Types of studies

Randomised controlled trials with parallel group design or randomised cross‐over trials with a washout period of 10 days or more.

Clinical trials of mouthrinse efficacy often use a randomised cross‐over design. Cross‐over trials are not considered appropriate when there is likely to be carry‐over of treatment effects across treatment periods. Due to its substantivity; there is concern that chlorhexidine may exert an effect beyond the 3 to 7 day washout period that is commonly used (Newcombe 1992; Newcombe 1995; Gehlen 2000). There is insufficient information in the published literature to accurately determine the length of the washout period that would be adequate to eliminate carry‐over due to chlorhexidine. However, a meta‐analysis of 12 4‐day plaque regrowth studies found evidence of carry‐over from chlorhexidine use after a 3 day washout period but no evidence of carry‐over when the washout period was 10 days (Newcombe 1995). Therefore for the purposes of this review, only studies with a washout period of 10 days or more will be considered for inclusion in the review (the washout period of 10 days or more applies to the cross‐over phase only, not to the study initiation phase).

Types of participants

Individuals of any age, gender or race:

• with gingivitis

• with periodontitis who have undergone treatment for periodontal disease and are in the maintenance phase including individuals who have had periodontal surgery providing that they are capable of performing normal mechanical oral hygiene procedures.

Types of interventions

The following interventions and comparisons, involving chlorhexidine mouthrinse, when used in conjunction with mechanical oral hygiene procedures (i.e. toothbrushing with/without the use of dental floss and/or interdental cleaning aids), will be considered for inclusion in this review:

• chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus mechanical oral hygiene alone

• chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus placebo mouthrinse

• chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus mouthrinses containing other plaque inhibitory, anti‐plaque or anti‐gingivitis agents

• chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus mouthrinses containing different formulations of chlorhexidine (e.g. with/without alcohol, with/without anti‐staining ingredients)

• one chlorhexidine concentration versus another at the same frequency of rinsing

• one frequency of rinsing (e.g. once/twice a day) with chlorhexidine mouthrinse versus another using the same concentration of chlorhexidine rinse.

Studies where chlorhexidine mouthrinse is applied locally e.g. with a brush or subgingival irrigation will be excluded.

The guidelines of the American Dental Association for acceptance of anti‐gingivitis chemotherapeutic agents specify a minimum study duration of 6 months (American Dental Association 2008). However, chlorhexidine mouthrinse is rarely used for a continuous period of 6 months. Therefore, in order to reflect the 'real life' use of chlorhexidine as an adjunct to normal mechanical oral hygiene procedures only studies where the duration of the intervention is 4 weeks or longer will be considered for inclusion in the review. This approach is consistent with the guidelines of the American Dental Association for the acceptance of adjunctive dental therapies for the reduction of plaque and gingivitis (American Dental Association 1997). Studies where the duration of the intervention is less than 4 weeks will be excluded.

Types of outcome measures

Electronic searches

Databases to be searched include:

• Cochrane Oral Health Group's Trials Register (to date)

• The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, current issue)

• MEDLINE via OVID (1950 to the present)

• EMBASE via OVID (1980 to the present)

• CINAHL via EBSCO (1980 to the present).

Searching other resources

The following journals have been identified as being important to be handsearched for this review for the period 1966 (or journal start date if later) to the present.  Where these have not already been searched as part of the Cochrane Journal Handsearching Programme, the journals will be handsearched by the review authors.

• Journal of Clinical Periodontology

• Journal of Periodontology

• Journal of Periodontal Research

• Journal of Dental Research

• Scandinavian Journal of Dental Research (1970 to 1994) / European Journal of Oral Sciences (1995 to date)

• Journal of Disability and Oral Health

• Special Care in Dentistry

• Community Dental Health

• Acta Odontologica Scandinavica.

The reference lists of relevant existing systematic reviews and trials meeting the inclusion criteria for the review will be checked for additional studies.

Selection of studies

The search results from the electronic search, the handsearch, the search of reference lists of relevant existing systematic reviews and the records of unpublished trials will be merged and duplicate records will be removed. Endnote will be used for reference management. The titles and abstracts of the records retrieved by the search will be examined by two review authors independently and in duplicate and obviously irrelevant reports will be removed. Eligibility criteria will be pilot tested on a sample of the reports and will be refined or clarified prior to the application of the eligibility criteria to the search results. Potentially relevant reports will be examined in full text and their compliance with the eligibility criteria will be assessed independently and in duplicate by two review authors. Where information relating to eligibility is not reported by the study authors or is unclear, study authors will be contacted to request the information.

