As such, we asked Dr. Amanda Sozer to come and talk to us about the issues of evidence collection, prioritization, et cetera, and some of the roadblocks that occur in that dynamic with an eye towards any recommendations that the Commission could come up with to help develop pilot programs that would, in fact, facilitate better coordination, more effective use of database sample collection and analysis.

Dr. Sozer received her undergraduate degree from Rikers University and a Ph.D. In biological sciences with a specialization in genetics and emphasis in biotechnology from the University of Tennessee, Oakridge Graduate School of Biological Sciences.

She began working with forensic DNA testing in 1990 at Cell Mark Diagnostics. She is currently the associate director at one of the major private DNA labs in the country, Fairfax Identity Labs, and is actively involved in collection and analysis of CODIS convicted offender samples from a number of different States.

So with that, thank you, Dr. Sozer, for being here.

DR. SOZER: Thank you for inviting me here today. Can everyone hear me? Okay.

I came today to share with you some of the difficulties encountered in the collection and assessioning of samples for CODIS.

When I talk about collecting samples, I don't only talk about putting a needle in someone's arm or taking a buccal swab, but actually trying to identify who you need to collect, actually obtaining the sample, and completing all the required paperwork to maintain the chain of custody and identify that sample.

When I talk about assessioning of samples, I talk about receiving the samples at a laboratory, logging those samples into a computer system, verifying the samples, and preparing the samples for testing.

During the talk today I'm going to tell you about the experience that our laboratory has in collection and processing of specimens for testing. Then I'm going to talk about the specific -- I'm going to go over the specific steps of processing a sample for CODIS testing.

I'm going to talk about the difficulties associated specifically with the collection and assessioning of those samples and I want to say that to my knowledge I don't think that this topic has ever really been formally addressed before this meeting.

Unfortunately I don't have exact numbers on how prevalent some of those roadblocks are. I can tell what our experiences are with the States that we have dealt with, especially Ohio and Louisiana. And if you would like, I would be happy to get some exact numbers at a later date and present those to you. I'm going to conclude my talk by presenting some suggestions and ways to streamline the collection of samples for CODIS testing.

Now let me start by telling you a little bit about what our laboratory has experienced then, so you know where we are coming from.

Our laboratory started DNA testing in the early 1990s. We were actually awarded the first DNA only state contract for paternity testing. So we were the first private laboratory to be doing DNA only testing for paternity.

We currently test about 40 to 50,000 samples per year for paternity testing. We do that by RFLP, by PCR, we check blood samples and we also collect buccal swab samples.

I think what's important with this is that when we are doing the testing, we're involved in the collection of samples and the receipt of those samples into our laboratory.

We do collections in a variety of different ways. We have mass collections where people all come to the same location, maybe at little variations in times, and they are collected en mass and the samples are shipped to our laboratory.

We do collections in accordance with court proceedings, too, so we have our employees at court taking samples when people are done seeing the judge. We coordinate with jails and prisons for jails and prisons to collect samples for us and send us samples to our laboratory.

And we also coordinate with jails and prisons to send our own employees into those facilities to collect the samples and send them to our laboratory. So we have specimens coming in from a lot of different locations.

In addition to our paternity testing work, we also do work for State agencies in conjunction with CODIS. We do this for a number of different States.

We do work for Ohio currently. We primarily receive and log in samples that they've collected in Ohio and they come directly to our laboratory for assessioning and resolution. We receive approximately 18,000 samples from Ohio.

We are currently working with the State of Kansas in doing RFLP testing on convicted offender samples that they were unable to process in their laboratory.

We were recently awarded a contract with the state of Florida to do STR analysis on their specimens that they had previously tested by RFLP.

And finally, we have a contract with the State of Louisiana crime laboratory to get their DNA lab on-line for STR testing. In addition, we provide assistance in getting their CODIS program started and running by next year.

Louisiana is very unique. They had a law that was passed very early on in the CODIS testing that was rescinded because of lack of funding. They have come back and they are one of the last States to have a law passed about rules and regulations for collection of samples.

And when they wrote their law, they wrote it very broad. So they are not only collecting samples from convicted offenders, but on arrest.

So this brings into light a whole new array of issues. And actually, the more we started working with Louisiana and the more I started talking to other States about what do you do in this situation to try to learn from other States, I realized that some of the problems that Louisiana will encounter in collecting samples upon arrest are really not that different than other States encounter now in collection upon conviction.

So this is when I really became aware of a lot of the issues. So why don't I start by going over the steps for processing specimens in general. One little line may take a whole week to do, so they are very short.

But first you need to ascertain which individual is to provide a sample. And usually this is mandated by the State's law. You need to get that individual to whatever collection location you're going to be collecting that sample at, and I will go into these in more detail.

You need to identify that individual and you need to collect a specimen from that individual.

Then you need to ship that sample to whatever laboratory you're going to use, and you will need to get that information into some type of data management program and verify that that sample is in fact a valid sample to test.

These are the areas that I'm going to talk about today. The remaining areas have been addressed, sometimes in great detail, but with a lot of research, with a lot of new automation techniques, and those would be to test and analyze the sample, to upload the profile into CODIS. And then into addition to that, you're going to need to search your database and do data management, including expungements, which the top issues can effect the expungements greatly, and what to do when you have a hit.

So why don't I go through each of the first initially steps, step by step. The first thing that you need to do is ascertain which individual is to provide the sample. As I said before, usually you identify which offenses and this would be mandated by the law.

You want to make sure that you don't collect from individuals with other offenses, and this actually is a big problem. Carl outlined yesterday that very often criminal records are not well maintained, and bottom line is that people who should not be collected are collected.

So you're collecting extra samples and those samples will come into a laboratory and you have to do something with that sample.

The other problem they can have is that you may duplicate samples. People say, oh, I'm not really sure it's been done or not, so they will do it again, then you have a second sample. Now you have selected a sample again that comes into the laboratory that you have to do something with. And each time it this happens, it's a tremendous amount of manpower that goes into resolving these issues.

