The pathogenesis of periodontal disease as a bacterial infection has fallen in and out of favor over time. In the late 19th century, bacteria were considered the obvious cause of periodontal disease while in the early to mid 20th century, occlusion and parafunctional forces were thought to be the predominant primary etiological factors responsible for causing periodontal disease. It was not until the late 1960’s that the role of bacteria as an etiologic factor in periodontal disease was actually accepted. At this time, Loesche (year) presented the non-specific plaque hypothesis stating that periodontal disease is a result of an increase in the quantity of plaque and that all bacteria had the same capacity to cause disease. This theory was challenged by Socransky’s specific-plaque hypothesis (year) in which he stressed that certain species are more involved in periodontal destruction. As periodontal disease was not a classical disease process caused by a single microorganism, Socransky’s group derived a new set of postulates that would help to implicate an organism as being a risk factor for periodontal disease. . Socransky’s postulates included that (i) the bacterium must be found more frequently and in higher numbers in disease sites (ii) can be isolated from diseased sites; (iii) upon elimination of the bacterium, the tissue returns to health ; (iv) the host mount a response against this specific pathogen; (v) the pathogen expresses virulence factor; and (vi) animal studies should prove its pathogenicity and potential for disease.

Over the years, research first focused on the identification of potential pathogens implicated in periodontal diseases. Then, the focus switched to studying these particular pathogens in detail so as to give clues on how to eradicate them effectively from the periodontal tissues. More recently,, molecular methods aimed at identifying bacteria associated with the different forms of the disease are becoming more “broad-spectrum” and science is moving from identifying the presence of known pathogens to once again, probing for unknown pathogens. In this review of post-1996 literature we aim to outline where periodontal research has gone in attempting to further elucidate the role microorganisms play in the pathogenisis of periodontal disease.

Bacteria implicated in periodontal disease:

Chronic Periodontitis:

In a breakthrough study, Socransky’s group (1998) was able to show that certain bacterial complexes associate with each other and that certain complexes are more likely to potentiate disease. A total of 185 healthy or periodontaly diseased patients were probed for 40 species using DNA-DNA hybridization; prevalence, community ordination and association with 0-3mm, 4-6mm and >6mm probing depths were assessed as well as pre- and post-therapy bacterial profiles. 5 clusters with >60% similarity were identified and named “red”, “orange”, “yellow”, “green” and “purple” complexes. The authors were able to show that in deeper sites that bled upon probing, “red complex” species were highly prevalent. Also, the “red complex” organisms were typically present with those from the “orange complex”. The reverse was not necessarily true, and thus the authors hypothesized that community ordination occurs whereby the communities build upon each other. In the healthy subjects, the less virulent “green” and “purple” complexes predominated whereas minimal amounts of “orange” and “red complex” organisms were seen. Ten years later, Haffajee’s group completed a similar study examining the relationship between the 40 bacteria identified via DNA-DNA hybridization. 187 subjects (age 22-74y, including those that were healthy and with periodontitis) provided full mouth supragingival plaque samples of unknown age. 55 provided full mouth samples at 1,2,4 and 7 days post scaling, i.e. short term plaque development. 93 provided long term plaque development samples over 3-24 months post scaling and root planing. DNA-DNA hybridization was used to determine the bacteria in the sample and the data were used to generate similarity coefficientsThe complexes found in the subgingival plaque are similar to those found in supragingival plaque samples. However, in this study, the red complex was changed by the addition of E. nodatum.

In an earlier study by Riviere et al., (1996), the authors evaluated the probability of finding oral treponemes at sites that harbored Pg. Similar to the results of Socransky, deeper periodontal pockets, increased the likelihood of co-infection.

In 2000, Ximenez-Fyvie et al., showed that species present in disease were also present in the healthy state but in lower proportions. Also, supragingival sites harbored the same species as subgingival sites, again in smaller proportions. The authors postulated that this may be a mechanism by which re-infection occurs and these “healthy” sites with “red” complex organisms may be at higher risk for progression. In fact, earlier studies showed that certain species were associated with disease progression. Similarly, Haffajee et al., (2008) showed that bacterial complexes in supragingival plaque were similar to those present subgingivally. In a Korean population, Choi et al., (2000) showed that healthy sites within advanced disease patients had higher prevalence of Aggregatibacter actynomycetemcomitans (Aa), Prevotella micros (Pm), Prevotella intermedia (Pi), Pg and Treponema sp. compared to healthy patients perhaps indicating their higher risk of progression.

