Jules A. Kieser, 1 Ph.D.; Wayne Laing, 2 B.D.S.; and Peter Herbison, 3 M.Sc. Lessons Learned from Large-scale Comparative Dental Analysis Following the South Asian Tsunami of 2004 ABSTRACT: The aim of this study was to examine the quality of the ante-(AM) and postmortem (PM) dental data that were submitted for entry into the PLASSdata system in Phuket, Thailand, following the Boxing Day (December 26) Tsunami, 2004. The investigators were two forensic odontlogists who were part of the New Zealand Disaster Victim Identification team that worked at Wat YangYao morgue and at the Information Management Center in Phuket. Our findings underline the usefulness of dental data in human identification, but point to a number of significant sources of error. Of the 78 PM records received, only 68% of radiographs and 49% of photos confirmed the accompanying dental charting. This underlines the value, particularly of photographs of the dental arches, in quality control. It also points to a large error component, which may have been due to inexperience of the operators, fatigue, poor conditions in the temporary morgue, or the problem of tooth-colored fillings. Of the 106 AM records received, 62% were of unacceptable quality and 64% were either not accompanied by radiographs or had poor quality radiographs. These results indicate that AM data collection ideally needs to be collated and checked by a forensically trained dentist(s) in the country of origin. KEYWORDS: forensic science, forensic odontology, human identification, dental records, tsunami On 26 December 2004, 1000 km of fault ruptured beneath the sea west of Sumatra, creating an earthquake that measured 9 on the Richter scale. The resultant tsunami was the third biggest nat- ural disaster in the past 100 years and claimed over 250,000 lives, including some 5,300 in southern Thailand (1). In response a multinational Disaster Victim Identification (DVI) Center was set up in Phuket to identify those killed. This newly assembled In- formation Management Center (IMC) processed firstly, postmor- tem (PM) data obtained from the 31 national teams involved in examining victims of the tsunami in temporary morgues estab- lished at Wat YangYao and Mai Kau; and secondly, antemortem (AM) data gathered from the numerous countries involved. These data were entered into the PLASSdata system (2) under standard operating procedures defined in Interpol Disaster Identification Manual (3). Where reconciliations could be made, these were presented to the Thai Reconciliation Commission, who if satisfied, then authorised the issue of a Thai death certificate. Three main categories of data were entered into the system; fingerprints, DNA, and dental records. In this paper, we document our own findings of the quality of dental records, obtained while we were part of the New Zealand DVI team based in Phuket, Thailand, 16 January to 7 February 2005. Method According to standard Interpol DVI protocols (3), PM dental data for each individual body were entered onto pink DVI forms. Additionally, bitewing and periapical radiographs were taken where necessary in accordance with standard operating proce- dures authored by J. Taylor and issued after 16 January 2005; ‘‘RADIOGRAPHS Standard for each body is bitewings and any other films as deemed appropriate; 1 or 2 periapicals for children for age determinations’’ (J. Taylor, Standard Operating Procedure document submitted to the Scientific Advisory Sub-Committee, Thai Tsunami Victim Identification Committee). Constraints on time, especially in the early weeks of the exercise, precluded tak- ing full mouth radiographic surveys of each PM case. Polaroid images of the maxillary and mandibular occlusal tables and an anterior edge-to-edge view of the incisors were also taken. These data were then entered into the Plassdata system at the IMC. To determine the quality of the PM dental data entered, we (J. K. and W. L.) checked the quality of each of the three sets of input; dental PM charting, radiographs, and photographs. Each charting was ranked as being either of good or poor quality, based upon features such as clarity of writing and drawing and also of conformation to the Interpol instructions. We then noted whether the charting con- formed to the written description entered on the pink form. Ra- diographs were rated according to criteria adapted from Helminen et al. (4) It was accepted that although a radiograph might be technically imperfect, it could still provide useful information, and hence quality indicators had to be pertinent to the realities of the situation we found ourselves in. Radiographs were taken as ac- ceptable when they conformed to the following criteria: 1. Image not too dark or light—enamel, dentine, pulp, alveolar bone distinguishable. 2. Periapical shows entire crown and root of tooth or teeth radi- ographed, bitewings show entire crowns as well as marginal bone lines. 3. There was no approximal overlapping of crowns such that the enamel of one tooth obscured the dentine of its neighbor. 4. There were no ghost images, stains, scratches, stripes, or artifacts. 1 Department of Oral Sciences, Faculty of Dentistry, University of Otago, Dunedin, New Zealand. 2 Family Dental Center, Allen Street, Morrinsville, New Zealand. 3 Department of Social and Preventive Medicine, Faculty of Medicine, University of Otago, Dunedin, New Zealand. Received 19 Mar. 2005; and in revised form 27 July 2005; accepted 20 Aug. 2005; published 26 Dec. 2005. Copyright r 2005 by American Academy of Forensic Sciences 109 J Forensic Sci, January 2006, Vol. 51, No. 1 doi:10.1111/j.1556-4029.2005.00012.x Available online at: www.blackwell-synergy.com