The current periodontal classification system, adopted in 2018 following the 2017 World Workshop, replaces the older “chronic” and “aggressive” periodontitis labels with a single disease category characterized by two dimensions: staging (how severe and complex the case is right now) and grading (how fast it is likely to progress and how it responds to risk factors). The framework was designed to work much like oncology staging, giving clinicians a shorthand that captures both the damage already done and the biological behavior driving that damage forward.
Why the Old Labels Were Retired
For nearly two decades, clinicians diagnosed patients with either “chronic periodontitis” or “aggressive periodontitis” under the 1999 classification. The distinction sounded clean, but it created problems. Chronic periodontitis was supposed to progress slowly and match the amount of plaque present; aggressive periodontitis was supposed to strike younger patients with rapid bone loss that outpaced what the plaque burden alone could explain. In practice, many patients fell somewhere in between, and the line separating the two diagnoses was blurry enough that different clinicians would classify the same mouth differently. Researchers also struggled to compare studies when their patient pools were sorted by unreliable categories.
The 2017 World Workshop concluded that the evidence no longer supported treating chronic and aggressive periodontitis as biologically distinct diseases. Instead, the workshop grouped all plaque-associated periodontitis under one umbrella and built the staging-and-grading matrix to capture the variability that the old binary labels had tried, and failed, to describe.
What Staging Tells You
Staging answers two questions at once: how much periodontal destruction has already occurred, and how complex will management be? There are four stages, and the initial assignment is driven primarily by the worst site in the mouth, not the average.
Stage I represents early periodontitis with modest attachment loss and no tooth loss attributable to the disease. Stage II reflects moderate destruction but still without major complexity factors. Stage III marks severe disease: deeper attachment loss, potentially some teeth already lost, and management complications such as deep pockets, vertical bone defects, or furcation involvement in multi-rooted teeth. Stage IV adds functional consequences on top of the Stage III criteria, such as significant tooth mobility, drifting, flaring, bite collapse, or the loss of enough teeth to compromise chewing ability.
Staging is determined by the worst affected site, but a case is also described by its extent. If fewer than 30% of teeth are involved, the case is “localized”; 30% or more makes it “generalized.” A case can also be labeled with a molar-incisor distribution pattern, which effectively captures what the old system tried to flag as “aggressive” when bone loss clusters at molars and incisors in younger patients.
How Grading Adds a Second Dimension
Where staging is a snapshot of current damage, grading estimates the trajectory. A patient at Stage III could be someone whose disease crept along over 30 years (slow progression, good prognosis with treatment) or someone who reached the same level of destruction in just a few years (rapid progression, harder to stabilize). Staging alone cannot distinguish these two people. Grading can.
There are three grades. Grade A indicates slow progression, Grade B moderate progression, and Grade C rapid progression. The primary way to estimate this is indirect: compare the percentage of bone loss at the worst site to the patient’s age. If destruction is proportionally low relative to how long the patient has been alive, progression has been slow. If destruction is disproportionately high for someone that young, progression has been fast. When available, direct evidence like sequential radiographs taken years apart showing the actual rate of bone loss can override this estimate.
Once a provisional grade is assigned, it can be shifted upward by two specific risk factors: smoking and diabetes. These “grade modifiers” reflect the strong evidence that both conditions accelerate periodontal breakdown and worsen treatment outcomes. Smoking fewer than 10 cigarettes per day shifts a patient to at least Grade B; 10 or more per day shifts them to Grade C. For diabetes, a well-controlled patient (HbA1c below 7%) is shifted to at least Grade B, while an HbA1c of 7% or higher pushes the grade to C.
How the Pieces Fit Together in Practice
A full diagnosis under the 2018 system reads something like “Periodontitis, Stage III, Grade C, generalized.” That single line tells the clinician that the patient has severe disease with complexity factors at multiple sites, that the disease is progressing rapidly or is being driven by significant risk factors, and that most of the mouth is affected. It communicates far more than the old “chronic periodontitis, generalized, severe” ever did, because the grading dimension captures biological behavior rather than just a severity label.
The diagnostic process itself follows a stepwise logic. First, the clinician screens for whether the patient has periodontal health, gingivitis, or suspected periodontitis. Second, if periodontitis is suspected, a confirmation step differentiates true periodontitis from other conditions that can cause attachment loss, such as trauma or orthodontic movement. Third, the case is staged. Fourth, it is graded.
Reliability Between Clinicians
A classification system is only as useful as its consistency. If two periodontists look at the same radiographs and clinical data and arrive at different diagnoses, the system’s value drops. Research on the 2018 classification has found that grading is the most reproducible component. One study reported agreement with a gold standard at about 82% for grade, compared with roughly 69% for stage and 76% for extent. Neither academic position nor years of experience significantly affected how well clinicians agreed with the reference diagnosis.
