By Bill Berger, Chief of Police, Palm Bay, Florida; Joe Chimera, General Manager and Laboratory Director, DNA Security, Inc., Burlington, North Carolina; and Major John Blackledge, Investigations Division, Palm Bay, Florida, Police Department
n December 2006, the Palm Bay, Florida, Police Department (PBPD) and DNA Security, Inc. (DNA:SI LABS), collaborated to develop a local agency databank of forensic DNA evidence for use on most crimes. The project was established to determine if mass collection of DNA from common crimes and subject reference samples, placed into a database, could be developed into an effective investigative evidence tool, identifying criminals and connecting seemingly unrelated crime scenes. The process had to be cost-effective, and the results from the laboratory had to be expedient. The results would also have to be accessible by the average road officer or detective in an easily understood format.
Although there have been several DNA projects and programs aimed at the local or regional agency level, these projects have been undertaken at large metropolitan agencies or regions with unique resources and atypical demographics.1 Most projects have focused on specific crimes, such as burglaries or violent crimes. Considering real-world conditions, where more than 80 percent of law enforcement agencies cover jurisdictions having populations of less than 250,000, these projects are not representative of the average police jurisdiction. Moreover, the bulk of crime that affects most communities is property related. It is common knowledge and an accepted criminology statistic that a small number of criminals commit more than 80 percent of crimes. Many of the offenders who commit violent crimes today have a history of petty crimes. The general premise of a localized forensic DNA database is predicated on these facts. In Palm Bay, juveniles commit more than 75 percent of the crime activity. The purpose of the Local DNA Index System (LODIS) is to bring forensic DNA technology down to the average city or county level.
The city of Palm Bay is approximately 100 square miles of suburban community consisting mostly of residential neighborhoods; about half are residential single-family neighborhoods, and the rest can be defined as rural residential and sparsely populated wooded areas. The PBPD has 166 sworn officers and 4 civilian crime scene investigation specialists. Recently, the city has experienced rapid growth, and with a population exceeding 107,000, police staffing has been a challenge. Despite the growing trend of violent crime in the Orlando region and across the United States, Palm Bay has thus far experienced few violent crimes. The bulk of the criminal activity there consists of burglary and theft cases, mostly committed by local offenders and tied to drug activity or juvenile delinquency.
DNA:SI LABS is a privately owned forensic DNA laboratory located in Burlington, North Carolina, that provides DNA processing services to prosecuting attorneys, defense lawyers, and agency and private investigators. The laboratory is accredited by the American Society of Crime Laboratory Directors/Laboratory Accreditation Board (ASCLD/LAB). DNA:SI LABS noted that there was a need to provide law enforcement agencies covering smaller jurisdictions with access to forensic DNA testing for investigating crimes that are more common.
In the Palm Bay LODIS project, DNA processing was made an option for virtually every crime scene, except for murders and rapes due to issues with access to the U.S. Federal Bureau of Investigation (FBI) national database, the Combined DNA Index System (CODIS). For six months, every officer, crime scene technician, and detective was encouraged to submit any number of swabs collected from any type of crime processed. Officers obtained “known” reference samples limited only by stated legal parameters.
The project has three phases. Phase 1 of this project, completed in November 2007, had as its primary goal the training of patrol officers in DNA collection. Officers were encouraged to collect samples at all crime scenes (see figure 1) to obtain six months’ worth of data, which could be used to determine the value of such evidence and refine the collection and analysis process for the local level. The PBPD is currently in phase 2, working with DNA:SI LABS to refine the automation portion that will allow officers to review DNA test results from a car computer over an encrypted, secure network (see figure 2). Another aspect of phase 2 is to automate the processing and tracking of submissions from the officer to the laboratory, reducing the time and cost even further. Phase 3 will assess the overall results in impact on crime and determine if the process is affordable for the average agency.
DNA evidence is not unlike fingerprint evidence in that it may be used both as an investigative tool and as evidence in court against a suspect. Palm Bay LODIS focuses on the value of DNA at the investigative stage as a method of early identification of suspects. One difference between DNA and fingerprints, however, is that when automated fingerprint identification systems (AFIS) were started, millions of fingerprints were available for input into its database; by contrast, for the foundation of a DNA database, reference samples—the “known” or identifiable suspect DNA profiles—must be acquired in sufficient numbers before the system can be useful.
