Sun, 14 Jun 2009

Information retained in the National DNA Database profile records

The Home Office consultation Keeping the right people on the DNA database (see Sentenced to genetic probation) explains in Annex B: 'All NDNAD records have the same structure consisting of 36 data fields. However, some of these fields are only relevant to subject profiles, and some are only relevant to crime scene profiles. Therefore, no record will have all data fields completed.' The list included is mostly correct but still manages to miss a couple of fields, and doesn't describe at all the second screen of information. Below is the complete list with explanations and a graphic guide that was sent by a reader of this blog.

The National DNA Database (NDNAD) holds both DNA profiles derived from samples taken from known individuals (referred to as subject samples) and DNA profiles derived from samples left at unsolved crime scenes from unknown individuals (referred to as crime scene samples). DNA subject profiles consist of a set of markers, currently ten, plus a gender marker.

When a suspect is arrested for a recordable offence and detained at a police station, two DNA samples are taken unless the arrestee already has a PNC record and it indicates that their DNA profile is already on the DNA Database. (Even when a DNA profile is already loaded, the police may exceptionally take a new sample.) The DNA samples and a form are put in a sealed bag and sent to a private forensic lab. The lab processes one of the samples to obtain a DNA profile and keeps the other one in a freezer. The DNA profile is then loaded onto the National DNA Database.

NDNAD fields

1. Sample status

1a. DNA sample barcode. Reference of the barcode sticker that was put on the sealed tubes holding each sample. The DNA sample barcode number for DNA samples of those arrested begin with a 9 (Police elimination sample barcode numbers begin with a 7). Following the introduction of a new PACE DNA sampling kit on 2005-04-01, all profiles from arrestees are of evidential standard; this can be recognised by the first two digits of the barcode number (96 and above).

1b. Arrest summons number (ASN). The ASN is one of the unique reference numbers used to identify the sample.

1c. Supplier. The forensic provider whose lab profiled the sample.

1d. Class code. Identifies how the sample is treated on the database and whether the sample was from an arrestee, a volunteer etc.

There are two classifications of criminal justice sample records on the NDNAD, namely 'SA' and 'SP'. In April 2004 the style of kits used to take DNA samples was changed to tie in with the implementation of changes to the Police and Criminal Evidence Act 1984 made under the Criminal Justice Act 2003. 'SA' is the code for samples taken prior to April 2004 and 'SP' is the code for those taken post April 2004 using the revised DNA kits.

1e. Police national computer (PNC) status. Reconciled or non-reconciled; denotes whether there is a link to PNC.

The consultation explains:

6.27 There are approximately 850,000 legacy profiles [of people arrested but not convicted or acquitted] of which approximately 500,000 have no linked PNC Record. This means it is not possible to tell whether the latter profiles relate to persons arrested and not convicted or subject to no further action, or to people who have been convicted.

Curiously, as I reported in The DNA database and you, the National Police Improvement Agency detailed in a Freedom of Information response last year that:

Data obtained from the Police National Computer (PNC) on 31 March [2008] indicates [...] 573,639 persons (14% of persons on the NDNAD sampled by forces in England and Wales) had no current conviction, caution, formal warning or reprimand recorded on PNC. The PNC records for the other 283,727 persons (7% of persons on the NDNAD) had been removed from the PNC.

That's a total of 857,366 legacy profiles as of end of March 2008, basically the same as the more recent recent approximate 850,000 overall number of the consultation. As we know that only a few hundreds profiles were deleted between these two data (e.g., 274 for the calendar year 2008; xls), the huge decrease in number of profiles of definite innocents (from 573,639 to 350,000) and corresponding increase in the number of unreconciled profiles (from 283,727 to 500,000) doesn't make sense. As we have shown in Sentenced to genetic probation that there are several mistakes in the consultation, the most like explanation that the data in the consultation is mistaken. From the data, on can hazard the guess that possibly the numbers of definite innocent profiles and unreconciled profiles were swapped.

The PNC record also includes data about the DNA profile. Here's an excerpt of the relevant section from a copy of my PNC record from 2006, with some explanations (in some of the entries I have replaced digits with the character 'n'):

DNA report summary

A/S ref: 05/0000/00/nnnnnnA. The arrest summons reference number, the first two digits are for the year.

DNA status: Confirmed. Marker indicating the status of the DNA sample and profile and whether the arrestee has been convicted.
(Note that the DNA status in this example is incorrect, 'Confirmed' indicates that the profile is on the database and a conviction has been achieved. I have never been convicted. At the time this PNC record copy was obtained, this field should have been set to 'Profiled' indicating only that my DNA profile was on the database.)

Sample barcode: 96nnnnnn. Barcode number of the sample.

Date of sample: 28/07/05. Date the sample was taken.

Sent to lab: Forensic Science Service. Forensic provider's laboratory to which the sample was sent to, i.e., FSS.

Sample type: Swab. The sample type, i.e., mouth swab.

DNA FS/Ref: 01MS/CUSnnnn/05. Custody record number for this arrest, i.e., 01 stands for the Met, MS for Walworth police station, CUSnnnn is the custody number, and 05 is the year.

1f. Criminal Record Office (CRO) number. Identifier given to an individual when fingerprints are taken (the last two digits (/nn) in a CRO number indicates year of issue. When the CRO number is retained, the minimum information that is held consists of the surname, forname(s), sex and date of birth of the subject.

