Evidence-based fact checking for selective procedures in restorative dentistry

Similar to other dental specialties, there are many clinical procedures in restorative dentistry that may or may not be supported by good evidence. Thus, the effectiveness of these procedures is uncertain. The aim of this paper is to reduce this knowledge gap by critically inspecting selective procedures in restorative dentistry and exploring if these well-established or widely advocated treatment modalities are necessary for improving treatment outcomes based on the best available evidence. A MEDLINE search was conducted to identify research on selective procedures while focusing on clinical trials and systematic reviews. Due to their practical relevance in the decision-making process, cost-effectiveness analyses were also included. Mixed results were identified regarding the included interventions. Some procedures had adequate evidence supporting them while others were mostly based on beliefs. A critical review of the available literature indicates that some common restorative procedures lack adequate support from high-quality research evidence. This paper attempts to highlight the need to critically examine the scientific validity of traditional knowledge and techniques through the context of current research evidence. This will not only help generate consensus between educators, clinicians, and researchers regarding restorative procedures but will also lead to improved patient care and outcomes.


Introduction
Evidence-based dentistry (EBD) is an approach to oral healthcare that requires judicious integration of systematic assessments of clinically relevant scientific evidence, relating to the patient's oral and medical condition and history, with the dentist's clinical expertise and the patient's treatment needs and preferences [1]. Many clinical procedures in restorative dentistry may not be supported by good research evidence. Hence, the effectiveness of these procedures remains uncertain. Evidence-based practice addresses the dynamic nature of healthcare as it requires clinicians to evaluate the appropriateness of a particular treatment based on the best available current evidence [2]. This ensures prioritization of the most effective treatment options thereby improving outcomes of clinical care. However, it cannot be assumed that newly generated evidence, however compelling, will immediately produce a significant change in clinical practice and this transition of scientific evidence to clinical practice is complex, slow, and sporadic [3]. Majority of roadblocks exist that fall in one or more of the three areas summarized below [4]: a. Don't know: Could be due to practitioner's general ignorance (lack of initiative to learn more or lack of access to high-quality evidence) or willful ignorance (practitioner chooses not to learn more perhaps because it challenges their current beliefs). b. Can't do: Could be due to practitioner's lack of confidence in performing a new procedure. c. Won't change: Could be due to practitioner's level of comfort with previous treatment methods and distrust of evidence-based research.
An essential starting point for managing the problem of "don't know" is the availability of high-quality, evidencebased guidance on best clinical practices [4]. Hence, the aim of this paper is to reduce this knowledge gap by critically inspecting current clinical methods for selected restorative procedures. Geared towards practitioners seeking to practice EBD, we sought to explore if these well-established or widely advocated treatment modalities were necessary for improving treatment outcomes in restorative dentistry based on the best available evidence.

Methods
All literature searches were conducted using the MEDLINE (National Library of Medicine) database via the Ovid interface. Controlled vocabulary (Medical Subject Headings) and keywords for individual topic areas were utilized in combination with Comparative effectiveness or Survival or Prognosis. The following medical subject headings were used: rubber dams, dental caries, dental cavity preparation, dental restoration, dentin sensitivity, dentin desensitizing agents, dental disinfectants, tooth non-vital, post-and-core techniques, crowns, dental restoration failure, and tooth bleaching. In order to maximize sensitivity of the searches, the following advanced limits were selected in the Clinical Queries section: humans, reviews, therapy, prognosis, economics, and costs. Searches were restricted to articles published in English only. In order to minimize inherent methodological shortcomings of other study designs and to provide highest quality evidence to guide clinical decision-making, only randomized controlled clinical trials (RCTs), non-randomized controlled clinical trials (NRCTs), systematic reviews (SRs) of RCTs/NRCTs, and evidencebased practice guidelines were included. Cost-effectiveness analyses were also included due to their practical relevance to the decision-making process. Other study designs (in vitro, narrative reviews, case series, cohort studies, etc.) were mostly utilized to provide comparative or background information. Reference lists of eligible studies and review articles were explored, and forward citation searches were performed using Google Scholar.
Formal risk of bias assessments was not undertaken for individual clinical studies. Some systematic reviews utilized the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach in addition to the risk of bias assessment for individual studies [5]. The GRADE rating was included in the findings if the original systematic reviews reported it. GRADE assesses the quality (certainty) of the collective body of evidence (QoE) with reference to the overall risk of bias of the included studies, the directness of the evidence, the consistency of the results (heterogeneity), the precision of the effect estimates, and the risk of publication bias. It represents confidence that the treatment effects are appropriate to inform the recommendations and is rated as high, moderate, low, or verylow certainty (Table 1) [5]. In addition, it is important for clinicians to be aware how much trust they can place into those recommendations. Additionally, one clinical practice guideline by the International Caries Consensus Collaboration [6,7] assigned levels of endorsements (Weak or Strong recommendation) for or against different treatments based on the quality and quantity of evidence (i.e., level of certainty). These levels of recommendation were also reported along with their corresponding findings to aid in decision-making.

