15 Measuring Dental Caries
This chapter describes the methods for measuring dental caries in human populations. Some historical measures from the early twentieth century include the proportion of first molars lost through caries21,22 and the percentage of erupted permanent teeth affected by caries.1 Both of these measures were useful when there was little information of any kind about the disease, though they lacked sensitivity. At the other extreme, the Bodeckers’ index of surfaces affected by caries, described in 1931,10 was sensitive but complicated. Dean and his colleagues,14 in their pioneering studies of the caries-fluoride relationship, counted the number of teeth in the mouth visibly affected by caries, which was a forerunner of the DMF count. The first description of what is now known as the DMF index is attributed to Klein, Palmer, and Knutson in their studies of dental caries in Hagerstown, Maryland, in the 1930s.32 Since then, the DMF index has received practically universal acceptance and is the best known and most widely used of all dental indexes.
Filled teeth were assumed to have been unequivocally decayed prior to restoration. The DMF score for any one individual can range from 0 to 32, in whole numbers. A mean DMF score for a group, which is the total of individual values divided by the number of subjects examined, can have a decimal value. The DMF index can be applied to whole teeth (designated as DMFT) or to surfaces (DMFS). Modifications can be made to the index for factors such as filled teeth that have redecayed, crowned teeth, bridge pontics, and any other particular attribute required for a given study. To save time in a large survey, the DMF index can be used half-mouth; that is, it can be applied to opposite diagonal quadrants and the score doubled, an approach which assumes that the carious attack is bilateral. The DMF index can be applied to all 32 teeth, although because of the widespread removal of third molars in young adults some prefer to record a score for 28 teeth. Either approach is acceptable as long as the method is clearly stated.
BOX 15-1 Limitations of the DMF Index
The DMF index has received remarkably little challenge over some 70 years of life, probably because it is simple and versatile. It was developed for use in children a long time ago, however, and accordingly it shows its age in a few areas. The principal limitations are the following:
The DMF index for permanent teeth is always signified by uppercase letters; the equivalent index for the primary dentition is the def index and its modifications.19 The teeth counted to derive this index were originally defined as follows:
Teeth missing because of caries are not recorded because of the complications of exfoliation and lack of knowledge as to whether missing teeth were carious prior to exfoliation. Modifications of this index are (1) the dmf index for use in children before the age of exfoliation, (2) the dmf index applied only to the primary molar teeth, and (3) the df index. Values for df and def should be numerically the same; def allows for two grades of caries, and neither index counts missing teeth. Both the def and df indexes may therefore understate the true extent of the carious attack; the trade-off is presumably greater reliability from ignoring missing teeth.
Because of the present-day skewed distribution of caries prevalence in the population, the Significant Caries Index (SiC Index) was developed.11 The SiC Index is not a new index, because it is based on the distribution of DMF values in a population, but is a way of expressing caries distribution that goes beyond mean DMF. It is actually the mean DMF score for the third of the distribution most affected by caries and is intended to be used alongside the mean DMF value to give a more complete summary of caries in the whole population. The more skewed the distribution, the greater the gap between the mean DMF score and the SiC value. A new global goal of an SiC score of 3.0 or less for 2015 has been suggested57 and, if adopted and used, would provide some further information on caries distribution.38
Other methods of measuring dental caries with a different philosophical base have been suggested. One is Grainger’s hierarchy, an ordinal scale designed to simplify the recording of the caries status of a population, which uses five zones of severity of the carious attack.18 This scale appears to be valid29,31,45 but has received little further use, probably because of low sensitivity. More recently, “composite” indicators have been suggested that attempt to measure health rather than disease by weighting healthy restored teeth differently from missing or decayed teeth.46 The first of these is the FS-T, which sums the sound and healthy restored teeth. The second is T-Health, which seeks to measure the amount of healthy dental tissue and assigns descending numerical weights for a sound healthy tooth, a filled tooth, and a decayed tooth. These are conceptually sound approaches to measuring dental health and function (rather than disease), and they deserve more attention than they have received.
Sealants had not arrived when the DMF index first appeared, but there are two reasonable approaches for dealing with sealants in the DMF system. One view holds that the sealed tooth is not restored in the classic sense and should therefore be considered sound. The other contends that it has required hands-on, one-to-one dental attention and so should be considered a filled tooth. Probably the best way to deal with sealed teeth is to put them into a category by themselves, S for sealed. The DMFS index would then become DMFSS. Depending on the purpose of a given study, the S teeth can be left separate, included with the F teeth, or regarded as sound.
With modern preventive and restorative technology, the DMF index is really outdated as a measure of caries attack; the index may be more valid as a measure of treatment received. It is philosophically questionable to use a disease index that is so dependent on the treatment judgments of many practitioners, and combining previous treatment (i.e., the M and F components) with current treatment need (the D component) is unsuitable for surveillance purposes (see Chapter 4). A measure of caries activity would be preferable for many purposes, but approaches to scoring caries activity are still based on clinical acumen.41 Until a more objective measure is developed and accepted, the DMF index will continue to be used. The results of its use, however, should always be interpreted with care.
There is no global consensus on the criteria for diagnosing dental caries, despite a vast quantity of words on the subject. Different traditions about defining a lesion in the gray area, where it is difficult to tell whether the disease is irreversibly established or not, have grown up and are still adhered to. Apart from the inherent problem of diagnosing a borderline lesion, the major philosophical issue is how to score an early carious lesion that has not yet become cavitated, whether diagnosed clinically or radiographically. Such a lesion appears as a discolored fissure without loss of substance, as a “white spot” on visible smooth surfaces, or radiographically as an early interproximal shadow. The issue is that not all noncavitated lesions progress to become dentinal lesions requiring restorative treatment; a good proportion of them remain static or even remineralize, especially smooth surface lesions.43 These lesions are thus reversible, as opposed to a dentinal lesion, which is generally considered irreversible. Because there are usually more noncavitated than cavitated lesions at any one time in both high-caries and low-caries populations,9,23,44 the decision as to whether to include or exclude them, and how to express them if included, can make a substantial difference in the oral health profiles obtained.
Examples of these two different approaches to diagnostic criteria for dental caries are shown in Box 15-2. Traditionally, European investigators have recorded caries on a scale that extends through the full range of disease from the earliest detectable noncavitated lesion through to pulpal involvement.3 The criteria in Box 15-2 are based on those first published by the World Health Organization (WHO) in 197954 and now usually referred to as the D1-D3 scale. Clinical researchers in Europe have expanded on this concept to produce a scale with up to 10 points, combining increasing depths of lesion development with clinical signs of activity or inactivity.37 On the other hand, investigators in North America, Britain, and the other English-speaking countries have traditionally recorded caries as a dichotomous condition, meaning that caries is diagnosed only as present or absent. (We refer to this as the dichotomous scale.) In dichotomous recording, caries is only noted when it has reached the level of dentinal involvement,20 that is, the D3 level. Use of the D1-D3 scale requires that the teeth be dried and be given a longer, more meticulous survey examination. Although there are more diagnostic decisions to make when the D1-D3 scale is used, adequate examiner reliability can be maintained when examiners have been trained in this system.43