The Truth Is Out There –Are You?

by Laurie Gelb, July 28, 2010

Health care surveys still ask “analog questions” in a digital world, limiting the impact of disease management, marketing initiatives and even transactional communiqués like EOBs. Besides evoking “socially acceptable” responses (who wouldn’t want to be healthily skeptical and savvy, unless you wanted to be stubborn and oppositional?), surveys in 2010 often still rest on “service as product” and “product = attribute bundle” paradigms, which apply poorly to medicine.

As interventions move into the social and mobile media, the risk of pouring more money and brand equity into misguided action increases. In fact, adding to “stimulus overload” can hasten patient and caregiver denial, apathy, fatalism, overkill – and it’s ever-easier to tell everyone in their social networks how and why they got to that point.

Health care realities that are often overlooked by forced choice (e.g. A vs. B scenarios, point allocations, rankings) and attribute-based questions include:

• N=1. No two patients have exactly the same personal/family histories and environments. Yet we ask everyone the same questions. Why? We have computers now so we can personalize questionnaires in real time, the same way we say we want to personalize interventions.

• Heuristics – shortcuts – are more necessary to making health choices than any other kind. You may be able to consider all the possible routes to work in the morning, but you can’t consider – ever—all the supplements you could be taking.

• Opportunities to re-evaluate choices like daily dosing, glucose monitoring, diet, exercise are infinite– unlike  the dishwasher that you’re basically stuck with for a few years

• Instability/unpredictability of product “attributes” – we don’t all define “effectiveness” the same way, yet we all know what “four bedrooms” means, and the drug you took with no issues yesterday can land you in ER today.

• Inability to create what everyone knows would be the ideal product (want a vitamin water that melts solid tumors?), unlike, say, the cereal industry (Apple Jacks with 12g sugar/serving)

• No single “health personality.” For me, popping a naproxen is nothing; for my son, it’s agonizing. Yet he’s blasé enough to have visited a chiro, whereas I never will. So if you ask the two of us the same questions about beliefs and recent care, you’ll miss why our choices differ.

• In for a penny, in for a pound. A plumber can unclog the kitchen sink with no effect on the bathroom, whereas treatment focused on one system often adversely affects another. And when weighing the zero-sum game of deductibles, co-pays and OOP limits, it’s easy to feel that there are no good choices.

By replacing traditional questionnaires with decision-centered designs in which no two respondents may see exactly the same questions, we can understand and track what our audiences believe they know and the extent to which these beliefs are associated with their choices. With dynamic surveys, domains, measures and thresholds are not pre-established but are provided to us by respondents (with whom we are conversing, not forcing them to abstract something that is very real). This on-the-ground data enables us to better address knowledge gaps, evolving expectations, epidemiology/behavior, barriers to action and more – often with interactive tools. Moreover, patient, clinician and payor thresholds often differ significantly, creating misaligned incentives. When we understand how these differ, we can realign programs for “win/win/win” scenarios that optimize health outcomes.