Problems and Future Approaches for Assessment of Periodontal Disease

Clinical diagnosis of periodontal disease is generally obtained by probing or attachment loss. Those have also been used in epidemiological investigations, though several problems have been pointed out, such as follows: infectious aggression by the dental instrument, time-consuming for both dental professionals and patients, and patient discomfort (e.g., pain or itching during probing). For standardization of assessment of periodontal disease, calibration of probing between inter-examiners is needed. A systematic review of the periodontal literature has been conducted to assess the extent of reporting of calibration of probing. Searches of four electronic databases (Medline, Embase/BIDS, PubMed, and Cochrane Library) were undertaken. Search constraints were: 1996–2009; clinical trials; human studies; English language. Also, hand searches of 12 periodontal/dental journals were undertaken for 14-year period. Then, a total of 1037 papers from 668 citations as the electronic search and 369 citations as the hand search were reviewed. Of those papers, only 28% reported examiner alignment for that assessment, while 42% gave no information on examiner alignment and assessment (1). Furthermore, in 288 papers that reported the statistical methods for determining examiner alignment and assessment, approximately 20% noted kappa value, a standard index used to show whether an adequate agreement between examiners was guaranteed (>0.8 better), while no description regarding such statistics were found in 66% of those articles (1). Although it is important to begin by considering the definition of periodontal disease, no uniform case definition exists. Gingival inflammation (gingivitis) (2), deepened pockets (3), periodontal bone loss (4), and a combination of different oral parameters (5) have been used to define periodontal disease. To further complicate the matter, a wide variation of cut-off values for these parameters have been utilized to identify the presence of periodontal disease, such as attachment loss greater than 3 mm per year (6), attachment loss greater than 3 mm and probing depth greater than 4 mm in at least 30% of sites measured (7), attachment loss greater than 4 mm in at least 30% of sites measured (8), and classification based on the proportion of teeth with greater than 4 mm of probing depth (9). Therefore, comparisons of results from different studies include uncertainty and it is difficult to make conclusions from metaanalyses. Thus, since the parameters used for diagnosis of periodontal disease are not straightforward, we consider that a probing method is inappropriate for determination of underlying disease status. We would propose the following four methods as alternatives to probing for use in epidemiological studies and populationbased community examination: (1) number of lost teeth, (2) bleeding from gingiva, (3) self-reported questionnaire of periodontal condition, and (4) combination of general medical assessment and noninvasive tests of the gingiva. Problems with these methods and future perspective are discussed below.