Rubber dam isolation for restorative treatment
Background information Successful dental restorations depend largely on isolating the tooth/restorative material against moisture during the procedure [8]. Rubber dam We are very confident that the true effect lies close to that of the estimate of the effect.

Moderate
We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect.

Very low
We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect.
use has been one of the most widely advocated isolation techniques taught in all North American dental schools for posterior teeth restorations [9, 10] that offer other potential benefits detailed elsewhere [11][12][13][14]. Additionally, most dental licensure regional testing agencies in the USA mandate its use during certain restorative procedures [15][16][17][18]. Many experts consider it to be an essential component of modern adhesive dentistry that can boost the quality of restorations [19][20][21][22]. Studies conducted on extracted teeth have observed significantly higher shear bond strength and reduced microleakage with rubber dam versus cotton rolls [23,24]. However, clinical trials assessing treatment outcomes with rubber dam use have yielded conflicting results [25][26][27][28]. Moreover, a survey of US general dentists estimated that only about 18% of dentists used a rubber dam for composite restorations [29]. A multi-center practice-based research study documented an even lower prevalence (12%) of rubber dam usage during operative procedures [30]. Furthermore, potential disadvantages of rubber dam use include concerns regarding patient acceptance, application time, cost, insufficient training, inconvenience, latex allergy, chipping of restorations/tooth structure, and mucosal damage [29,[31][32][33].
Findings from the evidence Multiple systematic reviews and meta-analyses have assessed the effect of rubber dam isolation on various types of restorations. Although one review concluded that rubber dam isolation may not be needed for longevity of anterior restorations if adequate relative isolation (cotton roll/saliva ejector) was ensured (21 studies) [34], other reviews found significantly better composite restoration retention rates in non-carious cervical lesions (NCCLs) (81 studies) [35] and significant reduction of fractures in Class II composite restorations with rubber dam isolation (59 studies) [36]. These three systematic reviews did not include any assessments regarding risk of bias or the collective quality of evidence and also utilized indirect comparison of parameters. Hence, the validity of findings is questionable. A 2014 systematic review found that rubber dam did not influence the longevity of restorations (4 studies; 2 studies in permanent teeth and 2 studies in primary teeth) [37]. Although this review assessed the risk of bias for individual studies, it did not assess the certainty of the collective body of evidence [7]. Additionally, NRCTs were incorrectly classified as being randomized.
A systematic review and meta-analysis by the Cochrane Oral Health Group found that the use of rubber dam may lead to a lower failure rate of NCCL restorations in permanent teeth compared with cotton roll isolation at 6 months (2 studies; GRADE QoE: low) but not at 12 or 18 months (1 study; GRADE QoE: very low). However, rubber dam use was associated with lower restoration failure rate of proximal atraumatic restorations in primary molars at 24 months (1 study; GRADE QoE: very low). Hence, the authors concluded that the effect of rubber dam use on restoration longevity remained undefinable and further high-quality research was required [31].
Complete vs. partial carious dentin removal Background information Traditionally, operative treatment of dental caries involves complete removal of carious tissue (infected and affected dentin) before restoration placement [38]. However, this nonselective dentin removal is often associated with irreversible consequences such as reduction in remaining dentin thickness, pulpal exposure, pain, and discomfort ultimately leading to weakened tooth structure and compromised pulpal health [39][40][41]. Contemporary alternatives to complete removal of carious tissue include no removal, selective removal, and stepwise removal of carious tissue [42,43].
No removal involves diverse techniques that do not remove any carious tissue. These include pit and fissure sealants for permanent molars, stainless steel crowns on primary molars that do not involve any tooth preparation (Hall technique), and other non-restorative cavity control strategies such as prescription fluorides and fluoride varnish [43]. Selective removal involves excavating carious tissue to achieve hard enamel and dentin at the peripheries of the lesion while leaving soft or firm carious dentin in the deepest pulpal aspect [41]. Stepwise removal involves selective removal of carious tissue and provisional restoration placement. Six to 12 months later, the provisional restoration and carious tissue are removed followed by placement of a definitive restoration [43].
Findings from the evidence A Cochrane systematic review concluded that in symptomless, vital, carious primary or permanent teeth, selective and stepwise removal of carious tissue reduced the risk of pulp exposure over complete removal (2 studies for selective removal; 4 studies for stepwise removal; GRADE QoE: moderate) [41]. Another systematic review found that selective and stepwise removal in deep carious lesions not only reduced the risk of pulpal exposure (7 studies; GRADE QoE: moderate) but also prevented development of post-operative pulpal symptoms (6 studies; GRADE QoE: low) compared with complete removal [44].