The review authors will not be blinded to the journal name, authors, institution, results or sources of funding. Multiple results of the same study will be identified in order to avoid inclusion of duplicate publications in the analysis. Disagreements about whether a study should be included will be resolved by discussion between the review authors. Where agreement cannot be reached on the inclusion of a study, a third review author will arbitrate to reach agreement.

Data extraction and management

Data from eligible reports will be extracted independently and in duplicate by two review authors using a previously piloted data extraction form. Any discrepancies between review authors in data extracted will be resolved by discussion. Where agreement cannot be reached, a third review author will arbitrate to reach agreement. Where the reported data are unclear, the authors of the studies will be contacted for clarification.

Data extraction will include the following.

• Eligibility of the study for the review and reason for exclusion (if applicable).

• Methodology: Study design, study duration, sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other concerns about bias.

• Participants: Total number at baseline, age, sex, study setting, baseline level of periodontal disease, baseline level of plaque, baseline level of gingivitis, smoker/non‐smoker, inclusion/exclusion criteria for participation in the trial.

• Intervention: Total number of intervention groups, details of test mouthrinse, placebo or comparison mouthrinse, concentration and volume of mouthrinse, rinse frequency and duration, length of intervention, no baseline prophylaxis/pre‐baseline prophylaxis/baseline prophylaxis, details of pre‐baseline/baseline oral hygiene instruction (OHI), supervised or non‐supervised mouthrinsing, details of mechanical oral hygiene, timing of toothbrushing in relation to mouthrinsing, length of washout period for cross‐over trials.

• Outcomes: Gingivitis or dental plaque measured using any appropriate index, details of the indices used, timing of outcome measurement (in particular, the time between cessation of the intervention and outcome measurement), reported adverse effects.

• Results: Number of participants allocated to each intervention/comparison group, number of participants in each group completing the trial, missing participants (total and in each group), reasons for drop out of participants, means and standard deviations of gingival and plaque scores, analysis includes all sites or only sites with inflammation at baseline, intention‐to‐treat analysis, paired/unpaired analysis (cross‐over trials), analysis based on differences between groups or differences between individuals (cross‐over trials), first period data only reported/data from all treatment periods (cross‐over trials).

• Miscellaneous: Funding source, key conclusions of the study authors.

Assessment of risk of bias in included studies

Risk of bias in the included studies will be assessed independently and in duplicate by two review authors using the Cochrane Collaboration's tool for assessing risk of bias as outlined in the Cochrane Handbook for Systematic Reviews of Interventions version 5.0.2 (updated September 2009) (Higgins 2009). Risk of bias will be assessed and judged for six separate domains:

• Sequence generation: Was the allocation sequence adequately generated? (For cross‐over trials: Was the order of receiving treatments randomised?)

• Allocation concealment: Was allocation adequately concealed?

• Blinding of participants, personnel and outcome assessors: Was knowledge of the allocated intervention adequately prevented during the study? Due to the propensity of chlorhexidine to stain the teeth and the oral mucosa, it is recognised that it is not possible to achieve adequate blinding in trials where chlorhexidine is compared to placebo or to agents that do not produce such staining. Therefore such trials must be judged by the authors not to have adequate blinding of participants, personnel or outcome assessors even if they are described as blinded by the study authors.

• Incomplete outcome data: Were incomplete outcome data adequately addressed?

• Selective outcome reporting: Are reports of the study free of suggestion of selective outcome reporting?

• Other sources of bias: Was the study apparently free of other problems that could put it at a high risk of bias? (For cross‐over trials: Can it be assumed that the trial was not biased from carry‐over effects?) Industry sponsorship will be noted here as a potential source of bias.

Each study will receive a judgement of 'Yes' (indicating low risk of bias), 'No' (indicating high risk of bias) or 'Unclear' (indicating either lack of sufficient information to make a judgement or uncertainty over the risk of bias) for each of the six domains. The authors of the included studies will be contacted for missing information if necessary. The overall risk of bias for each included study will be assessed with three possible judgements:

• The study is judged to have a low risk of bias

• The study is judged to have a high risk of bias

• The study is judged to have an unclear risk of bias.

