Are You Positive?

by Stephen Davis

  “Would you please explain this chart, Dr. Tillman?”

  “What it shows is that out of the forty-four times we processed this same blood sample, 26 times we found four proteins reacting on the HIV Western Blot test; ten times there were three proteins; twice there were two proteins; five times we found only one protein; and most unbelievably, once there were no proteins at all that reacted.”

  “And what did this tell you?”

  Tillman is obviously ashamed of these results, and almost appears apologetic. “Well, the first thing I did was to see what lab tech did what test, expecting to find one particular person screwing up some of the results. But that wasn’t true. Some of our best techs were getting different results each time, and no one was consistently wrong. So I literally had to rule out human error as the cause of such diverse results.”

  “If it wasn’t human error, what was it?”

  Sitting up very straight, Tillman answers abruptly. “Wait. Let me qualify what I just said. There could have been some human error in one or two of the results, but not enough to explain this entire chart. And since it wasn’t human error, the only other possibility is that the test itself is not reproducible. In other words, you can get different results at different times doing exactly the same thing.”

  Looking at the jury, Campbell suggests, “That doesn’t sound like a very reliable test.”

  “No, it isn’t. But the most shocking thing about it was that the results often mean life or death to someone. In this case, about 80% of the time this person is going to be told they are confirmed HIV-Positive. But more than 15% of the time, they would be told they are Indeterminate, and once they would be told they are HIV-Negative. And it all depended on which day we did the test. That doesn’t speak well for the accuracy of the test itself. If we can’t get the same results every time with the same blood sample, or at least 98 or 99 percent of the time, then we’re causing a lot of people some very undue pain and heartache, and sometimes even death.”

  Campbell is not only willing, but anxious for this witness to express his emotional reaction as well as his scientific expertise. It’s time, he thought, to make this real to people, to show the very human side of it all. Unfortunately, he has to keep his questions focused on the science. “Did you use the same criteria to interpret the results each time?”

  “Yes, we used the CDC criteria.”

  “And if you had used the FDA criteria?”

  “Let me see.” Tillman pulls a card out of his pocket and then looks at the chart on the screen. “Well, it’s even worse. The results would have been Positive only 26 times, or about 60% of the time, and the rest Indeterminate; and of course, with one Negative still.”

  “So are you saying that the results of an HIV Western Blot test are so unpredictable that it literally depends on who processes the test and what day it is?”

  “Unfortunately, that’s exactly what I’m saying.”

  “How did you feel when you discovered this?” Campbell sneaks a quick peak at the Solicitor, hoping Armand doesn’t object to this line of questioning. Looks like he’s going to sit still, at least for a while.

  Tillman sits back in his chair. “I later found out that this person had been told they were confirmed HIV-Positive based on the test results from another laboratory, and that didn’t sit well with me. I mean, anywhere from 20-40 percent of the time in our lab, they were not Positive; and I was very uncomfortable that we were giving people a death sentence based on these results.”

  “So what did you do?”

  “I began to do my own private research to see what was happening in other labs with these HIV Western Blot tests. Naturally I was concerned that there might be something very wrong in our lab if we were the only ones to come up with these kinds of results.”

  “And what did you find out?”

  “That we were definitely not alone. This kind of thing was happening everywhere.”

  “Can you give us an example?”

  “I believe, Mr. Campbell, you have another chart you can put up on the screen?”

  “Which one?”

  “The one with nineteen different Western Blot strips.”

  Campbell shuffles through some transparencies until he finds the one Tillman wants.