I can give you some numbers. In Ohio I think that the duplicate samples that they get in are probably relatively low as compared to some other States, and there may be a variety of different reasons. It may be that they are just starting their program. They started collecting samples for a little over a year now. The longer you collect samples the more duplications you're going to get. So they have about a 4 percent duplication of samples.

In Florida they say it's about 10 percent that they duplicate their samples, and 2 percent of those samples -- when the samples coming into the lab they look for duplicates. 2 percent get by. They have a different corrections number, they have a different name, they have a different Social Security Number, they have a different date of birth.

So there are a large number of samples that go all the way through the testing process and it's not until they put them into their database that they realized they have duplicates. So I think in some States it's even higher.

COMMISSIONER SMITH: What percent are individuals not covered by the statute?

DR. SOZER: In Ohio I probably have the best feel, and it may be less than 1 percent. But even a very small number of people that are there and shouldn't be there, it's a tremendous amount of work. Someone has to request that that sample be expunged, someone has to research that, someone has to physically get the sample, destroy the records, destroy the sample, and it's a tremendous amount of work. So even a very, very small amount of samples really can translate into a tremendous amount of work.

You can go into a collection facilities where they have, by mistake, 20 samples or 30 samples in one day taken that they shouldn't have. It's lot of manpower.

COMMISSIONER SMITH: When you say somebody has to request the expungement, somebody other than your own personnel?

DR. SOZER: We don't know. When samples come into a laboratory sometimes you don't know why they are there. Someone will have to identify that that person was collected. Maybe it will be the individual themselves who say according to the law I shouldn't be collected, and they will talk to somebody in the department of corrections. So there may be four or five letters that are associated with that sample before it's expunged.

COMMISSIONER BASHINSKI: Is that because you are getting the sample directly and it's not going through the law enforcement agencies? They are not doing any screening, right? You just get everything.

DR. SOZER: Right. For Ohio we are taking everything. And actually we report back to -- Ohio had a very new program. We report back to Ohio on who has been collected and then they will go back and verify that everything has been done correctly. So there is one more step. But actually, if we were the testing lab initially, you will bring us samples in and look the samples up.

COMMISSIONER BASHINSKI: Do they have thumbprints?

DR. SOZER: Yes.

So you identify which individuals need to be collected from. And then the next step is you need to get those individuals to a collection location.

There are a variety of different places that someone can be collected, that that sample can actually be taken.

You can do it upon arrest, and this is what is -- again, these are not convicted offenders yet, but down in Louisiana there are many different Sheriff departments where individuals can come through, so there are many different locations where there will be collections.

Some will come in multiple times. They will come in this Friday in that jail, and next Friday they will come in at another jail. And there is a tremendous problem with collecting duplicate samples.

And the third problem is correct identification on arrest. You're not really sure who this person is when you're arresting them. Again, these are issues that transcend, actually, through a whole variety of different collection locations, they're just really accentuated upon arrest.

The next place that you could potentially collect people is at the time of sentencing. Very often people will be arrested, they will go into court, they will be either put on -- I'm not an attorney so I may not always be using the correct term -- they will be put on probation, they will be let off for time served while they were in the county jail, and they will walk out of the courtroom.

And once they are out of that courtroom, once they are out of some facility where you have great control over someone, it's very difficult to get someone to come back in. And actually it will go down to this area down here during probation, and I will talk about some of the issues.

You can collect a sample while at intake, so if they are sentenced, convicted and moved on. A lot of States have intake units where people stay for about a week where they have medical screening, psychological screening, and you can collect sample there.

And that seems to be a very efficient place to collect specimens. They are a small number, and there is a lot of record information that is maintained. And I think the States that I've had that can take while people are at intake have a good collection record and a small number of errors.

You can collect people at the prison while they're incarcerated, and there are a whole list of problems with this. For example, in Texas people are moved in the middle of the night and you don't know where they're going to be moved to and you don't know when they're going to be moved.

And this is actually a problem that we see in paternity collections. We do collections for the State of Texas. We'll have a phlebotomist ready to go in, will call the day before, is the inmate there. Yes, he is. We have to call the next morning, but they're moved in the middle of the night to prevent the association of gangs. So it's very fluid in the State of Texas.

In many of the department of corrections, the nursing staff can't help in the collections. They've taken an oath that they will not do this. So you need to send someone else to go in to do the collections, and I will talk about the collections in a little bit.

You can collect them from the local jail. Again, there are a large number of local jails, each run differently in different areas.

You can collect upon release. Again, this can be difficult, but I think that they have -- when do you something like this you do have a fairly good collection rate. Again, you're looking at records and people are very careful about what they are doing.

Then you can collect on probation or parole. And this, in my experience in the States that we've deal with, is almost impossible to do. It's extremely time consuming.

You have to be able to identify who this person is. This person has no reason to want to go give a sample. You can send him a letter, you can send them a court order, please go get a sample. And very often this is the area where you have a lot of miscollections where someone will send their brother, send their uncle, send their best friend in to do collections.

You can have someone go with them to the collection sites, but then you have a parole officer who spent their whole day driving across town to the local lab to give one specimen.

So this is really very high. And again, if you have the wrong person go in for a draw, even if it's less than 1 percent of the time, when you're talking about millions of samples that we need to collect, that is a big, big issue. It's really a big issue to resolve these problems.

COMMISSIONER GAINER: Doctor, when you talk about the problems, except for misidentification on these probation ones and maybe upon release, are you really talking about administerial cost effective profit margin problems?

DR. SOZER: All of those, right. You're talking about logistic problems. When somebody says oh, it only takes ten minutes to do this. We say ten minutes times one billion, that's a tremendous amounts of time.

So I'm going to talk about some solutions to this so I don't want to only be the bearer of bad news.

