The bleeding site: a multi-level analysis of associated factors

Aim To evaluate the association between the probability of a sulcus/pocket to bleed on probing (BoP) and patient related as well as site-specific characteristics. Methods Data from 88960 sites were retrospectively derived from the clinical record charts of 601 adult patients. BoP (positive/negative) had been recorded at the initial periodontal visit after probing pocket depth (PPD) assessment. To analyse the influence of patient-, tooth- and site-related factors on the probability for a site to be BoP+, a logistic, 3-level model was built with BoP as the binary outcome variable. Results (i) The mean probability to be BoP+ for a site with PPD=3mm was 18%, and the log odds increased by 0.69 for each 1mm increment in PPD; (ii) a significantly higher risk for BoP+ was observed for inter-proximal versus approximal surfaces, posterior teeth versus anterior teeth, females versus males, while a significantly lower risk was observed for smokers versus non-smokers; (iii) when controlling for the significant covariates, different BoP+ probabilities could still be observed among sites in patients with a different susceptibility to BoP. Conclusions The probability of a site to be BoP+ was associated with either site-specific (i.e. PPD, tooth aspect, tooth type) or patient-related factors (i.e. gender, smoking status).