Introduction

Periodontal disease results from inflammation of the supporting structures of the teeth in response to chronic infections caused by various periodontopathic bacteria [1]. The main goals of periodontal therapy are to eliminate bacterial deposits and niches by removing the supragingival and subgingival biofilms [2, 3] and to restore the biological compatibility of periodontally diseased root surfaces for subsequent attachment of periodontal tissues to the treated root surface [4]. Generally, these objectives are achieved by mechanical scaling and root planning (SRP), which consists of hand- or electronic instrumentation of the periodontally affected sites.

Although SRP produces significant clinical improvements in patients with chronic periodontitis (CP), the complete elimination of bacterial deposits can be difficult to accomplish [5]. Indeed, mechanical therapy alone may fail to eliminate pathogenic bacterial niches in the soft tissue and in areas that are inaccessible to periodontal instruments, such as deep pockets, furcation areas, and root depressions [6, 7]. Moreover, the use of SRP in the treatment of CP may result in a moderate and temporary shift in the composition of the microbial flora [8, 9], particularly in deep pockets where periodontopathic bacteria can persist after SRP. This situation can lead to the recolonization of treated sites [10, 11].

To overcome these limitations of conventional mechanical therapy, several adjunctive protocols have been developed. Among these, the use of lasers has been proposed for its bactericidal and detoxification effects and for its capacity to reach sites that conventional mechanical instrumentation cannot [1]. In particular, studies [12, 13] have shown that the application of the diode laser (DL), with a wavelength between 655 and 980 nm, can accelerate wound healing through the facilitation of collagen synthesis, promotion of angiogenesis, and augmentation of growth factor release. Furthermore, DL displays in vitro bactericidal and detoxification effects [14] and can prevent ablation of the root surface [15], which theoretically reduces the risk of normal root tissue removal.

Despite these potential beneficial effects, controversial clinical results for DL have been reported by studies conducted on humans comparing the adjunctive use of DL (SRP + DL) to SRP alone [16, 17]. Furthermore, important issues with paramount clinical implications remain to be defined, such as the effectiveness of adjunctive DL on microbiological outcomes and its clinical safety [17, 18]. Given the contrasting results of the studies and the absence of any previous meta-analyses, there is a great need to assess the literature systematically.

The aim of the present meta-analysis is to evaluate scientific evidence concerning the effectiveness of SRP + DL compared with SRP alone in the treatment of patients affected by CP.

Materials and methods

The present meta-analysis was conducted according to guidelines of the Cochrane Collaboration [19] and Preferred Reporting Items for Systematic Reviews and Meta-analysis [20].

Search

The following databases were searched from their earliest records through 18 March 2012: MEDLINE, Cochrane Controlled Clinical Trial Register, Cochrane Database of Systematic Reviews, CINAHL, Science Direct, ISI Web of Knowledge, and SCOPUS. The following search algorithm was used to explore databases, using Boolean operators and the asterisk symbol (*) as truncation: (“Periodontitis” [Mesh] OR “Chronic Periodontitis” [Mesh] OR “Periodontal Diseases” [Mesh] OR “Periodontal Pocket” [Mesh] OR “Periodontal Attachment Loss” [Mesh] OR “Tooth Mobility” [Mesh] OR periodontitis OR periodontal disease* OR periodontal pocket* OR attachment loss OR alveolar bone loss OR pocket depth OR clinical attachment level OR periodontal non surgical treatment OR periodontal non surgical therapy OR scaling root planing OR dental scaling OR periodontal treatment OR periodontal therapy OR calculus remov* OR calculus debridement OR dental debridement OR periodontal debridement OR “Dental Scaling” [Mesh] OR “Root Planing” [Mesh] OR “Dental Prophylaxis” [Mesh]) AND (“Lasers” [Mesh] OR “Laser Therapy” [Mesh] OR “Laser Therapy, Low-Level” [Mesh] OR “Lasers, Semiconductor” [Mesh] OR diode laser OR semiconductor laser OR low power laser OR gallium aluminum arsenide OR gallium arsenide). In the CINAHL, SCOPUS, ISI Web of Knowledge, and Science Direct databases, the MeSH terms were not used.