Tanner et al., (1998) conducted a longitudinal study using culture and DNA-probes to characterize bacteria in subgingival samples in healthy, gingivitis and active periodontitis sites within a patient. 56 healthy patients were longitudinally monitored for 18 months. Average pocket depth was 2mm and minimal attachment loss was detected. Clinical measurements included clinical attachment levels (CAL) and probing depths (PD) in duplicate and plaque, redness, bleeding on probing (BOP) and subgingival temperature on 6 sites per tooth (all but 8’s). CAL and PD were measured twice within two weeks to assess variation in probing measurements since such small differences in CAL were being measured. Clinical measurements were repeated at 3-month intervals for 12-18 months. Sites were considered active if they showed CAL: 1) greater than that due to probing variations and 2) ³ 1.5mm from baseline. Cluster analysis showed that 8/9 “active” patients had microbial profiles that were similar; higher Bacteroides forsythus (re-named Tanerella forsythesis-Tf), Campylobacter rectus (Cr)Selemonas noxia (Sn), and lower levels of Pm and Capnocytophaga gingivalis (Cg) were seen. This study suggests that TfCr and Sn are “candidates” to initiate periodontal loss. These results were verified for Tf by Tran et al., (2001) as well as Van Winkelhoff et al., (2002). In the Tran study, 44 subjects with low prevalence and severity of periodontal disease were followed over 2-years; some had loss of attachment (how much) while others had gain or both loss and gain of attachment. Patients that had persistent positive tests for Tf had a 5.3 fold increase in loss of attachment compared to those in which Tf was not continuously detected. In the Van Winkelhoff study, 116 periodontitis patients were compared to 94 healthy patients and both Pg and Tf were shown to be associated with destructive disease. The odds of Pg detection in diseased versus non diseased was 12.3 and for Tf was 10.4.

Similarly, when comparing healthy to well-maintained to periodontitis cases, Haffajee et al., (1998) showed that there was an increase in sites colonized with TfTreponema denticola (Td) and Pg in the periodontitis cases when compared to well maintained and healthy cases, again stressing that these pathogens have an etiologic role in chronic periodontitis. (Can you please ck to make sure this summary is correct. Was not written properly and did not have time to check.

When comparing periodontal pathogens from different geographic locations, variation is seen in most studies. Haffajee et al., 2004 looked at 300 adults with chronic periodontitis from America, Chile, Brazil and Sweden and showed that red complex species showed a lot of variation among the four countries. These differences were found after adjustment for age, gender, mean PD and smoking status. Interestingly, the Chilean subjects showed the most prevalence of Pg (mean percent of 11.7%) followed by Americans (5.9%) and Swedes (0.9%). Furthermore, Herrera et al., (2008) compared bacterial profiles of Chileans, Columbians and persons from Spain showing variability in colonization. The Columbian population demonstrated greater severity of periodontitis and they also had higher total counts of bacteria. However, in a study comparing subgingival microbiota in patients from Northern Cameroon, Ali et al., (1997) showed similar bacterial profiles compared to Western populations in both periodontaly diseased and healthy patients. Actinomyces and Streptococci were the most prevalent in healthy subjects while PgTfAaFusobacterium nucleatum (Fn) and Td were most prevalent in diseased cases.

Aggressive Periodontitis:

In a study looking at early-onset periodontitis, Albandar et al., (1997) considered the association of 7 species with its onset and progression. 266 patients were classified as healthy (40.3%), having localized (10.5%) or generalized (25.8%) EOP or incidental EOP (15.7%). Compared to healthy subjects, in the three diseased groups, PgPiFnCr and Td were increased. Among the three diseased groups, the generalized EOP group had the highest levels of these species compared to the other groups. In fact, Pg was 3X higher in generalized vs localized and 16X higher compared to healthy patients. Bacterial samples were again obtained 6 years after the onset of the study from 248 patients. PgTd and Pi were associated with disease progression; Pg was more associated with disease progression than Td and Pi. In terms of Aa, higher levels were seen in EOP patients compared to healthy, but this result was insignificant. Also, twice as many localized EOP patients had Aa than generalized EOP. Overall, this study showed that Pg has an important role in the onset and progression of generalized EOP while Aa was less associated with EOP. However, Aa was more associated with localized vs generalized disease. Faveri et al., (2008) also looked at presence of Aa but in generalized aggressive periodontitis . Using 16S rRNA analysis, they showed that 7 of 10 generalized aggressive periodontitis patients had Aa but that the predominant species was Selemonas thereby indicating that the latter is also associated with the disease. These results are in contrast to a study of an untreated Mexican population with generalized aggressive disease. Comparing the bacterial composition of generalized aggressive patients to chronic periodontitis patients, Ximenez-Fyvie (2006) showed that there was no difference between the bacterial compositions of the aggressive and chronic patients and that Aa did not play a role in either generalized aggressive of chronic periodontitis in this Mexican population. . Furthermore, Herpes viruses have been implicated in aggressive periodontitis. Yapar et al., (2003) conducted a cross-sectional study on 17 patients with aggressive periodontitis and 16 healthy controls. Plaque samples collected from the deepest pockets were positive for cytomegalovirus (HCMV) and Epstein Barr virus (EBV); only one pocket was positive for EBV from healthy subjects. Following treatment, a statistically significant decrease in virus prevalence was seen in diseased patients.

