The INTERPOL DVI Dental Process and Reconciliation

INTERPOL divides a DVI operation into five phases, and understanding where dental evidence collection and comparison fits within each phase is the basis of operational planning. The phases are not strictly sequential in a large operation, multiple phases run in parallel, but the classification helps commanders allocate resources and define milestones.

1. Phase 1: Scene examination
   Recovery and documentation of human remains and personal effects at the disaster site. Forensic odontologists may be deployed at this phase to assist with recognition and recovery of dental remains, including isolated teeth or jaw fragments, which may be missed by non-specialist teams.
2. Phase 2: Post-mortem examination
   Systematic examination of recovered remains at the DVI mortuary. Each PM case receives a full dental examination: charting, radiography, description of restorations and prostheses, and photography. Yellow PM dental forms are completed. Data is entered into the DVI database.
3. Phase 3: Ante-mortem data collection
   AM teams, typically working at a separate family reception centre, collect dental records from families and their dentists, photograph personal effects including dental prostheses, and complete the pink AM dental forms. Data is entered into the same DVI database. This phase runs in parallel with Phase 2.
4. Phase 4: Reconciliation
   The database is interrogated to generate candidate match lists. Expert reviewers compare candidate PM/AM pairs, and the reconciliation board evaluates proposed identifications. This is the most intellectually intensive phase and the one where errors are most likely if protocols are not followed rigorously.
5. Phase 5: Debriefing
   A formal review of the operation's conduct, findings, and any errors. Lessons documented here feed into updated national and international DVI protocols. Many improvements to the INTERPOL guidance emerged directly from post-operation debriefs.

The INTERPOL DVI dental forms are designed on a mirroring principle: every data field on the pink AM form has a corresponding field on the yellow PM form. This parallelism is not accidental. It means that when a reviewer sits down with one AM and one PM record, the comparison is a field-by-field exercise rather than a translation between two different recording conventions. A yellow PM chart showing tooth 36 with an MOD amalgam restoration can be compared directly to a pink AM chart showing tooth 36 with an MOD amalgam restoration, because both forms recorded the finding the same way.

The dental section of each form records: the dental chart (all teeth, using FDI notation, marked present/missing/unerupted/unknown), restoration details for each tooth (material, surfaces involved using the standard surface abbreviations: M=mesial, D=distal, O=occlusal, B=buccal, L=lingual/palatal), presence of crowns, bridges, implants, and partial or full dentures, and any notable anatomical features. The PM form also has fields for condition of the remains and degree of preservation, which inform the interpretation of discordancies.

The forms also include a section for radiograph quality and availability. In the AM form, the odontologist notes whether periapical films, bitewing films, a panoramic radiograph, or other images are available and their image quality. In the PM form, the quality of the radiographs taken at the mortuary is noted. This matters for interpreting the comparison: a weak comparison between two poor-quality radiographs is less weight than a strong comparison between two good-quality films.

The FDI two-digit system (International Organization for Standardization standard 3950) assigns each tooth a two-character code. The first digit is the quadrant: 1 is upper right, 2 is upper left, 3 is lower left, 4 is lower right, for permanent teeth. Primary teeth use 5-8 in the same quadrant order. The second digit counts from 1 at the midline outward to 8 at the wisdom tooth. So tooth 16 is the upper right first molar; tooth 36 is the lower left first molar.

Restoration surfaces are coded with standard abbreviations that are the same regardless of notation system: M (mesial), D (distal), O (occlusal for posterior teeth), I (incisal for anterior teeth), B (buccal), L (lingual or palatal), F (facial). A restoration described as tooth 36 MOD is a lower left first molar with a restoration covering the mesial, occlusal, and distal surfaces. This combination, its material (amalgam, composite, gold, ceramic), and whether it is a primary or secondary restoration, constitutes a specific data point that can be compared field-by-field between AM and PM records.

The codes also record tooth status beyond restorations. A tooth recorded as X (or the equivalent symbol on the form) is missing, either because it was never present, extracted, or not recovered. Recording a tooth as missing in the PM record when the AM record shows it present (or vice versa) can be an unexplained discordancy or an explained one depending on clinical history. A tooth extracted after the AM records were made should ideally be documented as a subsequent treatment, which is itself a form of AM update.

Before purpose-built DVI software existed, large operations managed AM/PM comparison with paper sorting and manual cross-referencing. The 1977 Tenerife disaster, which killed 583 people, was managed with card indexes. By the 1990s, proprietary database tools were being tested by national DVI units, and the 2004 tsunami accelerated their adoption by demonstrating that no paper-based system could handle thousands of simultaneous cases across dozens of contributing nations.

DVI System International (marketed commercially as DVI2000 and related versions) from Plass Data Software AS in Norway became the dominant platform. It manages all three primary identification streams in a single integrated database: fingerprints, dental records, and DNA profiles. Each stream has its own input forms mirroring the INTERPOL paper form structure, its own comparison engine, and its own candidate ranking system. The dental module encodes each PM and AM dental record as a structured profile of tooth presence and restoration data, then compares PM records against AM records using an algorithm that counts concordant features and flags discordancies, returning a ranked candidate list for each PM case.