In addition, a manual search was performed of issues from the last 15 years of the following journals: Lasers in Medical Science, Lasers in Surgery and Medicine, Photomedicine and Laser Surgery, Journal of Periodontology, International Journal of Periodontics and Restorative Dentistry, Journal of Clinical Periodontology, Journal of Dental Research, Journal of Periodontal Research, Periodontology 2000, Journal of Dentistry, Journal of the American Dental Association, Journal of Clinical Dentistry, and Clinical Oral Investigations. To be as inclusive as possible, no restrictions were applied with regard to the publication year or language of the studies. The references of all selected full-text articles and related reviews were scanned.

Study selection

Screening was performed independently by two blinded reviewers (F.S. and M.S.). Interreviewer reliability in the study selection process was determined by the Cohen k test, assuming an acceptable threshold value of 0.61 [21, 22]. In cases of disagreement regarding the inclusion or exclusion of studies, the issue was discussed until consensus was reached.

Eligibility criteria

The study selection process was performed by two blinded reviewers (R.G. and A.M.) in two phases. In the first phase, the studies were analyzed according to the following inclusion criteria (A): (1) randomized clinical trials (RCTs), (2) studies comparing SRP + DL with SRP ± Placebo DL, and (3) patients with diagnosed CP.

Only studies that met all inclusion criteria in (A) were admitted to the second phase, which consisted of analysis of the preselected studies according to the following exclusion criteria (B): (1) studies not reporting data as mean and standard deviation (SD), (2) inclusion of patients with systemic disease or who were assumed to be taking medications that are known to affect periodontal tissue and/or treatment or who had undergone periodontal therapy within the last month, (3) follow-up of <6 months, (4) duplicate or ancillary studies, and (5) primary outcome of interest not analyzed.

Outcome variables

The primary outcomes were changes in clinical attachment level (CAL) gain (in millimeters) and probing depth (PD) reduction (in millimeters). Secondary outcomes were changes in plaque index (PI), gingival index (GI), bleeding on probing (BOP), microbiological changes, adverse events, and cost/effectiveness ratio. All outcome variables were analyzed at 6 months of follow-up.

Data extraction

Data were collected by two independent reviewers (F.S. and R.G.). The following data were extracted from the included studies: year of publication, country, study design, demographic characteristics of participants, number of smokers per intervention arm, laser parameters, frequency and type of adverse events, and follow-up length. If data were presented both numerically (in tables or text) and graphically (in figures), only numeric data were extracted. The reviewers cross-checked all extracted data. Disagreements were resolved by discussion until consensus was reached.

Assessment of risk of bias

The quality of the methodologies of the included studies (Table 1) was analyzed independently by two blinded reviewers (F.S. and A.M.) according to the revised recommendations of the CONSORT statement [23]. The level of agreement between reviewers was calculated as described above. After determining the score of each study, the overall plausible risk of bias (low, moderate, or high) was estimated for each selected study. A low risk of bias was estimated when all of the criteria were met, a moderate risk was estimated when one or more criteria were partly met, and a high risk of bias was estimated when one or more criteria were not met [19].

Table 1 Categories used to assess the quality of selected studies
Full size table

Quantitative analysis

Measure of effect size

Data were combined for meta-analysis with the RevMan statistical software package, version 5.0 (The Nordic Cochrane Center, Cochrane Collaboration, Copenhagen, Denmark). The effect size was estimated and reported as the mean difference (MD), and the 95 % confidence interval (CI) was calculated. Due to the expected interstudy heterogeneity, a random effect model [24] was used. The pooled effect was considered significant if p was <0.05. Forest plots for each meta-analysis present the raw data (i.e., means, SDs, and sample sizes), point estimates (displayed as blocks), and CIs (displayed as lines) for the chosen effect, heterogeneity statistic (I 2), total number of participants per group, overall average effect (MD and Z-statistics) in the random effect model, and percent weight given to each study.

