Lasers Med Sci DOI 10.1007/s10103-014-1661-x
REVIEW ARTICLE
Low-level laser therapy for orthodontic pain: a systematic review F. J. Li & J. Y. Zhang & X. T. Zeng & Y. Guo
Received: 12 February 2014 / Accepted: 11 September 2014 # Springer-Verlag London 2014
Abstract This review aimed to evaluate the clinical outcome of different lasers management on orthodontic pain. Cochrane Library (Issue 7, 2014) and MEDLINE (1966–2014.7) were searched to collect randomized controlled trials on lasers for orthodontic pain. Studies meeting the inclusion criteria were systematically evaluated. The Cochrane Collaboration tools RevMan5.1.7 and GRADEpro 3.6 were used in this systematic review and meta-analysis. As a result, 11 randomized controlled trials (RCTs) studying on low-level laser therapy (LLLT) for orthodontic pain control were included. Metaanalysis and risk of bias assessment were implemented using RevMan5.1.7, and level of evidence assessments was measured by GRADEpro 3.6. In the outcome of the score of the most painful day, the comparison of laser versus placebo (pain associated with tooth movement) demonstrated that LLLT reduced the pain score significantly compared with placebo groups (MD=−4.39, 95 % CI range −5.9–−2.88, P50 %); otherwise, the outcomes were assessed by fixed effects model. For VAS continuous variable, mean difference (MD) with Fig. 1 PRISMA flowchart of the study selection process
Trials searched through MEDLINE n=63
95 % CI was adopted. The statistical significance for the hypothesis test was set at α=0.05. GRADE level of evidence Level of evidence assessments was implemented using GRADEpro 3.6. Factors of downgrade quality of evidence include risk of bias, inconsistency, indirectness, imprecision, and publication bias; factors of upgrade quality of evidence include large-effect, plausible confounding, and doseresponse gradient. Level of evidence would present “high”, “moderate”, “low”, and “very low” by the change of factors of downgrade and upgrade quality of evidence.
Data extraction The included trials were reviewed using a list of predefined questions that concerned the basic information of patients, patients’ inclusion criteria, random group methods, blinding methods, intervention methods, outcome data, and so on. Data extraction was conducted by two independent investigators (F. J. Li and X. T. Zeng). Data sheets were compared and discussed with a third review author (Y. Guo). The authors of the studies included will be contacted for missing information if necessary. To decrease the risk of overly positive answers, we will use open-ended questions when asking trial authors for information about study design and conduct. For example, to obtain information about blinding, a request of the following form might be appropriate: “Please describe all measures used, if any, to ensure blinding of trial participants and key trial personnel from knowledge of which intervention a participant had received.” To obtain information about the randomization process, a request of the following form might be appropriate: “How did you decide which treatment the next Trials searched through Cochrane Library n=15
RCTS (Randomized controlled trials) n=28 Trials unrelated to orthodontic pain n=7 Trials of checking duplicate n=9
RCTs after evaluation of effective n=12 Data duplication n=1 Eligible RCTs for systematic review n=11
76
60
59
88
60
84
37
120 64:56
Turhani et al., 2006
Tortamato et al., 2009
Domínguez et al., 2013
Kim et al., 2013
Nóbrega et al., 2013
Harazaki et al., 1997
Eslamian et al., 2014
Marini et al., 2013
NA not available, VAS visual analog scale
12:25
27:57
NA
23:65
19:40
18:42
30:46
NA
39
Lim et al., 1995
Orthodontic tooth movement
Orthodontic tooth movement
Orthodontic tooth movement
Orthodontic tooth movement Placing separators
6, 24, 30, 72, 96, 120, 144, 168
336
2, 6, 24, 72, 120
