Periodontal Myths and Mysteries Series (IV) – Guesswork and a Mental Note

An American Academy of Periodontology Task Force (AAP-TF) has recently reported about a planned update of the 1999 Classification of Periodontal Diseases and Conditions [1]. I have written about it here. As the update is announced for 2017, I suppose that the AAP-TF report is meant to initiate some discussion as “[c]oncerns had been expressed by the education community, the American Board of Periodontology, and the practicing community that the current classification presents challenges for the education of dental students and implementation in clinical practice”. The AAP-TF report focuses in particular on  attachment level, chronic versus aggressive, and localized versus generalized periodontitis. In my previous post, I had raised some concerns about the task force’s intention to keep the current differentiation between aggressive and chronic periodontitis and referred to an interesting essay by Baelum and Lopez (2003). As these authors have just published a harsh comment on the AAP-TF report, it’s interesting to see that we agree and utterly disagree in certain matters.

Attachment level measurements

The 1999 Classification indeed categorized severity of periodontitis only by amount of clinical attachment loss, slight, moderate, and severe (i.e. 1-2 mm, 3-4 mm, and 5 mm or more, respectively). Albeit attachment level measurements are important “for the scientific advancement of the knowledge of periodontitis”, the AAP-TF recognizes that attachment level measurements are challenging, time-consuming, difficult and “may involve some guesswork when the CEJ [cemento-enamel junction] is not readily evident via tactile sensation.” Consequently, they advocate new guidelines for determining severity, slight or mild, moderate, severe or advanced, of periodontitis which include, in addition to those based on clinical attachment level measurements, probing depths (>3 mm & 5 mm, >5 & <7 mm, and ≥7 mm, respectively), and radiographic bone loss (up to 15% of root length or 2-3 mm, 16-30% or 4-5 mm, and more than 30% or 6 mm or more). Bleeding on probing has always to be present if a diagnosis of periodontitis is to be made.

The task force also recognizes the necessity of defining a condition of a reduced but healthy periodontium after periodontal treatment if probing depths are within 1-3 mm and bleeding on probing is absent.

Despite the attempt of the task force to be overly precise in giving the range of probing depths when periodontal probes have at most just mm markings, all of this makes a lot of sense.

The concept of “clinical” attachment level is ill-defined. It has to be differentiated from histological, or true attachment level, i.e. the level of the most coronal fibers attached to root cementum. That cannot be determined clinically. When it comes to periodontitis, what is being treated is mainly pockets [2]. What is of major interest is pocket reduction and possibly ensuing recession. It may suffice to record pocket depth reduction and gingival recession after treatment. Proper treatment of pockets may lead to resolution of inflammation with greater resistance to (similarly ill-defined) probing. Seemingly, “attachment” has been gained. But when it comes to practicing dentists or periodontists, except for certain lesions where regenerative treatment modalities were applied (the ensuing result being entirely ill-defined in the general absence of histological evidence), the two endpoints of any periodontal treatment are pocket reduction until probing depths of 4 mm or less are achieved and no bleeding on probing is noted (in the absence of pressure control another ill-defined periodontal measure) [3].

There is hardly any dental school which would advocate full-mouth clinical attachment level measurements before and after periodontal treatment; and I am not aware of any practicing periodontist who routinely records clinical attachment levels of all teeth. When it comes to clinical periodontal research, however, clinical attachment level measurements seem now to be indispensable. As a reminder, it was Sigurd Ramfjord at the University of Michigan in Ann Arbor who first advocated attachment level measurements in periodontal research when describing the Periodontal Disease Index (PDI) in 1959 and 1967.

The PDI index  [sic] is primarily concerned with an accurate assessment of the periodontal status of the individual person. Emphasis is placed on recording of the attachment level of the periodontal tissues relative to the C-E [cemento-enamel] junction. Such accurate measurable assessments are essential for longitudinal studies of periodontal disease and as a scientific basis for clinical trials in Periodontology.

Ramfjord did recognize the challenges of identifying a subgingival cemento-enamel junction.

