Monday, October 21, 2002

INTEGRATING SYSTEMS

MR. SCHMITT: I want to move us now to the second numbered paragraph under this topic heading, that being integration systems. As we discussed this on the staff level, it seemed like there was a great deal of support for promoting integrated IT at different levels in the criminal justice system, and I want to test that hypothesis again here today.

Take a look, if you would, at the last sentence of the numbered paragraph, the notion linking DNA databases with other criminal justice databases such as IAFIS and NCIC. I know that the Bureau is in fact linking CODIS with both of those databases or they're in that process, and I'm sure Joe will comment on that as we go forward, and states perhaps might want to link their local DNA database with their own state center on NCIC.

How crucial is this to do at this juncture in the process? I guess also think broader than just connecting databases. Should this be connected with other agencies and allow you to share electronically information that you might otherwise share in paper or over the phone?

MR. COFFMAN: We're in the process right now of our most labor intensive part of our process. We're getting about 5 to 6 thousand samples a month for our convicted offender program. It's not the DNA analysis. It's the up-front crime lab tech. Data entry portion because you have several thousand agencies involved in collection. They all fill out the form to a different level of completeness. We want the form completed totally because we want to have this integrated with our criminal history.

We're in the process of hiring a company to - it's called our integrated criminal history effort to where at the collection site - we want to move the collection site, first of all, away from all of these other agencies and have it in the courtroom because everybody has to be pronounced guilty in the courtroom at some point, and then we have a live scan unit in each courtroom.

What will happen is the individual will put his thumbprint down, it will hit our integrated criminal history, and it will let the collector know if someone is needing a sample collected. If they are, it prints out I think it's a 2-D bar code that represents their fingerprint. That gets put on the form. It comes to our lab. We'll bar code it, and it dumps his information from the criminal history right into offender data so we won't have any more data entry and looking for missing information.

We right now look for duplicate samples so we don't waste resources, and we saved about $4-1/2 million in 12 years by not working samples that have been collected before. I mean we have people that are in and out of there two or three times a year, and if it's a different agency collecting it, you're going to get it again, and that just seems like a waste of resources to keep reanalyzing these guys.

Basically long story short, I think it's important that we do integrate some of our efforts. Right now I have six people to track down the juvenile justice criminal history, the Department of Corrections. This will get it down to where we think one to two people can handle our whole case load just by using a bar code scanner and processing it that way.

MR. FERRARA: I, too, think this is a critical avenue we do have to go down more so recently in Virginia. In January we started taking samples upon arrest for violent felonies and burglaries. Almost all of the crimes that we currently take samples from are people who are convicted. Now, implicit in that arrestee statute is that a person who is arrested for a felony and then subsequently the case is null pros'ed or dismissed, that sample has to be expunged in the records. That means obviously I have to have access to criminal record histories, which we've never needed before. Our laboratory is not part of a law enforcement agency. It just serves them all.

So I have found a critical need right from the magistrate when he or she issues an arrest warrant to have language indicating that this is an offense for which the law requires a sample to be taken, for the person taking that sample to check a system to determine if in fact a DNA sample has not already been drawn from that individual because keep in mind if we're taking samples on arrest, the idea is we're not going to take them on conviction.

Then once the sample has been taken we need to enter a field in a criminal history record somewhere whereby there is an indication that a sample has been taken from that individual. As Dave mentioned, our State Police, before they allow that data to go in state criminal record history is going to want to make sure there are fingerprints of that individual to verify that that DNA sample is from that individual.

Then when the case is adjudicated - and this is one of the most difficult problems that most of our laboratories have had - we would love to be able to give you lots of data on what is the disposition of the last 950 cold hits we have had, how many convictions, how many dismissals, et cetera, et cetera. We don't have access to that kind of information, and particularly in trying in Virginia to distinguish between juveniles who are adjudicated delinquent as opposed to juveniles who are convicted offenders because I think only those that are convicted as adults can I put into NDIS, even though my state law says that those adjudicated delinquent.

