Your Honor, if we could voir dire the witness?

All right. VOIR DIRE EXAMINATION

BY MR. UELMEN:

Mr. Matheson, how many PGM sub-type tests has your laboratory performed?

Well, I have specific figures on the number that I have run. I determined that I've done approximately a thousand items by the PGM sub-type method. The only information I can supply you as far as overall laboratory is since we started computerizing our statistics, which was the first part of 1990, and since then the laboratory as a whole has done a little over 1800 items by that method.

Now, you indicated that you've done about 6500 ABO typing tests?

Yes, that's correct.

But you've done less than a thousand PGM sub-type tests?

Slightly over a thousand, that's correct.

All right. So you do a PGM sub-type in about one out of six of the tests where you do an ABO type?

That's approximately how the math works out, yes.

How do you select which cases you're going to do a PGM sub-type?

Well, it depends an awful lot on how much sample there is to begin with. One of the -- our protocol has tended to be to do the ABO first, and then go on to the electrophoresis system. I mentioned earlier this group 1 system, which includes PGM as one of the enzymes in it, that you normally try to run first and determine what the PGM type is. Not the sub-type. And if you don't get a PGM type, then there's no reason to expend further sample running the sub-type system on it. So many times we don't run it.

Well, would you say then that there's some sort of random selection of the PGM sub-types from the larger group of ABO types that you do?

You mean as far as the ones that appear in our frequency charts or in the ones that are run?

What I'm saying is that the individual cases where you have determined to run a PGM sub-type are not a random sample of the ABO type tests that you've done.

I'm not totally sure what you mean by random. I do know that -- I mean, we follow certain criteria as to whether we run a particular test, and it depends on the quality of the sample. The data that appears in our frequency charts is based strictly on whole blood samples, not evidence stains, and we will not run a particular marker and whole blood sample if we don't have something to compare it to in the stain. So, you know, I'm not sure if that's random or not, but sometimes we run it and sometimes we don't.

And the frequency chart that you maintain for PGM sub-type, that's just based on the PGM sub-type tests that you have done in your laboratory?

That's correct.

And you compute that by race of the subject that you're testing?

No, the statistics that we keep in our laboratory, the general population statistics that I use to calculate this do not take race into account.

So you have no race specific frequency data for PGM sub-type?

No, that's not true. The chart that I used in this case and that we use regularly on all of our analyzed evidence reports does not break it down by race. However, we do maintain reference books in the laboratory that does if it's needed.

And these reference books would refer to frequency data collected in other laboratories?

That's correct.

But you just used your own data?

Yes. We always do.

And your data is race specific for neither the ABO nor the PGM sub-type?

No, we don't determine race on any of our specific -- on any of our frequency determinations that we have. We don't feel it's our job to determine what race left the stain. We just have a stain of an unknown origin. So we choose to report a general population statistic.

But yet you're aware that there are dramatic differences in the frequency of blood type, for example, among Blacks or African Americans as compared to whites.

Whether it's dramatic or not depends on the particular marker. I do have data here on the racial break down on the three genetic markers that I did run in this case, in case that information was needed.

Well, are you aware, for example, that other data suggests that type B is twice as frequent among the black population as it is among the white population?

Oh, I don't know specifically what it is. I'd be happy to look it up and confirm it, but I'll accept the fact that there can be major differences, particularly in the ABO system. In the PGM sub-type system, it doesn't vary quite as much as that.

Oh, it wouldn't surprise me that if in, you know, specific populations it can vary by that much. That's one reason why up to this point I've specifically mentioned that these frequency determinations were approximations.

And what you've done in order to come up with this .43 percent, or 1 in 200, is simply take your calculations of the frequency of each of the markers and then multiply them together to come up with 1 in 200?

Well, what is the frequency of type A that you used to multiply?

For the data that I have there, I used 33.7 percent.

And for the PGM sub-type 2+2-?

1.6 percent.

And you simply multiplied those together to get the .43 percent?

Well, also included in there is the esterase D type 1, which I used 79.6 percent.

Now, you have made no effort to correlate the frequency of PGM sub-type with the frequency of a particular result in the ABO test, have you?

Oh, you mean to see if there's any sort of correlation between the two?

Exactly.

No, in our laboratory we have not done that. I have learned throughout my training and have been advised that they are independent.

Been advised by who?

I don't have any specific name or reference. It's just been something that's been -- occurred during the course of the last -- last, oh, 12 years of my serology training. It's my understanding that they are independent factors.

But none of the tenting that you have done or the accumulation of data that you have done attempts to correlate how frequently you would have a 2+2-, for example, in an A blood sample as compares to a B blood sample nor an O Blood sample nor an AB blood sample?

No, within our laboratory we have never done those calculations.