Bleeding gums, and what they are actually telling you

Gingivitis is inflammation of the gum tissue alone. Remove the biofilm consistently and the tissue returns to health with nothing permanently lost. Periodontitis is what happens when that inflammation has reached the attachment apparatus, the fibres and bone that hold the tooth in place. That loss is not reversed by better brushing, or by any treatment. The distinction matters because it decides everything that follows, and it cannot be made by looking at the gums.

The measurement that separates them is the probing depth combined with where the attachment sits relative to the enamel margin. A deep pocket with intact attachment is swollen gum. A deep pocket with attachment loss is disease that has already taken bone. Two mouths can look identical across a room and sit in different categories on the chart.

Since 2018 periodontitis has been described with two numbers rather than one label. The stage, I to IV, records how much has already been lost and how complex the case is to manage: attachment loss relative to root length, teeth already lost to the disease, pocket depths, furcation involvement, mobility. The grade, A to C, records how fast it appears to be moving, using direct evidence where earlier radiographs exist and indirect evidence where they do not, with smoking and diabetes acting as grade modifiers.^[2]

The practical value of writing both down is that the next examination has something to be compared against. Without a baseline, every visit reports that the gums look inflamed, and nobody can say whether the last two years of treatment held or not.

Non-surgical periodontal therapy means removing calculus and biofilm from the root surfaces below the gum line, under local anaesthesia, with enough time given to each quadrant to actually finish it. An American Dental Association expert panel reviewed 72 randomised trials and found an average clinical attachment gain of about 0.5 mm from scaling and root planing, with a range of adjuncts adding between 0.2 and 0.6 mm on top of that.^[1] Half a millimetre sounds small until you consider that the alternative is continued loss.

What follows is a reassessment, not a booking for the next stage. Sites that closed are recorded as closed. Sites that did not are looked at individually, because a pocket that stays deep after thorough non-surgical treatment usually has a reason: an anatomical groove, a furcation, a residual bony defect, a fractured root. That reason decides whether a surgical stage is worth doing, and it is a different conversation from the first one.

A retrospective analysis compared teeth carrying a residual vertical bony defect against the matching tooth on the other side of the same mouth, in the same patients, through periodontal maintenance. The teeth with defects were about half as likely to hold a closed pocket at the last follow-up, and carried roughly 2.3 times the odds of being lost to periodontitis. Diabetes, stage IV disease, higher mobility and a higher risk score each carried greater odds again.^[4]

The same paper found those teeth could still be maintained long term. That is the honest reading: a residual defect is a reason to watch a specific tooth more closely and to name it in the plan, not a reason to remove it. Odds ratios from one retrospective series describe a group, and no number in that paper describes your tooth.

Active treatment ends. Supportive periodontal care does not. It is a scheduled re-examination at an interval chosen by risk, with the chart repeated, the bleeding sites recorded, and any site that has moved treated at that visit rather than at the next one. In a ten-year follow-up of 108 periodontally compromised patients kept under supportive care, 1.8% of all sites still measured 6 mm or deeper at re-examination.^[6]

What the evidence does not yet settle is which version of supportive care is best. A Cochrane review of randomised trials found no trial comparing supportive care against monitoring alone, and insufficient evidence to prefer one protocol or adjunct over another.^[7] So the interval is set from your own risk profile and the chart, and it is revisited when the chart changes rather than held because it was written once.