Refractory Periodontitis:

Refractory cases of periodontitis are cases that do not respond to treatment. Colombo et al., (1998) showed that periodontopathogens were equally or less prevalent in refractory cases. An attempt to discriminate refractory patients using clinical and laboratory parameters was also undertaken (Colombo et al., 1999). The 88 patients in these two papers had SCRP +/- surgery and were followed. Patients that had attachment loss or >3 sites with >2.5mm new attachment loss were placed into the refractory category. 27 subjects were refractory and 61 were successfully treated. At the start of this study, plaque, gingival redness, BOP and suppuration were measured, as were PD and CAL. Plaque samples were collected and content of 40 bacterial species (DNA-DNA hybridization) was assessed. Discriminate and logistic regression analysis was used to assess relationships between combinations of factors and development of refractory disease. Age, gender, race and smoking history were not statistically correlated with refractory disease. However, the presence of Selemonas constellatus (Sc) was associated with refractory disease as were Capnocytophaga species (16X). Actinomyces naeslundii (An) associated with improvement after SCRP.

Virulence factors implicated in disease:

Virulence factors are factors that make a bacterium more potent and associated with disease. According to a review by O’Brien and Simpson, there are several common virulence factors between the red complex organisms and Aa. Virulence factors include endotoxin (LPS) present on the outer membrane, proteases able to cleave host material, fimbriae used for motility and invasion, pili used for adhesion. Flagella are reserved for motile organisms and leukotoxin is a virulence factor only seen in Aa. In vitro, molecular and animal model studies have verified the pathogenicity of these virulence factors. Interestingly, clonal heterogeneity exists between different strains of the same species. For example, Pg fimA gene has 5 variants of which type II is the most prevalent and associated with severe forms of the disease. According to Amano et al., (2000), when comparing healthy and periodontaly affected patients, both harbor Pg (36.3% healthy and 87.1% diseased) but the healthy had 76.1% type I fimA while the diseased patients had 66.1% type II fimA. Fujise et al., (2005) evaluated the involvement of the Pg fimAgenotype in treatment outcome after non-surgical therapy and found that persistence of fimA associated with higher prevalence of incurable deep pockets with BOP.

Human Gingival Fibroblasts (HGF) exposed to Pg supernatant in vitro were analyzed for MMP and TIMP expression. Zhou et al., (2006) aimed to extend the understanding of Pg’s role in MMP and TIMP activity thereby shedding some light on Pg-mediated collagen destruction. The authors were able to show that the Pg-supernatant exposed fibroblasts had increased MMP and decreased TIMP protein levels. mRNA expression of these factors was affected. Therefore, Pg has the ability to increase collagen breakdown by affecting MMP-TIMP balance. Pg’sdestructive and inflammatory capacity was verified by exposing venous blood samples from six healthy patients to Pg strains causing a significant increase in pro-inflammatory cytokines in a dose dependent manner. Type II fimA strain induced the most potent inflammatory response (Bodet et al., 2006). High levels of inflammatory mediators and pro-catabolic proteins contribute to the progression of periodontal disease.

Bacterial invasion into gingival cells is also a feature of some periodontopathogens. Colombo et al., (2006) showed that the mean prevalence of Pg and Td in epithelial samples from diseased sites was 43% and 46% respectively. In healthy sites, the frequencies of these organisms in the gingiva were 25 and 28% respectively. This study showed that these species have invasive capacity and sites that are not diseased harbor these organisms as well. It is thus likely that these healthy sites are more likely to progress given that the bacteria are already invaded into the host. Assessing prevalence/invasion in healthy sites from healthy subjects may have been of use to compare.


