Periodontal health status and bacteraemia from daily oral activities: systematic review/metaanalysis
This document is a pre‐print version Article Published in Journal of Clinical Periodontology on September 15, 2011
You can find the original article at the following link: http://onlinelibrary.wiley.com/doi/10.1111/j.1600‐051X.2011.01784.x/abstract
TITLE: Periodontal health status and bacteraemia from daily oral activities: systematic review/meta-analysis RUNNING TITLE: Periodontal status and oral bacteraemia AUTHORS: Tomás I1, Diz P1, Tobías A2, Scully C3, Donos N4
1-Special Needs Unit. School of Medicine and Dentistry. Santiago de Compostela University. Santiago de Compostela, Spain 2-Institute of Environmental Assesssment and Water Research. Spanish Council for Scientific Research, Barcelona, Spain 3-Oral Medicine and Special Needs Dentistry Unit. Eastman Dental Institute. London. United Kingdom 4-Periodontology Unit. UCL Eastman Dental Institute. London. United Kingdom . NOTE: The authors declare that they have no conflict of interests
CORRESPONDING AUTHOR Inmaculada Tomás Carmona School of Medicine and Dentistry. Santiago de Compostela University C./ Entrerrios s/n 15872 Santiago de Compostela, Spain Tel: +34 981 563100 ext: 12344 Fax: +34 981 562226 e-mail:
[email protected]
1
ABSTRACT Aim: To investigate the robustness of the observations on the influence of oral hygiene, gingival and periodontal status on the development of bacteraemia from everyday oral activities (B-EOA), analysing its prevalence, duration, magnitude, and bacterial diversity. Material and Methods: This systematic review/meta-analysis complies with PRISMA reporting guidelines. MEDLINE-PubMed, the Cochrane Library and Embase were explored for detecting studies on B-EOA. Results: There were 290 potentially eligible articles, of which 12 papers on B-EOA fulfilled the inclusion criteria and were processed for data extraction (seven on toothbrushing, one on dental flossing and four on chewing). Evaluating the influence of plaque and gingival indices on the prevalence of bacteraemia following toothbrushing, the pooled odds ratios were 2.61 (95% confidence interval [CI]=1.45-4.69) and 2.77 (95% CI=1.50-5.11), respectively. None of 5 studies on bacteraemia following dental flossing and chewing revealed a statistically significant association between oral hygiene, gingival or periodontal status and the development of bacteraemia. Conclusions: Meta-analysis showed that plaque accumulation and gingival inflammation scores significantly increased the prevalence of bacteraemia following toothbrushing. However, systematic review showed no relationship between oral hygiene, gingival and periodontal status and the development of B-chewing, and there is no evidence that gingival and periodontal health status affects B-flossing. KEYWORDS: systemic review; meta-analysis; bacteraemia; everyday oral activity; oral hygiene; gingivitis; periodontitis
2
CLINICAL RELEVANCE Scientific rationale for the study: Everyday oral activities (EOA) can provoke bacteraemia. The clinical relevance of these bacteraemias remains unknown, but EOA may represent a greater risk than dental procedures because of cumulative exposure. The purpose of this study was to investigate the influence of oral hygiene, gingival and periodontal status on the development of bacteraemia due to EAO. Principal findings: Plaque accumulation and gingival inflammation significantly increase the prevalence of bacteraemia from toothbrushing. Oral hygiene, gingival and periodontal status didn´t favour the development of bacteraemia from chewing, and there is no enough evidence that they affect bacteraemia from flossing. Practical implications: Maintenance of optimal oral hygiene may reduce the risk of bacteraemia from toothbrushing.
3
INTRODUCTION Bacteraemia of oral origin is defined as presence of viable oral bacteria in the bloodstream following dental procedures or everyday oral activities (Carmona et al. 2002). A feature that is unique to the oral bacterial biofilm, particularly the subgingival plaque biofilm, is its close proximity to highly vascularised tissues. Any disruption of the natural integrity of the subgingival epithelium, which is at most about 10 cell layers thick, could lead to bacteraemia (Parahitiyawa et al. 2009). This can lead to seeding of microorganisms in different target organs, resulting in focal infections such as infective endocarditis (IE) (Tomás Carmona et al. 2002). In the latest guidelines published by the British Society for Antimicrobial Chemotherapy (Gould et al. 2006), the American Heart Association (AHA) (Wilson et al. 2007), the National Institute for Health and Clinical Excellence (NICE) of the United Kingdom (National Institute for Health and Clinical Excellence 2008), and the European Society of Cardiology (European Society of Cardiology et al. 2009), the emphasis for the cause of IE has shifted from procedure-related bacteraemia to cumulative bacteraemia due to everyday oral activities (B-EOA). NICE (National Institute for Health and Clinical Excellence 2008) considered that it was “biologically implausible” that a dental procedure would lead to a greater risk of IE than regular toothbrushing. B-EOA, such as toothbrushing and chewing, can also play an important role in the development of cardiovascular disease by promoting the formation of atherosclerotic plaques in blood vessel (Drangsholt 1998; Olsen 2008). Evidence from epidemiological and interventional studies suggests that periodontal infections are independently associated with atherosclerotic vascular disease and that periodontal treatment generally results in favourable effects on subclinical markers of this disease (Kebschull M et al. 2010). Oral bacteria have been detected in atherosclerotic plaques, heart valves and aortic aneurysms (Gaetti-Jardim et al. 2009; Nakano et al. 2009; Pucar et al. 2007),
particularly
including
periodontopathogens
such
as
Aggregatibacter
actinomycetemcomitans, Treponema denticola and Porphyromonas gingivalis (Gaetti-Jardim et al. 2009; Nakano et al. 2009; Pucar et al. 2007). Experimental mechanistic in vitro and in vivo studies have established the plausibility of a link between periodontal infections and atherogenesis, and have identified biological pathways through which these effects may be mediated (Kebschull M et al. 2010). Everyday oral activities can provoke bacteraemia possibly because these activities produce small movements of the tooth within the socket, causing intermittent positive and negative pressures that cause micro-organisms to gain access to the bloodstream (Roberts 1999). Studies on B-EOA in both children and adults have focused principally on bacteraemia after toothbrushing (Parahitiyawa et al. 2009; Diz Dios et al. 2011). It has been estimated that the prevalence of bacteraemia from toothbrushing is 0% to 62% (Bhanji et al. 2002; Hartzell et al. 2005) and from dental flossing is 0% to 41% (Carroll & Sebor 1980; Crasta et al. 2009); the maximum risk occurs with the use of oral irrigation devices (range, 27%-50%) (Felix et al. 1971; Berger et al. 1974) and the lowest risk with
4
chewing (range, 0%-17%) (Cobe 1954; Murphy et al. 2006). The maximum duration of B-EOA does not usually exceed 15 minutes and the bacteria most frequently isolated in positive posttoothbrushing blood cultures are Streptococcus spp. (45%), followed by obligate anaerobes (19%) and Staphylococcus spp. (15%) (Diz Dios et al. 2011). In general, B-EOA is of low intensity: the median level in the majority of studies published to date is 0.97 CFU/ml (range, 0.01-32 CFU/ml), although this intensity is considerably higher than that of the baseline bacteraemia reported in those same studies (median baseline bacteraemia in the majority of studies published to date, 0.02 CFU/ml; range, 0.01-0.05 CFU/ml) (Diz Dios et al. 2011). At the present time, the clinical importance of B-EOA is based on the concept of cumulative exposure to bacteraemia (Gutheroth 1984; Roberts 1999). Gutheroth (Gutheroth 1984) estimated a collective exposure of 5370 minutes of bacteraemia in a one-month period in dentate patients as a result of EOA compared with only six minutes of bacteraemia associated with a single tooth extraction. Other authors have estimated that the risk of cumulative bacteraemia with toothbrushing twice a day was 154,219 times greater than with tooth extraction (Roberts 1999). In common inflammatory conditions such as gingivitis and chronic periodontitis, the periodontal vasculature proliferates and dilates, producing an even greater surface area and facilitating the entry of microorganisms into the bloodstream (Parahitiyawa et al. 2009). Consequently, it has been assumed that the patients with periodontal disease may have a much higher risk of developing BEOA and, hence, systemic diseases of oral origin (Li et al. 2000; Gendron et al. 2000; Carmona et al. 2002; Sakamoto et al. 2007; Olsen 2008; Bolger 2009). Even some Expert Committee guidelines concurred with the premise: “Maintenance of optimal oral hygiene and periodontal health may reduce the incidence of B-EOA and is more important than prophylactic antibiotics for a dental procedure to reduce the risk of IE” (Wilson et al. 2007; National Institute for Health and Clinical Excellence 2008). However, there is limited information regarding the impact of oral hygiene, gingival or periodontal status on B-EOA, and findings are inconsistent (Silver et al. 1977; Madsen et al. 1974; Schlein et al. 1991; Lochkart et al. 2009). In an age of evidence-based medicine and dentistry, a systematic review/meta-analysis of the situation is required. The purpose of the present systematic review/meta-analysis was to investigate the robustness of the observations on the influence of oral hygiene, gingival or periodontal status on the development of B-EOA, analysing its prevalence, duration, magnitude and bacterial diversity. One hypothesis was formulated: oral hygiene, gingival or periodontal status represent risk factors for the development of B-EOA. MATERIAL AND METHODS This systematic review/meta-analysis complies with PRISMA reporting guidelines (Liberati et al. 2009).