Heterogeneity

Heterogeneity was assessed by the χ 2-based Q-statistic method and I 2 measurement, with significance indicated by p < 0.1.

Publication bias

Publication bias was investigated for each outcome of interest by two methods. Visual detection was used to analyze the funnel plot [25]. Quantitative analysis was performed by the regression asymmetry test [26] and the trim-and-fill method [27]. Publication bias was assessed with the Stata 10 Intercooled statistical software package (StataCorp LP, College Station, TX).

Results

Search results

A total of 1,501 articles were found through the electronic and manual searches: 865 in MEDLINE, 77 in Cochrane Database of Systematic Reviews, 155 in Cochrane Controlled Clinical Trial Register, 25 in CINHAL, 1 each in Science Direct and SCOPUS, and 377 in ISI Web of Knowledge. After removing duplicates, 1,131 articles qualified for screening, which resulted in the elimination of 1,094 studies via the Abstract and Title. Therefore, 37 articles [4, 14, 16, 17, 28–60] were evaluated for eligibility. A total of 22 studies [4, 16, 17, 29, 30, 35, 36, 38, 39, 41–43, 45–47, 49, 50, 53, 55, 58–60] were excluded for not satisfying one or more inclusion criteria, and 10 studies [32, 37, 40, 44, 48, 51, 52, 54, 56, 57] were excluded for meeting one or more exclusion criteria (Table 2). Finally, five studies [14, 28, 31, 33, 34] were included in the systematic review and meta-analysis (Fig. 1).

Table 2 Studies excluded and reason for exclusion
Full size table
Fig. 1
figure 1

PRISMA flowchart of study selection process

Full size image

Description of included studies

Four [14, 28, 31, 33] of the five included studies adopted a split-mouth design, whereas one study [34] used a parallel design. The included studies used different definitions of CP. The PD threshold value for inclusion of CP patients was PD of ≥5 mm in two studies [14, 28], PD of >4 mm in one study [33], and 4 mm < PD <6 mm in one study [31]. One study [34] included patients with moderate to advanced CP. Two studies [33, 34] reported the inclusion of smokers but only one [34] included a clear definition of smoking status.

All studies used DL, but the applied wavelength varied among included RCTs (range, 808 to 980 nm), as did the fiber tip diameter, tip angulation, and power settings (Table 3). The protocol of laser application also differed among the studies: one study [34] performed three sessions of DL in 1 week; one study [31] performed ten sessions of DL in 5 weeks; and one study [28] performed two sessions of DL in 1 week. In two studies [14, 33], the number of laser application sessions was not clearly stated.

Table 3 Design and demographic characteristics of the included studies
Full size table

Microbiological outcomes were analyzed in three studies [14, 28, 33]. None of these studies reported significant differences between SRP and SRP + DL in terms of the number of subjects positive for the investigated pathogens at 6 months. One study [33] did not find significant differences in the reduction of the total bacterial count. Another study [28] failed to observe any significant differences between the SRP and SRP + DL groups for the CFU of the investigated periodontal pathogens. Only one study [34] analyzed adverse events, reporting that no adverse events related to the application of DL occurred during the study period.

Risk of bias in included studies

The CONSORT-based quality analysis revealed that only two studies [28, 33] were at low risk of bias, whereas three [14, 31, 33] showed a high risk of bias (Table 4). The most frequent methodological shortcoming was related to sample size calculation (criteria A), which was missing in three studies [14, 31, 33]. Methods of randomization (criteria B), the use of masking (criteria F), and adequate statistical analysis (criteria G) were other criteria that were not satisfied by two studies [14, 31]. Two studies [28, 33] satisfied all of the quality analysis criteria.