Currently, many of the state-funded crime laboratories are not capable of processing the high volume of DNA evidence that can result from collecting evidence material from virtually all crimes. The state laboratory offers the benefit of submission and comparison of DNA profiles with State DNA Index Systems (SDISs) and the National DNA Index System (NDIS), both components of the CODIS hierarchical system of databases. This is beneficial for determining DNA matches of suspects previously convicted of crimes over a broad geographic region.
Furthermore, for the most part, CODIS does not include the profiles of the common property crimes offender and has almost no juvenile offenders. Therefore, CODIS databases, in their current state of implementation and funding, are not effective as an investigative tool for local agencies to identify the criminals committing most of the crimes in their communities. LODIS was designed specifically to provide local agencies with a system to create local DNA databases, which are flexible to meet the unique investigative needs of local law enforcement agencies. LODIS gives local, average-sized agencies the ability to deploy CODIS at their agencies and to be used in conjunction with other investigative techniques on more commonly committed crimes. As such, it provides an approach for implementing the local DNA index system (LDIS) component of CODIS on a broad scale and independent of any limitations in DNA testing capacity at the state laboratory level. As more LODIS-type systems are deployed, more law enforcement agencies will benefit from the CODIS hierarchical design, which provides state and local agencies with the flexibility to configure CODIS to meet their specific investigative and legislative needs.
Formulating the Project
In the Palm Bay project, PBPD and DNA:SI LABS jointly developed the procedures and put into place the methodology for collecting and handling both evidence and subject reference samples. DNA:SI LABS created the database and underlying programming necessary to determine if there was a match or “hit” worthy of further investigative effort. All PBPD uniformed officers, investigative staff members, crime scene personnel, and property and evidence staff members were trained in the proper techniques of collecting and packaging DNA evidence. The department also provided training regarding legal aspects of collecting reference samples from possible suspects and victim/witness elimination samples.
DNA:SI LABS committed to processing samples in 10 days or less after receiving them. The agreement was that the DNA profile data in the system are the property of the PBPD. This provided an option for future input into other databases, including state-level CODIS databases. Meetings were held with the local state attorney’s office, where the two partners discussed the use of the private laboratory for prosecutorial decisions and issues relating to court testimony.
DNA Databases and the Law
The case law on the collection and retention of individual DNA profiles from criminal suspects is almost nonexistent, because until now no one had created such an extensive database. Until recently, CODIS has been limited to a reference sample base primarily of convicted offenders and known profiles of those having committed very serious crimes. (The enabling federal law for CODIS also allows for local DNA databases.) Many of these offenders have been incarcerated; therefore, they are not free to commit crimes in the community.
In the Palm Bay LODIS project, DNA profiles are obtained from a wide variety of subjects: persons developed as suspects during stop-and-frisk scenarios, arrested subjects, and victims and witnesses for the purposes of eliminating them as suspects in the case of a mixed sample. Most samples were obtained from persons who had given consent. In some cases, where there was a solid criminal suspect or a person was under arrest, samples were taken through normal booking processes or from discarded items, such as drink bottles or cigarette butts. Although the option was available to officers, no samples were taken by court order during the first phase of the project.
One concern raised about the project was the issue of personal privacy. The DNA profile laboratory process used by DNA:SI LABS for this project (Identifiler) does not reveal any aspect of a person’s medical history or condition. The only specific identifier that is readily apparent is gender. The profile allows only for statistically conclusive identification and does not jeopardize any personal information. Florida Statute 760.40 prohibits the use of DNA without a person’s informed consent. This statute is designed to protect against the malicious and unlawful use of a person’s DNA profile and provides criminal penalties for violations. (However, there is an exemption specifically for the purposes of criminal prosecution.) While the purpose of the project was to focus exclusively on the prosecution of suspects, the statute prompted the PBPD to adopt a conservative approach in the collection of known reference samples. Officers were made aware of the statute to prevent any abuse of the process. Before known subject reference samples were submitted to the laboratory and processed for inclusion into the database, the project manager reviewed the circumstances surrounding the collection of each sample.