Two other systematic reviews reported that selective removal resulted in greater success of maintaining pulp vitality (5 studies) and better restorative outcomes (10 studies,) compared with both stepwise and complete removal [45,46]. However, the superiority of selective removal over stepwise removal was only supported by a low number of studies consisting of limited-quality patient-oriented evidence [45].
In 2016, the International Caries Consensus Collaboration published clinical recommendations for carious tissue removal. The Collaboration recommended that there was no need to remove bacterially contaminated or demineralized tissues close to the pulp. In teeth with deeper lesions (radiographically extending into pulpal third of dentine) and sensible (vital) pulps, preserving pulpal health should be prioritized. In primary and permanent teeth with deep lesions, selective removal of carious tissue should be performed (strong recommendation). In permanent teeth with deep lesions, stepwise removal might also be an option (strong recommendation). However, the second removal step may be omitted, as it increases the risk of pulpal exposure while adding additional cost, time, and discomfort to the patient [42].
Management of operatively exposed dentin before restoration placement Background information Despite recent advances in adhesive dentistry, some aspects still present as vexing issues for clinicians and researchers. Though bonding of glass ceramic indirect restorations has gained popularity over recent years, adhesion to dentin remains a challenge leading to debonding or fracture of restorations. Post-operative thermal sensitivity (POS) and secondary caries with resin-based composite restorations (RBCs) are commonly encountered in clinical practice. Incidence of POS after 24 h of RBC placement could be anywhere between 5 to 26% depending on the cavity configuration [47]. According to current concepts, POS has been linked to the stresses arising from composite resin polymerization shrinkage that may cause debonding and/or cuspal deflection; or to microleakage that leads to hydrodynamic fluid flow within the dentinal tubules or causes bacterial by-products to reach the pulp through the enlarged tubules [48,49]. Long-term enzymatic degradation of the adhesive-dentine interface by matrix metallo-proteinases (MMP) activated by acidic resin components applied during dentin etching eventually leads to in vitro microleakage [50] and secondary caries [51][52][53] formation with RBCs. Various solutions have been proposed and tested to address these issues and primarily involve some sort of management of operatively exposed dentin before restoration placement. These management strategies include use of liners/bases, dentinal desensitizers, cavity disinfectants, MMP inhibitors, and immediate dentin sealing.
Majority of North American dental schools teach the use of a liner/base for restoring teeth with deep carious lesions [9,10]. A survey of general dentists in the UK revealed that most practitioners (83%) always placed a liner under posterior composite restorations except in shallow preparations [54]. Most commonly used materials are hard-setting calcium hydroxide, conventional glass-ionomer or resinmodified glass-ionomer cement, or flowable composite resin [9, 10,54,55]. Apart from providing a better seal of exposed dentinal tubules while reducing microleakage and hydrodynamic fluid flow thereby reducing POS, use of a liner/base is purported to have several other advantages such as pulpal protection from toxic effects of restorative materials, stimulating reparative dentin formation in deep lesions, exerting a sedative effect on the pulp, and potentially improving restoration longevity [39,49,[55][56][57][58]. Researchers have argued that the clinical benefits of liners/bases may not live up to their theoretical value even in deeper restorations [59] thereby making their use an unnecessary step for composite restorations [49,60,61]. Furthermore, some have stated that not only are liners/bases unnecessary but they may limit the available surface area for bonding and reduce the thickness of the restorative material thereby affecting the biomechanical properties of the restoration [61][62][63].
Use of dentinal desensitizers was also proposed to prevent POS with RBCs. The desensitizer (35% HEMA/5% glutaraldehyde) could be either pre-applied to cavity walls or incorporated in the adhesive itself [64,65]. In vitro studies have revealed that HEMA physically seals exposed dentinal tubules while glutaraldehyde inhibits bacterial growth and exerts a fixative effect on dentinal fluid proteins thereby preventing fluid movement within the tubules and reducing POS [64].
Strategies for preventing hybrid layer degradation in vitro involve pre-treatment of cavity surfaces with MMP inhibitors that can inhibit bacterial growth in cavity walls and thus prevent the development of an inflammatory response thereby preventing POS with resin composites. In vitro studies have reported an increase in bond strengths by using these materials [66][67][68]. Most commonly used agents include chlorhexidine, proanthocyanidins (collagen cross-linkers), sodium hypochlorite, and diode lasers [68][69][70].
Immediate dentin sealing (IDS) has been proposed by some investigators for improving bonding of glass ceramic indirect restorations to dentin [71,72]. IDS involves applying an adhesive immediately after tooth preparation prior to impression taking and also before restoration cementation versus just before cementation. The proposed advantages include increased bond strengths, reduced marginal gaps and bacterial leakage, and prevention of POS [73][74][75][76]. Some investigators have also recommended using IDS in routine clinical practice given its potential advantages and no documented scientific reasons preventing clinicians from applying the technique [77,78].

