Koshy 2005.

Study characteristics
Methods	Study design: RCT with 3‐arm parallel design Recruitment period: unclear Setting: university dental clinic, Japan Number of centres: 1 Funding source: grant from Scientific Society
Participants	Inclusion criteria: diagnosis of chronic periodontitis with PD ≥ 5 mm. All participants in good general health Exclusion criteria: SRP in past 6 months or on antibiotics from 6 months before or during study, smokers, pregnancy, allergic to iodine Age: 34–66 years Sex: 23 F (FMD: 8; FMS: 7; control: 8) and 13 M (FMD: 4; FMS: 5; control: 4) Smokers: 0 Number randomised: 36 (12 per group); all Japanese Number evaluated: 36 (12 per group)
Interventions	Comparison: FMS vs control; FMD vs control; FMS vs FMD FMS group: (FMS + water): FMS 1 session US scaling with water (duration 2–2.5 hours) FMD group: (FMS + povidone): FMS 1 session US scaling with 1% povidone iodine (duration 2–2.5 hours), participants rinsing with CHX 0.05% twice a day for 1 month, tongue brushing Control group: (QMD) QRP 4 sessions US scaling with water at 1‐week intervals (duration 40–50 minutes each) OHI before study start: yes Instruments used: US instruments Time per Q: unclear Maintenance: every month Retreatment: none Duration of study: 6 months
Outcomes	Primary outcome: PPD (6 sites per tooth) Secondary outcomes: PAL, BOP (6 sites per tooth). Manual probe with stent for all measurements Teeth: whole‐mouth recordings (baseline, 1, 3 and 6 months). Data split in single‐/multi‐rooted teeth and initial moderate (PPD 5–6 mm) and deep pockets (PPD > 6 mm) Pocket depth at baseline: moderate (5 to < 7 mm), deep (≥ 7 mm) Outcome time reported: 6 months used Other outcomes: PI, mean pain VAS score (0–10), body temperature, number of analgesics, microbiology
Notes	PAL is equal to RAL.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The random sequence was computer generated, with no stratification or balancing of factors".
Allocation concealment (selection bias)	Low risk	Quote: "The subjects chose a sequentially numbered opaque, sealed envelope, which enclosed the code for the treatment protocol they were to receive. The number of envelopes was same as the number of subjects".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The treatment groups were coded so that only the operator was aware of the protocol and the examiner remained blinded throughout the study".
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants completed study.
Selective reporting (reporting bias)	Low risk	Data reported on all primary and secondary outcomes.
Other bias	Low risk	Baseline balance good for pocket depth. No apparent other biases.