Literature search
5
Three internet databases were selected to search for suitable papers: the National Library of Medicine, Washington DC, USA (MEDLINE, PubMed), the Cochrane Library and Embase. No date or publication status restrictions were imposed. Only English-language articles were considered (Moher et al. 2000). All databases were researched up to 31 July, 2010. Studies that satisfied the following inclusion criteria were selected: a) Studies on B-EOA b) Studies analysing the influence of oral hygiene, gingival or periodontal status on the development of B-EOA c) Study group(s) formed of subjects aged 15 years or older d) Study group(s) having a minimum of 11 subjects e) Determination of baseline bacteraemia When the title and the abstract of a paper failed to provide sufficient information to assess its eligibility, the full report was obtained. Papers with no abstract but whose title suggested that they were related to the objectives of this review were also selected so that the full text could be screened. Data that were published twice or more were selected only once. Unpublished data described in the full text of published papers were included in the analysis. Reviews containing comments on B-EOA were excluded after screening for additional studies or unpublished data. The search was supplemented by reference checking.
Search strategy
The electronic search was performed in accordance with the most recent recommendations published by the Centre for Reviews and Dissemination, University of York (Centre for Reviews and Dissemination, University of York 2008). The primary outcome variable was the development of B-EOA. The databases were searched using the following strategy and keywords: (Everyday oral activities) Toothbrushing [MeSH Terms] OR toothbrushing [All Fields]; mastication [MeSH Terms] OR mastication [All Fields] OR chewing [All Fields]; mouth [MeSH Terms] OR mouth [All Fields] OR oral [All Fields]; irrigation [MeSH Terms] OR irrigation [All Fields]; equipment and supplies [MeSH Terms] OR equipment [All Fields] AND supplies [All Fields] OR equipment and supplies [All Fields] OR device [All Fields]; dental devices, home care [MeSH Terms] OR dental [All Fields] AND devices [All Fields] AND home [All Fields] AND care [All Fields] OR home care dental devices [All Fields] OR dental [All Fields] AND floss [All Fields] OR dental floss [All Fields]; toothpick [All Fields]; daily [All Fields] AND activities [All Fields]; oral hygiene [MeSH Terms] OR oral [All Fields] AND hygiene [All Fields] OR oral hygiene [All Fields] AND methods [Subheading] OR methods [All
6
Fields] OR procedures [All Fields] OR methods [MeSH Terms] OR procedures [All Fields] AND nursing [Subheading] OR nursing [All Fields] OR home [All Fields] AND care [All Fields] OR home care [All Fields] OR home care services [MeSH Terms] OR home [All Fields] AND care [All Fields] AND services [All Fields]) OR home care services [All Fields] OR home [All Fields] AND care [All Fields] AND methods [Subheading] OR methods [All Fields] OR procedures [All Fields] OR methods [MeSH Terms] OR procedures [All Fields]. AND (Outcome) bacteraemia [All Fields] OR bacteraemia [MeSH Terms] OR bacteraemia [All Fields] OR bacteraemia [MeSH Terms] AND spontaneous [All Fields] AND mouth [MeSH Terms] OR mouth [All Fields] OR oral [All Fields] AND random allocation [MeSH Terms] OR random [All Fields] AND allocation [All Fields] OR random allocation [All Fields] OR random [All Fields].
Screening and selection of papers
Two reviewers (IT and PD) independently inspected each reference identified by the search, scanned the full texts of relevant studies, applied the inclusion criteria and extracted the data. Any disagreement between the reviewers was resolved after additional discussion. The following factors were evaluated in order to investigate the heterogeneity of the studies: a) Study design b) Methodological study quality assessment a)
Study design
Two different types of study design were found: 1) Studies based on an analysis of the influence of oral hygiene, gingival or periodontal status on the development of B-EOA in control subjects (with good gingival and periodontal health) and/or patients with a diagnosis of gingivitis and/or periodontitis (comparing at least two groups). 2) Studies based on an analysis of the influence of oral hygiene, gingival or periodontal status on the development of B-EOA in patients with different oral hygiene, gingival and periodontal index values, with an initially known or unknown periodontal diagnosis. b)
Methodological quality assessment of the studies
The quality assessment of studies was performed in accordance with the most recent recommendations published by the Centre for Reviews and Dissemination, University of York (Centre for Reviews and Dissemination, University of York 2008), evaluating the following parameters:
7
- Description of possible selection bias (inclusion and exclusion criteria) - Description of the specific characteristics associated with performing each activity - Definition of control subjects and patients with gingivitis or periodontitis - Description of the clinical parameters recorded to evaluate oral hygiene, gingival and periodontal status - Microbiological methods used to determine exposure to B-EOA - Statistical methods used to analyse exposure to B-EOA - Reference to the calculation of sample size.
Quantitative outcome analysis
The principal results obtained on the influence of oral hygiene and gingival or periodontal status on the development of B-EOA were analysed in terms of prevalence and/or duration and/or intensity and/or bacterial diversity of the bacteraemia. In the different studies, the post-activity blood samples (excluding the first blood sample drawn after the activity) were considered for the analysis of the duration of the bacteraemia.
Data extraction and statistical analysis
Cohen’s Kappa coefficient was used to measure the agreement between the two readers at each stage. First, the eligibility criteria were applied, and the factors evaluating heterogeneity of the primary outcome studies were then analysed. Initially, extracted data were expressed in the same way that they were presented in the different studies. However, in studies that had the same design but varied in their presentation of results, the results were processed for data extraction.