Liners and bases
A systematic review by the Cochrane Oral Health Group concluded that there was inconsistent evidence regarding the difference in POS between RBCs placed with or without liners (3 studies; GRADE QoE: low). A benefit in favor of liners was shown only at 24 h (1 study) but this difference was not maintained during the 1-week or 1-month follow-up visits (3 studies). Additionally, there was no difference in restoration failure with or without the use of liners (4 studies; GRADE QoE: low) [49]. Another systematic review and network meta-analysis concluded that there was insufficient evidence to recommend cavity lining or the use of any specific liner based on their antibacterial effects (14 studies) [79]. A recent RCT not included in the above reviews also found that application of a liner did not influence the clinical performance of deep RBCs [80].

Desensitizers
A recent RCT did not find any significant differences in spontaneous or stimuli-induced POS with Class I and II RBCs with or without prior application of a glutaraldehydecontaining desensitizer [81]. Additionally, two NRCTs found no difference in POS in Class I and Class II RBCs with either the pre-application of a dentinal desensitizer or with the use of an adhesive incorporating glutaraldehyde [82,83].

MMP inhibitors/cavity disinfectants
Only one RCT found that chlorhexidine significantly decreased immediate POS in posterior RBCs at the 1-day follow-up visit. However, this effect was noted at subsequent follow-up visits (1 week and 1 month) [84]. Multiple other RCTs failed to demonstrate a beneficial effect of MMP inhibitors (chlorhexidine or proanthocyanidins) on the clinical performance of RBCs either when pre-applied to the cavity or when they were incorporated in the adhesive [69,85,86]. On the other hand, one RCT found that the incorporation of proanthocyanidins into the adhesive solution impaired the longevity of RBCs [87].
One systematic review reported that majority of included studies did not show any significant differences in outcomes with the use of different MMP inhibitors (4 studies). Two of the included studies had worse clinical outcomes whereas only one study demonstrated better outcomes [68]. Another systematic review that also included a meta-analysis concluded that there was insufficient evidence to recommend or refute hybrid layer degradation inhibitory cavity pre-treatment prior to placing RBCs (10 studies; GRADE QoE: very low) [51]. Additionally, the International Caries Consensus Collaboration stated that cavity disinfection procedures currently have no evidence of patient benefit to support their use (weak recommendation) [6].

Immediate dentin sealing for bonded indirect restorations
Only 2 RCTs were identified, and both failed to demonstrate any significant differences in clinical outcomes (restoration success, survival, POS, or patient satisfaction) with or without use of IDS [88, 89].