Phase 1 of the project was conducted considering the “real-world” environment. Officers were provided approximately five hours of monthly in-service training on the project. While officers were strongly encouraged to participate in the collection process, participation was not mandatory, nor did the agency keep track of how many specimens an individual officer submitted; the agency was interested primarily in officer reaction to the process. Some officers were enthusiastic and actively participated in the project, while others collected no DNA. Officers, including crime scene unit personnel, also received in-depth training at the Palm Bay facility from DNA:SI LABS forensic scientists.
Employees, officers, and civilians were asked to submit their own DNA reference samples for elimination purposes in the advent of inadvertent contamination; over 85 percent complied, the first being the chief of police. A random number was issued to each employee so that the profiles were anonymous.
Evaluation of Evidence Collection and Crime Scene Training
In phase 1, officers obtained 832 DNA reference swabs from suspects, victims, and witnesses. Virtually all of the reference samples produced a profile. Additionally, 635 DNA profiles were obtained from over 1,648 swabs and evidence items collected from crime scenes. Most of the evidence submissions were in the form of swabs of evidence items officers collected while processing crime scenes; however, some evidence found at crime scenes was sent directly to the laboratory without an intermediate swabbing of the item. One of the most unique cases involved a 10-pound chunk of asphalt (from a pothole) used to break a window and gain entry in a car burglary.
The ability to obtain a DNA profile is highly dependent on the collection or swabbing technique as well as the target evidence item and its surface. This is especially relevant when attempting to collect DNA profiles from evidentiary items that have been handled or touched by a suspect but do not have any visual indication of a body fluid, such as blood or semen. The DNA sample collected in this situation is sometimes referred to as “touch” or “contact” DNA and includes evidence generated by actions likely to cause skin cells to abrade and be retained on the evidence (see figure 3). The LODIS evidence training program focused intentionally on the collection of touch DNA, so that many officers can work with victims to identify virtually everything that a suspect may have touched or with which the subject may have come into contact.
During phase 1 of the project, many DNA profiles were obtained from contact DNA collected from swabs of such items as automobile steering wheels, bags containing illicit drugs, entry points to residential and commercial buildings, gun grips, light switches, doorknobs, and handheld electronic devices such as cell phones. These items were submitted for analysis; such items often yielded a good DNA profile. In addition, items with dried saliva, such as beverage containers, drug paraphernalia, and tobacco or marijuana cigarette butts are even more likely to yield a good DNA profile.
On the other hand, other items that did not contain any visible signs of body fluids or were not likely to cause skin abrasion did not yield adequate amounts of DNA to produce a profile. These items included such things as wall light switches, which occasionally yielded DNA, and driver’s licenses and credit cards, which rarely produce DNA since they are unlikely to abrade skin cells when handled.
Overall, the results indicate that the training and techniques for patrol officers and others processing scenes through swabbing and handling of evidence was adequate for capturing DNA from typical items found at the scene. In an effort to evaluate the effectiveness of the evidence training program and to improve the cost-effectiveness of the project, an evaluation was conducted to determine if evidence items can be accurately culled based on the likelihood that they will provide a DNA profile.
In a blind evaluation, expert DNA analysts at DNA:SI LABS ranked 1,648 evidence collections by Palm Bay officers with a high (H), medium (M), or low (L) likelihood of producing a positive DNA profile using only evidence item descriptions. The expert opinion ranking for each item was compared with the actual result for each item yielding a DNA profile. The blinded comparison revealed that 80 percent of items ranked H by the expert were positive for DNA; 50 percent of the items classified as M were positive for DNA; and only 15 percent of evidence items classified as L were positive for DNA. Thus, Palm Bay officers were successful in triaging and collecting DNA profiles from the most appropriate evidence items, and the training program was effective.