It's also hard to quantitate, because you have a lot of different departments involved. You have department of corrections, you have the crime lab, you have probation departments, you have the courts. You call up and talk to somebody for ten minutes over here and you talk with somebody for ten minutes over there, and they send a letter over here. It's a tremendous amount of time. And when you go between departments it's hard for someone to say it's this many dollars over here.

COMMISSIONER GAINER: Are you using, for instance, in Ohio the 4 percent duplication? Are you using that as your blind efficiency testing?

DR. SOZER: We actually do. They are logged in and we know that they have been tested before.

We get duplications in for two reasons. You will get duplications in because the first time the sample came in the chain of custody isn't adequate, the thumbprint was not adequate. This is actually a big problem I think in a lot of states where you can't read the print so you need to go back and recollect that specimen.

We will have problems with paperwork. We will get samples in from well-meaning people who just don't send any paperwork besides a name on the tube and you just get a tube in like that.

And I think Ohio is very good about responding immediately to this, calling back the collection location saying this is unacceptable. And usually they kind of come and go in spurts. You will get some samples in, Ohio notifies the collection agency, and then it stops.

COMMISSIONER SCHECK: Which states have you been dealing with people on parole or probation? How frequently is that actually?

DR. SOZER: That from our lab, none. We are not getting any samples in that I know of. We get a few samples in from Ohio from smaller locations, but the main information that I have gotten from this is from Florida and Kansas.

COMMISSIONER SCHECK: So these are things you have been told about?

DR. SOZER: These are things that I've been told.

COMMISSIONER SCHECK: In those instances where somebody comes in and gives blood or a buccal swab for someone else, do you know if they're criminally prosecuted or if there is any of that.

DR. SOZER: I don't know. And actually Cindy Schuler (phonetic) is running the CODIS program in Kansas. She said that all of their mis-samples have come in from those types of situations. They're usually when the parole officer doesn't understand how important it is to identify this person, the parole department is given a kit, and the parole officer will give it to the parolee and say, here, take this down to First and Main and have your sample collected, and give the kit to them.

Again, it's an educational problem that is ongoing. You can educate people today and have new staff next week. And so it's an ongoing problem. But that has been the area that they have had the most problems.

Actually, they have sent people from their forensic labs into the facilities to have these samples collected initially to make sure they're done properly.

COMMISSIONER SMITH: Do you know if any of those jurisdictions whether the parole agents themselves collect the samples?

DR. SOZER: I think in -- somebody else may be better off to answer that question. I know that there are some areas where they will do buccal swabs, I believe, at the parole office. And then those samples will be sent. And I thought it might have been California.

COMMISSIONER BASHINSKI: No.

DR. SOZER: Okay. Very often when you have people who are not involved in the end product in the collections, and we see this in a lot of wide variety in testing, we see it in paternity testing where people take in the initial information and don't understand how important it is or people doing intake collections that when it says fill it out correctly, that means fill it out correctly. That means everything is correct.

And when they don't have that understanding and if it's not accurate, you almost wish you never even got it in the first place because it takes so much longer to resolve an issue. And very often it ends up that you have to get another sample collected.

COMMISSIONER GAINER: Just one more question. Based on that 4 percent, if you're using it for blind proficiency, have you had zero error rate in that? Do you ultimately match up the 4 percent.

DR. SOZER: They haven't gotten as far as testing. The reason why they're collected, at least in Ohio, from my feeling of when we look at when they come in, is when they have moved from one prison to another prison. Usually we don't get duplication from an inmate in the same prison. It's when they move to a different prison that we get it.

When you talk about when people say oh, we get them in twice, you can use them as blind proficiencies, that is a good use for those samples coming in.

But there is also a lot of investment in the collection process. And if you're talking about 10 percent of duplicates and 10 percent of the time you have to identify these as a duplicate sample, it won't go in the database the second time, we won't run it but we will compare it, at some point it's really not a cost effective measure.

Some of the duplicates that we will get into are acceptable duplicates where the first time the sample was not acceptable for testing, the second time they will come in and it will actually replace the first sample.

We talk about the identification of the individual and the collection of the specimen. And I've just listed that the collections can be performed by a number of different agencies and very often in a state a number of different agencies will perform collections at a different time.

The department of corrections can do that, the crime laboratory actually can do that. This is what Kansas did. They actually send someone in from the crime lab to collect the samples, so they understood what you needed to get for the end product.

Some states say that it's just the agency responsible for the care of the individual. Whoever has that individual under their care will be responsible for collecting that specimen. And actually that is what Kansas does now. If they are on parole, it's the parole office, in it's the local jail, the local jail will be responsible for the collections. The crime laboratory assists in that, but the agency is the primary person involved.

Then some states will contract out and have a contractor come in and do the collections. Sometimes this may be a legal lab or hospital. In fact, Kansas uses that also, where a hospital will have kits and people can go to the hospital and get collected. Other people may send in independent phlebotomists to go into the prison to do the collections.

I think that this is something very similar that has been in paternity for a long time. We just send someone in and we may have five or six individuals in that prison and we will collect them all on that same day.

But again, it's a lot of time for someone going to the prison, get through the door, collect the samples, go home. It's a tremendous expense.

Some things that you have to keep in mind. Is this the correct individual. Is this the individual who you think they are. Is all the paperwork complete and accurate. If it's not, if you have an SID number or your corrections numbers and you're missing one digit, that is a tremendous problem. You have to get back on the phone when the sample comes in, why is this a digit off, should it be this, should it be something else.

And again, people don't understand that when we say paperwork needs to be complete and accurate, that means it needs to be complete and accurate. That doesn't mean that it's just complete because you have written something in the line. It really means that it's accurate.

You need to make sure that a sufficient sample must be obtained. Sometimes this is hard. In some of the prisons they will use force to get the sample. In Kansas they will send a team of 12 individuals that will hold the inmate down and hold the arm behind the back and collect a sample. 12 people to collect one sample. That is very inefficient to do that.

The sample must be identified properly. That means that the information on the paperwork, whatever that submitting paperwork is, needs to match the information on the tube. Again, this can be a training and conceptual problem for some people to understand what that means.