Restoration of endodontically treated teeth (ETT)
Background information Two main approaches commonly utilized to restore ETT are direct restorations (conventional fillings) or indirect restorations (onlays/crowns) [90]. Clinicians often face a significant predicament regarding the choice of restoration due to the varying amounts of tooth structure loss exhibited by ETT. In general, direct restorations are indicated in teeth with minimal coronal tooth structure loss whereas crowns/onlays are considered when larger amount of tooth structure is missing. Advantages of direct restorations include preservation of sound tooth structure, ease of placement, reduced treatment time, and costs [90,91]. On the other hand, crowns provide better protection and reinforcement of the remaining tooth structure thereby considerably enhancing the ability of ETT to withstand occlusal forces [90,92]. In certain clinical situations, posts (fiber or metal) may also be needed to provide retention for the final restoration [93,94].
Conflicting information regarding treatment considerations complicates restorative decision-making for ETT. Studies have emphasized the importance of ferrule (circumferential tooth structure that is 1.5-2.0 mm high and 1.0 mm thick) and if possible, preservation of coronal walls (residual coronal structure at least 3 mm high above the gingival level) for improving the stress-bearing ability and long-term survival of ETT [95][96][97][98][99][100][101][102]. However, some studies have indicated that the ferrule effect or the number of remaining walls does not significantly influence restoration survival or success [91,102,103].
Although crowns can provide sufficient protection against fracture, extensive removal of sound tooth structure during crown preparation can further weaken the tooth and may not significantly improve survival statistics of ETT when compared to conventional fillings [91,104]. The low modulus of elasticity exhibited by prefabricated fiber posts (Fib-P) while beneficial to prevent root fractures can lead to increased post-flexion during function causing post-debonding or postfracture. Contrastingly, rigid metal posts and in particular cast post-and-core (Cast-PC) due to its better adaptation can decrease post-debonding/fractures but cause more root fractures [105,106].

Ferrule and number of remaining coronal walls
Some studies have demonstrated that the presence or absence of ferrule did not affect the prognosis of ETT [96,101,103]. However, most studies concur that in addition to the ferrule, retention of coronal walls has a significant influence on survival and success [95,96,100,101,107,108]. ETT retaining all four coronal walls did not exhibit any failure regardless of the restorative procedure and failures increased as the number of coronal walls decreased [96,101]. These findings were also corroborated by a systematic review of RCTs which stated that the failure rates of the treatments on ETT may depend on the amount of remaining tooth structure and restorations in teeth without ferrules exhibited a high failure rate (4 studies) [109].

Direct versus indirect restorations
RCTs have demonstrated similar short-term (< 5 years) clinical performance when ETT with Fib-P were restored with direct restorations versus crowns [91,104]. However, those with indirect restorations demonstrated higher success and lower need for re-intervention [91]. Although lower quality studies suggest that crowns may lead to better longterm (10-20 years) survival, success, and cost-effectiveness [110,111], a systematic review from the Cochrane Oral Health Group found no significant differences in restorative success between direct restorations or crowns in the short term (≤5 years). However, due to the very-low quality of evidence, the authors stated that there was insufficient reliable evidence to determine the comparative effectiveness of a crown or a conventional filling until more evidence becomes available (1 study; very-low certainty) [90]. This was further supported by a more recent systematic review that included 1 additional RCT (2 studies; GRADE QoE: low) [92].

Use of posts
Evidence suggests that posts are efficacious in improving survival and success of ETT when substantial tooth structure is missing (less than two remaining walls; lack of ferrule) [96,101,107,111,112]. However, posts do not affect the prognosis of ETT when adequate tooth structure is present (two or more walls; adequate ferrule) [95,107,[112][113][114][115]. Similar findings were reported by a systematic review of RCTs concluding that post-retained crowns were associated with the most favorable outcomes in ETT with one to two remaining coronal wall, whereas post-free crowns were superior when greater tooth structure was available (4 studies) [109].
Regarding the influence of post material (metal versus fiber) on the clinical performance of ETT, individual studies have reported mixed results. Most studies demonstrated that there were no significant differences in outcomes with either material [95,[115][116][117] while some indicated that Fib-P might result in improved outcomes [105,118]. Although systematic reviews of RCTs revealed no difference between the post types in the short to medium term (6-10 years), the evidence was considered insufficient and unreliable highlighting the need for longer term studies [93,94]. Metal posts were the most cost-effective followed by Fiber-P and then Cast-PC [119].
Similarly, method of post-fabrication (prefabricated vs. customized) did not have a significant influence on survival or success in most studies [95,115,116,120]. Some studies indicated that active metal posts and Cast-PC may be associated with a higher incidence of catastrophic failures (root fractures) [105,106]. Additionally, reduced treatment time and procedural complexity along with associated cost savings may favor the use of passive prefabricated posts over Cast-PC.