Proteases are important in the breakdown of host tissue and the potentiation of bacteria. Several proteases have been described in detail. Hamlet et al., (2008) conducted a study on Tf protease gene expression (prtH) in patients with attachment loss compared to healthy controls. RT-PCR analysis showed that higher levels of prtH were predictive of future attachment loss.

Periodontal disease & bacterial levels in Smokers:

Smoking increases the odds of periodontal disease 2.6-6.0 times Should we then expect higher levels of “red complex” species in smokers? Several studies have attempted to answer this question. Haffajee et al., (2001) conducted a retrospective study looking at never, former and current smokers who presented with periodontitis, health, or that were well maintained. Each of the study groups had never smokers, past smokers and current smokers and sites were divided into healthy (<4mm) ordiseased (>4mm). The authors suggest that the greatest difference between the smokers and never/former smokers were in pockets <4mm where they found a higher proportion of red/orange complex groups in smokers.. Pockets >4mm had similar proportions regardless of smoking status. Similarly, Bostrom et al., (2001) compared 33 smokers and 31 non-smokers with moderate to severe periodontitis and showed that both groups had the same frequency of the 12 species investigated including the red complex organisms. Also, the authors noted that differences seen by other groups might be a result of study populations, sampling strategies or detection techniques.

Several groups showed differences in microbial profiles between smokers and non-smokers. Gomes et al., (2006) used RT-PCR to quantify periodontal pathogens in smokers compared to non-smokers and showed that while smokers had significantly higher CAL than non-smokers, there was no difference in the total number of bacteria between the two groups. Smokers did however have higher levels of Dialister pneumosintes (Dp) and Mm. The authors concluded that periodontal health and microbiologic profile is worse in smokers. Van der Velden (2003) conducted a study on 30 smokers and 29 non-smokers after periodontal treatment and showed that while clinical measures improved, smokers showed no significant change in the prevalence of all periodontal pathogens. Pg did however decrease in the smoking group after treatment perhaps attributing to the improvement in clinical features. Regardless, more patients remained culture positive in the smoking group leading to the less favorable outcomes after treatment. Similar to these findings, van Winkelhoff (2001) showed that smokers treated for periodontal disease had a higher frequency of detection of TfPm and Cr. These bacteria were analyzed as a cluster, which was 3.3x more likely to be in smokers than non-smokers (p<0.0001). The results confirmed that the microbiological profile of the subgingival flora differs between untreated and treated smokers and is different from non-smoking patients. Furthermore, Zambon et al., (1996) used immunofluorescence to assess bacteria in current, former and non-smokers. Significantly higher levels of AaTf and Pg were seen in smokers compared to non-smokers. Controlling for periodontal disease severity, former and current smokers were 1.5 X more likely to be infected with Bf compared to non-smokers. Also, current smokers were 3.1X and 2.3X more likely to be infected with Aa and Tf respectively than were former and non-smokers thereby concluding that smoking increases the likelihood of subgingival infection with periodontal pathogens. Shiloah et al., (2000) compared a group of healthy non-smokers to smokers and showed that smoking and age were independent risk factors for infection with eight periodontal pathogens. The amount and duration of smoking determined the presence of target species as patients with a >5 pack-year history were 18 times more likely to harbor these pathogens.

Methods in identifying bacterial species associated with periodontal disease:

Culturing of bacterial samples has been the gold standard method for bacterial analysis. One of its main advantage has traditionally been in assisting with the discovery of new microorganisms. However, there are many problems with this method. For one, only viable bacteria grow, therefore stringent transport and storage conditions are needed. Also, many species in the oral cavity are fastidious and require rigorous growth conditions. The sensitivity of culturing is much lower than newer molecular methods. As a result, species that may associate with periodontal disease avoid detection. Culturing is still useful in assessing bacterial sensitivity to antibiotics and for verifying the presence of known species. It is also the only way by which species can be characterized, and thus is not entirely obsolete.