Since the surfaces of the enamel and the cementum have different inclines, the change in direction of the movement of the point of the probe is detectable when it moves from the enamel to the cementum. The texture or surface characteristics are also different since the cementum is distinctly rougher than enamel. […] After the distance from the free gingival margin to the cementum enamel junction has been measured, an attempt should be made to move the probe along the cemental surface. This of course can be achieved only if there has been loss of periodontal attachment. If calculus covers the cementum enamel junction it has to be removed before the C-E junction can be localized. Occasionally it is also necessary to remove heavy deposits of supragingival calculus to gain access to the gingival crevice. (Emphasis added.)

Probably due to the above difficulties Ramfjord advocated, by the way, partial recording. The buccal and mesiobuccal surfaces of the so-called Ramfjord teeth (upper right first molar and central incisor, upper left first premolar, lower left first molar and central incisor and lower right first premolar) were scored only. An utterly pragmatic approach which has only recently been shown to be largely inappropriate in epidemiological studies.

Direct attachment level measurements

In case of gingival recession, attachment level can easily be measured by considering the two relevant landmarks, the cemento-enamel junction (CEJ) and the gingival margin, in one pass [4]. Problems arise when the cemento-enamel junction is covered by gingiva. In that instance, the pocket depth may be measured first followed by a second pass, when the CEJ is identified subgingivally. This is not so straightforward. As Ramfjord already noted, see above, calculus may obscure the CEJ. It may also have been destroyed by dental caries or cavity preparation for a restoration. It is pure speculation that “enamel and cementum have different inclines”, which could be noted when the probe was moved apically. That texture and surface characteristics could reliably be differentiated is speculation as well. The clinician will probe with slight movements, back and forth, but whether (s)he was able to identify the CEJ is largely uncertain. Guesswork.


There is no clinical gold standard available, so validity of the measurement cannot be determined. Consequently, in a clinical situation, attachment level measurements may be easy in, usually, a minority of sites when recession has occurred and the CEJ is clearly identifiable. In that case, one pass of probing may suffice and probing depth and clinical attachment level are measured at the same time. In the majority of sites, the CEJ would not be visible and has to be “determined” in a rather questionable way in a second pass. “Recession” has to get “signed” (if true, it strangely gets a minus sign) and the two measurements are “subtracted” [5].

In a rather vitriolic comment on the AAP-TF report, remarkably published in the European Journal of Oral Sciences, Lopez and Baelum [6] make the point that they are “aware that clinicians who have been brought up in an era ascribing a pivotal role to pocket depths, may find it difficult to adopt different recording practices,” as if the AAP-TF was about to abandon attachment level measurements.

It is not only natural that these clinicians may find it more difficult, challenging, and time consuming to make clinical attachment level measurements than to make the usual probing depth measurements. Clinical attachment level measurements can be achieved in two different ways: by the direct method (i.e. adding [sic] the measured, and signed distance between the CEJ and the gingival margin to the probing depth recording); or by the direct method, as the distance (sliding) between the CEJ and the bottom of the probeable pocket (4). The claim regarding time consumption can therefore only be valid if clinicians choose to record clinical attachment level using the indirect method.

What a misconception. As mentioned above, clinicians would apply different routines in different situations. What is expressed by Lopez and Baelum is a desire to standardize an utterly problematic assessment probably in order to address a concern raised by statisticians, who instantly think of additional measurement error when two independent measurements are subtracted. I suppose, however, that clinicians, who are usually no statisticians, intuitively try to reduce errors by applying the indirect method when required by the situation, i.e. a subgingival CEJ in the presence of deep pockets. And, they would do that only if there was a special interest, for instance, in the response to their treatment. Clinicians probably know for decades that relevant gains of “clinical attachment” would occur only in deep pockets of, say 7 mm or more. They know that 4-6 mm deep pockets may gain just about 1 mm “clinical attachment”, so why bother about such an amount. They would rather be interested in pocket reduction to 3 mm or less and no bleeding on probing.