Bleaching techniques
Background information Professional bleaching techniques can be broadly classified into at-home, in-office, and a combination of the two. At-home bleaching involves application of low-concentration (10%) carbamide peroxide (CP) or hydrogen peroxide (HP) in a custom-fitted tray that is worn overnight or during the day for 2 to 6 weeks. On the other hand, in-office bleaching uses a higher concentration of HP (30-40%) which is applied in the office for 45-60 min in 2-3 sessions. The combination technique, also called as the "jump-start" technique, can be used to boost the bleaching effect obtained through a single session of in-office bleaching by adding in the at-home component [121][122][123]. Results obtained during the in-office session can also be used to motivate patients to comply with the subsequent at-home bleaching regimen. Some studies have also suggested using light with in-office bleaching which can increase the dissociation of HP [124,125] and, consequently, increase the availability of free radicals that can boost the whitening effect [126].
Good long-term outcome stability without noticeable side effects has already been demonstrated with at-home bleaching [127][128][129]. However, long-term clinical trials evaluating the efficacy of in-office bleaching are scarce. Although athome bleaching has been considered more effective than inoffice bleaching, [130]. a NRCT reported similar outcomes with both techniques [131]. It has been suggested that the combination technique may result in improved color stability of in-office bleaching especially when the sole application of in-office bleaching does not result in satisfactory outcomes. [122,130,132]. However, in the past, studies have failed to demonstrate significant differences in outcomes with the combination technique versus sole in-office or at-home techniques [123,133]. More recent RCTs have observed that the combination technique resulted in significant improvement of the participants' perceived oral health and satisfaction [134,135]. in addition to reducing the treatment time by an average of 3.7 days [135]. Despite the marketing claims of improved whitening from light-activated bleaching, this association has been strongly questioned [136].
Obvious disadvantages of at-home bleaching are reliance on patient compliance and longer treatment time. On the other hand, disadvantages of in-office bleaching are increased tooth sensitivity and less stable bleaching outcomes [132,135]. There also have been concerns regarding the safety of higher concentrations of in-office bleaching products which could diffuse into the pulp chamber [137]. resulting in pulpal tissue changes and post-treatment sensitivity [138][139][140]. In teeth with thin enamel, such as mandibular incisors, it potentially could lead to pulpal necrosis [138]. Additionally, contact of higher concentration HP with the oral mucosa may cause chemical burns resulting in severe patient discomfort. Studies have suggested using low-(less than 20%) or medium-concentration (between 20 and 30%) HP products to decrease the risk and intensity of tooth sensitivity associated with the in-office component of the combined bleaching technique [141][142][143].

At-home bleaching
A Cochrane systematic review of RCTs found that athome bleaching with CP was more effective than placebo with mild adverse effects of transient tooth sensitivity and oral irritation (6 studies; GRADE QoE: very low). The same review found that at-home bleaching with HP was also more effective than placebo (1 study; GRADE QoE: very low). CP gel was more effective for whitening but showed similar tooth sensitivity and oral irritation as compared to HP gel (6 studies; GRADE QoE: very low). When different concentrations of CP with or without a desensitizer were compared, higher concentrations of CP resulted in lighter shade whereas the inclusion of a desensitizer showed significantly less teeth sensitivity compared to the groups without the desensitizer (7 studies; GRADE QoE: very low, 1 study; GRADE QoE: low). However, these findings should be viewed with caution due to the low or very-low quality of evidence because of low number of studies with high risk of bias [144]. Another systematic review concluded that 10% CP has a significantly lower risk and intensity of teeth sensitivity compared to higher CP concentrations without jeopardizing color change. Authors cautioned that the results should be interpreted with caution because most studies included in the meta-analysis were at unclear risk of bias [145].

