not as absurd as the great-uncle example. 

Here’s my theory: health care market research is a victim of its own success. We have found the patients, we have found them with rare conditions and we have asked them every question we have. I’m certainly proud of that but it also scares the heck out of me. Don’t get me wrong, it’s all positive. But each time we push further, future expectations are pushed further and I often wonder if perhaps we are pushing ourselves TOO far. Success comes with responsibility and boundaries.

What will that mean? Let’s break the discussion down into three parts – the patient, the micro-condition and the questions – and see where we end up.

Part 1: The patient

Back in the pre-modern days of market research, “health care” meant PCPs and people who took any prescription medication (the word “therapy” was not in heavy use yet). HIPAA was relatively new; managing it made us all nervous. GDPR did not even exist yet but many countries already had thorny privacy standards we needed to learn. But “knowing” health care was simple by today’s standards. It meant understanding your country’s national health care system, which HCP treats specific conditions and maybe the basics of a few diseases.

Then we took a leap. Slowly, conditions like diabetes and arthritis crept in. I can remember when we feared recruiting rheumatoid arthritis patients but gleefully took on osteoarthritis patients because they were “easier.” It was so obvious to the sponsors of the research but without directly asking, we had respondents we could not use because they were not asked properly. No one’s fault; we didn’t know what we didn’t know. There were times even that question was not on a screener because it was assumed everybody knew the difference, same for a discussion guide based on that assumption.

But we learned as we went along. I know I went from, “It will take max effort to find rheumatoid arthritis patients; can we use osteoarthritis if needed?” to “No problem at all.” We found resources, we educated ourselves and we tried to foster full partnerships among all parties in the research. 

Then the goalposts started moving. Suddenly it was rheumatoid arthritis patients using a specific therapy, perhaps one relatively new therapy, a much less common therapy or an insanely expensive therapy. And then there was this: “You want us to research WHAT KIND OF PATIENT? Only stress-incontinent patients? Why would anyone agree to discuss something so personal with strangers?” The answer to that came quickly: lots and lots of them. In both cases, once we jumped the first hurdle, these patients became as common to recruit as people who eat cereal for breakfast.

How did we jump the hurdle? We looked at how we had developed what we already had – panels, qual or quant, discussion guides and materials used for very specific non-health care respondents – and adapted them. For instance, recruiters added conditions to their main questionnaire database-gathering. Easy enough. How did we educate ourselves? One example was an aha moment that seems so quaint today: We realized could find patients through their comorbidities. In the mid-1990s, HIV/AIDS was “newish” to research. An added challenge was patients even admitting they had the virus because of the stigmas. Once we learned that many with HIV/AIDS also had hepatitis C, we found one more path to the patients. We had done research with hep C patients but if HIV/AIDS was not part of the discussion we didn’t know if they had the virus or not. Not only did we find more patients but we found them to be quite open to HIV/AIDS research as they had to been to hep C. 

Up next, we went from leaps and hurdles to the long jump. Now came cancer, multiple sclerosis, leukemia and others. I am not meaning to minimize arthritis by saying that. Everyone’s conditions are of vital importance them. We only live in our own bodies. But there are differences in how research is done with what we first called “hard patients.” It was next-level!

Cancer did not stay cancer for long.  It was separated out into breast cancer, non-small cell lung cancer, prostate cancer, colorectal cancer and so many others. Then it was stages of cancer. 

But two very important changes came of this leap.

• The first was a top-to-bottom, full-industry rethink. Moderators, agencies and the like had to become experts in specific conditions to speak to patients intelligently and thoughtfully. Over at the pharma companies there was a shift to a more fearless approach to market research that pushed boundaries beyond imagination. Again, not bad, just unexpected. 
• The second happened on the recruiting side. How do we find these patients? Specialty recruiters and referrals were early innovations. More than anything was the seismic shift on which the whole industry agreed: Caregivers are valuable too. Not just “too” but often more valuable. Now we could reach audiences like pediatric patients, neurological patients and late-stage oncology patients by speaking to their caregivers, people who often know more about the daily issues their loved ones encounter. They often have a more rounded perspective, so we happily spoke to them.

The progression is clear. Every time we made the impossible possible, a new wave of impossibility followed. And then the long jump became a full-on decathlon.

Part 2: The micro-condition

As I write this in mid-2023, I have had 17 projects this year where we were speaking to patients with conditions only maybe 100,000-500,000 people have worldwide. That’s not rare, that’s not even super rare, that’s incomprehensively rare. Can we even give them a 30-minute questionnaire? We can make assumptions based on similar patients but the answer will only arrive when we attempt it. Now we have moved away from using what we have learned over time to learning it on the fly with the rarest patients we will ever encounter.

But that’s only half of it. Let’s revisit our cancer patients. They were segmented by cancer type when we last discussed them. But now they were segmented by type, then therapy, then length of diagnosis, then specific experiences and on and on from there. Oh, but don’t forget the demographics. We also need them to be of a specific age or ethnicity or use a certain insurance (at least in the U.S.). The screeners and guides each stretched to hundreds of pages. I kept asking questions like this: “If there are only 17 people in France on this therapy, does age actually define them and thus have to be on the screener?”