So all of these are examples of the need for communication between now the laboratory and all of these criminal information systems. This is making it all the more clear that this is the direction where laboratories are going to have to go.

MS. JONES: I wanted to know how your courts and your law enforcement agencies are receiving this information that it's going to be their responsibility now to integrate this process into an already overburdened system, either of you, both.

MR. COFFMAN: We've had a memorandum of understanding with the sheriffs and just said look, we've had enough hits. We're getting enough each month that they finally can't say what is this going to do for us. They know what it's going to do for them. So we have a memorandum of understanding with the sheriffs to agree that they take on the responsibility of collection in the courthouse.

One of the most difficult parts has been in the past we've collected blood. Well, in 2001 we switched to oral swabs, which made it easier for them. We increased our collections 1,000 samples a month for people we weren't getting it before by going to oral swabs onto FTA paper. So it's fully automated. Well, not fully. I have a guy that makes a robotic noise when he puts it on the paper.

We talked to the sheriffs association this year, and they seem very positive on it. I don't know whether they will actually do it when it comes around, but the plan really is to have this live scan unit. It's supposed to get back results in less than a minute that says a sample needs to be taken. They're already fingerprinting them. Their case prints are already being taken, so all we're asking for is another couple of thumbprints. They don't have to fill out the form anymore other than the name because all of his information will be housed in a bar code that prints out, which represents his criminal history record.

Then once we get it and we say we haven't seen this sample before, then when it dumps information into our system, it will update the criminal history to say the sample is already collected. We were trying to stop duplicate collections. We reached 50% duplicates a year, you know, if we weren't careful.

So I think it is tough, but I think they're willing to do it, especially if we can limit where it's being done and we can make it as easy for them as possible.

MS. JONES: Do you think it largely has to do with getting law enforcement to embrace this technology as their own and not just a tool for the prosecution and defense anymore? For the courts basically.

MR. COFFMAN: Right. I think there are not many states that they can't embrace it as their own. There has been so many of them that have had amazing cases solved, and this is the only way they did it. As I said, in our state we're doing volunteers now. We get 5,000 volunteers a year, and that just amazes me, but they do just out of the goodness of their hearts.

MR. SCHMITT: I've got ask what motivates them to volunteer?

MR. COFFMAN: Well, I think it is - we were asked to do this by local law enforcement because they didn't want to wait until they were convicted or someone was convicted, so they brought them in.

MR. SCHMITT: I know I'm going to commit a crime, so get me now?

MR. COFFMAN: The only thing I can figure is they think the more cooperative they are with law enforcement, that it's going to go easier on them. They're just not thinking of the big picture, but they sign a consent form, and we have had 18 hits. Believe me, this is the one time I was not wanting hits because I didn't want this to expand statewide because it's hard to plan our budget. One county will send 2,000 a year; the other will send three, and it's hard to plan, but it's unfortunately working.

MR. CLARKE: I just wanted to address the linking of some of these systems like IAFIS and CODIS and NCIC. We have been working for a couple of years now, especially in regards to CODIS and NCIC, and our approach was not to have everyone who has CODIS access have NCIC access or vice versa, but it's important for the investigator to know if he's searching a criminal history record in NCIC that a CODIS record is available because he may have evidence that he may not submit unless he knows a CODIS record is available.

So the idea is to put some sort of indicator into NCIC if a CODIS record is available, and an indicator also would include where that profile resides and who to contact about that profile.

So that's the approach we're taking, and we have been in the process of doing that now for a couple of years. It's long bureaucratic process to make changes in some of these databases, but we are integrating some of them.

MR. SCHMITT: Is that linkage made only with respect to the samples that are uploaded into the national database?

MR. CLARKE: We haven't gotten that far yet, but probably that's the way it will be.