And the chain of custody must be maintained. That means when collected, you seal it up. When the samples get to the laboratory for testing, they need to be assessioned and put into the computer and received and they need to be -- let me backtrack.

First the samples need to go from the point that they were collected to the laboratory. So places go by U.S. mail, sometimes they are hand carried in, sometimes they will go by some overnight carrier. Usually it will depend on how urgently the sample needs to be tested. This is a big expense also. When you talk about shipping a sample overnight at $15 a package you don't want to do that all the time.

Hand carried is time. Again sometimes these things are hard to evaluate because it may be another department who says, well, we just send somebody over to the laboratory every time we have a specimen. That may be two hours, three hours of time to send samples in.

Verification that the sample is received by the laboratory is also important. You want to make sure that the samples you collected were received. The samples that weren't collected were never received, so you don't have other samples coming in to the laboratory that were not collected.

So nobody is sending in any other samples. And this is important. We see this in paternity testing, where samples are collected at a draw site and we are a lab and there is a lab in Richmond that gets the same type of packages that we do. We have to be very careful. We actually track our specimens when they're collected, and say that we know that we are getting this tracking number, this tracking number, this tracking number, and this tracking number, and they don't end up in some other lab. There are a lot of labs in the United States that do testing, but they don't have to worry about it.

The samples come in, they just bring them all in. They open them up and two days later they say this doesn't look right. We need to make sure that what you have collected actually gets to the agency, so there needs to be talk back to the individuals who scheduled the collection.

We need to ascession and verify the sample. You need to make sure that there were no signs of tampering. You need to make sure that the paperwork is consistent. You need to look for duplicate samples. You need to verify that the samples should be tested.

Very often fingerprints are verified and if there is any problem with any of these, the submitting agency is contacted, and usually that can take multiple phone calls and multiple weeks to resolve an issue. Why was the sample collected. Well, I'm not really sure. That person is on vacation for two weeks. We will get back to you.

Florida actually has a computer in their lab where they can go back and look up records for the Department of Corrections and in fact when we were down in Florida at a CODIS meeting a few years ago, I picked up on that and we now started to do that for paternity testing, or on-line with some of the agencies, where we can pull up on a computer, so we don't have to go through voicemail. You don't have to wait for somebody to call you back. You don't have to wait for something in writing, you can download it off of a computer. But if you have a problem with any of these, it's extremely time-consuming. We talked about the time for testing a sample can be minimal compared to the time for resolving any one of these issues.

Let me kind of summarize some of the roadblocks and problems that are associated with collecting and assessioning samples for CODIS. The individuals are hard to locate to get them to a collection site. It's very time consuming to bring someone in. The paperwork may not be complete and accurate, so you go through the whole thing again. You have just done it once. Now you have to do it again.

It may be difficult to determine if the sample has already been collected on this individual, because of poor record maintenance. Any verification that you do can be very time consuming. When people say I was talking at the CODIS meeting last week and they said that they just wanted to run our samples down to our fingerprint guys and I was standing next to someone. This kind of comes from being in a private lab. My eyes opened up and I said, you just run them down? That is ten minutes down. It's ten minutes back. It's ten minutes there. That is 30 minutes. We could have just put that whole sample through the CE in the time that you've just verified the fingerprint on that. So it's really very time-consuming. So what can we do about it?

I have some suggestions. My suggestions are to talk about collecting samples for CODIS. That the individual should be properly identified, collected, and tested one time. You can put duplicate samples in and there are always going to be duplicate samples for QC, but not at high numbers. Let's gets them in once. Let's get them right. And let's get them in the database. Let's make sure that our collections are done rapidly with the proper chain of custody and accurate and complete paperwork. Again, we are doing it right, we are doing it fast, and we are doing it once. This is a lot easier said than done, so we are going to need to do something about that.

My recommendation is that funding should be made available for the following. We need to have funding for good education, continuing education of the training of the Court and the correctional personnel. Then we need to develop some type of instrumentation for rapid identification and collection of individuals. And I don't think that this is very far off.

We need to have this instrumentation that will be readily available for courts and it will involve minimal human intervention to operate this collection.

And I'm saying that this is not far off because when we went into Louisiana and we look at some of these prisons and this was a big issue, where you're going to have people coming in multiple times. People saying that they're not who they are. A lot of collections.

We started looking at this machine, here. This is what is used down in Louisiana. The Live-Scan AFIS machine, where they go in, their fingerprints are taken. You know that they are good fingerprints, because the machine tells you if they're good or not.

The information goes in to the central agency, and you come back out with the suspect's rap sheet. There is supposed to be a very nice picture of someone here, and it's been taped out. But based on their fingerprints, you know who they are, you can see who they are, you know their name, you know any information that you have about them. And hopefully soon in Louisiana we will know based on what they were arrested for, if you should take a sample or not.

If don't need to take the sample, you won't take one. If you do need to take a sample, you will. You won't have people sitting with a list that says is this the right crime or is this the right crime for collecting a sample.

And I don't think -- these are very big machines that people use and they are very expensive. And in Louisiana they are in all of the major sheriffs departments where all of the bookings are taking place and they are in the major department of corrections.

But I don't think it's very far off to have these all on laptops where you can take a fingerprint and you can identify who your individual is. And we are not far off from machines that will take a finger prick for you. The Laser Lance-It by Transmedica. Stick your finger in and you can get a sample.

And what I would like to see, my envisionment of these collections is a machine where you can put your finger in, you can read the fingerprint, you can identify if you need to take the sample or not based on the fingerprint, based on the computer records of the crimes that you're going to take a sample for, based on if the sample has been collected or not.

And if you do need a sample, a sample would be collected. The information will come out on the card, and no one will touch that. It will all be done automatically. And I don't think that this is very far off.