Placing separators
Orthodontic tooth movement Placing separators
Placing separators
5 min, 1, 6, 12, 24, 48, Placing separators 72, 96, 120, 144, 168
2, 6, 24, 48, 72, 168
No exact time point
6, 30, 54
0, 24, 48, 72, 96
24, 72, 720
Mean 23.01 12, 24, 36, 48, 72, 96
11–32
11–34
12–26
Mean 22.7
Mean 24.3
Mean 15.9
Mean 23.1
21–24
12–23
G4(23): LED treatment G1 (20): control side and did not receive laser but placebo G2 (20): experimental side and receive laser Within each subject, a multigroup pretest-posttest design (randomly laser treated for 15, 30, 60, and 30 s placebo) was used. G1(38): LLLT (lower level laser), maxilla (19); mandible (11); both jaws (8) G2(38): control (placebo), maxilla (21); mandible (10); both jaws (7) G1 (10): experimental group maxilla (EU) G2 (10): experimental group mandible (EL) G3 (10): blind group simulated laser maxilla (BU) G4 (10): blind group simulated laser mandible (BL) G5 (10): control no laser maxilla (CU) G6 (10): control no laser mandible (CL) In Ovation C brackets (29): one arch was irradiated with the laser; the opposite dental arch was placebo treated. Equilibrium brackets (30): one arch was irradiated with the laser; the opposite dental arch was placebo treated. Laser irradiation group (28) LED placebo group (30) Control group (30) Intervention group (30): laser irradiation Control group (30): placebo infrared light radiation Pseudo-irradiation group Blank control group The choice of the first molar to receive a laser irradiation or to be a placebo was determined randomly. U group (60): laser group (20), placebo group (20), control group (20) L group (60): laser group (20), placebo group (20), control group (20)
G2
G3(12): laser treatment
G1 (17): no treatment G2 (38): placebo group
8:12
Placing separators
Doshi-Mehta et al., 2012 20
2, 24, 48, 72, 96, 120
39:16
55
Esper et al., 2011
24.1±8.1
No. Gender (M:F) Age (years) Pain evaluation time (h) Pain induction methods Group methods (patients)
Study
Table 2 Basic information of included studies
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
Pain measure
Lasers Med Sci
InGaAlP
Semiconductor (aluminum gallium arsenide) diode Class 3B Ga-As-A1 diode
Esper et al., 2011
Doshi-Mchta et al., 2012 Lim HM et al., 1995
Twin Flex Evolution®, MMOptics, Classe 3b, registry ANVISA 80051420014 LA3D0001.1; LAMBDA S.p.A., Vicenza, Italy
Model (manufacturer)
30
25
Treatment time per point(s)
ED energy density
P-Laser System International, Egedalsvej, Vekso Each subject being exposed to three treatment time 15, 30, and 60 s Turhani et al., 2006 Semiconductor laser Mini Laser 2075 [dental version], Helbo 30 Photodynamic Systems GmbH & Co KG, Linz, Austria Tortamato et al., 2009 Class 3B gallium-arsenic-aluminum BioWave LLLT, Kondortech, São Carlos, Brazil 16 diode laser Domínguez et al., 2013 GaAlAs laser Photon Lase II® DMC Equipamentos, 44 s/arch São Carlos, Brazil Kim et al., 2013 semiconductor laser device with an Not reported 30 AlGaInP diode Nóbrega et al., 2013 AlGaAs device (class 3B) Kondorthec—São Carlos, Brazil 25 Harazaki et al., 1997 He–Ne gas laser Not reported 1 min/tooth, 12–24 min/patient Eslamian et al., 2014 Ga-Al-As laser DMC Equipamentos, São Carlos, Brazil 20 Marini et al., 2013 GaAs diode laser Lumix 2 HFPL, Fisioline, Verduno, Italy 34
Laser type
Study
Table 3 Technical features of laser for pain used in the included trials
800
660
Wave length (nm)
0.5 80
Not reported Not reported 1or 2 5 Not reported Not reported 2 Not reported Not reported 10.88
30 100 6 40.6 Not reported 100 160
Not reported
4.8
2.25
4.2
75
810 910
830 632.8
635
830
830
670
0.45, 0.95, 1.8 830
0.042
0.735
0.9, 1.79, 3.58
5.0
4.0
Total dose per tooth (J)
30
0.7
30
Output power ED (J/cm2) (mW)
Lasers Med Sci
Lasers Med Sci
patient should get?” More focused questions can then be asked to clarify the uncertainties remained.