Immunofluorescence (IF) methods are used to identify known species. These methods depend on specific primary and secondary antibodies grown in rabbits against known pathogens. IF may be more sensitive than culturing methods especially if the bacteria are present at less than threshold levels. IF does not allow for antibiotic sensitivity testing but can be used to validate presence or absence of species seen in culture studies. In a 1998 study by Tanner et al., a rapid-chair-side DNA probe test was compared to IF and culturing. IF was used to “reconcile” results from culturing and the PCR test thereby reducing false +ve’s and –ve’s. Overall, the ability of each test to detect Tf and Pg was similar; sensitivity and specificity increased with “reconciliation” using IF. Also, Noiri et al., (1997) showed that IF has the potential to “locate” the precise position of periodontal pathogens in pockets. Extracted teeth were exposed to IF and showed Pg predominant in deeper pockets and in close approximation to epithelium. Actinomyces viscosus (Av) was seen in shallow, mid and deep pockets and were more densely packed and associated with plaque. IF is also useful in assessing host levels of periodontal pathogen specific antibodies (Mooney et al., 1997).

Molecular analysis for the identification of oral bacteria became popular in the 1990’s with closed-ended techniques such as whole genome probes and PCR methods to indentify specific bacteria. While useful in identifying presence or absence of specific species, whole genome probes allow for cross-reactivity between species, and thus has sensitivity comparable to culture methods. Newer probes against oligonucleotides or species specific regions of the 16S rRNA gene are more specific. Nevertheless, probes provide an analysis of species that are known to be periodontal pathogens. Haffajee in 1992 presented a study using genomic probes against specific bacteria in localized vs generalized disease. The conclusions of this study were that the probes provided a substantial but incomplete estimation of the microbial composition of subgingival plaque. In a study comparing culture methods with species specific probes, Van Steenbergen et al., showed high sensitivity and specificity with a mock community of species (PiPg and Aa); these results did not hold up for clinical samples as the probes were less sensitive and less specific potentially due to cross reactivity against other species. Other studies (Tsai et al., 2003) showed that a DNA probe kit was highly sensitive for Bf but less sensitive and specific for Pg compared to culture methods.

Newer methods using DNA-DNA hybridization showed higher sensitivity and specificity against 40 known species. Cross reactivity of probes was minimal and differences were validated between healthy and diseased patients by the presence of “red” and “orange” complexes in diseased patients. Socransky et al., (2004) showed that this method detected 104 species and that adjusting detection to 103 bacteria can modify its sensitivity. PCR methods have also been extensively used in the qualitative analysis of bacterial presence. Meurman et al., (1997) showed that PCR methods detected Tf in 52 of 58 patients whereas culture detected Tf in 22 of 58 patients. Their conclusions were that PCR methods provide a more accurate representation of bacteria in periodontal pockets. This is likely due to a lower threshold of detection as well as higher specificity and sensitivity of PCR compared to culture. Basic PCR methods do have limitations in that they qualitatively assess bacterial presence. Often, bacterial quantification is necessary and Real-Time (RT-) PCR analysis allows for this. Boutaga et al., (2005) compared RT-PCR to culture methods and showed that culturing resulted in lower prevalence of bacteria in comparison to the molecular methods. The discrepancy was mostly a result of RT-PCR +ve and culture –ve samples likely due to the lower CFU necessary for bacterial detection by RT-PCR.

All in all, molecular methods of the 1990’s and early 2000’s focused on identifying (or verifying) the presence of known periodontal pathogens and did not consider other potentially clinically relevant species. The older molecular methods provided a closed-ended approach to bacterial identification. It was not until the studies of Paster and Dewhirst that the research community understood the breadth of complexity of the “oral microbiome”. Paster et al., 2001 used PCR amplification, cloning and sequencing of 16S rRNA segments to identify 347 species from 9 phyla present in healthy or periodontaly diseased patients. The authors postulated that 68 additional unseen species were present for a total of 415 species. The nine phyla included: Obsidian pool OB11, TM7, Deferribacteres, Spirochaetes, Fusobacteria, Actinobacteria, Firmicutes, Proteobacteria and Bacteroidetes. The authors also compared bacteria between health and disease and were able to show unique species present in disease that were absent in health. Besides known and cultivable species such as Pg and Tf, other novel clones were identified as part of the diseased microbiome. Ten of the phylotypes were yet-to-be cultivated clones that may be clinically relevant to the disease process. In 2010 the Dewhirst group assessed 16S rRNA gene sequences of 36,043 clones and >1000 isolates for the purpose of identifying novel clones and also generating the human oral microbiome 16S rRNA reference set. 98.5% sequence similarity cutoff was used to determine different phylotypes. From the 1000 isolates, 619 taxa were identified and 65.5% of these have been cultured; the 6 major phyla accounted for 96% of these taxa while remaining phyla contained the other 4%. From clonal analysis, another 434 named and un-named taxa were identified in this study. It is important to note that by 2010, there was an increase in the number of phyla from 9 that were described in 2001 to 13 in 2010. It is estimated that >22 Phyla exist in the oral cavity. These studies showed that the current methods of studying bacteriology and periodontal disease had quite a narrow scope and that a greater diversity that what is appreciated exists. Perhaps, the research community is keeping clinically relevant species from identification by the use of closed-ended molecular techniques. In 2003, Kumar et al., used 16S rRNA analysis to show that in the diseased state, novel clones were as prevalent as known pathogens. Also, in this study, Gram +ve species seemed to dominate the diseased state; the opposite is the currently accepted ideology of the pathogenesis of periodontal disease.