Unit of analysis issues

Poor reporting of methodological issues in the design, conduct and presentation of results in cross‐over trials has been identified (Mills 2009). The data required to include paired analyses from cross‐over trials in a meta‐analysis are often not reported by the study authors. Where such data are missing, the authors of these cross‐over trials will be contacted to request the data required to conduct a paired analysis. Where additional data cannot be obtained from the study authors, consideration will be given to either imputing the missing data to approximate a paired analysis or to analysing the data as if the study was a parallel group trial.

Dealing with missing data

Study authors will be contacted to request missing data where necessary. In situations where the required data cannot be calculated from data reported in the study (see sections 7.7.3.2 to 7.7.3.8 in the Cochrane Handbook for Systematic Reviews of Interventions version 5.0.2 (updated September 2009) (Higgins 2009)) and attempted contact with the study author has failed, missing data may be imputed using methods described in section 16.1.3 in the Cochrane Handbook for Systematic Reviews of Interventions version 5.0.2 (updated September 2009) (Higgins 2009).

Assessment of heterogeneity

If the studies are considered similar enough, study results will be combined using meta‐analysis. Fixed‐effect or random‐effects meta‐analysis will be chosen to combine study results as appropriate based on the Chi² test (a low P‐value < 0.1 will be interpreted as providing evidence of heterogeneity of intervention effects), the I² statistic and the magnitude and direction of the intervention effects. The interpretation of the I² statistic will be based on section 9.5.2 in the Cochrane Handbook for Systematic Reviews of Interventions version 5.0.2 (updated September 2009) (Higgins 2009):

• 0% to 40%: might not be important

• 30% to 60%: may represent moderate heterogeneity

• 50% to 90%: may represent substantial heterogeneity

• 75% to 100%: considerable heterogeneity.

Assessment of reporting biases

Where there are sufficient studies (10 or more), and where it is considered appropriate, publication bias will be assessed using a funnel plot. For continuous outcomes with intervention effects measured using mean differences, the test for funnel plot asymmetry proposed by Egger (Egger 1997) will be used.

Data synthesis

Statistical analysis of the study results will follow the statistical guidance in the Cochrane Handbook for Systematic Reviews of Interventions version 5.0.2 (updated September 2009) (Higgins 2009). Many of the indices used to measure gingivitis and plaque use a short ordinal scale. The data are often presented as continuous data, e.g. a mean and a standard deviation (or standard error), however it is possible that binary data may also be reported. Where meta‐analysis is considered appropriate, data presented as continuous will be summarised using the mean difference or the standardised mean difference and associated 95% confidence intervals. Where binary data are presented, data will be summarised using the pooled risk ratio and associated 95% confidence intervals.

Subgroup analysis and investigation of heterogeneity

Data will be summarised in evidence tables and studies will be grouped separately according to the comparison being made i.e. chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus mechanical oral hygiene alone; chlorhexidine mouthrinse (used at any concentration, volume, frequency or duration of rinsing) versus placebo mouthrinse etc.

Potential sources of heterogeneity will be investigated using subgroup analyses and meta‐regression if sufficient studies are available.

Potential sources of heterogeneity include but are not limited to:

• Child (0 to 18 years) or adult (> 18 years)

• Baseline level of plaque

• Baseline level of gingivitis

• Baseline level of periodontal disease

• Smokers/non‐smokers

• Chlorhexidine concentration

• Volume, duration and frequency of rinsing with chlorhexidine

• Study design (parallel group/cross‐over)

• No baseline prophylaxis/pre‐baseline prophylaxis/baseline prophylaxis

• Pre‐baseline/baseline oral hygiene instruction

• Supervised/unsupervised mouthrinsing

• Study duration

• Time between cessation of the intervention and outcome measurement

• Use of different indices to measure plaque and gingivitis

• Timing of toothbrushing in relation to mouthrinsing

• Industry funding.

Sensitivity analysis

If there are sufficient studies a sensitivity analysis will be conducted to examine the effect of sequence generation, allocation concealment and blinding on the effect estimates. The effect of including unpublished trials will also be examined if there are sufficient studies. The impact of including cross‐over trials in the review and the effect of different washout periods on the effect estimates will be examined if possible.