MR. SCHMITT: Will there be at some point a modification to NCIC to allow the local inputters to note in the national database that there is a local DNA file in the local database?

MR. CLARKE: I'm sure with the CODIS redesign it will make it much easier for some of the state and locals to know or agencies with CODIS profiles available or that may be available in another agency or another laboratory. So I think the CODIS redesign will facilitate that.

MS. NARVESON: Arizona just went through a very interesting exercise, and I hope it remains only an exercise, but the Rules of Criminal Justice Procedure are being revised, and one interpretation is that the laboratories will be required to complete all requests for analysis within 30 days of arraignment.

Even though there are times when we don't think we're very technologically advanced, we have a LIMS system that can't get the data from the other computer systems that exist, and I would say that police agencies very often have very antiquated systems. Ours is one that's a 10, 12-year-old proprietary system. We have to talk to that, which has to also be able to at some point provide data to county attorneys, city attorneys, U.S. attorneys, et cetera.

These systems need revamping. We can't even get an arraignment date that we can rely on. There is no automated mechanism. In my mind it's such an easy thing technologically to do, and I end up talking to the hand. They don't want to hear about it because they're not ready and it's not their idea, but this truly needs to be an integrated effort across law enforcement and the courts and the laboratories because we are in a leadership role right now where our technology is more advanced than theirs as far as communication is concerned. We can't meet the demands of the courts and the investigations without having that easy and free, but secure flow of information back and forth.

MR. SCHMITT: What is the preference among people here in terms of how this sort of technology would be developed? That we would fund it specifically to allow states to develop their own or we would fund two or three sites and from that come up with one or two applications that would then be made available to other states to allow their systems to be integrated?

In case you don't know, NIJ has been working in this area of criminal justice systems integration, not necessarily the DNA area, but other information systems in integrating them for some time, and we have had great successes in coming up with hardware and software that allows disparate systems that have no relation with one another to actually be able to share information and not just record identities, but actual pictures of their records back and forth among agencies and among data systems.

MR. COFFMAN: I would like to see maybe you form maybe very small working groups to maybe come up with different solutions. We've learned a lot in the LIMS system that we created. A year ago I thought it was the worst decision I ever made. Now I think it's the best thing I ever did. It's a real learning process, and if everybody goes out on their own and does it, you're going to get 50 different versions of some garbage.

I really think it would be better to get more minds in the mix, not too big of groups, but maybe fund two or three smaller working groups to work in their states to develop something and then use that as a model and maybe merge it later on.

Also I was going to mention about the automation. Another thing that comes to mind, we've tried to automate since 1996, and another personnel need pops up when you start doing it, and that's an IT professional to be in your lab. Most IT people in our agencies out there, first of all, will never talk to another IT person. If it's not their idea, they're not going to accept it, and also when you have someone on site - and I have been very fortunate to have that since 1996 - they learn what you're doing. Even though they're not scientists, I mean DNA scientists, they learn what you're doing and so when you ask for something, they give you what you need, not what you asked for, and that is very valuable.

So I think the more complicated our computer systems have gotten, our LIMS and robotics - I have been very fortunate to have a guy who is really good with that kind of stuff, but I know a lot of labs don't have that.

MR. SCHMITT: Any comments on this approach of two or three working groups, Dave's idea to come up with a system that would then be kind of a state-of-the-art system that people could use?

MS. CROUSE: We actually talked about this a couple of years ago, and the idea was to have a small group and use it as a template, but use it as a sophisticated enough template where we could find what some of the issues were really going to be, but it ranked everything from booking all the way through the court system. It was the entire system that was going to be interlinked, not just the DNA system to the ballistics system or to the NCIC. It was everything, and for some reason the idea never got off the ground, which means the template never formed.