A separate study using kappa statistics, which adjust for the agreement you would expect by chance alone, found almost perfect inter-examiner reliability for grading and extent but only moderate reliability for staging. The difficulty tends to concentrate at the boundary between Stage III and Stage IV, where subjective judgments about complexity factors like bite collapse, tooth drifting, and masticatory dysfunction can tip the diagnosis either way. Interestingly, some of these complexity variables themselves showed low agreement between examiners: bite collapse and drifting, in particular, were hard for clinicians to evaluate consistently.
Does the Classification Predict Tooth Loss?
The real test of any disease classification is whether it tells you something about the future, not just the present. Several long-term studies have examined whether the 2018 staging and grading system predicts which patients will lose teeth to periodontitis over time, both during active treatment and during the maintenance phase that follows.
The evidence is encouraging on this front. A long-term retrospective study found that patients initially diagnosed at Stage IV had roughly 3.7 times the risk of losing teeth to periodontitis compared with Stage I patients, and those at Grade C had about 4.8 times the risk compared with Grade A. Both stage and grade independently predicted tooth loss after adjusting for other variables, though the extent of disease (localized versus generalized) did not add significant predictive power in that particular analysis.
Another study confirmed that patients with Stage IV or Grade C lost significantly more teeth during supportive periodontal therapy. It also found that patient adherence to maintenance visits had a major impact on whether the classification’s predictions held true, which makes intuitive sense: even the most aggressive disease can be slowed if someone shows up for their appointments. That study noted, however, that neither the 2018 classification nor the older 1999 system, when used alone, showed particularly high discriminative accuracy for tooth loss, with areas under the curve just under 60% for both. The staging and grading system is better thought of as one input into prognosis rather than a standalone crystal ball.
Systemic Diseases and Risk Modifiers
The 2017 World Workshop made a deliberate decision about how to handle conditions like diabetes and smoking within the classification. Rather than creating separate disease categories (such as “diabetic periodontitis”), the workshop treated these as modifying descriptors that attach to the single periodontitis diagnosis. The reasoning is that diabetes and smoking do not cause a different disease; they alter the severity and progression of the same disease.
Diabetes is the most studied systemic modifier. It affects the inflammatory response, impairs wound healing, and is associated with worse periodontal outcomes at every stage. The grading system captures this by using glycemic control thresholds to shift the grade upward, as described earlier. Smoking operates through a different biological pathway, primarily by reducing blood flow to the gums and suppressing the local immune response, but the classification handles it the same way: as a grade modifier rather than a separate diagnosis.
Beyond these common modifiers, the classification also acknowledges rare systemic conditions that have a major, sometimes devastating, effect on periodontal tissues. Conditions like Papillon-Lefèvre syndrome and leukocyte adhesion deficiency cause severe periodontitis in childhood because the immune system cannot mount an adequate defense against the bacteria in dental plaque. The workshop recommended that when periodontitis is clearly driven by a systemic condition, the systemic disease should be identified as the primary diagnosis with periodontal manifestations listed as a consequence, rather than the other way around.
Acute and Necrotizing Conditions
Not everything fits neatly into the staging-and-grading matrix. The classification also addresses acute periodontal conditions that demand immediate attention: periodontal abscesses, necrotizing periodontal diseases, and endo-periodontal lesions (where infection from the tooth’s root canal system and the surrounding periodontal tissues overlap).
Periodontal abscesses are classified according to their cause, the most common being an abscess forming in a pre-existing deep pocket. Necrotizing periodontal diseases, which cause rapid tissue destruction with characteristic painful ulcers and necrosis of the gum papillae, are strongly linked to the patient’s immune status. They tend to appear in people with compromised immune function, whether from HIV, severe malnutrition, or extreme psychological stress. Endo-periodontal lesions, meanwhile, are classified based on features that directly affect their prognosis and treatment: whether there is a root fracture or perforation, and whether periodontitis is present alongside the endodontic problem.
Peri-Implant Conditions
The 2017 World Workshop did not stop at natural teeth. It also established clear definitions for conditions around dental implants, which had previously been described inconsistently across the literature. Three states are recognized: peri-implant health, peri-implant mucositis, and peri-implantitis.
Peri-implant health is defined by the absence of redness, bleeding on probing, swelling, and pus. One practical complication is that there is no universal range of “healthy” probing depths around implants the way there is around natural teeth, because implant designs and placement depths vary widely. Health can exist even around implants that have already lost some bone support, as long as the soft tissue is currently free of inflammation.