Results Search results Figure 1 describes the procedure of selecting eligible trials. After review of the titles and abstracts, one trial [43] was excluded due to data duplication with authors’ another trial [21]. Finally, eleven studies [21, 23, 24, 30, 38–45] were included. Description of the included studies Eleven RCTs studied on the low-level laser for orthodontic pain were included. The eleven studies were conducted in Brazil, India, Japan, Korea, Iran, Singapore, Austria, and Colombia, respectively. All of the 11 trials used the visual analog scale (VAS) for the evaluation of pain intensity, which was considered as an effective method to balance the change of pain scale after the usage of LLLT. A 10 cm VAS tool with graded linear horizontal scale was used to assess the pain, which was a method previously used in many literatures, and had been shown to be a reliable and accurate tool in evaluating subjective experiences, such as pain, to meaningfully compare groups or individuals. This method provides additional variety in responses and gives a more sensitive and accurate representation of pain intensity than do descriptive pain scales [46]. So, we extracted the data on the VAS to describe the detailed application effect of different LLLT system. A brief qualitative analysis of the evidence was provided in the narrative forms (Tables 2 and 3).
Fig. 2 Risk of bias graph: review authors’ judgments about each risk of bias item presented as percentages across all included studies
patients knew whether LLLT or placebo had been given in this trial. Five trials (Turhani D 2006, Domínguez A 2013, Kim WT 2013, Eslamian L 2014, and Gauri Doshi-Mehta 2012) used single-blind setting; the operator who applied the laser treatment and the placebo could distinguish between them, but the patients were blinded to the difference. One trial (Esper MA 2011) did not describe the double-blind setting clearly. None of the 11 trials stated the existence of selective reporting clearly, so they were considered as unclear in this item. In one trial (LIM HM 1995), dental students were selected for the study, which reduced the emotive and attitudinal variables of pain responses and thus provided a more objective assessment of the treatment. So, the conclusions drawn from this sample might not fully apply to the general orthodontic patient
Results of methodological quality assessment All the 11 studies were presented as randomized. However, one trial (LIM HM 1995) revealed that the sequence of application of the four treatment parameters was randomized among the subjects by the Latin square method. A randomly assigned incomplete block split-mouth design (Doshi-Mehta G 2012) was used to prevent interindividual biologic variation. Two trials’ randomization procedures were performed by using a computer program (Domínguez A 2013, Nóbrega C 2013). One trial’s randomization procedures were performed by using a coin flip and envelope containing the number of the group that each subject was assigned to (Marini I 2013). The purpose of these procedures was to ensure maximum allocation concealment. Double blinding was described in five trials (Harazaki M 1997, LIM HM 1995, Nóbrega C 2013, Marini I 2013, and Tortamano A 2009); neither the personnel nor the
Fig. 3 Risk of bias summary: review authors’ judgments about each risk of bias item for each included study
Lasers Med Sci Fig. 4 Forest plot of comparison: laser versus control. Outcome: score of the most painful day— pain associated with placing separators
population, and we considered it as a potential source of bias. (See Figs. 2 and 3 for overview). Effect of LLLT for orthodontic pain The score for the most painful day Continuous data for “the score for the most painful day” was available in five RCTs [24, 38, 39, 41, 44]. Depending on the difference of pain induction, we separate the articles into two groups: first, those of pain associated with tooth movement and second, those of pain associated with placing separators and then evaluate the articles. The results of meta-analysis and GRADE level of evidence were shown in Figs. 4, 5, 6, and 7 and Tables 4, 5, 6, and 7. Figure 4 showed that the score for the most painful day (pain associated with placing separators)
