Meaningful Data Analysis and Adaptive Patient Recruitment Tactical Efficacy

Meaningful Data Analysis and Adaptive Patient Recruitment Tactical Efficacy

By Joan F. Bachenheimer on Wed, Jun 25, 2014

describe the imageThis is the second part of a two-part series by Joan F. Bachenheimer, founding principal, and Matt Kibby, principal, technology and innovation, BBK Worldwide.

Even if you were able to correlate advertising referrals to consented patients with a high degree of accuracy (and then to screen-fail and randomization) the fact of the matter is, "So what?"

Four scenarios spring to mind with regard to the quest for this correlation:

  1. What are the metrics used to justify giving research sites money to run their own media outreach? Typically sponsors cannot track if a site used the funds to place media, or whether it was used to enroll their study or a competing study. But they know that sites want it because it works for something.
  2. What about the effect of patient-facing support materials and a broad-based media campaign on a site’s sense of support and morale – and how is this captured in our referral-to-consents data? This intuitively offers a real effect on the way sites screen their patients, but it doesn’t appear in any referrals-to-consents data.
  3. Does this mean that efforts such as those mentioned above can’t be understood or measured? Or that they have no discernable value? Of course not.
  4. What caused the following turnaround? In a rescue situation, screening activity at sites often improves prior to the launch of advertising while the materials are in EC review because of the new scrutiny and energy applied to recruitment and site intervention. No referral-to-consent data in this situation.

The real questions to ask are; "Did this effort help to accelerate recruitment timelines?" or "Have I accelerated my time to market for this compound?" This move to newer, more relevant metrics is driving the construct of adaptive recruitment and fostering a new appreciation for the power of BBK Worldwide’s TrialCentralNetSM and its ability to deliver strategic insight for better enrollment planning and mid-course correction, as needed.

I can still hear the clinical teams asking, "Why are recruitment firms not able to measure referrals and referral sources to consents accurately?"

There isn’t a single recruitment firm that maintains control over all of the data from referral to consent. The necessary interference of the site and the need for multiple systems interjects multiple data matching problems. Consider:

Sponsors are increasingly less inclined to collect data in the IVR system that can be useful for tracing a patient back to a consumer referral. The reason for this mirrors BBK’s concern about patient information privacy. You just don’t need it. In 9 out of 10 data feeds, we can see that:

    • Gender is no longer recorded
    • Initials are no longer recorded – or are entered by the site user as a simple code such as AAA, BBB, or CCC etc.
    • Date of Birth is no longer captured – or is not recorded as an accurate date and rather as a simple date such as 1/1/1956, 1/1/1970, etc.
Relying on site reported information is inaccurate and misleading. See my blog about Money(ball) Recruitment to understand why sites can misperceive the macro-level effects of a recruitment campaign.

By every discernable metric, recruitment campaigns have a combined effect of raising screening rates at participating sites by 15-50 percent, depending on other factors such as the protocol and the therapeutic area and condition. You just don’t need to jeopardize patient confidentiality to understand the effect of what you are doing. It’s like a crazy Pavlovian conditioning to want a metric that’s not valuable or necessary – when better information is there to assist the strategic decision-making process.

mHealth + Patient Engagement eBook

Topics: BBK Worldwide, Patient Recruitment, Technology