Peri-implant mucositis is the implant equivalent of gingivitis: inflammation of the soft tissue without bone loss. Bleeding on gentle probing is the hallmark sign, and strong experimental evidence points to plaque as the direct cause. Peri-implantitis is the more serious counterpart, involving both soft tissue inflammation and progressive bone loss around the implant. It parallels periodontitis in natural teeth, but the tissue architecture and blood supply around implants differ enough that peri-implantitis tends to progress faster and can be harder to treat. The classification does not apply the same staging-and-grading matrix used for periodontitis to peri-implant disease; instead, it relies on case definitions based on probing depth changes, bleeding, suppuration, and radiographic bone loss relative to baseline.
Mucogingival and Structural Conditions
The classification also updated how gingival recessions and related soft-tissue conditions are described. Under the old system, recession was classified mainly by how much tissue had been lost relative to anatomical landmarks. The 2018 version adds several clinically relevant details: the severity of the recession, the thickness of the remaining gum tissue (sometimes called gingival biotype), whether there is root surface decay or non-carious cervical wear at the exposed root, whether the patient has tooth sensitivity, and whether the patient has an esthetic concern about the appearance.
These additions sound granular, but they matter for treatment decisions. A shallow recession on a thick tissue biotype with no sensitivity and no esthetic complaint can often be monitored. A deep recession on thin tissue with exposed root decay and a patient who is unhappy with how it looks calls for a completely different approach. The updated classification captures these distinctions in a way the old one did not.
Salivary Biomarkers and the Future of Grading
One of the acknowledged limitations of the current grading system is that it leans heavily on indirect evidence, specifically the bone-loss-to-age ratio, to estimate disease progression. Direct evidence like serial radiographs is ideal but rarely available at a first visit. Researchers are actively exploring whether biomarkers in saliva could make grading more objective and less dependent on historical records.
A systematic review identified several promising host-derived salivary markers for periodontitis, including MMP-8 (an enzyme involved in tissue breakdown), IL-1β and IL-6 (inflammatory signaling molecules), and hemoglobin. MMP-8 has attracted particular attention because it reflects active tissue destruction and can correlate with both the staging and grading of periodontitis, with reported sensitivity ranging from about 65 to 87% and specificity from about 48 to 87% depending on the study and the cutoff used. These are not yet part of routine clinical practice, but they represent a plausible path toward chair-side tests that could supplement or refine the current grading criteria.
The grade modifiers themselves are also under scrutiny. A recent review revisited the specific thresholds used for smoking and diabetes, noting that the cutoffs (10 cigarettes per day for the smoking boundary, HbA1c of 7% for diabetes) are drawn from association studies linking these levels to worse periodontal outcomes. Whether these thresholds are optimally placed, or whether finer gradations would improve the system’s predictive accuracy, remains an open question.
Artificial Intelligence and Automated Staging
The staging-and-grading system requires measuring bone levels, identifying the cemento-enamel junction on every tooth, detecting furcation involvement, and synthesizing all of that into a diagnosis. It is time-consuming, and as the reliability studies show, subjective in places. This makes it a natural candidate for artificial intelligence tools.
Several research groups have trained deep-learning models to detect bone levels and the cemento-enamel junction on panoramic radiographs, then use those measurements to assign periodontal stages and grades automatically. One recent system achieved precision of 0.95 to 0.97 and recall of 0.94 to 0.96 for bone level and cemento-enamel junction detection, with an interactive interface that lets clinicians review and adjust the AI’s output. The tool does not replace clinical judgment, particularly for complexity factors that require probing and intraoral examination, but it can handle the radiographic measurements that form the backbone of staging.
A meta-analysis comparing AI diagnostic performance across different imaging types found that panoramic radiographs delivered the highest accuracy for staging classification, at about 89%, along with the highest specificity at roughly 85%. Periapical radiographs, which show individual teeth in greater detail, had higher sensitivity at about 76%, making them better suited for catching early disease. The practical implication is that panoramic images work well for screening and initial staging, while periapical films add value when the question is whether a borderline case has crossed into a more advanced stage.
Communicating the Diagnosis to Patients
A classification system designed by and for specialists does not automatically translate into something patients understand. Research into whether the 2018 system improves patient comprehension of their periodontal diagnosis has produced underwhelming results. While the structured framework gives clinicians a more precise language for talking to each other, its effectiveness as a communication tool for patients remains uncertain.
This gap matters because patient behavior, especially adherence to maintenance visits, significantly influences long-term outcomes. If someone does not understand that their “Stage III, Grade C, generalized” diagnosis means their disease is severe and progressing quickly, they may not appreciate why regular recall appointments are important. Some clinicians have adopted simplified visual aids or analogies (comparing staging to the stages of cancer, for instance) to bridge this gap, but there is no standardized patient-facing translation of the system yet. Given that adherence has been shown to affect whether the classification’s prognostic predictions hold true, this is not a trivial shortcoming.