Meta-analysis
Meta-analysis could be only performed on papers on bacteraemia following toothbrushing: four papers on the influence of the plaque index (PI) (Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Lockhart et al. 2009) and four papers on the influence of the gingival index (GI) (Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Lockhart et al. 2009). In those studies, the influence of plaque and gingival indices was analyzed through a comparison of different cut-off values (Madsen et al compared 0-0.50, 0.51-1.00, 1.01-1.50, 1.51-2.00 and 2.01-2.50; Silver et al, 0-0.75, 0.761.50, 1.51-2.25 and 2.26-3.00; Schlein et al, 0-1.50 and ≥1.51-3.00; and Lockhart et al, 0-1.99 and ≥2.00) (Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Lockhart et al. 2009). In order to compare those studies in the present meta-analysis, the influence of the plaque and gingival indices was analyzed comparing values 0-1.50 vs. ≥1.51, except for the paper by Lockhart et al where the values used were 0-1.99 vs. ≥2.00 (Lockhart et al. 2009). The odds ratios (OR) and 95% confidence intervals (CI) were calculated for the data extracted from each study to estimate the association between these clinical parameters and the prevalence of bacteraemia following
8
toothbrushing, excepting the paper by Lockhart et al (Lockhart et al. 2009), in which the OR and 95% CI were already reported. In consequence, we used the OR as the summary statistic. We used the DerSimonian and Laird Q statistic to test for significance and homogeneity. To quantify the heterogeneity, the proportion of total variance caused by between-study variance (the I2 statistic) was calculated (Takkouche et al. 1999). The study-specific adjusted log ORs were weighted by the inverse of their variance to compute a pooled OR and its 95% CI. On the basis of the results on homogeneity, we calculated and used fixed-effects (F-effects) pooled estimates. The fixed-effects model assumes that there is no between-study variance (i.e. that the results of the studies used in the meta-analysis are homogeneous and their variation is purely a result of sampling). Due to the small number of studies included in the meta-analysis, the evaluation of publication bias was not considered relevant. All analyses were performed using the Stata 10.0 software. The forest plots for each meta-analysis show the following data: a) The raw data (OR and 95% CI) for each arm of each study included b) Point estimates and 95% CI for the selected effect measure, as blocks and lines, respectively c) Heterogeneity statistics (I2) d) The total number of participants per group e) The overall average effect in the F-effect model; and f)
Percent weight given to each study.
RESULTS The combined MEDLINE PubMed, EMBASE and Cochrane CENTRAL searches resulted in 290 potentially eligible articles (Figure 1).These articles were screened by title and abstract for eligibility. The screening resulted in 34 articles that qualified for full-text reading (inter-reader agreement, k=0.90 ± 0.05). Screening of reviews (n=35) for additional studies and unpublished data gave additional information (electronic appendix). After full-text reading, 22 papers were considered to be unsuitable and were therefore excluded (inter-reader agreement, k=1), leaving 12 articles for analysis. The studies that were excluded are listed in electronic appendix, together with the reason for exclusion. Thus 12 articles on B-EOA fulfilled the inclusion criteria and were processed for data extraction: seven articles on bacteraemia from toothbrushing (B-toothbrushing) (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009), one on bacteraemia from dental flossing (B-dental flossing) (Crasta et al. 2009), and the remaining four on
9
bacteraemia from chewing (B-chewing) (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Fine et al. 2010).
Study designs
In four (34%) of the 12 selected studies, study groups of subjects with a specific gingival or periodontal diagnosis were compared (one paper on B-toothbrushing: patients with gingivitis vs. patients with periodontitis; one paper on B-dental flossing: patients with gingivitis vs. patients with periodontitis; two papers on B-chewing: one paper including control subjects vs. patients with gingivitis vs. patients with periodontitis and one paper including patients with gingivitis vs. patients with periodontitis) (Madsen 1974; Murphy et al. 2006; Forner et al. 2006; Crasta et al. 2009). In the remaining eight papers (66%), the study groups comprised subjects with different oral hygiene, gingival and periodontal index values, with known or unknown periodontal diagnosis (six papers on B-toothbrushing and two papers on B-chewing) (Robinson et al. 1950; Sconyers et al. 1973; Silver et al. 1977; Schlein et al. 1991; Hartzell et al. 2005; Kinane et al. 2005; Lockhart et al. 2009; Fine et al. 2010).
Quality assessment
The different aspects of the quality assessment of the selected studies are described in Tables 1 to 4.
Description of possible selection bias (inclusion and exclusion criteria) (Table 1) These criteria were expressed as exclusion criteria, which were grouped into four categories: 1) Systemic infections and other systemic diseases or conditions (principally haematological disorders, cardiovascular diseases, congenital or acquired heart defects, diabetes and pregnancy) were considered exclusion criteria in 10 papers (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). 2) Previous pharmacological treatment (mainly antibiotic and/or immunosuppressive medication, although the period of use considered was heterogeneous, ranging from two weeks to six months) was applied as an exclusion criterion in 10 papers (Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Hartzell et al. 2005; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). 3) Certain oral or dental factors (mainly the presence of at least 10 teeth and/or teeth with caries or pulpal/periapical infections) were considered as exclusion factors in nine papers (Sconyers et al. 1973; Madsen 1974; Kinane et al. 2005; Hartzell et al. 2005; Forner et al. 2006; Murphy et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010).
10
4) Other factors, such as previous problems with venepuncture, were considered in three papers (Kinane et al. 2005; Murphy et al. 2006; Fine et al. 2010).
Description of the specific characteristics associated with the performance of each activity With respect to toothbrushing, the time of brushing varied from one to four minutes (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009). Manual brushing was used in five studies (Madsen 1974; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009), while powered brushing was used in the other two (Sconyers et al. 1973; Silver et al. 1977). The activity was performed by the study subjects in six studies (Sconyers et al. 1973; Madsen, 1974; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009) and in one was performed by the investigator (Silver et al. 1977). In the only article selected that considered dental flossing (Castra et al. 2009), the flossing was performed by the investigator on all teeth in accordance with the 2008 recommendations of the American Dental Association (American Dental Association 2008). With respect to chewing, the duration of the activity varied between two and ten minutes and the products chewed were of different consistency in each study (apple, paraffin, chewing gum) (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Fine et al. 2010).
Definition of control, gingivitis and periodontitis Varying diagnostic criteria were used to define gingival and periodontal health and to establish the diagnoses of gingivitis and periodontitis in the studies in which the patient groups comprised individuals with a specified initial diagnosis of gingival or periodontal disease (Madsen 1974; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Crasta et al 2009; Fine et al 2010). There was only one paper in which no details were given of the clinical and radiographic criteria for the diagnosis of periodontitis (Sconyers et al. 1973).
Description of the clinical parameters recorded to evaluate the oral hygiene, gingival and periodontal status Parameters related to oral hygiene, gingival or periodontal status were recorded in all studies included in the meta-analysis. The parameters most frequently evaluated were degree of gingival inflammation (10 papers) (Robinson et al. 1950; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Hartzell et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010), levels of plaque (nine papers) (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010), pocket depth (six papers) (Sconyers et al. 1973; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009) and degree of tooth mobility (six papers) (Robinson et al. 1950; Sconyers et al. 1973; Kinane et al. 2005; Murphy et al.
11
2006; Lockhart et al. 2009; Crasta et al. 2009). Some of the selected papers included an evaluation of other clinical parameters, such as the degree of gingival recession and loss of clinical attachment, presence of bleeding on probing and after performing a specific activity, and the levels of calculus (Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Castra et al. 2009). In nine of the 12 studies selected, the authors did not comment on the examiner characteristics (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009; Fine et al. 2010). In one study, the clinical parameters were recorded by a single general dentist (Forner et al. 2006) and in the remaining two studies the examiners were calibrated to establish reproducibility (Murphy et al. 2006; Crasta et al. 2009).