If there is a development reason that only these 17 people can be on this therapy, age is suddenly as important to know ahead of time as being on the therapy. 

If we are asking age because market research has been asking age since we first asked cavemen what they thought of the wheel, we can safely break the habit for this screener without any ill effect on the research. 

If we are asking age in screening because it might be interesting to see how these 17 people might be thinking because they are different ages, it’s not a definer. It’s perfect for a qual interview but it does not actually affect the research. Put it in the discussion guide, not the screener.

If there are only 17 people on a therapy in France, we are beyond fortunate to find any of them; we should all be thrilled with that result.

We found the patients and we have found them with very rare conditions.  That much makes our entire industry the envy of all others.

Part 3: The questions

Now we come to what we actually want FROM the respondents. That will be different for every study, as is should be. But based on the path we took from the first arthritis patient research to the cancer study with only 17 people on treatment in France, we can at least set up guidelines that in total will keep us on that path of forever learning and adapting. 

Following are some suggestions and requests from clients, recruits and even respondents. If you are in any of those groups, and we all are, call me, e-mail me or send me a carrier pigeon so I can add your suggestions to this list.

• Remember that every respondent needs to be qualified, reachable, available and interested.
• Try to anticipate patient needs. We all have conditions, how do they affect us? When we look at very small populations of rare patients, we need to consider this far more than usual. For example, Alzheimer’s patients. Even with a mild diagnosis, what could be the pitfalls? They are able to function and be aware. In theory, at least. They may perfectly match OUR standards but if THEY have their own standards, self-perception wins.
• Be sensitive to timing. The truth is that the ideal time for the respondent is when the interview should happen. It may be ideal for the interview to happen immediately following an infusion visit, for example. That may not be possible, even with the best of intentions. Or the respondent may have a job that does not allow for a midday interview. How can we find a way to put the respondents first and still get amazing results? We need to adapt. 
• Again, stop and self-check. Is what we are discussing vital to the respondents? Years and years ago I received a furious call from a massive international cancer support group. They were outraged that we dared to ask stage 4 cancer patient about advertising, arguing that many of the patients may not live to be affected by it. Hey, ad testing is part of our business, it’s very important to our end clients. But the advocate wasn’t wrong. We can work around and satisfy both sides. I promise!
• Not everyone wants to do market research. It’s not an incentive issue, it’s not a scheduling issue, it’s a personal issue. They may simply not be up to it. Or just not like the idea of it. The universe of possible respondents is never the number of people with a condition. Even if you are doing a quant study of 10,000, that will never be the universe. 
• Impressive recruiting methods do not always reach everyone. They are amazing and I am constantly thrilled seeing the passion we all have for perfection. But there are aspects out of our control no matter how hard we try. What if you post an ad in an online support group and 50 people who would definitely qualify simply do not visit the site during the recruiting period?
• This one is going to be unpopular, so I’m tossing it out there and ducking for cover: with rare diseases, throw out the concept of incidence. Qual or quant, it has no accuracy. It’s a plain fact, albeit one a lot of people are uncomfortable discussing. We are not speaking to people who wear a certain brand of footwear. Incidence matters there. The population is so enormous, reality can equal prediction. But what is a 40% incidence among rare patients? How could any of us know? Is it based on the number of prescriptions a physician writes? That’s a measure of physician behavior, not patient behavior. Is it based on information from advocates or support groups? Maybe, but like HCPs, they only have access to people with whom they interact. If the medical community cannot say for sure how many patients have a rare disease, we don't even have a universe amount, the most basic number that goes into incidence.
• “Best effort” is best effort. This is not only a recruiting issue; it may also be a timing issue for moderators. It may be the end clients are not able to provide a list as they hoped or be stuck with a schedule even they hate.
• Always keep learning. I may be overzealous but the first time I see a condition, I dive into an investigation. At our firm, we consider that a part of the job. If you want to take a few hours to even learn the basics of a condition, go for it. This has always been my methodology anyway. Feasibility and honesty depend on facts; learn as many as you can.
• Partnership is key. Everyone involved in making a project happen, end clients to respondents, is an expert in their sphere. Put them all together and the chances of success are greater.
• Don’t say “no” without “but.” Possible solutions are key. They may not work but offering shows dedication and thought. When discussing caveats and feasibility that my clients do not necessarily want to hear, I always say, “Nothing would give me more pleasure than being wrong but let’s prepare for the worst.”

More than we need?

My original theory offered at the beginning of this piece was that we are victims of our own success in rare-disease work. It can now be posed as a question with a specificity that this whole industry loves: Based on the path we continue to travel together, are we so good that we’re getting to a point of deep digging that is more than we need? 

Not a judgement, just an honest question we can ask ourselves every so often as we learn and learn and learn about our place in rare-disease market research. 

By the way, I looked it up and it seems there are no recorded cases of people becoming fish or of fish counting traffic lights. But should either change, marketing research will be the first to know!