Commentary

The basic goal of periodontal therapy is to help dental patients preserve a functional natural dentition. With significant advances in the fields of genetics, immunology and microbiology we understand better the aetiology and pathogenesis of periodontal disease. Despite these advances, however, we continue to use the same basic therapy: this was even described by Abu'l-Qasim in the first century AD.1 Disruption of the plaque biofilm and removal of bacterially retentive factors remain the key to preventing, and stopping progression of, periodontal infections. Initial therapy, nonsurgical therapy, sanative phase, scaling and root planing, and mechanical debridement are simply different names for this process.

Suvan has presented a thorough review of the published reviews on the subject of nonsurgical periodontal therapy, presenting recommendations for clinical practice and future research based on the available evidence. One of the more interesting things in this paper is not the result of her review; she has, appropriately, probed deeper, examining the value of some statistical methods and questioning the type of outcomes used in clinical trials. As an example, one of the problems with interpreting evidence from periodontal trials prior to the early 1990s is the assumption of independence of data. Each periodontal probing site was seen as independent of all other sites on the same tooth, and in the same mouth, and thus pooled for analysis. Yet, as early as Hirschfeld and Wasserman's study nearly 30 years ago2 we have known that the majority of disease is present in the minority of sites, and in a minority of patients. As a result, pooling of the data from every single site measured in a study overestimates the treatment's true effect. Additionally, some of the earlier studies focused on outcomes using single rooted teeth, thus ignoring the areas where disease is most likely to occur!

Clinical studies have always measured outcomes that are most important to clinicians, and in as quantifiable way as possible. Most patients, however, do not care if the disto-lingual of the lower right second molar decreases by 1.3 mm. They want to know if the treatment will help them keep their teeth longer, prevent them from needing surgery and how much it will cost.

In extrapolating research results we must keep in mind the difference between a treatment that is efficacious, that is, beneficial under controlled circumstances (as in most clinical trials), effectiveness, that is, beneficial under most day-to-day circumstances, and efficiency, in which case the benefits outweigh the risks. Risks can include side effects, as well as the use of time (both the patient's and the clinician's) and costs. Thus, efficiency is the most useful information to help patients and their clinicians make the most appropriate clinical decisions. Unfortunately, few clinical trials address the issues of patient-based outcomes or opportunity costs.

Despite the lack of evidence supporting the effect or efficacy of mechanical nonsurgical periodontal therapy, clinical experience tells us that this is the best single therapy option available. Until such time as there is sound evidence of efficiency, we must continue to use the information we have from clinical research, together with our best clinical judgment (with all its inherent biases) as to when, where, and in whom mechanical therapy is required.