Now I heard this morning that there are problems with interacting between States, but I think that the same principle that was applied to identifying the core STR loci why there were two companies that both had products that could be used, where guidelines were set so that these companies are working together, I think that the same thing could apply to AFIS, and identify what information you need and how this information could apply to someone else.

COMMISSIONER SCHECK: Are you suggesting that the blood go on the card?

DR. SOZER: On an FDA card.

Are there any questions?

COMMISSIONER REINSTEIN: When I saw that Laser Lance-It, how much does one of those cost?

DR. SOZER: The last time I looked, they are probably about $3,000. It's kind of ballpark.

And I think once you start, the more you buy, the more you use, the less expensive they could be.

But if you start thinking about how much you're going to spend, think about how many people you miss by not getting them in courts. If you have some piece of equipment, I know -- I think in -- actually in Washington D.C. one of my neighbors works for a company that designs these. And he said, oh, we are just getting these new computers in all of the police cars and they can identify information on a car or vehicle or on a driver right while they are there. This isn't far off.

But if you think about the time that you spend for a parole officer to spend a day not doing anything else, for the most part, but taking this person to get their sample collected, these could pay off for themselves when you talk about the numbers, talk about the times that samples come in and they're not right.

COMMISSIONER GAINER: I would at least point out on the fingerprint identification card it may be debatable how well the technology is there, but I think we're a long way off. We have been talking about getting there for a very, very long time. Now the in-car computers that are not much different than this, you can do a name check and license check and some of those things.

But the automation of driver's license information and fingerprints. And we have been talking about it for a long time. NCI C2000, it was all supposed to be there.

And I suggest, and I think the Sheriff would affirm, that if we see it before our mandatory retirement days, be happy.

DR. SOZER: I was down in Louisiana, there when they are doing the bookings, they are reading the fingerprints and these are coming back, the information is coming in a couple of minutes. And yes, they go down and yes, there are problems.

COMMISSIONER GAINER: On the large scan they come back in a couple of minutes?

DR. SOZER: Yes.

COMMISSIONER FERRARA: With respect to Ohio and Louisiana and your work there, how do you, Fairfax Identity Laboratory, assure the law enforcement agency, the citizens of those States, that the samples that you have in your laboratory are protected with respect to all the private genetic information that is derived.

DR. SOZER: You're talking about information associated with that sample, or are you talking about the sample itself for other testing?

COMMISSIONER FERRARA: Both.

DR. SOZER: Sure. And these are the same issues that occur in State labs and in any other private labs. We have policies and procedures that we follow, that we enforce strictly, and security measures that we have.

And we are in the business of doing this type of testing and not in the business of giving these samples to or the information to anyone else.

There are some very -- you say, oh, it's a paternity testing lab. Paternity is a very, very private matter. And a lot of the other issues that are associated very often much more private than some forensic cases. So we were very well trained and very acute following these procedures.

COMMISSIONER SCHECK: In that connection, do you retain samples or DNA extracts from the cases that you type?

DR. SOZER: Let me say something else. The other issue with private laboratories is storage. And when we are done testing a sample, I don't want to store that sample. I don't want to store the records. It costs money to store records. I don't want to store the sample. It costs money to store the sample.

Everything goes back. Everything will be back to the submitting agency or to the agency that is contracted with us. In fact, this is very different than paternity testing. Usually in paternity testing we'll keep some of the records. In our CODIS work they all go back.

It's a tremendous amount of work to try to copies those. It's a tremendous amount of work to maintain those records. So those all go back. So ultimately there won't be anything left in the laboratory.

COMMISSIONER SCHECK: What goes back? Do you send both blood sample --

DR. SOZER: Whatever the agency wants. Each State will work a little bit different. You may sent the original blood tube back, you may send any extracted DNA back. All records, all radiographs, all computer printouts will go back also.

COMMISSIONER FERRARA: With respect to the samples, however, let's take Ohio for example. If they are collecting samples themselves or let's say you're doing the work for a State laboratory or the State has collected the samples.

The samples, when they come to you, therefore, are they not free of any personal information at all?

DR. SOZER: No, they do have personal information.

COMMISSIONER FERRARA: So those States are providing you samples with personal information? If they did not, would that not be an additional safeguard against the improper dissemination of any --

DR. SOZER: I would say if they did not, that would be -- you know, you have States that are coming on-line with DNA testing. You have States that have a lot of work to do. And I've just talked about a big problem in getting samples in, getting them right, triaging samples when they come into the lab. That's a tremendous amount of work for a lab to bring on-line.

So the State that's just coming on-line, like the State of Louisiana that is just coming on-line with testing, that is a big new area that someone needs to challenge that a laboratory like ours is used to doing. We are used to receiving samples in, we are used to examining those samples, and I do not feel that there is any issue about confidentiality.

COMMISSIONER SMITH: The procedures that you've put in place, the rigorous need for that, I don't know if this is answerable, but I wonder what sort of breaches to those protections do you worry about?

There must be something because of the need to put the restrictions in place. But for myself anyway, being in a different business, I can't offhand think of who it is that would be coming in your place of business looking for this stuff.

DR. SOZER: Actually, nobody comes to our place of business looking for it.

COMMISSIONER SMITH: Would you lose anything by having no restrictions on it then?

DR. SOZER: I don't know if I'm following exactly what you're saying.

COMMISSIONER SMITH: Basically what the market is for the information that you're retaining.

COMMISSIONER GAINER: Insurance.

COMMISSIONER SMITH: Is there any evidence that you know of that the insurance companies come by the lab themselves?

DR. SOZER: We have tremendous amounts of samples that we receive in for paternity testing, and our lab isn't the biggest laboratory, and I have never heard of this from any laboratory.

We are very protective of the samples we get in. The samples, we maintain confidentiality, we maintain privacy. The samples are destroyed under a sample storage policy that maintains confidentiality and privacy.

COMMISSIONER SMITH: So you aren't aware of any attempts to breach these protections.

DR. SOZER: No, I'm not.

COMMISSIONER FERRARA: The reason I raised the question is that you can call, we have members of our own Commission who have expressed concern about outsourcing samples particularly, and what protections that there were available to precluded that.