Fig. 5 Forest plot of comparison: laser versus placebo. Outcome: score of the most painful day— pain associated with tooth movement
Fig. 6 Forest plot of comparison: laser versus placebo. Outcome: score of the most painful day— pain associated with placing separators
Fig. 7 Forest plot of comparison: laser versus LED. Outcome: score of the most painful day—pain associated with placing separators
wasn’t significantly reduced in the laser group (MD=−1.68, 95 % CI range −4.21–0.86, P=0.19), however with high heterogeneity (χ2 =14.04, P=0.0002, I2 =93 %). Table 4 showed the GRADE level of evidence of comparison: laser versus control (pain associated with placing separators) with a very low quality for five downgrade factors. In the comparison of laser versus placebo (pain associated with tooth movement), all two trials showed reduced pain score (Fig. 5) (MD = −4.39, 95 % CI range −5.9– −2.88, P
REVIEW ARTICLE
Low-level laser therapy for orthodontic pain: a systematic review F. J. Li & J. Y. Zhang & X. T. Zeng & Y. Guo
Received: 12 February 2014 / Accepted: 11 September 2014 # Springer-Verlag London 2014
Abstract This review aimed to evaluate the clinical outcome of different lasers management on orthodontic pain. Cochrane Library (Issue 7, 2014) and MEDLINE (1966–2014.7) were searched to collect randomized controlled trials on lasers for orthodontic pain. Studies meeting the inclusion criteria were systematically evaluated. The Cochrane Collaboration tools RevMan5.1.7 and GRADEpro 3.6 were used in this systematic review and meta-analysis. As a result, 11 randomized controlled trials (RCTs) studying on low-level laser therapy (LLLT) for orthodontic pain control were included. Metaanalysis and risk of bias assessment were implemented using RevMan5.1.7, and level of evidence assessments was measured by GRADEpro 3.6. In the outcome of the score of the most painful day, the comparison of laser versus placebo (pain associated with tooth movement) demonstrated that LLLT reduced the pain score significantly compared with placebo groups (MD=−4.39, 95 % CI range −5.9–−2.88, P50 %); otherwise, the outcomes were assessed by fixed effects model. For VAS continuous variable, mean difference (MD) with Fig. 1 PRISMA flowchart of the study selection process
Trials searched through MEDLINE n=63
95 % CI was adopted. The statistical significance for the hypothesis test was set at α=0.05. GRADE level of evidence Level of evidence assessments was implemented using GRADEpro 3.6. Factors of downgrade quality of evidence include risk of bias, inconsistency, indirectness, imprecision, and publication bias; factors of upgrade quality of evidence include large-effect, plausible confounding, and doseresponse gradient. Level of evidence would present “high”, “moderate”, “low”, and “very low” by the change of factors of downgrade and upgrade quality of evidence.
Data extraction The included trials were reviewed using a list of predefined questions that concerned the basic information of patients, patients’ inclusion criteria, random group methods, blinding methods, intervention methods, outcome data, and so on. Data extraction was conducted by two independent investigators (F. J. Li and X. T. Zeng). Data sheets were compared and discussed with a third review author (Y. Guo). The authors of the studies included will be contacted for missing information if necessary. To decrease the risk of overly positive answers, we will use open-ended questions when asking trial authors for information about study design and conduct. For example, to obtain information about blinding, a request of the following form might be appropriate: “Please describe all measures used, if any, to ensure blinding of trial participants and key trial personnel from knowledge of which intervention a participant had received.” To obtain information about the randomization process, a request of the following form might be appropriate: “How did you decide which treatment the next Trials searched through Cochrane Library n=15
RCTS (Randomized controlled trials) n=28 Trials unrelated to orthodontic pain n=7 Trials of checking duplicate n=9
RCTs after evaluation of effective n=12 Data duplication n=1 Eligible RCTs for systematic review n=11
76
60
59
88
60
84
37
120 64:56
Turhani et al., 2006
Tortamato et al., 2009
