Zhang · Journal of clinical periodontology 2013 · within-subject comparative study · n=30

Incidence and magnitude of bacteraemia caused by flossing and by scaling and root planing.

Cited 84 times in the scientific literature.

Level 3 - non-randomized controlled study

Within-subject comparative clinical study without reported randomization

PubMed 23137266 · doi:10.1111/jcpe.12029 · record verified 2026-08-28

What was done

Thirty patients with chronic periodontitis underwent full-mouth flossing and single-quadrant scaling and root planing (SRP) at separate visits. Blood samples were drawn at baseline, 30 seconds, and 10 minutes after flossing, as well as 5 minutes into SRP and at 30 seconds and 10 minutes post-SRP. Samples were evaluated for total bacteremia and viridans streptococcal bacteremia (VSB) incidence, magnitude (CFU/ml), and associations with clinical parameters.

What was found

Total bacteremia occurred in 30% of patients after flossing and 43.3% after SRP (p = 0.21). Both procedures produced an identical 26.7% incidence of VSB. Mean magnitude of total bacteremia was 7.4 ± 16.2 CFU/ml for flossing versus 2.0 ± 3.4 CFU/ml for SRP (p = 0.2). Mean magnitude of VSB was 1.2 ± 1.6 CFU/ml for flossing versus 0.4 ± 0.2 CFU/ml for SRP (p = 0.09). Viridans streptococci represented 11.4% of flossing isolates and 7.6% of SRP isolates. Gingival inflammation was significantly associated with the incidence of total bacteremia (p = 0.01) and VSB (p = 0.001) following SRP, but no clinical parameters correlated with bacteremia from flossing or bacteremia magnitude from SRP.

Why it matters

Routine oral hygiene procedures like flossing introduce bacteria into the bloodstream at rates and magnitudes comparable to professional scaling and root planing. This supports ongoing evaluations regarding the necessity and scope of antibiotic prophylaxis before dental procedures to prevent infective endocarditis.

Limits

The study had a small sample size of 30 patients, tested only individuals with chronic periodontitis (limiting generalization to healthy populations), did not report whether intervention order was randomized, and assessed surrogate microbiological markers rather than clinical infection endpoints.

Cited by