DR. SOZER: I look at a contracting laboratory as just an extension of the forensic laboratory and they -- you know, we are governed by the contract, we're governed by what the forensic laboratory asks for, and are very well versed in handling many samples and maintaining confidentiality of those samples.

COMMISSIONER GAINER: To your knowledge have you received any subpoenas of any sorts on paternity information from a criminal investigation?

DR. SOZER: There are times when individuals will come for a paternity test and they are not the correct individual. So they will -- someone will send their friends in to give a paternity sample and this is discovered.

And there may be a criminal action brought against the individual who we have a sample for. Before we talk to or discuss the case with anyone, before we answer any question, I'm sorry, I can't admit or deny that we have any case in-house and I will be happy to take your name and number and that is the end of the conversation.

We will contact the agency who submitted that sample, and very often the agency will say, yes, I forgot to tell you we are prosecuting this individual for fraud and we're bringing criminal prosecution against this individual.

And sometimes we'll get paternity cases that are cases of incest and they say, oh, this may go on to criminal proceedings.

But there are always very strict policies and procedures that are in-house for discovering any information from the laboratory.

COMMISSIONER GAINER: Is your paternity information aggregate so that you could search an unknown case against your paternity cases.

DR. SOZER: An unknown convicted offender?

COMMISSIONER GAINER: An unknown suspect case.

DR. SOZER: The convicted offender database and the paternity database are two separate databases.

COMMISSIONER GAINER: I follow that. Just out of curiosity, is your paternity database, is it searchable?

DR. SOZER: Our paternity database, is it searchable for a certain profile? Yes. On our DNA View, paternities are done either by PCR or by RFLP, and on one of the features, I believe, and I don't really know if we've ever used it, that you can search your entire database to see if anyone in your database could be the potential father of this child. And it was for statistical analyzes when they say your database is so large can anyone else be. So one could make a database of all your samples. It would take a very, very long time, and potentially go in and you would have a lot of matches.

COMMISSIONER GAINER: But what you're saying is it's not set up now, but in theory one could come in and run an unknown sample against it.

DR. SOZER: In theory one could, but it's not set up and it would take a lot of time to do that.

COMMISSIONER CLARKE: Do you have the same kinds of markers in both databases? DR. SOZER: No. In RFLP we don't do all of the same genetic markers. PCR we do, but the analysis is done in two separate -- it's done very differently.

COMMISSIONER CLARKE: So I guess what I mean is the paternity cases are tested using a certain set of markers, right.

DR. SOZER: The paternity cases are tested differently, so they are kept separately in a laboratory. Some of the probes are the same. They are analyzed differently. At least for RFLP we use two different imaging systems. We use the digitized for paternity and we go into DNA View, we use RFLP scan and we go into CODIS Life for convicted offenders, and for PCR we may utilize the 310, we will go into DNA View for paternity, it will go into databanks for convicted offenders if it's on the FM Bio it will be done on the FM Bio and it will go into DNA View.

So it really is two different types of analyses. And once samples are in the laboratory for testing, they may be, you know, they are only identified by numbers. The numbers may be the crime laboratories so they are only identified as batches by numbers. And then paternity, they have a whole differently numbering systems.

COMMISSIONER BASHINSKI: But you don't have any criminal history or record information available to you?

DR. SOZER: No.

COMMISSIONER BASHINSKI: Just the number which is another major issue I think.

MADAM CHAIRMAN ABRAHAMSON: As I understood the question and response, could we use the paternity data for somebody who is a suspect or an offender. And I gather that they are very different and they are not compatible. But I'm not sure I understand, is your answer either an affirmative clear yes or clear no?

COMMISSIONER GAINER: I think you heard it exactly right. It was neither a clear yes or a clear no. I mean, if I came in with a warrant, with an unknown sample and said run this against your paternity, notwithstanding you wouldn't like to, notwithstanding you might run to the court to try to stop me, theoretically could you?

DR. SOZER: You could do that anywhere.

COMMISSIONER GAINER: As it's set up, absent trying to redo your system.

DR. SOZER: You're going in and you're asking me to re-run a particular sample.

You could come in with a profile on paper, you could say it's these alleles and these alleles and these alleles, and say okay, now put them into your database and run them against a paternity case and I could do that also. I could go into Paul's lab and do that or I could go into California and do that.

COMMISSIONER GAINER: We would do it in ours. The difference between the public and private is I can run my sample against my lab. The issue I'm just trying to get at from a civil libertarian point of view was what likelihood would law enforcement have to come in and search your files? Technically, if we could do it, aside from the legal issue?

DR. SOZER: I don't see it any different at a private lab.

COMMISSIONER FERRARA: You couldn't, Chief, in Virginia. I'd say no, it's against the law.

COMMISSIONER GAINER: That is why I said legal issue aside. I was wondering if the way, the technical way paternity information is stored is some individual way where it's not aggregate, where it's too cost prohibitive to search.

COMMISSIONER FERRARA: Amanda, aren't there some common loci?

DR. SOZER: Yes. But what I was saying is that actually in our lab, and it could be different at any other lab, that the actual set gel that is run is a little different, the amount of DNA, they are done separately. We have forensic analysts that will do CODIS work, we have other individuals that will do paternity work.

COMMISSIONER SCHECK: So if I understand this, when you do the analysis for CODIS for a jurisdiction, after you finish you will send back all the extracted DNA, all the biological sample, and all the paperwork. Don't you keep something? Don't you keep a profile?

DR. SOZER: Profile, we keep some minimal. For example in Kansas we have a checklist of what we sent out. We sent out this document, this document, this document, these are the individuals that reviewed it. But after a while it gets very expensive to keep files.

COMMISSIONER SCHECK: In your computer, for example, do you have a DNA profile that relates to some identifier, be it the name of the inmate or the inmate number or something.

