HFMA VIEWS

Scott MacStravic, PhD

The book How to Lie with Statistics by Huff and Geis has been around for decades, certainly since I went to graduate school in the ‘70s. We used to refer to it as the “little orange book,” since it was a small paperback the size of your hand, back then, though I noticed that the latest edition appears to be blue. It is the most amusing statistics ever written, would be my guess, though that may sound like being “damned by faint praise.”

The content described the incredible variety of ways that people have found or accidentally used when describing statistics about things, usually making them appear to be much bigger than they really are. I found a common approach to this exaggeration in a “Weekly Research Recap” e-mailed by a company on July 5. It included each of the following headlines as “teasers” for various research reports:

• "Increase Customer Clearance Speed 2x…”
• “Improve “Wrench Time” (the time equipment is not working) by 2x…”
• “Require 2x Fewer Full Time Employees…”
• “Reduce Data Loss at 2x the Industry Norm…”
• “Decrease the Cost per Hire 2.5x…”

None of these headlines was actually used in the titles or content of any of these reports, only in the teasers, but all represent an unfortunately common way to describe how much something that sales prospects might wish to be reduced can be reduced. And none of these two to two-and-a-half degree reductions are mathematically possible, after all. While something can be increased by two or more times, it cannot be so decreased.

Since a “1x” decrease equals total elimination, i.e. a 100% decrease in whatever is to be reduced, there is no way to reduce it by more than 1x, and what the above teasers are probably trying to say is that the costs, time, or whatever else is to be reduced can be cut by one-half or more. I’m guessing that a “2.5x” reduction amounts to 60%, i.e where the old cost per hire is 2.5x the new one, so the new one is 40% of the old. A “3x” reduction would presumably amount to a two-thirds decrease, and I have seen this described in another context, as well.

Many hospitals are in the midst of reducing the prices they charge to uninsured patients, often by half or more, compared to their “list price” used in negotiating or describing discounts given to third-party payors. We can only hope that in addition to ending the egregious and highly unpopular (among consumers) practice of expecting uninsured patients to pay exaggerated list prices, hospitals will avoid the equally misleading use of “2x”, “3x” or any similar expression of how much their price reductions amount to. I have no idea how this latter practice started, but it is certainly a mistake to be prevented.