Table 4 Quality assessment of selected studies prior to and after contact (parentheses) with corresponding authors
Full size table

Effects of intervention

Primary outcomes

No significant differences in PD reduction (MD = 0.10, 95 % CI range: –0.11 to 0.31, p = 0.35) (Fig. 2) or CAL gain (MD = 0.02, 95 % CI range: –0.39 to 0.44, p = 0.91) (Fig. 3) were observed at 6 months between the SRP and SRP + DL groups. No evidence of within- or among-study heterogeneity was observed for PD reduction (χ 2 = 1.67; p = 0.80; and I 2 = 0 %) or CAL gain (χ 2 = 1.72; p = 0.42; and I 2 = 0 %).

Fig. 2
figure 2

Forest plot for PD reduction at 6 months for SRP + DL vs. SRP

Full size image
Fig. 3
figure 3

Forest plot for CAL gain at 6 months for SRP + DL vs. SRP

Full size image

Secondary outcomes

No significant differences in changes in PI [61] (MD = 0.08; 95 % CI range, −0.16 to 0.32; and p = 0.51) (Fig. 4) or GI [62] (MD = 0.07; 95 % CI range, −0.23 to 0.36; and p = 0.66) (Fig. 3) were observed at 6 months between the SRP and SRP + DL groups (Fig. 5). No evidence of within- or among-study heterogeneity was observed for changes in PI (χ 2 = 1.74; p = 0.63; and I 2 = 0 %) or GI (χ 2 = 0.00; p = 0.97; and I 2 = 0 %).

Fig. 4
figure 4

Forest plot for Pl changes at 6 months for SRP + DL vs. SRP

Full size image
Fig. 5
figure 5

Forest plot for GI changes at 6 months for SRP + DL vs. SRP

Full size image

Publication bias

Visual inspection of the funnel plots for PD reduction (Fig. 6), CAL gain, PI changes, and GI changes did not suggest the presence of publication bias. Although the trim and fill analysis indicated that one study was missing for PD reduction (Fig. 7) and PI changes, the Egger’s regression asymmetry test revealed that the differences between the original estimate and the adjusted effect were nonsignificant for primary outcomes and PI changes. The regression asymmetry test could not be performed for GI changes, because only two studies [14, 31] were pooled (Table 5).

Fig. 6
figure 6

Funnel plot for PD reduction outcome at 6 months for SRP + DL vs. SRP

Full size image
Fig. 7
figure 7

Funnel plot for PD reduction outcome at 6 months for SRP + DL vs. SRP, adjusted with the trim-and-fill method

Full size image
Table 5 Quantitative analysis for publication bias assessments
Full size table

Discussion

Summary of main results

The results of the present meta-analysis indicated that the adjunctive use of DL to SRP did not provide additional benefits in terms of the improvement of clinical parameters. These findings are consistent with those provided by a previous systematic review [17], which, however, only included one RCT [48] with a short (3 months) follow-up time. Our findings are also consistent with those achieved by the included studies: four [14, 28, 31, 33] studies failed to observe significant differences at 6 months between SRP and SRP + DL, while only one study observed a significant additional benefit of DL use in the primary outcomes of interest. Furthermore, no study [14, 28, 33] reported an additional positive effect of DL on microbiological outcomes.

Although the I 2 test did not reveal the presence of heterogeneity (i.e., it gave a result of 0 % for all investigated outcomes), important differences in the use of different wavelengths, fiber tip diameters, power settings, laser application times, number of DL sessions, etc., were observed among the included studies. It has been suggested that conflicting results in the literature could be due to the lack of standardization of the reported irradiation parameters and to the inappropriate specification of dosimetry (power, beam area, time, dose, contact, or defocused irradiation mode) [16, 63]. In the present meta-analysis, no sensitivity analysis or meta-regression investigating the effect of different laser parameters on the clinical results could be performed, because only five studies were included. However, given that an optimal dosage has yet to be defined, and that laser parameters are related to the effectiveness of DL [34], future studies should be performed to specify the appropriate laser settings and dosimetry of this method.