I totally agree with Dave that a lot of Rue Goldbergs can happen. If everybody tries to do this themselves, it's really going to be very complicated, and if someone can just show that it works - we have an IT person that likes us because we're behind four doors and he can get locked in our lab and nobody else can touch him, and that's where he comes to rest, but while he is there he does some stuff. We have had to hire somebody on the outside to take care of all the MacIntosh issues because we're Mac driven for our technology, and we don't have any intention of switching over. So it gets more and more complicated, but I think a template would be fabulous to be able to hand somebody.

MR. KRESBACH: I think in a sense we've already got hundreds of agencies around the country that are doing their own thing already, whether it's very low tech. To the very high tech. Unfortunately, my system can't talk to my state laboratory. The local agencies around our jurisdiction can't talk to us. I can't even send an e-mail to our state laboratory without it getting bounced back at me, much less these bigger picture type of things.

I would think whatever is decided, whether it's small groups or a single group that would get together and start to formulate ideas that would work, that it needs to be done such that it could be universally applicable on similar - not the same, but similar platforms, similar, if not identical, software systems in a sense very similar to the way CODIS is now.

I can buy whatever hardware I want as long as it meets minimum specifications. I can buy certain types of software applications that I might want to use ancillary to CODIS, but at the end the core CODIS software is the same regardless of whether I'm a city lab, a state lab, a county lab. You name it; it's all the same. You're talking apples to apples unlike things like IAFIS.

If I've got one system, my system won't talk to the state system, won't talk to the county system, won't talk to whoever even though we're doing the same type of thing. So whatever it is that would come out potentially from these small groups, it needs to be something that's universally accessible and universally communicable. Otherwise, we'll still have these agencies veering off in 100 different directions, my own included.

Again, I can't send an e-mail to our state lab, but I could do everything internal to my own building; I just can't get it out and they can't get their information in. We've spent huge sums of money already for a system that's not yet in place.

MR. SCHMITT: Other comments?

MR. SELAVKA: I'm reminded that there is an assurance that we all have to take care of when we get grants from NIJ that involves the information technology administrator for the state. When I was in New York and then went to Massachusetts - that requirement kind of started when I got to Massachusetts - I realized that although there is a person with that name, they don't do what that name implies that they do, and I'm not sure if they do that in any state. I know that it doesn't happen in Massachusetts.

There is no integration. The administrator has no - although I think he might have the authority, and I really have to ask him that, in his own mind he may have titular authority, but I'm not sure he has it functionally, but he can't get it accomplished. The infrastructure doesn't allow for it, and unless I think you leverage funds to do other things, with the requirement that state's IT administrators make that integration happen as part of the requirements of these grants, that may be the only way it ends up happening. It's too big.

I think Cecilia stated it really well. From the beginning to the end it's so big and so complicated you would have to - this is almost one of those highway funds things. If you want highway funds, your blood alcohol will be .08. You may have to leverage something really big to get this to happen the right way. I don't think you can do it with pilot programs because they will just sit on a shelf. Everybody will say that's nice in Albuquerque, but it doesn't work in Boston.

MR. SCHMITT: Other comments? Moving on, if you will note Paragraph 3 under IT integration and information sharing, at the last meeting there was discussion about wanting to know what was going on in other places. That's part of the reason why we had the folks from South Africa here today. We also gave some thought to what is going on in the U.K. They have been in the news quite a bit, particularly about the tremendous successes that they have enjoyed. Of course, they don't have some of the similar procedural constraints as we do in the United States, which leads to why this is so helpful in that country for using this type of analysis. It is a little different than it is here or at least perceived to be here in the United States.

We thought it would be helpful to have a greater understanding of how DNA analysis is used in the U.K., and so Director Hart has engaged a research project funded by NIJ to look at this, and we expect a final report this coming summer; however, we're not going to make you wait that long for a preview, and tomorrow on the agenda at our working lunch Tim Schellberg and Lisa Hurst, who are here from the Smith Alling Lane firm, which is the entity that won the grant from us to do this work, will be here to give you an update on where their research stands. So that's mostly a commercial advertisement for what is coming tomorrow.