Based on these results, culling of evidence items prior to submission for DNA analysis can be an effective means of maximizing the number of usable DNA profiles. Culling also will improve the cost-effectiveness of the project by reducing unnecessary evidence item processing without significantly affecting the benefits of LODIS. The project also revealed that it is possible to obtain DNA profiles from contact DNA, collected from items likely to cause abrasion, and possible but less likely from items that do not cause abrasion. Many items ranked L (not likely to yield sufficient DNA for a profile) did contain sufficient DNA to obtain a profile, and some of these matched suspect profiles. Therefore, the culling of evidence for DNA processing based upon likelihood ranking should be considered by the officer performing the collection when submitting evidence items to the laboratory, and items ranked H should be processed first. The primary point is that officers should be encouraged to swab aggressively for contact DNA (in addition to such likely DNA sources as blood stains, items containing dried saliva, and cigarette butts) and allow the culling process to focus on the samples most likely to produce a profile, while allowing other samples to be available should those swabs assessed H or M not produce a profile.
The DNA profiles of 41 suspects were matched to the DNA profiles taken from items collected at the crime scenes. A high percentage of these were primary (one-to-one) matches between a suspect and the evidence. A few of these were extended matches (matches between suspects and multiple evidence items) from one or more crime scenes. These extended matches are particularly interesting because they often allow reconstruction of separate criminal episodes.
Some evidence items yielded matching DNA profiles that tied together otherwise unrelated crimes. When the same DNA profile is present at multiple scenes, it indicates that the same individual was present at each scene. In some cases, the evidence matches did not yet have a suspect or reference match. When a suspect profile is eventually linked by LODIS to the evidence profile, it will connect that suspect to all of the crimes. This scenario is illustrated in the first sample case provided.
In these cases, profiles from multiple crime scenes were positively matched to each other before the suspect was positively identified. The value of matches from otherwise unrelated criminal episodes where a suspect has not yet been developed could allow officers or detectives to draw the cases together and view the “totality” of the collective information to see if this perspective of the cases reveals leads or even identifies the criminal, even when there is no suspect profile match.
Nineteen reference specimens were submitted more than once. On some of these duplicate submissions, the names of the individuals from whom the specimens were obtained were slightly altered. These differences involved slight variations in spelling of the names, inclusion or exclusion of a middle name, and/or the use of a nickname. These may have been the result of suspect deception or officer error in recording on the label. One person provided two reference specimens using different names. Florida Statute 901.36 provides for a specific criminal charge when a suspect of a reasonable suspicion stop or an arrested person gives a false name to an officer.
Phase 1 of the Palm Bay LODIS project has demonstrated that widespread forensic DNA analysis through the support of local DNA databases can be a highly effective investigative tool, not only to identify criminals but also to connect otherwise unknown related cases.
In phase 1, test results were provided primarily to the agency in a paper format, which then had to be reviewed, deciphered, and discussed with submitting officers. This caused a notable delay in the investigation’s progress. Phase 2 changes were incorporated to include the implementation of a Web-based automated interface that began in January 2008. Phase 2 will also include a front-end interface with the PBPD’s property and evidence software that will allow electronic uploading to the laboratory. This will reduce cost, typographical errors, and the time involved to obtain profile results. The system will be preloaded with the e-mail addresses of the officers, detectives, and technicians and will notify the appropriate personnel of a LODIS match. The notified personnel can then access the secure site and review the LODIS results.
One key factor that must be considered in starting such a database is the acceptance that finding matches and solving cases will take time as the reference foundation is established. Another option that an agency should consider is to start the collection of reference samples from known criminal suspects in advance and provide those reference specimens at the initiation of the program. Collection of reference samples should follow rigidly set agency procedures that consider the current law. Many violent crimes including rape, homicide, and battery contain multiple evidence items that are excellent candidates for DNA analysis. Palm Bay LODIS initially excluded both evidence and reference DNA items in homicides and sexual battery cases from the project, as private laboratories cannot directly upload DNA results that they produce to state-level CODIS databases. Because of this policy, phase 2 of Palm Bay’s LODIS implementation will include the collection of duplicate samples to allow for both processes until the matter of open access to CODIS by private DNA laboratories is resolved.
Phase 1 allowed the PBPD to examine procedures and refine processes in an attempt to improve the success rate and reduce costs. As a result, phase 2 will proceed with the collection of DNA evidence at virtually all crime scenes, regardless of the type of crime. The collection of DNA must be allowed all the way down the investigative process to the first-responder level, where appropriate. Phase 1 clearly determined that the training of officers on both collection techniques and selection of evidence or surfaces to swab is critical to the success of obtaining profiles and improving cost-effectiveness. Additional video training is currently in production to further help officers rank evidence to help the laboratory determine which items are likely to produce the needed profile. By combining DNA analysis for all crime scene processing with a locally focused database, the connection between items of evidence and otherwise unrelated criminal incidents can be established. In addition, there have been numerous links among crimes, suspects, and evidence that can be established based on DNA profiles alone.