In the above quote, Lopez and Baelum (2015) refer to one of their own papers by Corraini et al. [7]. I have analyzed that paper in a previous post, see here. Corraini et al. describes the direct method of measuring clinical attachment levels as follows,

CALDIR was measured, as the distance from the cement-enamel junction (CEJ) to the bottom of the clinical pocket. When the FGM [free gingival margin] was located above [sic] the CEJ, that is, covering it, the periodontal probe tip was run along the tooth crown in apical direction until the CEJ was felt. At this point, a mental note was made of the position of the FGM on the probe, and the additional probing movement to the bottom of the clinical pocket, that is, the “slide” movement from the bottom of the clinical pocket to the identified CEJ position was subsequently recorded in mm, withdrawing the periodontal probe only once (Griffith et al. 1988). If the probe tip did not reach the CEJ, a CALDIR recording of 0 mm was made.

So, a “mental note” was made when the CEJ was passed. When considering the above (e.g. presence of calculus, spurious tactile sensations, de facto lack of any gold standard), that routine is likely to be invalid. Anyway, Corraini et al. (2013) claim that direct clinical attachment level measurements were slightly more reliable than indirect measurements. Their data contains some disturbing recordings, though. For instance, in Fig. 2c of their article one can identify one site where the difference between test and retest (in a huge number of untreated subjects with a perceived need for periodontal treatment measurements had been repeated after one week) of direct clinical attachment level measurements was 10 mm while the mean of test and retest was 7 mm. That would translate into 12 mm and 2 mm at test and retest (if retest was subtracted from test measurement). More sites with disturbing discrepancies between test and retest can be identified when direct CAL measurements had been done which apparently did not occur (at least to that extent) when indirect measurements had been made (Fig. 2d). For example, another site was measured at an average of 8 mm when the difference between the two measurements was -10 mm. That translates to 3 and 13 mm at test and retest.

Lopez and Baelum (2015) “have not been able to find evidence that direct clinical attachment level measurements are generally more time consuming than are probing depth recordings.”

There is evidence, however, that practicing periodontists tend to be less reproducible in their recordings of both clinical attachment level and probing depth, probably because they: “conduct thorough examinations as they search for furcations and craters, and tend not only to ‘read too much’ into their examinations, but to do so somewhat inconsistently” (5).

Well, it all depends on the focus of an examination. The everyday clinical situation in fact demands search for furcation involvements and “craters” (infrabony lesions) as they are a major challenge for any periodontist, who is usually not doing research. And, the common treatment endpoint is shallow pockets which do not bleed on probing, not ill-defined gain of “clinical” attachment (exceptions mentioned above). In their comment, AAP-TF explicitly addresses clinicians, teachers and undergraduates.



[1] American Academy of Periodontology Task Force Report on the Update to the 1999 Classification of Periodontal Diseases and Conditions. J Periodontol 2015; 86: 835-838.

[2] Concepts regarding the gingival sulcus, or crevice; or the periodontal “pocket” and its base are ill-defined when it comes to their clinical assessment with unspecified probing pressure.

[3] When asked by Maurizio Tonetti, during the closing ceremony of Europerio8, about the endpoints of periodontal treatment, Professors Lindhe and Lang seemed to agree upon probing depths of or below 4 mm and no bleeding on probing in the presence of very low levels of dental plaque. They did not even mention possible regeneration of periodontal tissues or gain of clinical attachment. Lindhe mentioned that pocket elimination has still to be considered one valuable measure to achieve this goal.

[4] Of course the CEJ may have been destroyed by abrasion, dental caries or cavity preparation. In that case, considering the CEJ at adjacent tooth areas or neighboring teeth is just guesswork.

[5] The “signed” measurements between the gingival margin and the CEJ had probably been introduced by Ramfjord.

[6] Lopez R, Baelum V. Focus. Periodontal disease classification revisited. Eur J Oral Sci 2015; 123: 385-389.

[7] Corraini P, Baelum V, Lopez R. Reliability of direct and indirect clinical attachment level measurements. J Clin Periodontol 2013; 40: 896-905.


2 December 2015 @ 7:36 am.

Last modified December 11, 2015.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s