The findings of this meta-analysis revealed that relevant differences existed for the presence of an important prognostic factor: namely, smoking. Smoking modifies the results of nonsurgical therapy [64] and wound healing [65]. Indeed, one of the studies in this meta-analysis [28] clearly excluded smokers, whereas another reported that none of the included patients were smokers [31]. In a third study, the inclusion of smokers was unclear [14]. Two studies [33, 34] reported the presence of smokers but only one [34] reported a clear definition of smoking and a stratified analysis. In this latter study, smoking patients receiving SRP + DL showed a greater PD reduction than smokers who received SRP alone. Unfortunately, the study did not indicate whether significant differences were present in the assignment of smokers to the treatment arms.

Interestingly, a wide range of CP definitions, mainly based on PD, were adopted by the included studies. Different levels of CP extension and severity (i.e., moderate, moderate to advanced, severe, etc.) and treatment stages (i.e., untreated patients and maintenance patients) were considered by the included studies. The effectiveness of SRP could depend on PD, because deeper pockets have more potential PD reduction [34]. Accordingly, this issue should be considered when interpreting the findings of the present meta-analysis.

No meta-analysis could be performed with regard to microbiological outcomes. Nevertheless, the studies addressing this issue [14, 28, 33] reported that no significant differences were detected in the number of patients positive for the investigated pathogens, such as Aggregatibacter actinomycetemcomitans and Porphyromonas gingivalis. Future studies should address the microbiological effects of adjunctive DL.

The present meta-analysis utilized rigorous inclusion/exclusion criteria and a wide search strategy with no language restrictions. In addition, publication bias was analyzed and, through qualitative and quantitative methods, was not found to be present. However, despite the adopted methodology, the current findings should be interpreted with caution. Additional studies are needed to assess the effectiveness of SRP + DL compared with SRP alone.

Quality of the evidence

The quality analysis revealed a high risk of bias for three [14, 31, 33] of the five included RCTs. The most frequently unsatisfied criterion was the lack of a sample size calculation. This methodological shortcoming could theoretically have contributed to the low power for three studies [14, 31, 33]. Four [14, 28, 31, 33] of the five included RCTs assumed a split-mouth design, in which the randomization was performed on the basis of the mouth sides rather than on the basis of the patients. This design potentially reduces the error variance of the experiment, assumes a higher statistical power [66], and requires fewer patients for the trial [67]. However, comparisons made on a within-patient basis have important disadvantages, because treatments may affect the experimental site in unpredictable ways (i.e., carryover effects) [17]. Furthermore, a recent review [68] highlighted that split-mouth trials showed deficiencies in reporting and in the application of correct statistical procedures. Therefore, unless a priori knowledge indicates that no carryover effects exist, then the reported estimates of the treatment efficacy should be considered to be biased [67].

Limitations of the meta-analysis

The main limitation of the present meta-analysis was the small number of included RCTs, three of which [14, 31, 33] had a high risk of bias. As a result, the findings of this meta-analysis should be interpreted with caution.

Implications for research

Future well-designed parallel RCTs with large sample sizes should be performed to assess whether the adjunctive application of DL to SRP could provide additional clinical benefit to SRP alone. Appropriate dosimetry and complete laser settings should be reported, and additional outcomes, such as microbiological outcomes, occurrence of adverse events, and the cost/effectiveness ratio of the adjunctive use of DL, should be addressed.

Implications for clinical practice

Based on the findings of the present meta-analysis, no clinical recommendation for the adjunctive use of DL with SRP can be given. Until high-quality RCTs become available, DL + SRP should be considered as an experimental and unpredictable treatment.