Eligibility criteria
The systematic review part was carried out according to the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analysis) guidelines[19]. The “PICO” strategy for systematic exploratory review provides guidance to the development of research contents[20]. This systematic review and meta-analysis aimed to answer the following question: whether CPP-ACP provides any remineralizing benefit superior to that of nonintervention or placebo? The PICO strategy in clinical research was: P (patients with early enamel carious), I (CPP-ACP therapy), C (non- CPP-ACP therapy applied or placebo), and O (WSLs regression). In clinical research,the PICO strategy was: P (artificial carious lesions of extracted teeth), I (CPP-ACP therapy), C (non- CPP-ACP therapy applied), and O (surface roughness microhardness and recovery degression).
Search strategy
An electronic systematic literature search covered the electronic databases: Cochrane Center Register of Controlled Trials, PubMed, EmBase, and Ovid in May, 2019 in English and without time restrictions. Additional records were identified by searching reference lists of included studies. The medical subject headings (MeSH) words and free text words were included during the search. “Casein phosphopeptide-amorphous calcium phosphate nanocomplex,” “CPP-ACP,” “GC tooth mousse,” “Recaldent,” “milk derivate,” “casein derivate,” “dental caries,” “enamel demineralization,” “white spot lesion,” “remineralisation,” “RCT,” “Randomized Controlled Trials,” “Controlled Clinical Trials,” “Equivalence Trial,” and “Pragmatic Clinical Trial” were used in combination with other strategies (more details in Additional file 1: Table 3).).
Based on the titles and abstracts, initial screening of the retrieved studies was carried out. After the removal of the duplicated and obviously irrelevant studies, full texts of potential interests were reassessed and only those meeting inclusion criteria were included. This work was accomplished by two reviewers (X.M. and X.L.), independently. When any disagreement occurred, a third reviewer (F.X.) was consulted and a decision arrived at by consensus after the issues solved.
Inclusion criteria
Study inclusive criteria: (1) randomized controlled trials, retrospective and prospective studies, which were placebo-controlled or blank-controlled and had a parallel-group design. (2) Participants for in clinical study including patients with early enamel carious had to be randomized to test or control groups. Participants for in vitro study using extracted human teeth had to utilize teeth free of any enamel defects, microcracks, or caries. (3) Interventions had to use remineralizing agents based on CPP-ACP. Any kind of product containing CPP-ACP could be included in this meta-analysis, such as MI Paste or Tooth Mousse.
Exclusion criteria
Studies containing the follow criteria were excluded in this meta-analysis: (1) Irrelevant studies. (2)The outcomes of studies were not quantitative primary outcomes but descriptive analysis. (3) Participants of in vitro studies were non-human animal teeth, such as bovine teeth and mouse teeth.
Quality assessment and data extraction
The Cochrane Collaboration methodology was used to assess the risk of bias of every retrieved study included. The assessment tool included random sequence generation, allocation concealment, blinding of assessment, incomplete outcome data, selective reporting, and other possible sources of bias so as to appraise the methodological quality of included studies. Bias in every study was classified as “low risk of bias,” “high risk of bias,” and “unclear risk of bias.” Cochrane Review Manager Version 5.3 (The Nordic Cochrane Centre, Copenhagen, Denmark) was used to generate risks of bias Figures.
The following information and data were extracted by two authors (X.M. and X.L.) independently from filtered studies, consisting of research features, contributor information, and major outcomes. The research features included publication date, the name of the first author, follow-up period, and type of intervention. The contributor information of in clinical experiments included demographic factors (sex and age), sample size in each group, and location of lesions, while as to the contributor information of in vitro experiments, demographic factors were replaced with tooth position. When a disagreement occurs, a third reviewer (F.X.) reaches a decision.
Statistical analysis
The Cochrane Handbook for Systematic Reviews of Interventions was used to conduct the statistical analysis [21]. This meta-analysis of randomized control trials (RCTs) was performed to evaluate the effectiveness of CPP-ACP for WSL treatment. The data type for the outcome measurement was mainly continuous data. To avoid errors caused by different measuring instruments, the SMD was used instead of MD with a 95% CI to generalize the effectiveness of treatment in each report. P-values were used to test the heterogeneity across studies. For P < 0.05, the data is considered significantly heterogeneous. In the meantime, the degree of inconsistency of the statistical analysis was assessed by I2 [22]. The new quantity I2 has the range 0% to 100%; the values 25%, 50%, and 75% represent low, moderate, and high heterogeneity, respectively [23]. If all the included studies showed good homogeneity, the fixed effects model was used. When the clinical and methodological heterogeneity was high or P < 0.05, we used the random effects models to combine the studies [23]. RevMan statistical software version 5.3 (The Nordic Cochrane Centre, Copenhagen, Denmark) was used to conduct the statistical analyses. If there were 10 or fewer studies, publication bias was not assessed, because more than 10 studies are required to check funnel-plot asymmetry [21]. Sensitivity analysis was performed by the leave-one-out approach in this review. The analysis was carried out using STATA version 14.1 (StataCorp, College Station, Texas, USA).
Abstract: 1. Do you restrict the included studies in English? If not, why there is a restriction in language in your method? (Method part, line 2).
Exclusion criteria Duplicate studies have been removed at identification stage already. The exclusion criteria is mainly for exclude studies after identification stage, like title and abstract screening, and full-text reading stage.