Domínguez et al., 2013
Kim et al., 2013
Nóbrega et al., 2013
Harazaki et al., 1997
Eslamian et al., 2014
Marini et al., 2013
NA not available, VAS visual analog scale
12:25
27:57
NA
23:65
19:40
18:42
30:46
NA
39
Lim et al., 1995
Orthodontic tooth movement
Orthodontic tooth movement
Orthodontic tooth movement
Orthodontic tooth movement Placing separators
6, 24, 30, 72, 96, 120, 144, 168
336
2, 6, 24, 72, 120
Placing separators
Orthodontic tooth movement Placing separators
Placing separators
5 min, 1, 6, 12, 24, 48, Placing separators 72, 96, 120, 144, 168
2, 6, 24, 48, 72, 168
No exact time point
6, 30, 54
0, 24, 48, 72, 96
24, 72, 720
Mean 23.01 12, 24, 36, 48, 72, 96
11–32
11–34
12–26
Mean 22.7
Mean 24.3
Mean 15.9
Mean 23.1
21–24
12–23
G4(23): LED treatment G1 (20): control side and did not receive laser but placebo G2 (20): experimental side and receive laser Within each subject, a multigroup pretest-posttest design (randomly laser treated for 15, 30, 60, and 30 s placebo) was used. G1(38): LLLT (lower level laser), maxilla (19); mandible (11); both jaws (8) G2(38): control (placebo), maxilla (21); mandible (10); both jaws (7) G1 (10): experimental group maxilla (EU) G2 (10): experimental group mandible (EL) G3 (10): blind group simulated laser maxilla (BU) G4 (10): blind group simulated laser mandible (BL) G5 (10): control no laser maxilla (CU) G6 (10): control no laser mandible (CL) In Ovation C brackets (29): one arch was irradiated with the laser; the opposite dental arch was placebo treated. Equilibrium brackets (30): one arch was irradiated with the laser; the opposite dental arch was placebo treated. Laser irradiation group (28) LED placebo group (30) Control group (30) Intervention group (30): laser irradiation Control group (30): placebo infrared light radiation Pseudo-irradiation group Blank control group The choice of the first molar to receive a laser irradiation or to be a placebo was determined randomly. U group (60): laser group (20), placebo group (20), control group (20) L group (60): laser group (20), placebo group (20), control group (20)
G2
G3(12): laser treatment
G1 (17): no treatment G2 (38): placebo group
8:12
Placing separators
Doshi-Mehta et al., 2012 20
2, 24, 48, 72, 96, 120
39:16
55
Esper et al., 2011
24.1±8.1
No. Gender (M:F) Age (years) Pain evaluation time (h) Pain induction methods Group methods (patients)
Study
Table 2 Basic information of included studies
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
VAS
Pain measure
Lasers Med Sci
InGaAlP
Semiconductor (aluminum gallium arsenide) diode Class 3B Ga-As-A1 diode
Esper et al., 2011
Doshi-Mchta et al., 2012 Lim HM et al., 1995
Twin Flex Evolution®, MMOptics, Classe 3b, registry ANVISA 80051420014 LA3D0001.1; LAMBDA S.p.A., Vicenza, Italy
Model (manufacturer)
30
25
Treatment time per point(s)
ED energy density
P-Laser System International, Egedalsvej, Vekso Each subject being exposed to three treatment time 15, 30, and 60 s Turhani et al., 2006 Semiconductor laser Mini Laser 2075 [dental version], Helbo 30 Photodynamic Systems GmbH & Co KG, Linz, Austria Tortamato et al., 2009 Class 3B gallium-arsenic-aluminum BioWave LLLT, Kondortech, São Carlos, Brazil 16 diode laser Domínguez et al., 2013 GaAlAs laser Photon Lase II® DMC Equipamentos, 44 s/arch São Carlos, Brazil Kim et al., 2013 semiconductor laser device with an Not reported 30 AlGaInP diode Nóbrega et al., 2013 AlGaAs device (class 3B) Kondorthec—São Carlos, Brazil 25 Harazaki et al., 1997 He–Ne gas laser Not reported 1 min/tooth, 12–24 min/patient Eslamian et al., 2014 Ga-Al-As laser DMC Equipamentos, São Carlos, Brazil 20 Marini et al., 2013 GaAs diode laser Lumix 2 HFPL, Fisioline, Verduno, Italy 34
Laser type
Study
Table 3 Technical features of laser for pain used in the included trials
800
660
Wave length (nm)
0.5 80
Not reported Not reported 1or 2 5 Not reported Not reported 2 Not reported Not reported 10.88
30 100 6 40.6 Not reported 100 160
Not reported
4.8
2.25
4.2
75
810 910
830 632.8
635
830
830
670
0.45, 0.95, 1.8 830
0.042
0.735
0.9, 1.79, 3.58
5.0
4.0
Total dose per tooth (J)
30
0.7
30
Output power ED (J/cm2) (mW)
Lasers Med Sci
Lasers Med Sci
patient should get?” More focused questions can then be asked to clarify the uncertainties remained.