Microbiological methods used to measure exposure to B-EOA When studying the influence of oral hygiene, gingival or periodontal status on the prevalence of BEOA, blood samples were drawn at different times, ranging from during the activity (Lockhart et al. 2009) to three minutes after the activity (Schlein et al. 1991); for the analysis of the duration of bacteraemia, samples were drawn between 10 minutes (Crasta et al. 2009) and 60 minutes (Lockhart et al. 2009) after the activity. In all studies included, the detection of bacteria in the bloodstream was performed using conventional microbiological techniques, although we found marked differences in the types of transport and culture media used, incubation atmospheres and incubation times (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). There were three papers in which bacterial quantification was performed using pour-plate (Sconyers et al. 1973; Fine et al. 2010) or lysis-filtration (Forner et al. 2006) techniques. Only Kinane et al (Kinane et al. 2005) made simultaneous use of conventional microbiological techniques and polymerase
chain reaction (PCR) analysis. Phenotypic
identification of the bacterial isolates was performed in 10 papers (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Murphy et al. 2006; Forner et al. 2006; Crasta et al. 2009), while Forner et al (Forner et al. 2006) and Lockhart et al (Lockhart et al. 2009) performed genotypic identification of the bacteria isolated in positive blood cultures.
Statistical methods used to measure exposure to B-EOA In the papers on B-EOA evaluating the influence of oral hygiene, gingival or periodontal status in subjects with a previous diagnosis of a gingival and/or periodontal condition, the analysis of the results was based on a comparison of the percentages of positive and negative post-activity blood
12
cultures between the different groups (Madsen 1974; Murphy et al. 2006; Forner et al. 2006; Crasta et al. 2009). In studies evaluating the influence of more than one clinical parameters, the data were analysed in several ways: comparison of the mean values of the different clinical parameters between subjects with positive and negative blood cultures (Sconyers et al. 1973; Fine et al. 2010), comparison of the percentage of subjects with and without bacteraemia in a given value range of a clinical parameter (Madsen, 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005), calculation of the risk statistics (e.g. OR and 95% CI) (Lockhart et al. 2009), or the use of statistics of association between the clinical parameters and the development of B-EOA (Crasta et al. 2009).
Reference to the sample size calculation Only five of the 12 selected papers (42%) performed a calculation of the sample size (Murphy et al. 2006; Lockhart et al. 2009; Hartzell et al. 2005; Crasta et al. 2009; Fine et al. 2010). RESULTS
Quantitative outcomes analysis
The results of the selected studies are described in Tables 2 to 4. Studies on B-toothbrushing (seven papers) All the studies on B-toothbrushing included in the systematic review evaluated the influence of the oral hygiene, gingival or periodontal status on the prevalence of bacteraemia (Sconyers et al. 1973; Madsen, 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009), two papers analysed the duration (Hartzell et al. 2005; Lockhart et al. 2009), and one paper determined the bacterial diversity (Silver et al. 1977). In five of the seven selected papers the authors found no statistically significant association between the state of oral hygiene, gingival or periodontal status and the development of Btoothbrushing (Sconyers et al. 1973; Madsen, 1974; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005). In patients with PI and GI scores greater than 1.50, Silver et al (Silver et al. 1977) detected B-toothbrushing percentages of 60% and 62%, respectively, compared to 35% and 25%, respectively, in patients with PI and GI scores between 0 and 1.50; those authors also demonstrated that positive post-toothbrushing blood cultures with isolation of three or more different bacterial species were significantly more common in patients with GI scores greater than 1.50 than in those with GI scores between 0 and 1.50 (28% vs. 2%). Lockhart et al (Lockhart et al. 2009), analysing the influence of a number of clinical parameters, found a significant effect of the following parameters on the prevalence and duration of B-toothbrushing: PI ≥2.00 (OR=3.78, 95% CI=1.41-10.16), calculus index ≥2.00 (OR= 4.43, 95% CI=1.60-12.25) and the type of bleeding (generalized vs. localized bleeding) after the activity (OR=7.96, 95% CI=1.49-42.56).
13
Studies on B-dental flossing (one paper) Castra et al (Castra et al. 2009) evaluated the influence of oral hygiene, gingival and periodontal status on the prevalence (at 30 seconds), duration (at 10 minutes) and bacterial diversity of Bdental flossing. They detected no statistically significant association between oral hygiene, gingival or periodontal status and the development of B-dental flossing.
Studies on B-chewing (four papers) All four studies on B-chewing included in the systematic review evaluated the influence of oral hygiene, gingival or periodontal status on the prevalence of B-chewing (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Fine et al. 2010), two papers analysed the duration (Murphy et al. 2006; Forner et al. 2006) and one paper determined the magnitude (Forner et al. 2006). None of the four studies revealed a statistically significant association between oral hygiene, gingival or periodontal status and the development of B-chewing.
Meta-analysis
Individual and pooled OR and 95% CI, the weight (%) of each study and I2-statistic values are shown in Figure 2. The studies by Silver et al (Silver et al. 1977) and Lockhart et al (Lockhart et al. 2009) presented the highest weight in the analysis of PI (F-effects=43.50% and 35.39%, respectively) and of GI (F-effects=48.59% and 39.91%, respectively). Evaluating the influence of the PI (0-1.50 vs. ≥1.51) on the prevalence of bacteraemia following toothbrushing, the F-effects pooled OR was 2.61 (95% CI=1.45-4.69) with statistically nonsignificant heterogeneity (I2=0%, p=0.576). Evaluating the influence of the GI (0-1.50 vs. ≥1.51) on the prevalence of bacteraemia following toothbrushing, the F-effects pooled OR was 2.77 (95% CI=1.50-5.11), with statistically non-significant heterogeneity (I2=17.6%; p=0.303). DISCUSSION
Eligibility criteria
The prevalence of bacteraemia after everyday oral activities (mainly toothbrushing) detected in children appears to differ from that of adults (Bhanji et al. 2002; Kinane et al. 2005); it would therefore be inappropriate to perform a direct comparison of the findings obtained in subjects aged 15 years or older. Baseline blood sample collection (prior to performing any everyday oral activity) is a standard methodological characteristic in studies of B-EOA (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005;
14
Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). In 2004, in a paper published jointly with the Royal College of Physicians (London), the British Cardiac Society (British Cardiac Society Clinical Practice Committee & Royal College of Physicians, 2004) stated that to obtain results on significant oral bacteraemia it is necessary to exclude those reports in which the investigators did not take a pre-activity blood sample. We therefore only selected those studies on B-EOA that included a determination of basal bacteraemia. Some authors investigated the development of B-EOA in a single group of subjects with healthy gingiva and periodontium, in patients with gingivitis or in patients with periodontitis, but they didn’t analize the influence of different oral hygiene, gingival and periodontal index values (Diener 1964; Romans et al. 1971; Berger et al. 1974; Ramadan et al. 1975). We decided to select only the studies on B-EOA that, using the same methodology, simultaneously analysed the influence of oral hygiene, gingival or periodontal status on the development of bacteraemia in subjects with different gingival and/or periodontal diagnosis or those with different oral hygiene, gingival and periodontal index values, as described in Material and Methods section (“Screening and selection of papers”) We recognise the risk of publication bias as the literature search was limited to English-language articles.