In-office bleaching
A recent systematic review and meta-analysis found that bleaching with lower concentrations of HP resulted in less tooth sensitivity and better color change as measured with a spectrophotometer. However, no difference was detected between the different concentrations regarding subjective color measurements with the shade guide (7 studies). Included studies were considered to be at low or unclear risk of bias [142] Several systematic reviews have investigated the effect of different types of light sources in conjunction to in-office bleaching with different concentrations of HP. In a meta-analysis by He et al., subgroup analysis revealed no significant differences in immediate (within 1 day) (3 RCTs) or short-term (1-4 weeks) bleaching effect (2 RCTs) with a high concentration HP gel (25-35%) between light-activated systems versus no light. However, with lower concentrations of HP (15-20%), the light-activated system produced better immediate (2 RCTs) and short-term results (1 RCT). Use of a light led to a significantly higher incidence (4 RCTs) and intensity of tooth sensitivity (3 RCTs). Risk of bias assessments revealed five out of the nine RCTs with a low risk of bias, four RCTs with a moderate risk of bias, and two NRCTs with a high risk of bias [136] Another systematic review and meta-analysis of RCTs compared in-office bleaching with HP (15 to 38%) using a variety of light sources including halogen, LED, laser, and plasma arc lamps. The efficacy of in-office bleaching was not influenced by the use of light, regardless of the HP concentration in terms of spectrophotometer measurements (11 RCTs; GRADE QoE: moderate) or shade guide measurements (15 RCTs; GRADE QoE: very low). There were no significant differences in the risk (9 RCTs; GRADE QoE: moderate) or the intensity of tooth sensitivity (10 RCTs; GRADE QoE: low) with or without the use of light activation, regardless of the hydrogen peroxide concentration. The authors stated that the results must be interpreted with caution since variations in the protocol were not considered, including number of bleaching sessions, commercial product used, and the type of the light [146].
More recent systematic reviews also revealed no difference in bleaching efficacy with or without the use of light activation [147,148]. Additionally, there was no difference in the incidence of tooth sensitivity [147]. The use of laser light compared with other non-laser lights also did not reduce the incidence or severity of sensitivity and had no significant increase in terms of color change [149].

At-home vs. in-office bleaching
A systematic review and meta-analysis of RCTs found no significant differences between at-home or in-office bleaching in terms of the primary outcomes of risk and intensity of tooth sensitivity (5 studies) or for secondary outcome of bleaching efficacy (4 studies). Authors cautioned that the low number of studies included in the meta-analyses and the high heterogeneity of the studies may have contributed to this similarity, in addition to the high variability of protocols used for both bleaching techniques [150].

Combined bleaching techniques vs. sole application
A recent systematic review and meta-analysis of RCTs found no significant differences in color change as measured with a shade guide or a spectrophotometer with the combined bleaching technique compared to sole at-home bleaching (4 studies; GRADE QoE: low to very low). When compared to sole in-office bleaching, no significant difference in color change was detected with a spectrophotometer but significantly better results were observed in shade guide units with the combination method (2 studies; GRADE QoE: low to very low). A higher risk (3 studies; GRADE QoE: low) and intensity of tooth sensitivity (4 studies; GRADE QoE: very low) was observed with the combined technique as compared to at-home bleaching alone but not when compared to sole in-office bleaching (3 studies; GRADE QoE: low) possibly due to the higher hydrogen peroxide concentration used during in-office bleaching. These findings should be interpreted with caution due to the low or very-low quality of evidence for most of the outcomes [151].

