The collection of dental information in health studies based in remote settings need not be compromised because of lack of access to conventional resources or examination facilities. It is possible to obtain light and portable dental chairs, head torches and disposable instruments to ensure high quality examinations take place. It is also possible to have cost-effective and portable recording equipment to enable reliable collection of clinical data.
In any given multidisciplinary health investigation such as the ABC study, it is critical that the materials, instruments and data collection tools used in the dental component are minimal, light, fool-proof, durable, expendable and able to be set down and packed up again quickly in a range of environments whilst still allowing the privacy of participants to be respected. A key issue for the ABC study was portability of study equipment, with travel including ferry, four-wheel-drive and charter plane. There was limited space and sometimes weight constraints on all transport modes, and equipment from other components of the study were bulky (for example, centrifuge and dry shipper for blood samples). Health investigations that involve a remote component such as the ABC study may face additional logistic challenges in terms of weather (heavy rain, floods, cyclones), transport (unavailable or inoperable charter planes or vehicles) or unforeseen road closures; it is important that the equipment used, including dental equipment, does not add any additional duress to these logistical elements.
Disposable probes were used for examination because routine sterilisation procedures were unable to be undertaken in the field due to the multiple and variable examination sites in the follow-up with poor supporting infrastructures as well as the cost involved and time available. Disposable probes have been used in other oral epidemiological studies and are sanctioned by the Australian Research Centre for Population Oral Health for use in settings that are not conducive to routine sterilisation procedures.
At a community level, logistical challenges in the ABC study usually involved locating a convenient location in which to conduct assessments, and organising and picking up a reliable vehicle to collect and drop off study participants. The conventional study, in comparison, faced markedly fewer logistical constraints. Postal techniques and telephone calls were largely successful, participants did not need to be picked up and dropped off, and examinations were conducted in dental clinics where infrastructure (such as electricity and privacy) was generally reliable.
There were a number of cultural aspects in the ABC study that needed to be acknowledged and addressed in order for recruitment and dental data collection to be successful. Participants were highly mobile, meaning marked effort was necessary in order for a) participants to be located and b) participants agreeing to present for data collection. While this is a feature of health research conducted among indigenous populations on the whole , it also reflects the transient nature of the early adult age-group being assessed. There was poor voluntary presentation, perhaps due to the unfamiliar and somewhat invasive nature of the data collection procedures (with many participants having never visited a dentist before) and possibly because the study team were predominantly non-indigenous . The study team needed to be sensitive to community circumstances such as deaths and funerals, as well as unforeseen events such government interventions . Using local advocates was essential to engender a spirit of trust in the community, to help build a rapport with participants and to enable study team members to communicate with those for whom English was a second language. Consent documents were not provided in any language aside from English, due to the large number of language groups used by study participants. If it appeared that English was not the first language of the study participant, and that the participant was struggling to understand the consent process, the locally-employed research assistants, who did speak the same first language, were asked to convey the consent information verbally. Other study participants with a solid grasp of English also assisted.
Gender separation was critical at times when female participants were present with children and when various skin-groups could not be present in the same vicinity . Timing in the day was also crucial; early morning was non-ideal for late-rising communities, and mid-afternoon no use in communities where card playing was a strong social feature . Because many participants were not familiar with the dental instruments, time needed to be taken to explain what was going to be done and how instruments would be used. When children were present, it helped to let them wear the dental headlamp for a time, to show them the dental mirror and to examine their own teeth if time permitted.
Many participants required assistance with the self-report component of the dental examination, with the questionnaire structure being unfamiliar to some (the use of Likert-type scales, for example) and many not wanting to lose face by showing that they did not fully understand the questions (literacy levels of indigenous Australians being markedly less than their non-indigenous counterparts ). For many participants, especially those remotely-located and for whom English was a second language, the concepts used in the questionnaire were frequently not something they were familiar with (items relating to dental fear, self-rated oral health for example). Because the dental component was one of nine components being measured (and was often the last), participants were often exhausted by the time they reached the dental station and were impatient to leave (especially if accompanied by young infants). A large number of research investigations being conducted in some communities also meant that participants were 'surveyed out', particularly if they felt that nothing was being returned to either them or the community as part of participating in such studies . Each participant received food and a 'thank you' bag containing, among other items, a toothbrush, toothpaste sample and pamphlet outlining the risk of smoking in periodontal disease and oral cancer. Towards the end of the follow-up wave urban participants received a local department store voucher to boost recruitment, however we were unable to show if this made a difference to presentation. Each community received a de-identified summary report of the community-level findings.
The cultural aspects of the conventional study, in comparison with the ABC study, were relatively less, although issues regarding English as a second language (13 percent of the sample spoke another language other than English at home ) and exhaustion at the end of the dental examination did exist.
While a human recorder may be the best way of capturing dental epidemiological data in conventional settings, this may not be possible in investigations such as the ABC study where logistical or financial constraints prevent the inclusion of another person (no room on the charter plane etc). For purposes of this study, digital voice recorders proved to be portable, cheap, easy-to-use, reliable and did not require much time to implement. Computer-based voice recording software were comparatively time-consuming, unreliable, logistically difficult and burdensome, although it may be that such software becomes less cumbersome and more dependable in the future.
Computer-based voice-recording software has proved successful in clinical settings such as endoscopy , but other areas of the medical profession adopting this approach for data recording reported similar problems as those experienced in the ABC study. For example, in a radiology setting, Gale et al.  found that the time required to produce a report using voice-recording software was significantly longer than that required by the traditional tape-transcription system, with the additional cost in time estimated to be, for their institution, $100,000 annually. In the dental literature, computerised voice recognition methods have been found to be significantly slower than traditional examiner/recorder methods, although differences in error rates have not been substantial . While it is likely that future systems will require less effort, in an investigation such as the ABC study, time limits for each dental examination need to be strictly adhered to. In this study, a digital voice recorder, which gave the examiner the option of playing back the recording at his or her leisure, was the preferred option when space limitations or other logistical constraints meant that inclusion of a human recorder was not possible.