Methodological study quality assessment
Almost all papers reported information on the different variables used to assess the methodological quality, although there was marked heterogeneity. With regard to possible selection bias, the parameters most frequently considered were the presence of systemic infections and/or other systemic diseases and previous drug treatment (Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010), followed by certain oral or dental factors (Sconyers et al. 1973; Madsen 1974; Kinane et al. 2005; Hartzell et al. 2005; Forner et al. 2006; Murphy et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). Parahitiyawa et al (Parahitiyawa et al. 2009) stated that the effect of routine personal oral hygiene measures and chewing on the development of bacteraemia is difficult to evaluate, mainly due to the different characteristics associated with the performance of each activity. The studies on Btoothbrushing included differences in brushing time (1-4 minutes), the type of brush used (manual or powered), the use of toothpaste or mouthwashes and whether the subject or investigator performed the activity (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009). Furthermore, in one paper the toothbrushing was complemented by the use of toothpicks (Madsen et al. 1974). With respect to the type of brush used, Bhanji et al (Bhanji et al. 2002) and Misra et al (Misra et al. 2007) demonstrated that powered toothbrushing produced a significantly higher rate of bacteraemia than
15
did manual toothbrushing. In the studies on B-chewing, differences were detected in the chewing time (2-10 minutes) and in the consistency of the material chewed (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Fine et al. 2010). On this subject, a series published by Cobe (Cobe 1954) showed that chewing hard candy provoked a higher incidence of bacteraemia than chewing gum (17.4% vs. 0%). Murphy et al (Murphy et al. 2006) stated more recently that the different consistencies of the various chewing mediums could contribute to the differences in bacteraemia rates reported. Differences were detected in the diagnostic criteria used for the definition of subjects with good gingival and periodontal health, patients with gingivitis, and patients with periodontitis (Madsen 1974; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Crasta et al. 2009; Fine et al. 2010). These discrepancies could be justified by the fact that, to date, there is no generally accepted epidemiological definition for periodontitis (Page & Eke 2007; Savage et al. 2009). The most frequently used parameters for the evaluation of oral hygiene, gingival and periodontal status were the plaque levels and the degrees of gingival inflammation (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Hartzell et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010) and, less frequently, the depth of periodontal pockets and degree of tooth mobility (Robinson et al. 1950; Sconyers et al. 1973; Kinane et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009). However, Lockhart et al (Lockhart et al. 2009) recognized that there is no consensus on which clinical measures of oral hygiene, gingival or periodontal status are most appropriate to study with regard to risk of developing bacteraemia. In 1992, Roberts et al (Roberts et al. 1992) studied the prevalence of post-extraction bacteraemia at different times after completing the operative procedure (at 10, 30, 60, 90, 120, 180, and 600 seconds). They found that the highest prevalence of bacteraemia occurred at 30 seconds after tooth extraction. However, we have found no studies in the literature that have evaluated the optimum time for analysing the prevalence in B-EOA. In the present review, the moment at which the blood sample was taken after performing the activity of daily living was variable, being during the activity (Lockhart et al. 2009) or up to three minutes later (Schlein et al. 1991). Most studies have applied conventional qualitative microbiological techniques for the detection and identification of bacterial isolates, and differences were detected in the techniques used (broth culture, pour-plate, lysis-filtration), transport and culture media, atmosphere, incubation times, and in the characteristics of the phenotypic identification (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). These differences in microbiological methodology were particularly noticeable when we compared studies reported in the period 1950-1977 (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Silver et al. 1977) with those reported in 2005-2010 (Kinane et al. 2005; Forner et al. 2006;
16
Lockhart et al. 2009), detecting greater precision in the detection and identification of bacterial isolates in the papers published more recently. However, some authors have therefore stated that “it is likely that oral bacteria recovered from blood by culture are probably only part of those present there” (Olsen 2008). It should be noted that molecular sequence-based methods have higher sensivities than do culture-based methods (Parahitiyawa et al. 2009). In particular, 16S RNAbased methods have brought renewed interest to this field for the specific detection and identification of oral microorganisms, as shown by the studies performed by Kinane et al (Kinane et al. 2005) and Lockhart et al (Lockhart et al. 2009). Recently it has been demonstrated that realtime PCR and pyrosequencing can identify microorganisms directly from positive blood culture bottles as accurately as culture-based methods (the two techniques were concordant for 97.8% of the bacteria) (Jordan et al. 2009). Hence, it is imperative that these molecular sequence-based methods be validated and used in prospective trials of oral bacteraemia, including B-EOA. In the studies on B-EOA in which different clinical gingival and periodontal parameters were recorded, varying statistical methods were used to measure the exposure to B-EOA. The most widely used was the comparison of the percentage of subjects with and without post-activity bacteraemia for a given value range of a clinical parameter (Madsen et al. 1974; Silver et al. 1977; Schlein et al. 1991, Kinane et al. 2005). Based on the design of studies of B-EOA, it is particularly useful to calculate the risk indices (Rutledge T & Loh 2004); however, the only study in which the OR and 95% CI were calculated was the one published by Lockhart et al (Lockhart et al. 2009). In the present review, only five of the 12 selected papers, all of them published recently, performed a calculation of the sample size (Hartzell et al. 2005; Murphy et al. 2006; Lockhart et al. 2009; Crasta et al. 2009; Fine et al. 2010). In agreement with some authors (Rutledge T & Loh 2004), we recognize that a suitable sample size is an important statistical consideration. It is well known that an apparent lack of statistical significance can often be attributed to design in a study based on a sample that is too small to detect a given level of difference (Rutledge T & Loh 2004). However, this dogma has been challenged by authors who consider that some information is better than none, and that even a small amount of inconclusive information may contribute to a later systematic review (Guyatt et al. 2008). Nevertheless, in our opinion, the results obtained in six of the remaining seven studies (in which no calculation of the sample size was performed) reporting “no statistically significant association between the state of oral hygiene, gingival or periodontal status and the development of B-EOA” should be interpreted with caution (Robinson et al. 1950; Sconyers et al. 1973; Madsen 1974; Schlein et al. 1991; Kinane et al. 2005; Forner et al. 2006). Influence of oral hygiene, gingival and periodontal status on the prevalence of bacteraemia from everyday oral activities Most studies included in this review reported very low percentages of bacteraemia under basal conditions (0%–2%) (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991;
17
Hartzell et al. 2005; Forner et al. 2006; Lockhart et al. 2009; Crasta et al. 2009), although some authors have reported figures of 6% to 8% of positive blood cultures (Robinson et al. 1950; Kinane et al. 2005; Murphy et al. 2006). Toothbrushing is the activity of daily living for which there is greatest evidence of an influence on the prevalence of bacteraemia depending on the state of oral hygiene, gingival or periodontal status (Sconyers et al. 1973; Madsen 1974; Silver et al. 1977; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005; Lockhart et al. 2009). In most papers (five of seven; 71%), the authors found no statistically significant association between the state of oral hygiene, gingival or periodontal status and the prevalence of B-toothbrushing (Sconyers et al. 1973; Madsen 1974; Schlein et al. 1991; Kinane et al. 2005; Hartzell et al. 2005). The results obtained in the present meta-analysis showed a significant influence of the plaque and gingival indices (0-1.50 vs. ≥1.51) on the prevalence of bacteraemia following toothbrushing (OR=2.61, 95% CI=1.45-4.69 and OR=2.77, 95% CI=1.50-5.11, respectively). There are few studies published on the influence of oral hygiene, gingival or periodontal status on the prevalence of bacteraemia after performing other activities of daily living (one paper on Bdental flossing and four papers on B-chewing) (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Castra et al. 2006; Fine et al. 2010). In those five studies there was no statistically significant association between the state of oral hygiene, gingival or periodontal status and the prevalence of B-dental flossing (Castra et al. 2009) or B-chewing (Robinson et al. 1950; Murphy et al. 2006; Forner et al. 2006; Fine et al. 2010). Influence of oral hygiene, gingival and periodontal status on the duration, magnitude and bacterial diversity of bacteraemia from everyday oral activities To date, there are no data indicating that a greater duration or magnitude of bacteraemia is more likely to cause a focal infection of oral origin (Lockhart & Durack 1999; Wilson et al. 2007). However, the bacterial inoculum size in individuals with healthy vs. diseased gingival tissues needs to be explored (Roberts 1999), as a higher magnitude of bacteraemia would take longer to clear from the bloodstream (Crasta et al. 2009). On the other hand, several studies have demonstrated the presence of periodontopathogens in atherosclerotic plaques and aortic aneurysms (Pucar et al. 2007; Gaetti-Jardim et al. 2009; Nakano et al. 2009), which indicates previous episodes of bacteraemia. However, only four studies evaluated the influence of the oral health status on the duration of the B-EOA (Murphy et al. 2006; Forner et al. 2006; Lockhart et al. 2009; Castra et al. 2009). Of these studies, only the one published by Lockhart et al (Lockhart el at. 2009) detected a significant association between PI, the type of bleeding after toothbrushing and the duration of Btoothbrushing. With respect to the magnitude of the B-EOA, only one paper (Forner et al. 2006) analysed the influence of oral hygiene, gingival and periodontal status on the magnitude of B-chewing, finding
18
no relationship. The influence of oral hygiene, gingival and periodontal status on the bacterial diversity of B-EOA was evaluated in only two papers, one on B-toothbrushing (Silver et al. 1977) and one on B-dental flossing (Crasta et al. 2009), with conflicting results. CONCLUSIONS Meta-analysis showed that plaque accumulation and gingival inflammation significantly increase the prevalence of B-toothbrushing. Systematic review showed no relationship between oral hygiene, gingival and periodontal status and the development of B-chewing, and there is no evidence that gingival and periodontal health status affects B-flossing. Further studies are needed to investigate the correlation between gingival and periodontal status and the development of B-EOA, and the analysis of other important parameters such as duration, magnitude and bacterial diversity of B-EOA should be considered.