1g. Police national computer (PNC) ID number. The PNCID is automatically allocated as a unique identifier when the PNC record is created. The PNCID and CRO number are two different numbers issued by two different systems although both can be held on the PNC.

2. Case details (relates to crime scene samples)

2a. Case lab code.

2b. Case year.

2c. Case number.

2d. Offence code.

2e. Job number.

2f. Item number.

2g. Crime Number.

3. Customer details

The customer is the police force that took the sample. The chief constable of that force is the owner of the DNA samples and profile.

3a. Police force code.

3b. Police force name.

3c. Station. A list of station code for the Met is available in the National Archives.

4. Personal details

4a. Name.

4b. Date of Birth.

4c. Alias 1. Aliases do not originate from the PNC record and are not often used.

4d. Alias 2.

4e. Gender.

4f. Ethnic code. Identity codes: 1 for White - North European, 2 for White - South European, 3 for Black, 4 for Asian, 5 for Chinese, Japanese, or other South East Asian, 6 for Arabic or North African or 0 for unknown.

4g. Ethnic appearance.

5. Profiling details

5a. Sample type. One or two digits code that refers to the type of sample provided (i.e., whether from saliva, hair, blood).

Usually a buccal scrape is taken. A sample is obtained via rubbing inside the suspect's inner cheek with a mouth swab to loosen and collect skin cells. The sample is then put into a plastic tube. The process is repeated with the other cheek so that two samples are taken. This is sample type 3.

Types of samples recovered from crime scenes are more varied. The quality varies and crime scene profiles often have fewer markers.

5b. Batch number. Laboratory batch identifier.

5c. Batch year. Laboratory batch identifier.

5d. Number in batch. Laboratory batch identifier.

5e. Track number. Laboratory batch identifier. Denotes the position of the sample in the gel.

5f. Date sample taken. The date the sample was taken from the subject. It originates from the PNC.

5g. Date sample loaded. The date the PNC stub record was uploaded to the database.

5h. Gel number. Laboratory batch identifier.

5i. Gel year. Laboratory batch identifier.

5j. Test method. DNA profiling technique used (SGM prior to 1999 and SGM+ since)

5k. Date batch first added. Date this batch of DNA profiles was uploaded to the database. (Used in 1990s but no longer used for loading profiles.)

5l. Date profile loaded. Date this specific DNA profile was uploaded to the database.

5m. Forensic supplier. Code for the forensic provider's laboratory.

5n. Whether record searchable. Yes or no answer to the question: is the profile currently searchable on the database?

NDNAD fields

6. Amplified sample/case details

6a. Sample barcode. Reference of the barcode sticker that was put on the sealed tubes holding each sample.

6b. Case lab. Relates to crime scene samples.

6c. Case. Relates to crime scene samples.

6d. Case year. Relates to crime scene samples.

6e. Batch. Laboratory batch identifier.

6f. Batch year. Laboratory batch identifier.

6g. No in batch. Laboratory batch identifier.

6h. Proc unit. Code for the forensic provider's laboratory.

6i. Gel. Laboratory batch identifier.

6j. Gel year. Laboratory batch identifier.

6k. Sample type. One or two digits code that refers to the type of sample provided (i.e., whether from saliva, hair, blood). See 5a.

6l. Sample description. 'Buccal cells', 'Hair roots', 'Blood', etc.

6m. Control flag. 'Control' for subject sample and 'Stain' for crime scene stain sample.

7. Profile results

The technique used to obtain the DNA profile from a DNA sample, called SGM+, only looks at specific areas known as short tandem repeats (STRs). STRs are places in the DNA where a short section of the genetic code repeats itself. People have varying numbers of repeats, which is how STRs can be used to identify individuals. Ten different STRs are analysed in each DNA sample. Because each STR is made up of two strands – one inherited from the mother and one from the father – this analysis produces 20 bits of information, known as alleles.

The DNA subject profile consists of a string of numbers indicating the number of repeats at each of the ten STRs plus a gender marker.

7a. Locus. Area of the DNA that is tested to create the profile.

7b. Low allele. The number of time the tested sequence is repeated.

7c. High allele. The number of time the tested sequence is repeated.

The custodian of the National DNA Database is the National Police Improvement Agency (NPIA), but the Forensic Science Service (FSS) is responsible for the operation and maintenance of the database. If your DNA profile is on the database and you want to check that the information recorded is accurate, you can obtain this information by sending a data subject access request to the attention of the data protection officer at the FSS. Kevin Reynolds did just that. You can check out how a real records – Kevin's – looks like in last year's post DNA retention of unconvicted people. His story shows how being an innocent with a DNA profile already retained on the DNA database didn't prevent the police from suspecting him of murder. They ignored his retained DNA profile when it should have eliminated him from suspicion.

To get off the DNA database, if you're innocent, write a letter to the chief constable of the force that arrested to reclaim your DNA. If you're successful but have any doubt, requesting a copy of your DNA profile is a way to double check that it has indeed been deleted. Last December, Kevin Reynolds did get confirmation that the deletion of all his samples and records had taken place. It's only when he contacted the FSS that he discovered that only the latter of his two DNA profiles and associated DNA samples had been deleted. He his still awaiting deletion of his earlier DNA profile.

You have until 2009-08-07 to respond to the Home Office consultation.

websiteblogblog archivenews feedfeedback