Results Search results Figure 1 describes the procedure of selecting eligible trials. After review of the titles and abstracts, one trial [43] was excluded due to data duplication with authors’ another trial [21]. Finally, eleven studies [21, 23, 24, 30, 38–45] were included. Description of the included studies Eleven RCTs studied on the low-level laser for orthodontic pain were included. The eleven studies were conducted in Brazil, India, Japan, Korea, Iran, Singapore, Austria, and Colombia, respectively. All of the 11 trials used the visual analog scale (VAS) for the evaluation of pain intensity, which was considered as an effective method to balance the change of pain scale after the usage of LLLT. A 10 cm VAS tool with graded linear horizontal scale was used to assess the pain, which was a method previously used in many literatures, and had been shown to be a reliable and accurate tool in evaluating subjective experiences, such as pain, to meaningfully compare groups or individuals. This method provides additional variety in responses and gives a more sensitive and accurate representation of pain intensity than do descriptive pain scales [46]. So, we extracted the data on the VAS to describe the detailed application effect of different LLLT system. A brief qualitative analysis of the evidence was provided in the narrative forms (Tables 2 and 3).
Fig. 2 Risk of bias graph: review authors’ judgments about each risk of bias item presented as percentages across all included studies
patients knew whether LLLT or placebo had been given in this trial. Five trials (Turhani D 2006, Domínguez A 2013, Kim WT 2013, Eslamian L 2014, and Gauri Doshi-Mehta 2012) used single-blind setting; the operator who applied the laser treatment and the placebo could distinguish between them, but the patients were blinded to the difference. One trial (Esper MA 2011) did not describe the double-blind setting clearly. None of the 11 trials stated the existence of selective reporting clearly, so they were considered as unclear in this item. In one trial (LIM HM 1995), dental students were selected for the study, which reduced the emotive and attitudinal variables of pain responses and thus provided a more objective assessment of the treatment. So, the conclusions drawn from this sample might not fully apply to the general orthodontic patient
Results of methodological quality assessment All the 11 studies were presented as randomized. However, one trial (LIM HM 1995) revealed that the sequence of application of the four treatment parameters was randomized among the subjects by the Latin square method. A randomly assigned incomplete block split-mouth design (Doshi-Mehta G 2012) was used to prevent interindividual biologic variation. Two trials’ randomization procedures were performed by using a computer program (Domínguez A 2013, Nóbrega C 2013). One trial’s randomization procedures were performed by using a coin flip and envelope containing the number of the group that each subject was assigned to (Marini I 2013). The purpose of these procedures was to ensure maximum allocation concealment. Double blinding was described in five trials (Harazaki M 1997, LIM HM 1995, Nóbrega C 2013, Marini I 2013, and Tortamano A 2009); neither the personnel nor the
Fig. 3 Risk of bias summary: review authors’ judgments about each risk of bias item for each included study
Lasers Med Sci Fig. 4 Forest plot of comparison: laser versus control. Outcome: score of the most painful day— pain associated with placing separators
population, and we considered it as a potential source of bias. (See Figs. 2 and 3 for overview). Effect of LLLT for orthodontic pain The score for the most painful day Continuous data for “the score for the most painful day” was available in five RCTs [24, 38, 39, 41, 44]. Depending on the difference of pain induction, we separate the articles into two groups: first, those of pain associated with tooth movement and second, those of pain associated with placing separators and then evaluate the articles. The results of meta-analysis and GRADE level of evidence were shown in Figs. 4, 5, 6, and 7 and Tables 4, 5, 6, and 7. Figure 4 showed that the score for the most painful day (pain associated with placing separators)
Fig. 5 Forest plot of comparison: laser versus placebo. Outcome: score of the most painful day— pain associated with tooth movement
Fig. 6 Forest plot of comparison: laser versus placebo. Outcome: score of the most painful day— pain associated with placing separators
Fig. 7 Forest plot of comparison: laser versus LED. Outcome: score of the most painful day—pain associated with placing separators
wasn’t significantly reduced in the laser group (MD=−1.68, 95 % CI range −4.21–0.86, P=0.19), however with high heterogeneity (χ2 =14.04, P=0.0002, I2 =93 %). Table 4 showed the GRADE level of evidence of comparison: laser versus control (pain associated with placing separators) with a very low quality for five downgrade factors. In the comparison of laser versus placebo (pain associated with tooth movement), all two trials showed reduced pain score (Fig. 5) (MD = −4.39, 95 % CI range −5.9– −2.88, P