19
CONFLICT OF INTERESTS The authors declare that they have no conflict of interests
FUNDING SOURCES This work was supported by Xunta de Galicia (Grant PGIDT 08CSA010208PR), Spain.
20
FIGURE 1. Flow-chart outlining the search strategy and results at the different stages Papers retrieved from MEDLINE PubMed research (n=242)
Papers retrieved from Cochrane Library research (n=28)
Papers retrieved from EMBASE research (n=20)
Papers retrieved from combined research (n=290)
Excluded during title/abstract screening (n=241)
Papers retrieved for further evaluation after title/abstract screening (n=69)*
Papers excluded during full-text reading (n=22) (Electronic appendix)
Reviews excluded after screening for additional studies and unpublished data (n=35) (Electronic appendix)
Papers appropriate for full reading and data extraction (n=12) (Tables 1-4)
Papers included in the meta-analysis (n=4) (Figure 2)
* Additional reviews (n=20) were added after reference checking of previously selected reviews.
21
TABLE 1. Inclusion and exclusion criteria applied in the selected studies expressed as exclusion criteria, in reversed chronological order.
General -no good general health (Fine et al.
PREVIOUS PHARMACOLOGICAL TREATMENT Antibiotics -history of administration (Fine et
2010; Schlein et al. 1991; Silver et al. 1977; Madsen 1974)
-in the previous 2 weeks (Lochkart
SYSTEMIC INFECTIONS AND/OR DISEASES OR CONDITIONS
-no medical factors that might have influenced the incidence of bacteraemia (Sconyers et al. 1973) -significant medical problems (Crasta et al. 2009)
-systemic diseases and chronic infections other than periodontitis (Forner et al. 2006)
-poorly controlled systemic disease (Lochkart et al. 2009)
Specified -history of rheumatic fever (Fine et al. 2010; Schlein et al. 1991)
-rheumatic heart disease (Schlein et al. 1991)
-at risk of endocarditis (Crasta et al. 2009; Kinane et al. 2005)
-previous endocarditis (Forner et al. 2006)
-heart defects (Fine et al. 2010; Crasta et al. 2009; Murphy et al. 2006; Schlein et al. 1991) -heart murmur (Fine et al. 2010) -prosthetic heart valves (Forner et al. 2006; Schlein et al. 1991) -sickle cell disease (Schlein et al. 1991)
al, 2010****) et al. 2009)
-in the previous month (Hartzell et al. 2005; Schlein et al. 1991; Silver et al. 1977) -in the previous 3 months (Castra et al. 2009; Murphy et al. 2006) -in the previous 6 months (Forner et al. 2006) -antibiotic prophylaxis needed (Fine et al. 2010; Crasta et al. 2009; Lochkart et al. 2009) -penicillin allergy (Lockhart et al. 2009)*
Anticoagulants -history of administration (Fine et al. 2010)
Immunosuppressants -history of administration (Fine et al. 2010, Murphy et al. 2006)
-in the previous month (Hartzell et al. 2005)
General medication -taking any medication at the time of the study (Kinane et al. 2005; Madsen 1974)
(Schlein et al. 1991)
-any pharmacological treatment other than oral contraceptives
-immunocompromised due to disease or medication (Fine et al.
(Forner et al. 2006) -drug abuse (Kinane et al. 2005)
2010; Crasta et al. 2009; Lochkart et al. 2009; Murphy et al. 2006) -haematological disorders (Fine et al. 2010****; Crasta et al. 2009; Murphy et al. 2006; Kinane et al. 2005; Schlein et al. 1991) -diabetes (Crasta et al. 2009; Murphy et al. 2006; Kinane et al. 2005)
2010)
-arteriosclerotic heart disease
-upper respiratory tract infections (Crasta et al. 2009 ; Murphy et al. 2006) -active viral infection (Lockhart et al. 2009)
-history of infectious disease
-taking medication for treatment of chronic conditions unstable in the previous 3 months (Fine et al.
ORAL OR DENTAL FACTORS
OTHER FACTORS
General -no dental factors that might have influenced the incidence of bacteraemia (Sconyers et al. 1973)
-previous problems with venepuncture (Fine et al.
Specified -less than 20 teeth (Fine et al. 2010; Kinane et al. 2005) -less than 10 teeth (Crasta et al. 2009; Lockhart et al. 2009)
-less than 9 teeth per arch (Sconyers et al. 1973)**
-less than 4 pairs of opposing posterior teeth (Murphy et al. 2006) -incompatible dentition (orthodontic bands, crowned surfaces, parcial dentures or teeth unsuitable for extensive ultrasonic scaling)(Fine et al.
2010; Kinane et al. 2005)
-no palpable vein in the antecubital fossa (Murphy et al. 2006)
-eating or toothbrushing within 1 hour before the study (Lockhart et al. 2009) -adverse events resulting from the use of oral hygiene products (Fine et al. 2010****)
2010****; Kinane et al. 2005***)
-teeth with caries (Fine et al. 2010) -teeth with pulpitis, apical periodontitis (Castra et al. 2009; Forner et al. 2006; Madsen 1974) or retained root fragments (Madsen 1974)
-periodontal disease (Fine et al. 2010)****
-facial cellulitis (Lockhart et al. 2009)
-desquamative oral lesions (Kinane et al. 2005) or oral ulceration (Castra et al. 2009) -history of periodontal treatment in the preceding 12 months (Crasta et al. 2009*****; Murphy et al. 2006)
-history of dental work within 30 days (Hartzell et al. 2005) -allergy to dental products (Kinane et al. 2005)
-participation in a dental plaque study within 30 days (Fine et al. 2010)****
-use of antimicrobial toothpaste, mouthwash or chewing gum on a regular basis (Fine et al. 2010)****
(Kinane et al. 2005) -malignancy (Kinane et al. 2005) -trasplants (Forner et al. 2006) -renal disease (Fine et al. 2010) -orthopaedic implants (Fine et al. 2010) -fever over 100.5 F (Lochkart et al. 2009) -pregnant (Crasta et al. 2009; Forner et al. 2006; Murphy et al. 2006) and/or breast-feeding women (Forner et al. 2006) In the paper published by Robinson et al (Robinson et al. 1950) there were no inclusion or exclusion criteria. *In this study, a group of patients underwent single-tooth extraction with antibiotic prophylaxis. **Except one patient who had five teeth in the maxillary arch. ***In this study, patients underwent ultrasonic scaling. ****This study only included a gingivitis group; the effect of an essential oil antiseptic mouthwash on the reduction of bacteraemia after chewing was analysed. *****This study included a periodontitis group.
