New targets need new solutions

Editor's note: Gennadiy Geyler is vice president of operations at MedSurvey, a Southampton, Pa., research firm. 

Over the past 15 years, the medical market research landscape has undergone dramatic changes. As the health care field has advanced and doctors have become more specialized, pharmaceutical companies have begun to seek survey respondents who are more specialized as well. A decade ago, a typical research project might target a list of 100 oncologists, broadly defined; today, projects may target 100 oncologists who specialize in a very specific condition, work with a very specific set of devices and are affiliated with a very specific subset of medical institutions at the forefront of their fields.

As pharmaceutical companies have become increasingly interested in targeting physicians who work in niche therapeutic areas at highly specialized facilities, clients’ requirements for survey respondents have been shifting. For many years, medical market research companies have worked largely with lists of specific physicians whom their clients wish to target. However, more and more frequently, instead of just providing lists of doctors, clients are supplying lists of centers of excellence or other medical facilities and asking to target any doctors affiliated with these institutions. Conducting market research based on a list of institutions rather than just a list of specific individuals presents a unique set of challenges that the medical market research field has yet to fully address.

When presented with a list of targeted physicians, many medical market research companies take rigorous steps to verify that a given doctor is, in fact, who they say they are. Methods for doing so include a broad range of strategies, often aided by recent technological advancements. For example, robust automated processes can check the location of respondents’ IP addresses against the state in which they claim to be located. Another effective means of verification is to use doctors’ NPI numbers to manually check doctors’ names and specialties against the information that is publicly available on the government CMS database.

Yet at present, there is no system in place to verify that survey respondents are actually affiliated with the hospitals or medical centers that they claim to be. There is a need to implement strategies that will ensure that survey respondents for these research projects actually align with the types of physicians whom clients wish to target.

For pharmaceutical companies, the lack of a robust verification system for doctors’ places of work is a problem that can lead to costly marketing decisions based on erroneous data.

Closing the knowledge gap

Let’s take a look at how this problem manifested for one client. For a certain project, a company wanted to target physicians who did not work at highly specialized cancer facilities. The client’s intention was to survey 80 doctors who had experience treating patients with a specific condition but who were not specialists in this very specific, niche therapeutic area. The goal? To compare the data they had already compiled on specialists’ knowledge with the newly collected data on non-specialists, in order to analyze differences between these two groups of doctors. This information, in turn, would help the pharmaceutical company to tailor its marketing campaign toward closing the knowledge gap between experts and non-experts. 

It should have been a relatively easy task to identify 80 physicians who were not affiliated with any of the specialized cancer centers on the client’s short list of facilities to exclude. After all, there are many more doctors nationwide who encounter patients with a given type of cancer than there are doctors who work at a handful of highly specialized facilities. Yet, when this research project went to field, nearly all survey respondents identified themselves as affiliated with one of the facilities on the client’s “blacklist.” As a result, the quota of doctors needed to complete the research project could not initially be reached.

So what happened here? Simply put, the survey as it had originally been designed included a screening question asking survey respondents if they were affiliated with one of the facilities on the client’s blacklist. In order to qualify for the study, doctors needed to indicate that they were not affiliated. However, many doctors (whom we might term “professional survey respondents”) are savvy enough to know that in most cases, marketing research companies are seeking to target specialists. Therefore, respondents will often falsely indicate that they are affiliated with a specialized facility in order to avoid being disqualified from the study.

This case study highlights an uncomfortable fact about studies that target doctors affiliated (or not affiliated) with specialized facilities. In the majority of cases, clients do wish to target specialists; yet as the study described above reveals, survey respondents are disproportionately identifying themselves as affiliated with the desired facilities, even when they are not. The result is that many seemingly successful research projects are in fact yielding erroneous completes and thus, inaccurate data. These data can, in turn, lead to expensive marketing mistakes for pharmaceutical companies.

Develop solutions

When it comes to verifying the medical facilities with which doctors are affiliated, medical market research companies will need to think outside the box and develop creative solutions. The following are just a few that can be very effective.

First, when designing any survey, it is critical to avoid asking leading screening questions whenever possible. This is not an earth-shattering proposition but it can require putting in some additional time and effort. If a survey asks doctors to choose the facility with which they are affiliated using a dropdown menu or checklist, it is posing one of the most leading questions that can be asked. The list immediately signals to doctors: “We are looking for survey respondents from one of these designated facilities; to qualify, all you have to do is select one of the options provided.” A much more effective means of determining a doctor’s actual affiliation would be to make the question open-ended. In other words, ask them to fill out a blank field with their place of work.

Of course, this strategy can lead to its own challenges. For example, there may be multiple ways of typing in the name of any given facility. Doctors affiliated with the University of Pennsylvania might indicate that they are from “UPenn,” “Penn,” “University of Pennsylvania” and so forth. Manually sifting through these responses in order to group doctors from the University of Pennsylvania together into one category can end up being arduous and time-consuming. Ideally, medical market research companies would take advantage of technology to simplify the process. They might employ a high-level programmer to create an automated solution that can decipher the multiple ways in which the same facility can be identified. Although this solution is not simple to put in place, it can be quite effective; and in the long run, it can save your research project from disaster. If the ultimate goal is to provide pharmaceutical companies with accurate data, asking doctors to manually enter the facility with which they are affiliated may be well worth the investment.

Another solution to the problem of physicians falsely identifying the medical facility with which they are affiliated is to keep the initial leading question but to manually verify responses on the back end. Again, this method requires an additional investment of time and effort but the payoff justifies the cost. Today, information about doctors’ specialties and places of work is readily available on the Internet. Simply conducting a search for each doctor’s name can be a simple and effective verification technique that may take only a few hours to complete. The end result is well worth the headache.

It is important to note that simply adding a disclaimer to a leading question, while easy to do, is not a reliable solution. It may seem reasonable to simply state upfront that physicians will not be paid if they misidentify their place of work and that their responses will be verified. Unfortunately, many survey respondents will either skip over “the fine print” or simply disbelieve the disclaimer when they do read it. After all, if they take surveys often, they will know that their responses have not been verified in the past.

Perhaps the most efficient solution that medical market research companies can employ is to create a very specific database of physicians, geared toward a specific research project. Various sources exist for determining which doctors are affiliated with which medical facilities. For example, hospital databases are publicly available and the doctors listed there can be identified by address, NPI number or another indicator. It is therefore possible for research companies to generate a project-dependent internal database of physicians affiliated with a given facility, which will ultimately simplify the verification process for survey responses.

Put a system in place

Just as it is critically important to verify the identities of survey respondents on client lists of targeted physicians, it is also imperative that medical market research companies put a system in place to verify that survey respondents have correctly identified the facility with which they are affiliated. If research companies fail to take these steps, they fail the pharmaceutical companies that rely on the data generated from their surveys. 

Operating on faith about the veracity of physicians’ survey responses is a surefire way to sabotage a research project. Pharmaceutical companies are counting on receiving accurate data to make important, often expensive marketing decisions. It is crucial for medical market research companies to implement creative solutions and streamlined verification processes that will empower pharmaceutical companies to make decisions with confidence.