This project has shown that a large quantity of crime scene evidence submitted for DNA analysis can be processed within days and compared with the DNA profiles of known suspects. Subsequent to the primary investigation, the evidence can be forwarded to the state crime laboratory to determine a possible link to crimes at the state and national level. This approach will allow the PBPD to maximize the benefits of DNA analysis as an investigative tool for combating crime. The following phases of the project will evaluate the impact of “forensic informatics” and computer reporting systems on managing the process of collecting and evaluating DNA evidence from large numbers of commonly committed crimes, demonstrating just how much benefit the law enforcement community stands to gain from this new approach. ■
Bill Berger is chief of police of the PBPD, a past president of the IACP, and the current chair of the IACP Forensics Committee.
Dr. Joe Chimera is the general manager and laboratory director for DNA Security, Inc., an ASCLD/ AB-accredited private laboratory, and has over 20 years of experience in the medicolegal laboratory services industry.
John Blackledge is the major of the Investigations Division at the PBPD and serves as the agency project manager for the LODIS project.
1See, for example, National Institute of Justice, “DNA in ‘Minor’ Crimes Yields Major Benefits in Public Safety,” In Short: Toward Criminal Justice Solutions series, November 2004, NCJ 207203, http://www.ncjrs.gov/pdffiles1/nij/207203.pdf (accessed March 6, 2008).
LODIS Case Examples
Case Example 1: In early 2007, the PBPD received a report of a burglary to the screened pool enclosure of a private residence in an upscale, gated community. The only item stolen was a child’s pool float. Entry was made by pushing through the screen and reaching inside to release the locking lever and opening the screen door. A police canine tracked and found the float discarded in a wooded area a distance from the burglarized home. Both the entry point and the float were processed using simple but aggressive swabbing techniques. The residents provided reference profiles for elimination purposes. Other homes in the area had been similarly burglarized, and minor items had been stolen. These initially appeared to be nuisance crimes—more vandalism than true burglary. However, the presence of slash marks on the stolen floats was cause for concern. Intuitive investigative thought led to the belief that the offender was likely a juvenile who was upset with the community or someone in it, acting out in anger. Within two weeks, LODIS reported that there were sufficient profile characteristics to make them valuable for identification. The profiles did not match the residents, but there was a match between the profiles from the point of entry and the float.
On February 6, 2007, there was a second burglary of the same residence, resulting in a second theft of an alligator-shaped float from the same screened enclosure. In all, there were four burglaries with similar MOs in a two-block area. Then, on February 12, 2007, an officer responded to a shoplifting case at a local discount store. The subject was detained by store security and requested that a trespass warning be issued but refused to prosecute the theft. The 22-year-old male detainee consented and provided a DNA reference swab. This swab was processed by DNA:SI LABS within days, resulting in a profile and a perfect match to the evidence profiles from the burglary of the pool enclosure.
A warrant was obtained, and the subject was arrested. During the resulting interrogation, the suspect confessed to additional crimes. Time lapse, mental state, and drug abuse negatively affected his ability to recall all of the events in which he was involved. In less than eight months, the subject was convicted and sentenced for these crimes.
Case Example 2: DNA evidence was collected from two vehicle burglaries committed at two distant locations in Palm Bay about 30 days apart. Both cases were closed without any solvability factors. A short time later, LODIS reported a match between the two incidents. A few weeks later, a reference sample was obtained voluntarily from a victim of a violent crime at a third location. When this sample was processed and entered into the database, LODIS showed that the two DNA evidence profiles from the vehicle burglaries matched the victim’s profile. This identified the victim of the violent crime as having been inside the burglarized vehicles. A detective and the investigating officer reviewed the results and approached the victim (who was now seen as a suspect) for an interview. The subject was subsequently arrested and confessed to those and several other more serious crimes.