22
TABLE 2– Characteristics, results and conclusions of the selected studies on the development of bacteraemia from toothbrushing, in reverse chronological order. TYPE OF SUBJECTS (n) AUTHORS, YEAR
CHARACTERISTICS OF TOOHBRUSHING
DIAGNOSTIC CRITERIA
MICROBIOLOGICAL PROCESSING
REGISTER OF PARAMETERS RELATED 1 TO OHS, GS AND/OR PS
STATISTICAL ANALYSIS
RESULTS
CONCLUSIONS
8
EXAMINER Lockhart et al., 2009
Kinane et al., 2005
Hartzell et al., 2005
New manual toothbrush without dentifrice, brushed by subject (2 min)
Supervised manual, brushed by subject (2 min)
Soft-bristle manual toothbrush with dentifrice, brushed by subject (1 min)
Subject without previous OHS, GS and/or PS diagnosis (98)
PI at 4 s/t (Silness & Löe 1964) CAI at 4 s/t (Ramfjord 1959) GI at 4 s/t (Löe & Silness 1963) PPD at 6 s/t Tooth mobility (Armitage 1990) GB after toothbrushing (yes, no) and type (generalized, localized) No reference about examiner
Sample size calculated
Untreated moderate-severe chronic PG (30)
PPD at 6 s/t BOP inter-proximally Recession at 6 s/t CAL at 6 s/t Tooth mobility
Blood sample of 28 ml collected immediately Aerobic and anaerobic BACTEC bottles BACTEC automated system (for 14 d) Subculture: CB agar (5% CO2 incubation 0 at 37 C for 14 d), FA agar (anaeroincubation at 37 C for 14 d), Phenotypic identification Bacterial detection by PCR (16S rRNA )
No reference about examiner
Sample size not calculated
Subjects without previous OHS, GS and/or PS diagnosis (96)
Blood sample of 20 ml collected at 30 sec and 20 min Aerobic and anaerobic culture bottles Automated system (for 5 d) Subculture and bacterial identification not specified
-ON PREVALENCE (at 30 sec after toothbrushing)
Sample size calculated
Basal=0%; Total G= 0%
Blood sample of 20 ml collected at 3 min 5 Aerobic (TS broth) and anaerobic (Columbia broth) culture bottles. 0 Incubation at 37 C for 7 d Subculture: TS agar (aero- and anaero0 incubation at 37 C for 48 h) Phenotypic identification
-ON PREVALENCE (at 3 min after toothbrushing) Basal=0%; Total G= 25% Values range of clinical parameters (subjects 7 with pos-BC) = PI 0-1.50 (23%); PI ≥1.51 (28%); GI 0-1.50 (29%); GI ≥1.51 (0%)
Non available
At least 1 site with PPD >6 mm in all quadrants
Non available GI using PSR (Nasi 1994) No reference about examiner
Schlein et al., 1991
New manual toothbrush without dentifrice plus rinse, brushed by subject (2 min)
-ON PREVALENCE (during toothbrushing) AND ON DURATION (immediately, 20, 40 and 60 min after toothbrushing)
Blood sample of 20 ml collected during, immediately, at 20 min, 40 min and 60 min Aerobic and anaerobic BACTEC bottles BACTEC automated system 3 Subculture : Blood agar, Choco agar, MacC II agar (aero-incubation); AB agar, 4 C-CNA agar and Blood agar (anaeroincubation) Genotypic identification (16S rRNA) Bacterial quantification by real-time PCR
Subjects without previous OHS, GS and/or PS diagnosis (20) Non available PI (Löe 1967) GI (Löe 1967)
6
Basal= 1% ; Total G (cumulative bacteremia)= 22%
YES, WITH ORAL HYGIENE STATUS AND TYPE OF BLEEDING (on prevalence and duration)
Range values of clinical parameters and incidence of B-toothbrushing (OR, 95% CI)= PI ≥2 (3.78, 1.41-10.16); CAI ≥2 (4.43, 1.60-12.25); Type of bleeding (7.96, 1.49-42.56) No significant association with GI ≥2, bleeding after toothbrushing, PPD and tooth mobility -ON PREVALENCE (immediately after toothbrushing)
NO (on prevalence)
Basal= 6% (culture) and 9% (PCR); PG= 3% (culture) and 13% (PCR) Range values of clinical parameters (subjects 7 with pos-BC)= PPD 3 mm and at least 1 interdental ulceration PI (Silness & Löe 1964) GI (Silness & Löe 1963) No reference about examiner
Sconyers et al., 1973
Powered medium toothbrush without dentifrice (4 min)
PG (30) Clinical and radiographic diagnostic criteria non described PI (Silness & Löe 1964) PPD Bone loss Tooth mobility Number of teeth present
Blood sample of 20 ml collected at 1 min Transport tubes containing serum broth and 1% SPS Culture: NS broth, SN agar, semisolid thioglycollate and Blood agar (aero- and 0 anaero-incubation at 35 C for 7 d) Subculture on solid media Phenotypic identification of streptococci
-ON PREVALENCE (at 30-60 sec after toothbrushing) Basal=2%; Total G= 43% Values range of clinical parameters (subjects 7 with pos-BC) = PI 0-1.50 (35%); PI ≥1.51 (60%); GI 0-1.50 (25%); GI ≥1.51 (62%)
YES, WITH ORAL HYGIENE STATUS AND GINGIVITIS (on prevalence and aetiology)
-ON BACTERIAL DIVERSITY (pos-BC with ≥3 7 bacteria) GI 0-1.50 (2%); GI ≥1.51 (28%) -ON PREVALENCE (at 1 min after toothbrushing) Basal= 0%; GG= 19%; PG= 54% Values range of clinical parameters (subjects 7 with pos-BC) = PI 0-1.50 (33%); PI ≥1.51 (36%); GI 0-1.50 (50%); GI ≥1.51 (50%)
NO (on prevalence)
-ON PREVALENCE (at 1 min after toothbrushing) Basal= 0%; PG= 17% Clinical parameters (mean values of subjects with pos-BC vs. subjects with neg-BC)= PI (1.91 vs. 1.85); PPD (3.71 vs. 3.63); Tooth mobility (1.25 vs. 1.17); Number of teeth (24.80 vs. 25.24)
NO (on prevalence)
Sample size not calculated Blood sample of 10 ml collected at 1 min 0 Aerobic and anaerobic incubation at 37 C (PRAS “E” agar, “E” medium, TS broth) Subculture: Aerobic pour plate (TS agar) Phenotypic identification Sample size not calculated