Discussion
The findings of this paper are important because the information presented can be used to identify gaps in the current literature regarding selected restorative procedures. Thus, research efforts could be directed towards filling these lacunae. The contents of this review can be further utilized to conduct further high-quality data syntheses studies such as systematic reviews and meta-analyses or to develop evidence-based clinical practice guidelines. Through the findings presented herein, this paper attempts to highlight the need to critically examine the scientific validity of traditional knowledge and techniques through the context of current research evidence. This will not only help generate consensus between educators, clinicians, and researchers regarding restorative procedures but will also lead to improved patient care and outcomes.
The paper has several limitations. Though we tried to cover multiple topic areas in restorative dentistry, this list is not comprehensive and may have left out other crucial topics that have been recently reported, such as the role of adhesion strategy on the POS in posterior RBC restorations [152,153]; the effect enamel bevel on the retention of cervical RBC restorations [154]; and the use of endocrowns to restore ETT [155,156].
Another drawback of including multiple topics is the inability for an in-depth analysis of individual topics. Although conducting a comprehensive systematic literature review would have allowed for more robust conclusions, it would have been very resource intensive. Hence, we decided to be selective with the topics that we felt were clinically important and concentrate on the highest quality of evidence on those particular topics. This may have introduced selection bias, the magnitude of which might be difficult to estimate. Thus, the conclusions presented in this paper warrant further in-depth analysis of individual topics by means of a more comprehensive and systematic guideline development approach. The evidence syntheses thus undertaken should strictly follow guidelines for systematically conducting, reporting, and appraising included studies [157,158].
Apart from the information presented in this review, another resource worth mentioning is a no-cost online clinical decision-making tool that was recently developed by a group of researchers interested in implementing evidencebased care [159,160]. The purpose of the tool is to provide instant access to clinically relevant information from evidence-based publications on the treatment outcomes of restorative and various other dental procedures. Userfriendly tools such as these could further facilitate access to evidence-based resources and help generate consensus between educators, clinicians, and researchers regarding restorative procedures thereby leading to improved patient care and outcomes. Although it was out of scope of this paper, we want to emphasize that besides comparison of clinical outcomes between treatment techniques, studies should also include patient-reported outcomes such as quality of life and cost-effectiveness of treatments to further enhance decision-making to maximally benefit the patient.

Conclusions
Within limitations of the evidence, the following conclusions can be drawn: 1. Rubber dam isolation: a Improved survival of non-carious cervical lesion restorations in permanent teeth (GRADE QoE: low); However, this difference is not observed beyond 6 months (GRADE QoE: very low). b Improved survival of atraumatic restorations in primary teeth (GRADE QoE: very low). c No effect on survival in anterior restorations (GRADE QoE: uncertain).

Carious dentin removal
a.
Selective removal of carious tissue over complete removal prevents pulpal exposure and maintains pulp vitality in primary and permanent teeth with deeper carious lesions and vital pulps (GRADE QoE: moderate). b.
Although stepwise removal might also be an option for permanent teeth, it may increase the risk of pulpal exposure while adding additional cost, time, and discomfort to the patient (GRADE QoE: moderate).
3. Management of operatively exposed dentin a. At-home bleaching with CP or HP is more effective than placebo (GRADE QoE: very-low certainty). b. CP is more effective than HP and results in similar tooth sensitivity and oral irritation (GRADE QoE: very-low certainty). c. Inclusion of a desensitizer with CP decreases tooth sensitivity (GRADE QoE: low certainty).

In-office bleaching:
a. Lower concentrations of HP result in better bleaching efficacy (spectrophotometer assessment) and lower tooth sensitivity (GRADE QoE: low certainty). However, there is no difference in bleaching efficacy between different concentrations of HP when assessed with a shade guide (GRADE QoE: very-low certainty). b. Adjunctive use of a light for in-office bleaching: No difference in bleaching efficacy, regardless of the concentration of the bleaching agent or the type of light used (GRADE QoE: low certainty). The use of light could lead to greater tooth sensitivity (GRADE QoE: low certainty).

5.3.
At-home versus in-office bleaching: No difference in bleaching efficacy, incidence, and intensity of tooth sensitivity (GRADE QoE: very-low certainty). 5.4. Combination method (in-office plus at-home) versus sole application (in-office OR at-home): a. Combination method versus sole at-home: No difference in bleaching efficacy (GRADE QoE: low to very-low certainty) but leads to higher risk and intensity of tooth sensitivity (GRADE QoE: low certainty). b. Combination method versus sole in-office: Combination method may result in better color change in shade guide units (GRADE QoE: very-low certainty) and similar risk and intensity of tooth sensitivity (GRADE QoE: low certainty).
Data availability Not applicable.
Author contributions Sanket Nagarkar -full literature review, wrote the manuscript and response to reviewers. Alessandro D Loguerciowrote the bleaching techniques section of the paper, edited the manuscript. Jorge Perdigão -research idea, edited the manuscript, revised all versions.
Funding No funding received.

Declarations
Ethical approval Not applicable.

Conflict of interest
The authors declare no competing interests of financial or personal nature.