Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Sunshine Heart And The Abuse Of Statistics...

|Includes: CHF Solutions, Inc. (CHFS)

Sunshine Heart (SSH) has found multiple ways to abuse statistics and I thought it might be instructive to look at some. The data comes from "Ambulatory Extra-Aortic Counterpulsation in Patients With Moderate to Severe Chronic Heart Failure," which describes the results of their "feasibility" trial of 20 patients:

www.sciencedirect.com/science/article/pii/S2213177914002728

Three patients died and the authors, on that basis, claim,"One-year survival was 85%." However, that is actually BALONEY -- not even very good BALONEY. They leave out a few minor details in their calculation. Specifically, they ignore that one patient discontinued c-pulse due to PIL problems, two patients left to get LVADs, two patients left to get heart transplants. It should be obvious that they can't claim these patients as one year survivors of the c-pulse, since NONE of them had the c-pulse for a year. So a more acurrate number would have been 12 out of 20 or a SIXTY PERCENT one year survival with c-pulse.

They do this yet again in discussing re-hospitalization rates:

"Over 12 months, 3 of the 20 implanted patients (15%) had 5 Clinical Events Committee-adjudicated heart failure hospitalizations."

Once again they IGNORE that there are NO LONGER 20 c-pulse patients, there are only TWELVE left in the study. They don't break the data down to indicate how many of the remaining 12 had heart failure hospitalizations, but obviously, 3 out of 12 (25%) is a substantially larger and WORSE number.

Ditto regarding "exit site infections," claiming 40% (8 out of 20), but once again not breaking down who had these infections. The problem is again that 20 did NOT complete the study, only TWELVE did. If it was those twelve that got line infections, then 8 0ut of 12 -- SEVENTY FIVE PERCENT -- is again obviously a significantly larger and WORSE number.

Of course , it's imposible to tell whether these statistical "distortions" were done purposely to decieve or simply out of ignorance is impossible to tell. Nevertheless, it demonstrates the importance of closely reading the data and checking to see what all has been left out of the calculations.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.