No reference about examiner OHS= Oral Hygiene Status; GS= Gingival Status; PS= Periodontal Status; PG= Periodontitis Group; PPD= Probing Pocket Depth; GG= Gingivitis Group; PI= Plaque Index; s/t= sites per tooth; CAI= Calculus Index; GI= Gingival Index; GB= Gingival Bleeding; BOP= Bleeding On Probing; CAL= Clinical Attachment Loss; PSR= Periodontal Screening and Recording; Choco= Chocolate; MacC= MacConkey; aero-incubation= plates that were incubated aerobically; AB= Anaerobic Blood; C-CNA= Columbia CNA; anaero-incubation= plates that were incubated anaerobically; 5% CO2-incubation= plates that were incubated at 5% CO2; CB= Columbia Blood; FA=;Fastidious Anaerobe; TS= Trypticase Soy; PRAS= Prereduced anaerobically sterile; BHI= Brain Heart Infusion; SPS= Sodium Polyanethol Sulfonate; NS= Nutrient Serum; SN= Semisolid Nutrient; Total G= Total Group; B-toothbrushing= bacteriemia from toothbrushing; OR= Odds Ratio; CI= Confidence Interval; pos-BC= positive Blood Cultures; neg-BC= negative Blood Cultures. 1- Parameters related to OHS, GS and/or PS were recorded after the clinical diagnosis was established; 2- The modification involved omission of instrumentation within the gingival sulcus; 3- Culture media description was detailed in a previous paper by Lockhart et al (Lockhart el al. 2008); 4- Supplemented with phenylethyl alcohol and kanamycin/vancomycin; 5- A Septi-Chek slide containing 3 agar media (chocolate agar, MacConkey agar and malt agar) was attached to the trypticase soy culture bottle; 6- This prevalence of bacteremiain basal conditions was described in a previous paper by Lockhart et al (Lockhart et al, 2008); 7These results were re-calculated based on those described in the original paper. 8- Conclusions section: “YES” means that there was a statistically significant association between oral hygiene, gingival or periodontal status and bacteremiadevelopment; “NO” means that there wasn´t a statistically significant association between oral hygiene, gingival or periodontal status and bacteremia development;
24
TABLE 3–Characteristics, results and conclusions of the selected studies on the development of bacteraemia from dental flossing, in reverse chronological order. TYPE OF SUBJECTS (n) AUTHORS, YEAR
CHARACTERISTICS OF DENTAL FLOSSING
DIAGNOSTIC CRITERIA
MICROBIOLOGICAL PROCESSING
REGISTER OF PARAMETERS RELATED 1 TO OHS, GS AND/OR PS
STATISTICAL ANALYSIS
RESULTS
CONCLUSIONS
4
EXAMINER Crasta et al., 2009
Recommended flossing action (ADA 2008) Calibrated force of 50 g Action repeated 3 times on each tooth surface (all teeth), performed by investigador
CG (30); Untreated chronic PG (30) ≤20% BOP and ≤10% sites with PPD ≥4 2 mm; at least 2 inter-proximal sites with CAL ≥6 mm with at least 1 of these sites with PPD ≥5 mm PI at 4 s/t (Silness & Löe 1964) MGI at all gingival papillae (Lobene et al. 1964) Papillary bleeding on flossing (Cartes & Barnes 1974) PPD at 6 s/t (mm) BOP at 6 s/t CAL Recession at 6 s/t Tooth mobility (Miller 1938)
Blood sample of 20 ml collected at 30 sec and 10 min BACTEC Aerobic and anaerobic culture bottles BACTEC automated system (for 14 d) Staff at laboratory were blinded Subculture: MacC agar (aero-incubation 0 at 35 C), HB agar and Choco agar (5% 0 CO2-incubation at 35 C), BHI agar 0 (anaero-incubation at 35 C) Phenotypic identification Sample size calculated
-ON PREVALENCE (at 30 sec after dental flossing) Basal= 2%; CG= 41%; PG= 40%
NO (on prevalence, duration and aetiology)
-ON DURATION (at 10 min after dental flossing) Basal= 2%; CG= 14%; PG= 27% No significant association with PI, GI, Bleeding on flossing, PPD, BOP, CAL -ON BACTERIAL DIVERSITY (Streptococcus spp. and periodontopathogenic bacteria pos3 BC) CG= 33% and 13%; PG= 43% and 25%
One examiner (CA and RE) OHS= Oral Hygiene Status; GS= Gingival Status; PS= Periodontal Status; CG= Control Group; PG= Periodontitis Group; BOP= Bleeding On Probing; PPD= Probing Pocket Depth; CAL= Clinical Attachment Loss; PI= Plaque Index; MGI= Marginal Gingival Index; GI= Gingival Index; s/t= sites per tooth; CA= Calibration; RE= Reproducibility; MacC= MacConkey; aero-incubation= plates that were incubated aerobically; HB= Horse Blood; Choco= Chocolate; 5% CO2-incubation= plates that were incubated at 5% CO2; BHI= Brain Heart Infusion; anaero-incubation= plates that were incubated anaerobically; BOP= Bleeding On Probing; pos-BC= positive Blood Cultures. 1- Parameters related to OHS, GS and/or PS were recorded after the clinical diagnosis was established; 2- On non-adjacent teeth and not on the same tooth; 3- These results were re-calculated based on those described in the original paper. 4- Conclusions section: “NO” means that there wasn´t a statistically significant association between oral hygiene, gingival or periodontal status and bacteraemia development.
25
TABLE 4– Characteristics, results and conclusions of the selected studies on the development of bacteraemia from chewing, in reverse chronological order. TYPE OF SUBJECTS (n) AUTHORS, YEAR
CHARACTERISTICS OF CHEWING
DIAGNOSTIC CRITERIA
MICROBIOLOGICAL PROCESSING
REGISTER OF PARAMETERS RELATED 1 TO OHS, GS AND/OR PS
STATISTICAL ANALYSIS
RESULTS
CONCLUSIONS
8
EXAMINER Fine et al., 2010
Three bites of an apple (2 min)
Mild to moderate GG (62) Presence of a PI ≥1.5 and modified GI ≥1.5 PI (Turesky modification of the Quigley Hein Index (Turesky et al. 1970) Modified GI (Lobene et al. 1986) No description of examiner
Blood sample of 2 ml collected at 2 min ± 30 sec Transport tubes containing 1% SPS 2 Direct plating method: TS agar (aeroincubation at 35 ± 2ºC for 1-2 d), SB agar (anaero-incubation at 35 ± 2ºC for 5-7 d) Staff at laboratory were blinded Phenotypic identification not specified Sample size calculated
Murphy et al., 2006
A standardized paraffin (4 min)
Plaque-induced GG (20); Untreated chronic PG (21) Previously described by Mariotti (Mariotti1999); At least 4 sites with PPD and CAL ≥5 mm and radiographic bone loss PI at 2 s/t (Silness & Löe 1964) GI at 2 s/t (Löe & Silness 1963) PPD at 6 s/t BOP at 6 s/t Recession at 6 s/t Tooth mobility (Miller 1938)
3
Blood sample of 20 ml collected at 2-3 min and 5 min BACTEC aerobic and anaerobic culture bottles BACTEC automated system Subculture: MacC agar (aero-incubation at 35ºC), HB agar and Choc agar (5% 4 CO2-incubation at 35ºC), BHI agar (anaero-incubation at 35ºC) Phenotypic identification
-ON PREVALENCE (at 2 min ± 30 sec after 5 chewing) 6 Basal= 50%; GG= 35% Clinical parameters (mean values of subjects with pos-BC vs. subjects with neg-BC)= PI (2.00 vs. 1.88); GI (1.75 vs 1.72)
9
NO (on prevalence)
-ON MAGNITUDE in CFU/ml (at 2-3 min after chewing) Basal= 1-2 6 GG= 19.50-25.70 -ON PREVALENCE (at 2-3 min after chewing) Basal= 7%; GG= 0%; PG= 0%
NO (on prevalence and duration)
-ON DURATION (at 5 min after chewing) Basal= 7%; GG= 0%; PG= 0%
Sample size calculated
One examiner (CA and RE) Forner et al., 2006
3 pieces of chewing gum (10 min)
CG (20); GG (20); PG (20) Distance between CEJ and alveolar bone