DR. SOZER: We would have on our computer program the inmate, or the inmate number that came in on a particular -- in a particular batch, and that would be maintained. But the information goes back on disk.

COMMISSIONER KENNARD: You would have no way of verifying if you were going to court to prove that what you sent to them is what you have on file and it's the same?

DR. SOZER: You know, we thought about that. We said we are going to send these records back. And you start talking about a million samples, if you talk about 50,000 samples and each sample has seven pieces of paperwork on it and you need to copy that, and we felt that it was most cost effective if they have questions about something that they have done, they can send the file back to us.

COMMISSIONER KENNARD: Don't you have a file, as Barry has mentioned, a profile that could relate what you have done with what you sent back to them so that if need be you could relate to a court of law that you did what you said you did.

DR. SOZER: If we are the custodian of records, those records were sent back to the submitting agency, and we can testify from our records that are housed at the submitting agency.

COMMISSIONER BASHINSKI: I think one point is that -- I'm assuming that Ohio does this the way that we do it, is if you do get a hit, that's always something that is verified first, at our labs, by redoing the analysis of the samples we have on hand and then by obtaining a sample ultimately from the individual that you may have hit on to verify that is the exact same profile.

So the ultimate testimony is probably from that last match to the second sample that is drawn from the individual. And you use the database only as a pointer, indicate to go to that person.

COMMISSIONER SCHECK: But in terms of in contrast when you do a paternity test, I take it that you're saving samples, you're saving a whole document, you're saving much more.

DR. SOZER: Right. For paternity tests the samples go -- may go back to an attorney, the samples may be back to a child support office.

COMMISSIONER SCHECK: You don't save the extracted DNA?

DR. SOZER: We do for a period of time, then it's discarded. We have a sample storage policy, then it's discarded. We don't save things indefinitely.

COMMISSIONER CLARKE: There are also two different scenarios. You may go to court on the paternity case where testimony may be needed, as opposed to CODIS tool, which is really just investigative as I understand it. You would never go to Court on a CODIS hit.

COMMISSIONER GAINER: You will.

COMMISSIONER CLARKE: In terms of a prosecution?

COMMISSIONER FERRARA: Our experience has been on cold hit cases either the prosecution or the defense is required to subpoena both the person who ran the original convicted offender sample, as well as the individuals who did the casework samples.

COMMISSIONER CLARKE: Don't you require an independent reexamination, though, independent test?

COMMISSIONER FERRARA: Then we do an independent re-test. But even despite that, you have to bring in the person who ran the convicted offender's sample in the first place.

COMMISSIONER CLARKE: For what purpose?

COMMISSIONER FERRARA: I don't know, but they get subpoenaed.

COMMISSIONER CLARKE: By defense.

COMMISSIONER FERRARA: Both the prosecution and the defendants will attack the integrity of the original databank, despite the fact that we have this re-analysis.

COMMISSIONER CLARKE: So it's used for a duplicate test basically.

COMMISSIONER FERRARA: It is. What I'm pointing out is that indeed we are getting our databank people or our contract laboratory person has to come and testify in a case.

COMMISSIONER CLARKE: Because that is not really the intent of the CODIS system, is it?

COMMISSIONER FERRARA: No. It's just lawyers.

I'm sorry I said that.

COMMISSIONER GAINER: I'm just trying to get -- maybe this one little bit of clarification.

In labs that do paternity testing, do they generally index their paternity results or not?

DR. SOZER: Generally index, what do you mean by that?

COMMISSIONER GAINER: In a CODIS fashion so it could be searched with little or no effort, or they're doing it for research purposes or analytical purposes.

DR. SOZER: Let me clarify. When I say can we search our data, you asked me theoretically could we do that. And I can say yes, I would. And then if you would say little or no effort, I would say no, I can't.

COMMISSIONER SCHECK: I think you said before in a paternity test you have a profile of the paternity contributor, you would run that against your database to see if you have seen that individual before.

DR. SOZER: No. I said theoretically you could do this. You said theoretically, barring this, this, this, this.

COMMISSIONER SCHECK: Don't you do that as a matter of course when you get into the logistics of the Court when you get into the frequency of the profile.

DR. SOZER: Right. We have a database and then we do look up frequency, but we don't update that database on a routine basis. We don't have all of our individuals in that database. So I can't just go in and search the database.

In fact, it would probably be easier for me to go to a CODIS lab and search. I could do that. And if you said I have a choice between a paternity lab and searching or go to a CODIS lab, I would go to CODIS, because they are all there. The data is much more readily available there than it would ever be.

COMMISSIONER GAINER: But is the size of the databank now greater in a paternity bank than it is in CODIS?

DR. SOZER: The size. It's a tremendous amount of work to make a database. It's a tremendous amount of work to culminate all these frequencies that one may have. Frequently allele sizes are then purged off the computer because you don't have enough disk space for all these -- all this information.

You see how much equipment goes into CODIS, how many computers go into CODIS, you don't have that for paternity.

If one were to go back and retrieve all of that data somehow, I don't even know if it would be possible, yeah, right now you probably would have more DNA data for paternity cases than do you for CODIS cases.

But the CODIS cases are in a format that is readily available to do this. So if I were to go to try to search something just in theory I would go to a CODIS database before I went to a paternity lab to try to get information.

MADAM CHAIRMAN ABRAHAMSON: Let me see if I can settle this and move on.

And maybe, Chief, I'm misunderstanding your question from you. I'm going to put it in kind of non-scientific, non-computer terms.

I think what is being asked is can you use this paternity database, in quotes, to do a search.

And I hear you saying theoretically, yes. But as a practical matter, no. Because it would take a lot of work to put that paternity database into a format or form that you could readily search with another kind of sample.

Am I anywhere close to what the responses -- Paul is nodding his head and, Chief, I think that is what you have been asking, right?

COMMISSIONER GAINER: Yes.

MADAM CHAIRMAN ABRAHAMSON: So it's theoretically possible, but as a practical matter it would take a lot of massaging of the data, as some techies say, to get it into a system. Does that satisfy everybody?