In a review of new molecular techniques to identify oral microbes, Pozhitkov et al., (2009) presented a diagram explaining where oral microbiologic studies are heading. The diagram clearly showed that we are moving from single to multiple to whole species identification in the oral cavity; a feat afforded to us with the advent of Next Generation Sequencing (NGS). NGS provides a new mechanism by which identification of novel phyla and species and further study the oral microbiome has become possible. NGS became popular in the mid 2000’s with the advent of clone-independent sequencing methodologies and the realization that whole microbiome identification could be completed in a short period of time. Several 1st, 2nd & 3rd generation NGS exist; 1st generation rely on DNA-amplification and are not conducive to single molecule detection whereas 2nd and 3rd generation are amplification-independent and can detect single molecules. Moreover, 3rd generation systems rely on nanopore technology and are not yet available to the scientific community. There are three 1st generation systems that function slightly differently; Illumina-Solexa is one of the 1st generation NGS systems. In short, DNA is fragmented and adaptor sequences are added to each end allowing for adhesion and immobilization on a microfluidic membrane. DNA templates are then copied via “bridge amplification” whereby the attached DNA fragment bridges over and hybridizes with a matching adaptor. Amplification yields a cluster of molecules with the same sequence; reverse strands are denatured and rinsed away leaving a ssDNA strand that is then sequenced with unique fluorophore-nucleotides. A single nucleotide attaches to its base-pair at a time and the fluorescent light emitted is captured by a charge-coupled device camera thereby identifying the added base. The reaction continues until the sequence is complete. This method allows for rapid sequencing of 16S rRNA genes from a multitude of bacterial samples. Thus far, Lazarevic et al., (2009) have used this method to identify salivary and oropharyngeal species from a small group of patients. They analyzed 1,373,824 sequences identifying 135 genera; Illumina Sequencing achieved a greater depth of coverage than previously used molecular methods. Also, Ahn et al., (2011) compared the use of NGS with the available human oral microbiome microarray (HOMIM) and showed broader spectrum capacity of NGS. NGS was able to detect 77 genera whereas 49 were detected with HOMIM; 37 genera were commonly identified.

Microbial Sampling and Testing:

While the clinical community understands bacteria play a role in the pathogenesis of periodontitis, the use of microbial testing in clinical practice is lacking. In a systematic review of the literature, Listgarten (2003) attempted to find evidence showing that the sampling and identification of periodontal pathogens can be valuable if the information gathered can affect diagnosis and/or treatment planning in a positive way. The literature was scanned for longitudinal studies with reference to outcome variables and bacterial identification. Longitudinal studies were lacking; case-series and case-controls were the majority of the studies. 31 papers were retrieved by hand search and pub-med search; 13 of these reported on the use of microbial analysis to guide patient treatment and 11 papers showed differential response to treatment depending on microbial detection. Due to a lack of controls, Listgarten could not compare microbial analysis (+/-) with diagnosis and treatment outcome. Also, reports were heterogeneous and so, data could not be pooled. Only one publication had a control group (I believe Levy study) and they showed that microbiologic testing modified treatment plans decreasing surgeries and increasing use of antibiotics. Many studies showed that absence of pathogens were indicators of health while presence did not indicate disease. Other studies showed that bacteria above a certain level were associated with worsened clinical parameters/recurrence of disease (don’t remember reading this? Just make sure). Of all the periodontopathogens, Pg was associated with poorer response to treatment and disease recurrence. Overall, Listgarten concluded that there is a lack of randomized controlled trials looking at microbial analysis and its value to diagnosis and treatment planning.