Other platforms exist and are used in various national operations. PlassData's WinID is an older dental-only comparison tool that some teams continue to use for dental-specific work. UVIS (Unified Victim Identification System) has been used in some US-led operations. The UK's IDENTIFY system is used by national police teams. The specific software matters less than the discipline of using a standardised one consistently within an operation: mixing multiple incompatible systems in the same operation is a data integrity problem.

The reconciliation board is the formal quality gate in a DVI operation: it prevents premature or incorrect identifications from reaching families, resolves conflicts between identification methods, and ensures that every confirmed identification has been independently reviewed by more than one qualified expert. The board meets regularly throughout the operation, with the frequency depending on the rate of candidate generation.

A standard reconciliation board includes: the senior forensic odontologist (or the dental team leader), the senior fingerprint examiner, the DNA coordinator, and the DVI operation commander. In some operations a forensic pathologist also sits on the board. The board reviews three categories of cases: confirmed candidates from all three primary methods (fast track to acceptance if all three agree), candidates where only one method has produced a match (requires discussion of whether the single-method evidence is sufficient), and cases where two methods conflict.

Conflict resolution is the board's most demanding function. A dental identification that conflicts with a DNA result is not automatically resolved in favour of DNA. Before the DNA result, the board asks: is the AM DNA reference sample definitively from the same person as the AM dental record? Could there be a sample labelling error? Is the DNA result definitive, or is it a partial profile? The board investigates the conflict source before accepting either result. In practice, conflicts between dental and DNA in large operations are rare but have occurred, and the resolution protocol is designed to catch labelling errors and sample contamination rather than assume one method is infallible.

• All-method agreement: dental, fingerprint, and DNA all confirm the same AM identity. The board confirms the identification and the DVI commander authorises family notification.
• Single-method identification: only dental (or only fingerprint, or only DNA) has produced a candidate. The board reviews the strength of the single-method evidence. A strong dental comparison with multiple concordant features and good radiograph quality may be accepted. A weak comparison with limited concordance goes back for further AM collection.
• Two-method agreement, one absent: typically accepted if both agreeing methods are strong. The absent method (often DNA in early-phase identifications) may confirm later, and the board documents the basis for the provisional confirmation.
• Conflict between methods: the case is not confirmed. The board investigates the source of the conflict before proceeding. Possible explanations include sample labelling error, transcription error in one record, or a genuine mismatch indicating one of the proposed AM matches is wrong.

The consequence of a misidentification in a DVI context is severe: a family receives remains belonging to the wrong person, buries or cremates them, and potentially closes a case that should remain open. Retractions, though rare, have occurred in large operations and are logistically and emotionally complex. The quality assurance steps embedded in the INTERPOL framework exist specifically to prevent that failure.

Within the dental stream, quality assurance operates at three levels. First, each odontologist's PM charting is reviewed by a second odontologist before data is entered into the database, specifically checking the FDI coding, restoration descriptions, and radiograph quality assessment. Second, when a candidate match is proposed by the software, the reviewing odontologist must examine the original radiographs, not just the encoded data, and document their comparison notes. Third, the reconciliation board provides the external check on the individual odontologist's conclusion.

The DVI operation also maintains a formal register of all confirmed identifications, the method or methods that supported each, and the date of confirmation. This register serves two functions: it tracks operational progress, and it provides the evidentiary record if any identification is subsequently challenged. In long operations, turnover of personnel means that the reasoning behind an identification from month two must be recoverable by a reviewer in month eighteen who was not present at the time.

INTERPOL maintains a DVI standing committee that reviews and updates the framework and coordinates training programmes. National DVI teams are expected to train their forensic odontologists specifically in INTERPOL DVI procedures, not just in general forensic dental examination. The specific skills required include: completing the pink and yellow forms correctly to INTERPOL standards, using FDI notation consistently and without transcription error, operating the DVI database software, and participating constructively in a reconciliation board.

The Disaster Victim Identification course offered by INTERPOL and by national academies such as the UK's NPIA (now the College of Policing) typically covers both the procedural framework and the practical skills. Simulated DVI exercises, sometimes called tabletop exercises and sometimes full-scale mortuary simulations with anonymised AM and PM records, are widely used to test team readiness. These exercises have consistently shown that teams familiar with the forms in training make fewer data entry errors in real operations.

Implementation varies by country. Some national forces maintain standing DVI teams with defined rosters, regular training, and pre-deployed equipment including portable dental radiography units. Others rely on ad hoc assembly of qualified practitioners after a disaster is declared. The consensus in the DVI community, driven by operational experience from events including the 2004 tsunami, the 2015 Nepal earthquake, and the 2016 Brussels bombings, is that pre-formed, regularly trained teams consistently outperform ad hoc assemblies on both speed and accuracy of identification.