COMMISSIONER FERRARA: Yes.

DR. SOZER: Yes. There are a lot of things that are theoretically possible, but not practical.

COMMISSIONER GAHN: I just have one comment. You touched on a number of concerns that Carl told us yesterday. One of the things I have been pondering since yesterday, these time served people that are missing and I don't really understand why. Probably the reason we're missing it is we've talked a lot about law enforcement, we talked about lab personnel, we're leaving a key person out of this, and that's the prosecutor. Because there is no law enforcement person with the prosecutor at the time of the time served, or lab personnel.

So I'm thinking, Chief, maybe in your group you could have just a section on the prosecutor's duties during that phase of insuring to get that sample once the conviction is in or guilty plea is in.

And what we can do is, I think, is prosecutors have been doing this for years for HIV testing. For years and years. We never let anybody get out on time served until there is a condition that you get down to the City health nurse and get your blood tested for HIV so we can make a comparison.

COMMISSIONER BASHINSKI: One can argue that is not crime scene evidence, but one can also argue it's a reference sample, so maybe we can throw that in.

COMMISSIONER GAHN: But also put a section on convicted offender databank and stuff. And I think this would be a part of that.

Weren't you going to talk about that?

COMMISSIONER BASHINSKI: We were going to talk about the new tools that make DNA of value as part of the introduction, and we can talk about that only being effective if you have enough samples from people, that kind of thing.

COMMISSIONER GAHN: Or somewhere, some group I think.

COMMISSION THOMA: And I agree with Norm as well. And I also agree with Jan. I don't know where it exactly fits in. But with this time served scenario with regard to the time between the actual conviction and the time of sentence, if the person is out of custody, for example, making that one of the first things the person has to do. So that when they come in for sentencing, they will want to do everything that is prescribed of them, obviously see a probation officer.

Obviously the person that is in custody you have a lot more control over, but I'm talking about the time served person who is literally on the street at conviction.

COMMISSIONER CLARKE: One example of how it breaks down there, is just a state example of California and a local example of how it operates with time served people or even individuals who are getting out shortly after that.

Our department of health in San Diego County literally had the facilities and personnel to draw blood to comply with our State statute for individuals who are released. And then the new statute was enacted which now requires a palm print instead of a thumbprint and the department of health is not qualified to do that.

So all of a sudden it stops and has stopped since January 1. Well, the effective date of the new statute.

But this is one example of how, as Amanda pointed out, all of these different procedures in place where so many mistakes can be made and the difficulty is because these people are being collected at so many different parts of the process. And it's just crying out for some uniformity in how this process is conducted nationwide.

We're dealing with 50 different jurisdictions, and then you get down to the number of jurisdictions within each individual State, and it's an absolute quicksand and quagmire, frankly.

COMMISSIONER SCHECK: But there does seem to be a certain consensus on when it is best and most efficient to do this.

In other words, you mentioned intake in prisons because that is how we did it in New York. And what we found was that a lot of time and effort went into the thought, how you collect these samples from the convicted prisoners at that point in time. Because you always take blood. Before any of these DNA came along you were always taking blood, for health purposes from prisoners. Now you're just taking an extra blood sample.

And you do it in that fashion, even though you're telling them what you're doing, there is less likelihood there is going to be resistance to it, and equally if you take it at the time of time served sentences.

If you have a certain uniformity initially, that's what I take it you're telling us, there are certain critical points in the procedure. It could be release dates, you said there was another good one where there is a certain chance of uniformity. You really can't get out of jail until you give your sample. And that's easily documented.

And you can focus on those points and suggest procedures along those lines.

DR. SOZER: I think what I was trying to emphasize is that we really need some type of automation, some type of procedure where there is minimal human intervention, some type of procedure where you have correct verification that you have the individual. That this is an individual who should be collected.

When you have technology out there that you can type samples in 30 seconds, but you're talking about collections that can take hours and hours to process.

This is really where a bottleneck is and why there is a lot of resources that are utilized among a lot of different departments where automation really -- I mean, you've got an end down, that's really good. Now you're moving on to the next area, which is the beginning.

And that is really what I wanted -- the take home message that I wanted to say. And that may involve identifying certain key areas where you're going to get most of your individuals.

But I think this is an area where I see really great progress can be made and I don't think it's that long away.

COMMISSIONER SCHECK: One last question. When I was in the UK and watched them doing their STR profiles for their databases and they were visualizing, reviewing the data, they had a not insignificant percentage where they knew they were getting wrong facts. In other words, they were getting certainty, but it was not -- they knew it was wrong and they knew they would have to re-test the sample again.

Are you getting that and in what percentage?

DR. SOZER: It depends on the sample type that you get in. It depends on the age of the sample. But even this is automated. You have programs that will say we have so much percent, we will take peak heights that are within this range. That is tremendously -- that is light years ahead of the collections. It's not even comparable. So to problems associated with collections.

COMMISSIONER SCHECK: It's funny, because Dave was telling us, and I literally saw it. Do you remember when you talked about stress, they literally meant people that were sitting by these computer machines and looking at artifacts and they had to change these peoples all the time because they were going nuts.

DR. SOZER: What I'm saying is in comparison to sample collections.

COMMISSIONER SCHECK: There is no question.

DR. SOZER: We have two people reviewing these independently. You have automation coming through that weeds out a lot, that will say here is kind of a general review of going in. We have two different people reading them. Those two reads are compared.

So there is quite a lot of work. I don't want to say that it's nothing to do this at all. I'm just saying there is a tremendous amount of work that has been spent on technology on the end, on the analysis.

MADAM CHAIRMAN ABRAHAMSON: Barry, if I could just interrupt, because I do want to move on.

Will you be here later today?

DR. SOZER: Sure.

MADAM CHAIRMAN ABRAHAMSON: So if we have time maybe we can bring you back and talk to you about this.

Previous	Contents	Next
Back to National Commission Main Page